Too much of a good thing? : examining the optimum level of organizational identification required for strategies to be implemented most effectively

(1)

Examining the Optimum Level

of Organizational Identification

Required for Strategies to be

Implemented Most Effectively

Full name of author: Sophie Anne Boog Student number: 10340173

Date of submission and version being submitted: June 29, 2017 - Final version

The qualification for which you are studying:

Executive Programme in Management Studies – Strategy Track Name of the institution:

University of Amsterdam / Amsterdam Business School Name of first supervisor:

(2)

full responsibility for the content of this document. I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

TABLE OF

ABSTRACT

Building on the theoretical argument that alignment of employees’

behavior with organizational objectives depends upon social identifi

cation, we bring together distinct literatures to examine the nature

of the relationship between organizational identification (OI) and

employee willingness to contribute to strategy implementation

(SI). We also hypothesize that this relationship is mediated by job

satisfaction (JS) and that job autonomy (JA) moderates the proposed

curvilinear OISI relationship such that the inflection point after which

the relationship disappears are lower for lowautonomous jobs than

they are for highautonomous jobs. Based on survey data obtained

from 261 public sector employees in the Netherlands, we find that

the relation between organizational identification and employee

willingness to contribute to the implementation of the organizational

strategy is linear as opposed to curvilinear, significantly explaining

that this willingness will grow as organizational identification

increases. The moderating effect of job autonomy was not found to

be significant. The results did confirm our hypothesis regarding the

mediating effect of job satisfaction. The theoretical and managerial

implications of our findings in an organizational context are discussed.

Key words:

strategy implementation; organizational identification; job satisfaction; job autonomy

(5)

Wellformulated strategies have little value unless they are effectively

implemented (Noble, 1999). Implementing their processes effectively

resembles an imperative of success for competitive advantage

(Gomes, 2009). Strategy implementation is always associated with a

certain degree of change (Alexander, 1985) and employees’ support

for organizational changes affects the successful implementation of

the desired strategy.

In this paper, we therefore use the concept of change willingness, a concept drawn from change management (Metselaar, 1997), to examine the willingness of employees to implement strategies that are related to an organizational change process. Social identities play an important role for the attitudes and behaviors of employees. Organizational identification derives from social identity theory (Turner, 1975) and influences the knowledge, attitude and behavior regarding the organizational objectives (Ashfort and Mael, 1989). The level of organizational identification among members directly affects change process effectiveness. (Cohen, 1999). Most prior research on organizational identification focused on its positive effects, however. Although it has attracted relatively less attention, its potential negative effects on both organization-related and individual-related outcomes have been studied.

Employees’ willingness to change is necessary for successful implementation of changes in an organization, and such willingness is only effective if it results in behavior that supports the change. Change willingness is a concept that has been extensively investigated by Metselaar and Cozijnsen (2002). Metselaar (1997) used Ajzen’s model of “planned behavior” as a starting point to explain willingness to change. Ajzen (1991) developed his theory for behavior in general, but Metselaar (1997) applied it to the area of change management. The attitude towards change

(6)

is, for the biggest part, determined by the factor of uncertainty that it so often brings forth. Lewin (1958) defined resistance as the employees’ development of counter powers against changes. Whenever a certain change is introduced in an organization, resistance against this change also emerges (Goldstein, 1988). Resistance can be regarded as a natural tendency to repeatedly wanting to maintain the status quo; alternatively, the approach of Cozijnsen and Vrakking (2003) to the concept of willingness to change is that resistance does not always have to be an opposing force of unwillingness. For example, this resistance may also come from different ideas as to how the new organization should take shape.

The social identity approach (see for example Haslam, 2004 for an overview) will serve as a theoretical background for our assumptions. Previous research suggests that social identification in an organizational context helps to understand an individual’s behavior. Organizational identification derives from social identity theory and it has been an increasingly important domain of inquiry for scholars (e.g. Mael & Ashforth, 1992; Elsbach, 1999; Pratt & Foreman, 2000; Riketta, 2005). Since the release of Albert and Whetten’s (1985) seminal publication, interest in organizational identification has grown increasingly. Organizational identification can be defined as an employee’s perceived “oneness” with an employing organization and a feeling that they belong to it (Ashforth & Mael, 1989). In general terms, organizational identification refers to the extent to which organizational members define themselves with reference to their organizational membership (Ashforth & Mael, 1989). Organizational identification has a potential capacity to generate a range of both positive and negative employee and organization outcomes. Therefore, the underlying aim of this research is to empirically identify the optimum level of organizational identification required for strategies to be implemented most effectively.

Despite the significance of strategy implementation (i.e. change willingness related to strategy implementation) and organizational identification, however, these concepts have not been examined simultaneously. Noble (1999), for example, argues that relatively little research has been undertaken into the

(7)

influence of firm factors, such as culture, on strategy implementation success. Prior implementation research has been limited to the examination of factors and outcomes at the organizational level as opposed to key factors that influence individual-level commitment and success in strategy implementation (Noble, 1999). In particular, it is surprising to find a lack of research on the impact of organizational identification on strategy implementation. This support for organizational change of employees is widely viewed as critical to the successful implementation of organizational change strategies (e.g. Bartunek, Krim, Necochea, & Humphries, 1999). Organizational identification is increasingly attracting attention in management research because it is seen as a fundamental psychological state reflecting the underlying link or bond between the employee and the organization and, thus, potentially facilitates predicting and verifying many important attitudes and behaviors in the workplace (Edwards, 2005). Processes of organizational identification are extremely helpful to ensure that employees work towards the interests of the organization (e.g. Edwards, 2005). Employees with relatively high organizational identification may be more likely to contribute to strategy implementation because they are more likely to agree with the practices, procedures, and policies of the organization and it strategies and, thus, are expected to behave accordingly.

Taking into account the relatively poor level of knowledge in the field of strategy implementation in general and bearing in mind that thorough empirical research concerning this issue with relation to organizational identification has been almost completely neglected, this study intends to contribute to the development of this field by providing relevant insights into a number of factors linked with strategy implementation. More specifically, the value of this study can especially be found in its combined investigation of job satisfaction and job autonomy as factors to successful strategy implementation. In particular, this research offers value by systematically addressing the following research question: What is the effect of job satisfaction (JS) and job autonomy (JA) on the relationship between Organizational Identification (OI) and employee willingness to contribute to employee willingness to contribute to strategy implementation (SI)?

(8)

In this study, it is hypothesized that the relationship between organizational identification and strategy implementation is best represented as curvilinear, as opposed to linear. Organizational identification is expected to have a positive impact on strategy implementation, but only to a certain extent, beyond which the positive effect turns negative. By reviewing current literature and quantitative research regarding organizational identification and strategy implementation, this study presents a general impression about organizational identification and its influential role within an organization as well as its relation to employee willingness to contribute to strategy implementation. This study further intends to clarify this relationship based on two important organizational behavior constructs: job satisfaction and job autonomy. There are theoretical reasons to expect the curvilinear relationship between organizational identification and strategy implementation to vary (depending on levels of job autonomy). Therefore, the mechanisms underlying this relationship between employees and organizations has theoretical and practical implications for management and organizational literature. We propose a set of testable hypotheses describing the organizational identification-strategy implementation relationship, the variables driving employees within organizations, plus factors that facilitate these relationships. Our hypothesized model was tested using data from 261 employees within a governmental organization in the Netherlands. Data were collected from employees in the operational core for the purposes of testing the model.

To address the research question systematically the paper is structured into five sections. After the introduction, section two offers a literature review focusing on change willingness, a concept drawn from change management (Metselaar, 1997) and organizational identification, a concept drawn from social identity theory (Ashforth & Mael, 1989). In section three, we explain the empirical research approach by describing the questionnaire, sample, respondents and the way the survey was conducted. Section four describes the empirical findings, organized so as to provide answers to the aforementioned research question. Finally, in section five, we present the concluding comments, discuss the implications and limitations of the study and suggest some directions for future research.

(9)

2

REVIEW OF THE

LITERATURE

In this section, we first explore the conceptualizations of willingness

to contribute to strategy implementation and organizational identi

fication separately. We then consider the contribution of organizational

identification to employees’ willingness to implement strategies.

Drawing on change willingness, a concept drawn from change management (Metselaar, 1997), and on social identity theory (Ashforth & Mael, 1989), we provide a rationale for examining this relationship between the organization and its members. Next, we seek to locate strategy implementation within a relevant framework for management. In particular, we consider the mediating role of job satisfaction as both an outcome of organizational identification and a determinant of strategy implementation. Third, we investigate the importance of an autonomous work environment in facilitating the organizational identification-strategy implementation relationship. Here we draw on previous research and the notion of work behavior constructs to support our rationale for expecting job autonomy to have important moderating effects on the link between organizational identification and strategy implementation.

2.1 WILLINGNESS TO CONTRIBUTE TO STRATEGY IMPLEMENTATION: CONCEPTUALIZATION OF THE CONSTRUCT

Despite the widespread acknowledgement of the role of strategy in mediating organization’s interaction with its environment (e.g. Ansoff, 1965;Alfred D. Chandler, 1962; (Miles, Snow, Meyer, & Coleman, 1978), the execution of strategy is not as clear and understood as the formulation of strategy (Hrebiniak, 2005). It seems that strategic management researchers have focused more on the content and creation of strategy from a variety of perspectives rather than considerations about implementation. In particular, there is a lack of research on factors that

(10)

influence individual-level commitment and success in strategy implementation (Noble, 1999). This is surprising since the support for organizational change by employees is widely viewed as critical for the successful implementation of organizational change strategies (e.g.Bartunek et al., 1999). In this paper, we use the concept of change willingness, a concept drawn from change management (Metselaar, 1997), to examine the willingness of employees to implement strategies that are related to an organizational change process. This reasoning can be clarified since strategic implementation always involves a degree of change (Alexander, 1991; 1985), and in turn, the support of employees to organizational changes can significantly affect the successful implementation of the desired strategy.

The change management literature consists of many different approaches, strategies interventions and actions through which change can be implemented (e.g. Burke, 2017). However, the literature mostly covers the distinction between planned and emergent process of change (Bamford & Forrester, 2003; Todnem By, 2005). Planned change occurs through a process of rational goal-setting and follows the advanced formulated change objectives that are implemented in a top-down fashion. Central to this theory is the belief that change can be successfully planned. According to planned change theories, often based on the seminal work of Lewin (1951), organizations must go through a number of phases in order to successful change to a desired future state (Burnes, 1996; 2004). The emergent approach to change is a more devolved and bottom-up way to implement change (Todnem By, 2005). The emergent change approach does not consider change as a linear process, or an isolated event, but sees change as continuous and unpredictable. Whereas the content of change is the starting point in the planned approach to change, the content of change in the emergent approach is rather the outcome of an emergent change process. Employees in this view are not seen as passive recipients of the organizational change, but are stimulated to actively contribute to the change process (Russ, 2008).

(11)

Although the planned and emergent change approaches differ considerably, they both stress that the willingness to implement a change by employees is crucial. The main issue dealt within this thesis is not the strategy as applied by management but the skills, motivation, and attitudes of employees towards this change that are the source of actions. These are the actions that ultimately determine the success or failure of organizational change processes. We focus on employees at the operational core of the organization because they have primary responsibility for incorporating newly developed strategies into their daily work activities (Metselaar, 1997). Therefore, employees’ willingness to change can be defined as “a positive behavioral intention toward the implementation of modifications in the organization’s structure or work and administrative processes, resulting in efforts from the employee’s side to support or enhance the change process” (Metselaar, 1997, p.42). This definition of organizational change emphasizes the implementation of change to enhance the effectiveness and efficiency of the organization.

2.2 ORGANIZATIONAL IDENTIFICATION: CONCEPTUALIZATION OF THE CONSTRUCT

In order to learn more about how organizational identification might contribute to individuals’ willingness to contribute to strategy implementation, we first need to verify what organizational identification entails. There have been many ways in which organizational identification has been defined. Although organizational identification scholars across disciplines seem to correlate in the comparison of the “self” with the organization, there is still a lack of consensus surrounding the notion. The different key contributors (i.e. Ashforth & Mael, 1989; Dutton, Dukerich, & Harquail, 1994; Pratt, 1998; Van Dick, 2001) all presented slightly different conceptualizations of this phenomenon as part of their reviews. In many of these reviews, however, social identity theory (SIT) and self-categorization theory (SCT) take a central place. From these perspectives, organizational identification can be associated with definitions of the self in relation to the organization as a whole, combining the “perception of oneness with, or belongingness to the organization” (Ashforth & Mael, 1989, p. 22). This definition of

(12)

organizational identification formulated by Ashforth and Mael is used the most often (Riketta, 2005).

Organizational identification can occur in many different forms, as there are various ways in which individuals can derive their identity from the organization. Kreiner and Ashforth (2004) propose a model of identification that enables people’s identification to be defined through their attachment to an organization. This model identifies four types: identification, dis-identification, ambivalent identification and neutral identification. Several researchers on social identity processes in organizations have also noted that a distinction should not only be made based on different dimensions of identification but also between foci of identification in the context of work-related attitudes (cf. Becker, Billings, Eveleth, & Gilbert, 1996; Knippenberg & Schie, 2000). These foci consist of one’s identification with (1) their own career, (2) their working unit or group, and (3) the organization as a whole. This is consonant with the differentiation of Self-Categorization Theory between different levels of self-categorization (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). In this research, we focus on identification with the organization as a whole.

2.3 STRATEGY IMPLEMENTATION AND RELATIONSHIP WITH OTHER VARIABLES

2.3.1 ORGANIZATIONAL IDENTIFICATION AND STRATEGY IMPLEMENTATION The level of organizational identification among members directly affects change process effectiveness (Cohen, 1999). Ashforth and Mael (1989) argue that if the social identity of employees is defined in terms of their membership of a particular organization then they are likely to strive with relevant co-workers to achieve positive outcomes for that organization, consequently perceiving both the successes (and failures) of that organization as their own. As a result, it seems crucial to understand how the perceived evaluation of change is affected by the relation between the individual and the organization. It seems likely that the relationship between organizational identification and the degree to which employees are willing to contribute to strategy implementation can be

(13)

both positive and negative. The reason for this assumption is that organizational identification can either increase or decrease the strategy implementation effectiveness. Employees with low level identification are assumed to be related with careless indifference about strategic change strategies, and are therefore less willing to help implement the strategy. While on the other hand, too much identification (“over-identification”) is expected to be associated with reluctance to change. The aim of this research is to identify the optimum level of organizational identification required for strategies to be implemented most effectively.

A large number of studies have identified organizational identification as a powerful predictor of employees’ commitment to a specific organization (e.g., Mael & Ashforth, 1992; D. Van Knippenberg & Ellemers, 2003) and their willingness to engage in acts of organizational citizenship (Organ, 1988; 1997). These two positive correlations suggest that a high level of identification should increase one’s willingness to contribute to strategy implementation. This is in line with the finding that employees with high organizational identification make decisions that are consistent with organizational objectives (Patchen, 1970; Barney & Stewart, 2000), and several outcomes of organizational identification suggest that organizational interests are likely being met (Whetten & Godfrey, 1998). As such, employees with high organizational identification will understand and accept a change strategy as they would regard it as being useful in favoring the organization. Individuals are therefore assumed to pursue organizational strategies when they identify with the organization.

Just as high organizational identification is a predictor of one’s willingness to commit to a specific organization, so low identification is a strong predictor of the desire to disengage from and exit the organization — either physically or psychologically (e.g., Abrams, Ando, & Hinkle, 1998; De Moura, Abrams, Retter, Gunnarsdottir, & Ando, 2009; Van Dick et al., 2004). Neutral, under-identification or misidentification is dysfunctional for the individual and the organization. For the organization, it loses valuable resources in this worker, who would be far

(14)

less likely to engage in superior performance or extra-role behaviors (Kreiner & Ashforth, 2004).

On the contrary, Dukerich et al. (1998) reasoned that too much identification (“over-identification”) is just as much of a threat to the individual and the organization as under-identification. Highly identified employees are prone to work hardest to achieve organizational goals because these individuals perceive the organization’s goals as their own, which in turn stimulates them to work harder in order to behave in congruence with the organization’s values and norms (Avanzi, van Dick, Fraccaroli, & Sarchielli, 2012). Yet, when individuals associate themselves with their organizations too strongly, they are likely to develop a dysfunctional attachment in the form of over-identification. Ashforth, Harrison and Corley (2008), for example, proposed various problems as a result of over-identification. These problems include that individuals may be reluctant to raise valid objections and be resistant to change.

Schneider, Goldstiein, and Smith (1995) also highlight the dangers of too much identification for strategy implementation. The authors argue that if organizations employ a growing proportion of employees who share similar values and beliefs with the organization, then the resulting homogeneity will cause the organization to implement strategies that do not match environmental conditions. Resistance to change may be an obstacle to successfully implement reinvention initiatives based on how individuals and organizations perceive their goals to be affected by the change (Trader-Leigh, 2002). These arguments would suggest that identification and the underlying factors of resistance may affect implementation outcomes. Therefore, it can be expected that a too strong level of organizational identification negatively affects the effectiveness of strategy implementation.

High and deep-structured identification can also be detrimental to the acceptance of change when employees experience a threat to their identity. This happens when change alters other features that make the firm attractive, as such an identification target (e.g. distinctness or high public reputation). Employees might

(15)

respond to change with lowered work motivation, stress, or anti-citizenship behavior, for example sabotage (Riketta, Van Dick, & Rousseau, 2006).

Indeed, many organizational scholars tend to view identification in uniformly positive terms: the more the better. However, Dukerich et al. also acknowledged the threat to over-identification to both the individual and the organization. Thus, we expect a challenge for organizations to determine and attain the optimal balance of inclusiveness (identification) and distinctiveness (disidentification; cf. Brewer, 1991). In this sense, we hypothesize that the association between organizational identification and employee willingness to contribute to strategy implementation will be curvilinear and we therefore predict an inverted U-shaped relation. This means that identification has a positive impact on employee willingness to contribute to strategy implementation, but only to a certain extent, beyond which the positive effect turns negative.

Hypothesis 1: The relation between organizational identification and employee willingness to contribute to strategy implementation will be curvilinear, i.e. there will be an inverted U-shaped relation, with an initially positive relation (more identification relates to a higher willingness to contribute to strategy implementation) that turns into a negative relation once a certain threshold in identification is exceeded.

2.3.2 THE MEDIATING ROLE OF JOB SATISFACTION

Job satisfaction has been described as “a pleasurable or positive emotional state resulting from the appraisal of one’s job or job experiences” (Locke, 1976, p. 1304). Job satisfaction is an effective response by people toward their jobs (Cranny, Smith, & Stone, 1992). In fact, multiple researchers (e.g. Knippenberg & Schie, 2000) suggest that organizational identifications and job-related variables such as job satisfaction mutually affect one other. In light of this study, we only focus on the effect of organizational identification on job satisfaction. Past research in the field of both communication and social psychology view job satisfaction as a consequence of identification and suggest that employee identification with an

(16)

organization affects job satisfaction (e.g. Mael & Tetrick, 1992; Russo, 1998; Van Dick, 2001; Riketta & Van Dick, 2005; Riketta, 2005). For instance, Alexopoulos, Palatsidi, Tigani and Darviri (2014) found that employees value their actual work situation more positively when they have strong identification with their organization. This, in turn, leads to higher job satisfaction (Van Dick et al., 2004). Employees interpret their job as verification of their organizational membership, and therefore as “validating those parts of their self that stem from this membership” (Van Dick et al., 2004, p. 353). Similarly, matching characteristics provide employees an enhanced status that contribute to their self-concepts, resulting in stronger identification with membership in the organization which in turn elicits greater job satisfaction (Dukerich, Golden, & Shortell, 2002). In line with this reasoning, we opt for the following hypothesis:

Hypothesis 2: Organizational identification is positively related to job satisfaction, such that with growing organizational identification increases job satisfaction.

Job satisfaction, in turn, has been regarded to be an anterior measure advancing the intentions of employees to perform well on the job (Shore, Newton, & Thornton, 1990). Lee (1971) suggests that, where identification with the organization is the form of loyalty, this relates to attitudes and behaviors that support or defend the organization. These attitudes and behaviors include “supporting the organizational objectives” (Lee, 1971, p.215). Thus, employees’ motivation, skills and attitudes toward change essentially influence actions and ultimately, the failure or success of the implementation of change processes. According to MacMillan (1978) people are motivated more by their perceived self-interest than by the organizational interest unless these are congruent. In their research into the motivation of employees to implement a certain strategy, Guth and MacMillan (1986) found that if employees believe that their self-interest is being compromised then they are likely to redirect, delay or even sabotage the implementation.

Uncertainties with regards the consequences of change and the resulting impact of transformations on their work are of primary concern for employees. Other

(17)

common issues include concerns about what aspects of the change process will benefit or hinder employees in the organization, or if the usefulness of their work position will be maintained after the change process (Gomes, 2009). One of the most important elements of job satisfaction is the work situation. The most common used prognosticators for job and change satisfaction are the employees’ responsibilities, level of challenge and the type of tasks, or the sort of interactions that an individual might have during a day-to-day work (Huselid, 1995). These kind of aspects relate with the basis of many organizational change processes. Many organizational changes require employees to adjust their responsibilities and tasks with the reorganization of workplaces.

Hypothesis 3: Job satisfaction is positively related with employee willingness to contribute to strategy implementation. A higher job satisfaction leads to a higher willingness to contribute strategy implementation.

Hypothesis 2 and 3 subsequently lead to hypothesis 4:

Hypothesis 4: Job satisfaction mediates the relationship between organizational identification and employee willingness to contribute to strategy implementation. Specifically, we expect that higher organizational identification of employees will be positively related to their job satisfaction which will to an individual’s’ increasing willingness to implement a strategy.

2.3.3 THE MODERATING ROLE OF JOB AUTONOMY

Past research clearly demonstrates the importance of taking the work characteristics into account when looking at organizational processes. We consider the notion of job autonomy as a key requirement for employees to be able to behave in a manner consistent with their attitudes and beliefs. Job autonomy is the extent to which employees have the discretion, independence and freedom over the way they schedule their work, make decisions and select methods to perform tasks (Hackman & Oldham, 1976). In this study, we suggest that an autonomous work environment is essential for employees’ organizational identification to

(18)

foster their willingness to contribute to strategy implementation. Therefore, we predict job autonomy to moderate the curvilinear organizational identification-strategy implementation relationship such that the inflection point after which the relationship disappears is lower for low-autonomous jobs rather than it is for high-autonomous jobs. This would suggest that high levels of organizational identification are more beneficial for strategy implementation effectiveness in high- rather than low-autonomous jobs. As such, we belief that higher levels of job autonomy have a stronger effect on the relationship between organizational identification and employee contributions to implementing the strategy.

It is likely that high autonomous jobs require relatively higher levels of organizational identification than do lower autonomous jobs for strategies to be effectively and efficiently implemented. For example, most jobs of high autonomy involve a certain degree of freedom, independence, and discretion to schedule work, make decisions, and select the methods used to perform tasks (Hackman & Oldham, 1975). A high level of job autonomy allows employees to decide how to perform their work (Fried, Hollenbeck, Slowik, Tiegs, & Ben-David, 1999; Troyer, Mueller, & Osinsky, 2000) but only those employees that are identified will think and act in congruence with their organization’s objectives (Mullins, End, & Carlin, 2010). Relaxing strict behavioral controls allows employees to reciprocate particularly supportive organizational relations with behaviors that are of benefit to the organization. In addition, a high degree of job autonomy permits employees to act upon their level of identification with the organization by responding with favorable or unfavorable behaviors. Hackman and Oldham (1976) showed that core job characteristics including autonomy generate a sense of responsibility for outcomes, meaningfulness, and knowledge of results, which in turn elicits intrinsic work motivation and job satisfaction (both indicators of strategy implementation). Therefore, we expect that employees with highly autonomous jobs want to help to achieve the goals and objectives of the organization.

Controversially, when employees’ job autonomy is curtailed through overly formalized rules and procedures (and they cannot exercise control in carrying

(19)

out their work) they are less likely to identify with and show commitment toward their organizations. More specifically, the less control individuals have in doing their work (low level of autonomy), the more dependent they are on their employer (Semmer, 2000). Consequently, when the level of autonomy is low, a lack of organizational identification may lead to employees to have unfavorable attitudes toward their job, organization and profession, which may reduce their willingness to contribute to the implementation of a strategy. Moreover, a low level of job autonomy coupled with a low level of organizational identification may lead individuals to perceive the organization that employs them in a negative way, which may decrease their willingness to contribute to a strategy.

Given these findings, one might expect that the relationship between organizational identification and the degree to which employees are willing to contribute to to strategy implementation will be stronger when employees are provided with substantial job autonomy by their employing organizations. On the contrary, low autonomy is expected to limit the extent to which higher levels of organizational identification translate into higher levels of willingness to contribute to strategy implementation. Based on this rationale, we advance the following hypothesis:

Hypothesis 5: Job autonomy moderates the relationship between organizational identification and employee willingness to contribute to strategy implementation, i.e. the inflection point for more autonomous jobs occurs at higher levels of organizational identification than the inflection point for less autonomous jobs.

The aforementioned rationales for the proposed relationships between the diverse constructs lead to the below hypothesized model (see figure 1). This model proposes two direct and one indirect determinants of employee willingness to contribute to strategy implementation. Organizational identification will have a direct, yet curvilinear, effect on employee willingness to contribute to strategy implementation. Organizational identification is also posited as an indirect predictor of employee willingness to contribute to strategy implementation

(20)

through the effect on job satisfaction. As such, a multidimensional measure of job satisfaction is proposed as a direct predictor of strategy implementation. Finally, job autonomy is proposed to moderate the effect of organizational identification on employee willingness to implement a strategy.

Figure 1

Conceptual model of analysis.

Job satisfaction Organizational identification Job autonomy Employee willingness to contribute to strategy imple-mentation H2 + H3+ + H5 H1 H4 U

(21)

We tested the aforementioned hypotheses in a study, using cross

sectional design in a sample of governmental employees. The

analysis builds upon a set of primary data. In this section, we

describe the data source and the most evident characteristics of the

collected sample. The variables included in the questionnaire and

corresponding reliabilities are discussed and a brief description will

be provided of the statistical analysis that tested for the expected

relationships as discussed in the previous chapter.

3.1 RESEARCH SETTING

The research was conducted in the context of a Dutch public sector organization. Studying a single case allows us to gather in depth information and to capture multiple variables as to identify how the complex set of the proposed circumstances come together. A public organization has been selected specifically over private sector organizations as the focus in this study. This because although change management research has increasingly addressed change management in public organizations (Fernandez & Rainey, 2006; Fernandez & Pitts, 2007) most literature on organizational change has concentrated on major changes that affect private sector organizations (Tummers, 2011). In addition, most of these studies were case-based designs using qualitative methods (Kuipers et al., 2014). Quantitative research can help in hypothesis testing and statistical generalization and therefore we are able to explore the relationship between the variables in this study.

3

DATA AND

(22)

To enhance the efficiency and effectiveness of their operation, the public organization in this study has recently undergone major organizational changes. Small and independent regional divisions have been merged into larger supra-regional organizations. The realization of the reorganization arrives from a strategic change strategy and entails forthcoming strategic objectives. The present study covers five local units within one district that is part of one of the ten supra-regional organizations. Each local unit consists of approximately 120 employees who are responsible for the main operational activities, in other words the operational core, which are directly or indirectly involved with the strategic objectives of the organization. Employees at the operational core of the organization are primarily responsible for fitting newly developed projects into their daily work activities (Metselaar, 1997). Data were collected four years after the reorganization program of five years had started.

3.2 DATA COLLECTION

This study was conducted in a general government sector organization operating in the Netherlands. Data were collected via online self-administered questionnaires. The questionnaires were sent to the entire population of employees that were directly or indirectly involved with the reorganization. To investigate the views of the involved employees, questionnaires were sent via the five team chiefs to the approximately 600 employees of the five local units. Questionnaire invites included an introduction to the survey assuring anonymity and voluntary participation. The research aim has also been mentioned briefly. Participation in this study was on a voluntary basis. The survey was conducted in the first quarter of 2017 and resulted in 261 (N = 261) valid returned questionnaires, an approximate response rate of 43.5%. There were no statistical differences in the response rate of the local units, which ranged from 40.83% for the local unit with the lowest level of responses to 56.67% for the local unit with the highest level of responses. The total response time for the respondents was three weeks after the invites.

(23)

3.3 SAMPLE

The sample consists of Dutch employees working on a part-time (20.3%) or fulltime basis (79.7%) for a governmental institution. Of the 305 employees that started filling out the questionnaire, 260 respondents fully completed the questionnaire (mean completion rate 85.2%). Participants who failed to complete the questionnaire, i.e. those with a progress level of less than 99% (N = 44), were excluded in further analyses. A total of 261 useable questionnaires were collected out of which 177 (67.8%) respondents were male and 84 (32.2%) were female.

Our sample has good diversification in relation to age and tenure. Most participants fell in the age range of 25-35 years (35.6%). The least number of participants were found in the age of 65 years and above (0.4%). The minority of the respondents reported an organizational tenure of more than 40 years (1.2 %). The majority (21.1%) reported an organizational tenure of 16-20 years (for further specifications on the diversification of both age and tenure see table 0). The positions ranged from entry-level to the highest position, with the majority of the respondents reporting a middle position (73.2%) followed by a high position (15.3%). To guarantee the anonymity of the organization we cannot further specify the positions. The sample obtained matches the aforementioned characteristics of the population as a whole. In other words, this sample is representative as it is a fairly accurate reflection of the population from which the sample is drawn.

Table 0 - Sample diversification (age, tenure)

Age level Organizational tenure

16-24 years 4,2% 1-5 years 16,5%

25-35 years 35,6% 6-10 years 19,2%

36-45 years 26,1% 11-15 years 19,9%

46-55 years 21,1% 16-20 years 21,1%

55-65 years 12,6% 21-25 years 8,0%

65 years and above 0,4% 26-30 years 6,5%

31-35 years 3,4%

36-40 years 4,2%

(24)

3.4 MEASUREMENT OF VARIABLES

Apart from questions about the respondents’ background, the questionnaire included four topics: organizational identification, one’s willingness to implement the strategy, job satisfaction and job autonomy.

3.4.1 EMPLOYEE WILLINGNESS TO CONTRIBUTE TO STRATEGY IMPLEMENTATION.

Ten items were built for accessing this variable. The questions were based on the DINAMO-questionnaire of Metselaar (1997), a questionnaire for the assessment of willingness to change. The theory behind the DINAMO supposes that one’s willingness to change is a key element for the successful implementation of change. The DINAMO originates from a social psychological model, referred to as “Ajzen’s model of planned behavior”. Ajzen’s model has often been used to clarify and predict various types of behavior and which already has proven its usefulness. The starting point for the construction of the DINAMO is built on the idea that willingness to change forms a positive behavioral intention, whereas resistance forms a negative behavioral intention. Sample item includes: “I am willing to cooperate in the realization of this strategy.” Measures were recorded using a seven point scale from 1 (Strongly Disagree) to 7 (Strongly Agree). Cronbach’s coefficient revealed good internal consistency (α = 0.908).

3.4.2 ORGANIZATIONAL IDENTIFICATION.

Organizational identification was measured by a Dutch version of the six-item organizational identification scale developed by Mael and Ashforth (1992) translated by van Knippenberg (2001). Responses were given on a 7-point scale. Items included: “I am very interested in what others think about (name of organization)”, “When somebody criticizes (name of organization), it feels like a personal insult,” and “When I talk about (name of organization), I usually say “we” rather than “they”. Cronbach’s coefficient revealed good internal consistency (α = 0.820).

(25)

3.4.3 JOB SATISFACTION.

The Job Satisfaction Survey (JSS) is a questionnaire used to evaluate nine dimensions of job satisfaction related to overall satisfaction. This instrument is well established among the other job satisfaction scales. A shorter version of original scale was desired. Thus, the JSS was reduced to include only those job satisfaction facets measured in the JDI (i.e. supervision, coworker, pay, promotion, and the nature of work). A Dutch translation by Blom (2015) has been used to assess job satisfaction. Because the Job Descriptive Index (JDI) is the most frequently used facet measure of job satisfaction (Rain et al., 1991), it is particularly important to investigate the personal variability of these facets. Therefore, JSS items measuring communication, contingent rewards, fringe benefits and operating procedures were removed from the scale for the present study. The removal of these items resulted in a 12-item scale for the questionnaire used in this study (Cronbach’s α = 821).

3.4.4 JOB AUTONOMY.

To measure job satisfaction, we used a Dutch translation of the nine items related to job autonomy based on Morgeson and Humphrey’s (2006) Work Design Questionnaire (WDQ) by Gorgievski et al.(2016). Sample item “The job allows me to make my own decisions about how to schedule my work”. Cronbach’s coefficient revealed good internal consistency (α = 0.921).

3.4.5 CONTROL VARIABLES.

Control variables were included in the analysis for ruling out the other possible explanations for the significant relationships. These control variables were age, gender, position, unit location, and length of tenure. Riketta (2005) has shown that variables like age and gender can influence an individual’s level of organizational identification.

3.5 STATISTICAL PROCEDURE / DATA ANALYSIS

In order to test the aforementioned hypotheses we performed various analysis, divided into three distinct steps. As part of the initial analysis, we performed a varimax rotation, a correlation analysis and a factor analysis. The factor analysis

(26)

was performed to test whether the factors (the questions in the questionnaire) represent the four variables of organizational identification, employee willingness to contribute to strategy implementation, job satisfaction and job autonomy. In the actual first step, we performed a quadratic regression analysis to test whether organizational identification and employee willingness to contribute to strategy implementation were non-linearly related. Secondly, we performed Baron and Kenny’s (1986) test to statistically check if job satisfaction is a mediating variable. In the final step, we tested the moderator relationship of job autonomy. The steps are described in more detail below, before presenting the results.

3.5.1 INITIAL ANALYSIS

Before applying hierarchical polynomial regression analysis in step 1, we performed both a varimax rotation and correlation analysis to test for the means, standard deviations, inter-correlations and internal consistency alphas for all of the variables. From the varimax rotation analysis, the categorization of the four factors Q6-15 (SI), Q16-Q21 (OI), Q22-Q33 (JS) and Q34-Q42 (JA) was clearly visible. Some questions had to be reversed and therefore three questions were recoded (Q26, Q27 and Q32). In this study, the variables demonstrated high internal consistency for all constructs, i.e. employee willingness to contribute to strategy implementation, organizational identification, job satisfaction and job autonomy (α = 0.908, α = 0.820, α = 0.821, and α = 0.914 respectively). The result of the Cronbach’s Alpha test revealed alphas far above 0.5. Therefore, none of the scale items had to be removed from the scales.

We also performed a confirmatory factor analyses in the form of data reduction and data structuring to test the measurement properties of organizational identification, employee willingness to contribute to strategy implementation, job satisfaction and job autonomy altogether in order to assess both the construct and discriminant validity of these factors. Finally, within this step, we checked for internal consistency in the final results. From this confirmatory factor analysis result, the data reduction test showed that 67.4% of the variance was represented in eight out of 37 Likert-scale questions arriving from the questionnaire (see

(27)

table 1 in the appendix). From the confirmatory factor analysis in the form of data structuring, we recognized that the factor loadings of all observed variables or items are adequate ranging from 0.525 to 0.826 (see table 2 in the appendix). The factor loadings or regression weight estimates of latent to observed variable should be above 0.50 (Hair et al, 2006; Byrne, 2010). This indicates that all of the constructs conform to the construct validity test and that all items belonged to the specified core values.

3.5.2 TESTS OF THE HYPOTHESES

In each of the three steps that we conducted and that we will further specify below, we analyzed the effect of adding one of the tested variables to the control variables on the standardized regression coefficients together with the squared correlation coefficients (R2_{) first. An incremental and unique variance (R}2_{) should ensue as a}

result of adding one of the four variables for the hypotheses to be confirmed.

Hierarchical polynomial regression analysis was used in testing the first step. Analyses were conducted for organizational identification and employee willingness to contribute to strategy implementation. In Equation 1a below, organizational identification was entered into a regression model predicting employee willingness to contribute to strategy implementation. Equation 1b represents the inflection point occurs at the following value of the predictor organizational identification (Weisberg, 2005). The quadratic term of organizational identification (square of the organizational identification score) used, represents the hypothesized curvilinear effect (cf. Cucina & Vasilopoulos, 2005). A statistically significant effect of the quadratic term found in these equations would provide support for Hypothesis 1. More specifically, a negative quadratic term would suggest an inverted-U-shaped relationship, supporting hypothesis 1.

For a polynomial regression model SI:

(28)

the inflection point occurs at the following value of the predictor OI (Weisberg, 2005):

OIinflection = B1/ 2B2. (1b)

As such, the inflection point depends on the values of B1, the regression coefficient

of organizational identification, and B2, the regression coefficient of the quadratic

term of the organizational identification.

In order to test the mediation effect of job satisfaction within the nonlinear relation, we used multiple regression analyses to compute the instantaneous indirect effect. In our analyses, we estimated the instantaneous indirect effect of organizational identification on employee willingness to contribute to strategy implementation through job satisfaction, conditional on the specific value of organizational identification. We performed the mediation test by Baron and Kenny (1986) for which four relationships must hold simultaneously. That is, there should be a positive relation between organizational identification and employee willingness to contribute to strategy implementation (Equation 2a), there should be a positive relationship between organizational identification and job satisfaction (Equation 2b), there should be a positive relation between job satisfaction and employee willingness to contribute to strategy implementation and (Equation 2c) and finally, when we add the mediator (job satisfaction) as another variable to the relationship between organizational identification and employee willingness to contribute to strategy implementation, the significance of organizational identification should either decrease (partial mediation) or organizational identification should lose its significance (full mediation).

SI = B00 + B01OI; and (2a)

SI = B10 + B11OI and (2b)

SI = B20 + B20JS. (2c)

Step 3 tests the moderator relationship of job autonomy for hypothesis 5, suggesting that the inflection point at which the effect of organizational identification on

(29)

employee willingness to contribute to strategy implementation disappears depends on levels of job autonomy, was examined only if the corresponding quadratic effects were found in step 1 described above. These hypotheses were tested by including interaction effects between (1) job autonomy and organizational identification and (2) job autonomy and the quadratic term of organizational identification.

The model in step 3 mentioned before can be presented as

SI = B0 + B1OI B2OI2 +B3JA + B4JA•OI + B5JA•OI2 + ε, (3a)

with JA being job autonomy. Equation 3a can be rearranged as follows:

SI = B0 + B3JA + (B1 + B4JA)OI + (B2 +B5JA•OI2) + ε. (3b)

Calling B*0 = B0 + B3JA, B*1 = B1 + B4JA, and

B*2_{= B}2_{+ B}

5JA, (3c)

we can then rewrite Equation 3b above in the same format as in Equation 3a:

SI = B*0 + B*1OI + B*2OI2 + ε. (3d)

Accordingly, the inflection point for the model in Equation 3e is

OIInflection = -B*1 /2B*2. (3e)

From Equations 3c and 3e, it can be seen that the inflection point depends on B*1

and B*2, which are functions of JA when B4 and B5 are different from zero. As such,

Hypothesis 5 would be supported when the either B4 or B5 (or both) is statistically

significant. In other words, the statistical significance of any interaction term in Equation 3a would indicate that the inflection point at which the effect of organizational identification on employee willingness to contribute to strategy implementation disappears depends on job autonomy, as hypothesized.

(30)

Table 1 provides the means, standard deviations, inter-correlations and internal consistency alphas for all of the variables. On average, individuals’ scores on the four variables of employee willingness to contribute to strategy implementation, organizational identification, job satisfaction and job autonomy were valued well above the midpoint of a seven-point scale (M > 3.50). Employee willingness to contribute to strategy implementation was positive (M = 5.23, SD = 0.97). On average, the strength of identification in the studied organization was scored the lowest of all variables. Individual scores on this variable was slightly positive (M = 4.27, SD = 1.16). Job satisfaction was perceived to be moderately positive (M = 4.68, SD = 0.73). Finally, job autonomy was valued most positive (M = 5.28, SD = 0.86).

In table 2, the standardized regression coefficients are shown together with the squared correlation coefficients (R2_{) for the control variables. The magnitude of the}

coefficients relative reflects the importance of the variables. As can be concluded from table 2, the control variables on their own do not provide significant variance. Table 1 - Means, Standard Deviations, Pearson Correlations and Scale Reliabilitiesa_(N=262*).

Variable M SD N 1 2 3 4 5 6 7 8 9

1. Age 25-35 1.131 261

2. Gender n/A n/A 261 -0,015

3. Position n/A n/A 259 .362** .070

4. Unit location n/A n/A 261 .179** .076 .138*

5. Length of tenure 15,61 9,68 258 .769** .097 .424** .181** 6. Employee willingness to contribute to strategy implementation 5,225 .965 261 .230** -.178** .248** -.076 .182** .908 7. Organizational identification 4,268 1,155 261 .141* -.12 .06 .036 .011 .369** .820 8. Job satisfaction 4,684 .732 261 .151* -.105 .235** .072 .067 .411** .346** .821 9. Job autonomy 5,277 .856 261 .103 -.033 .219** .080 .106 .244** .157* .528* .914

a_{Reliabilities (coefficient alphas) for each scale are indicated in Italics.}

* p < .05, ** p < .01.

(31)

Table 2 - Regression analysis for control variables.

Dependent variable N R2 _{Independent variables} _β _P

Employee willingness to contribute to stra-tegy implementation 262 .140 Age .165 .044 Gender -376 .003 Position .135 .001 Unit location -.084 .036 Length of tenure -.001 .902

The first step of our data analyses revealed that, unlike predicted in Hypothesis 1, there was no significant quadratic effect to be found between organizational identification and employee willingness to contribute to strategy implementation (see table 3 and 4 below). The incremental and unique variance (R2_{) that should}

ensue as a result of adding organizational identification as a variable as the next step in the model for hypothesis 1 to be supported, is not found. This can be concluded based on the variance that did not increase remarkably (from .140 in table 2 to .243 in table 3). Neither did we find a significant quadratic effect between organizational identification and employee willingness to contribute to strategy implementation (p=0.456) in the regression analysis that we performed in step 1, see table 4 below.

Table 3 - Regression analysis for control variables, including organizational identification. Dependent variable N R2 _{Independent variables} _β _P

Employee willingness to contribute to stra-tegy implementation 262 .243 Age .083 .286 Gender -312 .018 Position .118 .003 Unit location -.089 .018 Length of tenure .007 .470 Organizational identification .276 .000

Table 4- Regression analysis of hypothesis 1.

Employee willingness to contribute to stra-tegy implementation

262 .138 Organizational identification .534 .082

(32)

Instead, we found that the relation between organizational identification was linear as opposed to curvilinear, significantly explaining that there was not an additional variance beyond the linear relation. The linear relation between organizational identification and employee willingness to contribute to strategy implementation is depicted in Figure 2. As can be seen, this grows along with the increasing level of organizational identification.

OI 7,00 6,00 5,00 4,00 3,00 2,00 1,00 SI 7,00 6,00 5,00 4,00 3,00 2,00 1,00 y=3,91+0,31*x R2_{Linear = 0,136} Page 1 Figure 2

Linear relation between organizational identification and employee willingness to contribute to strategy implementation. Note: The straight line represents the linear relation between variables.

When we next added job satisfaction as a variable to the control variables, the variance (R2_{) that ensued as a result increased slightly (from .140 in table 2 to .238}

in table 5).

Table 5 - Regression analysis for control variables, including job satisfaction.

Employee willingness to contribute to stra-tegy implementation 262 .238 Age .111 .146 Gender -.300 .010 Position .076 .057 Unit location -.095 .011 Length of tenure .005 .574 Job satisfaction .470 .000

The results of the regression on the instantaneous indirect effect of organizational identification on employee willingness to contribute to strategy implementation through job satisfaction as tested in step two, are presented below. Organizational identification has been found to be significantly positively related to employee

(33)

willingness to contribute to strategy implementation (β= 0.369, p = 0.000), see table 6 below. Organizational identification was also found to be positively related to job satisfaction (β= 0.348 p=0.000), see table 7, as well as job satisfaction was found to have a significant and positive impact on employee willingness to help implement the strategy (β= 0.403, p = 0.000), see table 8. Results confirmed our hypothesis for the mediator, with organizational identification, which had a significant impact on job satisfaction, linearly but not nonlinearly, with the positive relationship between organizational identification and job satisfaction as well as the positive relation between job satisfaction and employee willingness to contribute to strategy implementation and finally, the significance of organizational identification lost its significance when we added the mediator to the relationship between organizational identification and employee willingness to contribute to implementation (see table 9). Job satisfaction there was found to fully mediate the relation between organizational identification and employee willingness to contribute to strategy implementation, confirming hypotheses 2, 3 and 4.

Table 6 - Regression analysis of hypothesis 4.

Job satisfaction 262 .121 Organizational identification .348 .000

262 .162 Job satisfaction .403 .000

(34)

The incremental and unique variance (R2_{) that should ensue as a result of job}

autonomy as a variable in the control variable model for hypothesis 4 to be suppor-ted, is not found. The variance changed from .140 in table 2 to .153 in table 10. Table 10 - Regression analysis for control variables, including job autonomy.

Employee willingness to contribute to stra-tegy implementation 261 .153 Age .160 .047 Gender -357 .004 Position .110 .009 Unit location -.090 .024 Length of tenure -.001 .899 Job autonomy .215 .001

For the continuation of the third and final step, we only tested the moderation effect by including interaction effects between job autonomy and organizational identification. We did not perform tests on the interaction effects on job autonomy and the quadratic term of organizational identification because we could not find a curvilinear relation as predicted in Hypothesis 1 in the first place. The moderation effect by between job autonomy and organizational identification was not significant. The statistical significance of an interaction term in Equation 4a did not indicate that an inflection point at which the effect of organizational identification on employee willingness to contribute to strategy implementation disappears depends on job autonomy, as hypothesized, therefore Hypothesis 5 is not supported. Surprisingly, we could not find job autonomy to have any other, i.e. a lesser or more positive effect of organizational identification on employee willingness to contribute to strategy implementation, either. The betas in table 11 did decrease but not significantly.

Employee willingness to contribute to stra-tegy implementation 261 .131 Organizational identification -.042 .887 Job autonomy -.035 .879 Organizational identification*Job autonomy .062 .263

(35)

In sum, we can conclude that our proposed model is by no means perfect. Both hypotheses 1 and 5 were not supported. However, considering that our intention was to identify the optimum level of organizational identification required for strategies to be implemented most effectively, we alternatively could conclude for hypothesis 1 that the relation between organizational identification and employee willingness to contribute to strategy implementation is linearly related. Therefore, our model is sufficiently interesting to investigate further.

(36)

5.1 GENERAL DISCUSSION

Several conclusions concerning the research question and the corresponding hypotheses can be drawn. A first conclusion is that of job satisfaction mediating the relationship between organizational identification and the degree to which employees are willing to contribute to strategy implementation. Results, as summarized in table 12, confirmed our hypothesis for the mediator, with organizational identification having a significant impact on job satisfaction and, subsequently, a positive relation between job satisfaction and employee willingness to contribute to strategy implementation as well. This means that the more employees perceive organizational identification, the great their job satisfaction and the more positive their attitude to contribute to the organizations’ strategy implementation.

Surprisingly, the expected moderation effect of job autonomy on the relation between organizational identification and employee willingness to contribute to strategy implementation was not significant. Neither did we find significant support for job autonomy to strengthen the relationship between the organizational identification and employee willingness to contribute to strategy implementation.

Contrary to our expectations, this study could not provide support for the assumption of curvilinearity in the relationship between organizational identification and employee willingness to contribute to strategy implementation. The underlying aim of this research was to empirically identify the optimum level of organizational identification required for strategies to be implemented most effectively. Unlike expected, there was no threshold of organizational identification to be found that would alter the positive relation (more identification relates to a higher willingness to contribute strategy implementation) into a negative