The impact of organizational culture on management control systems

(1)

Amsterdam Business School

The impact of Organizational Culture on Management Control

Systems

Name: Kevin Kradolfer Student number: 10884289

Thesis supervisor: dr. ir. S.P. van Triest Date: 22 January 2017

Word count: 13068

MSc Accountancy & Control, specialization Control

(2)

Statement of Originality

This document is written by student Kevin Kradolfer who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

This study uses the survey method to investigate the influence of the perception of organizational culture on the use of management controls systems, specifically the use of performance measures. The results indicate that individual employees subscribe to different organization cultural values. All four cultural types on the flexibility/control and internal/external continuum exist within one single organization. The results also reveal that these different cultural perceptions don’t have a significant impact on the choice of financial, nonfinancial or subjective performance measures by employees.

(4)

1 Introduction

Organizational culture is relevant according to the management literature (e.g., Henri, 2006; Dent, 1991). However, empirical evidence to support a relationship between organizational culture and management control systems (MCS) is sparse (Bhimani, 2003). Henri (2006) is one of the first to empirically investigate whether organizational culture impacts MCS. His study aims to articulate and test the relationships between organizational culture and one component of MCS, namely performance measurement systems (PMS) (Henri, 2006, p 77).

This study builds on and extends early work by Henri (2006) who seeks and succeeded to establish an association between organizational culture and PMS. Henri investigates the relationship between organizational culture and two attributes of PMS, namely the diversity of measurement and the nature of use. Henri (2006, p 78) founds that organizational culture has a direct effect on the diversity of measurement of performance measurement systems. Since this early work, relatively little progress has been made in identifying the impact of this contingency variable on the design and use of MCS.

This thesis extends previous management accounting literature by examining the influence of organizational culture on the use of control systems. Henri (2006) operationalizes organizational culture using the control versus flexibility orientation. This study extends that orientation with the other main dimension of the Competing Values model of Quinn and Rohrbaugh (1983): internal versus external orientation. This leads to the operationalization of organizational culture in four types of cultural orientation, namely group, development, hierarchical and rational culture. For the operationalization of MCS Henri (2006) focuses on the diversity of measurement of performance measurement systems. This study investigates another aspect of the MCS, namely the subjective or objective use of PMS. This is an important characteristic of most incentive contracts for evaluating and rewarding employees. According to Gibbs et al. (2004) theoretical research has suggested various plausible reasons for the use of subjectivity in the assignment of bonuses. Empirical testing of these theories is rare. This study helps to fill this gap.

According to Ittner & Larcker (2002) relatively little empirical evidence exists on the factors associated with the choice of performance measures in incentive plans and most studies have focused on CEO bonus contracts or business unit performance evaluation and have ignored worker-level incentive plans. This study answers to their call for future studies to

(7)

examine the determinants of worker incentive plans focusing on the perception of organizational culture.

There are few accounts in PMS literature where lower and middle-level employees consider PMS as something that supports them and that they can use for their own purpose (Wouters and Wilderom, 2008). For this research, survey data is collected within a large Dutch insurance company. This offers the opportunity to link perceptions of organizational culture to the subjective use of PMS at the level of the individual employee. Specifically, this thesis will adopt a contingency approach and use empirical analysis to identify the influence of organizational culture on the subjective use of performance measures. Therefore this thesis examines the following research question: Does the perception of organizational culture influence the

subjective use of performance measurement systems? The measured variables include the perception of

organizational culture and the subjective or objective use of performance evaluation.

The remainder of this thesis is organized as follows. The next section describes the underlying theory and literature background. The following two sections present the hypotheses development and the research methodology including the sample definition, data collection and measurement of constructs. After that the results are analyzed followed by the discussion, limitations and conclusion. This paper ends with the references and appendices.

(8)

2 Theory and Literature Background

Management control is a critical function in organizations. Management control failures can lead to large financial losses, reputation damage, and possibly even to organizational failure. Therefore is widely accepted that good management control systems are important (Merchant and Van der Stede, 2012, p 5). One of the most well known theories underlying management control is the agency theory of Jensen and Meckling (1976).

In this thesis a contingency approach is adopted. There is no universally best management control system that applies to all organizations. The contingency theory of Chenhall (2003) describes the impact of contingency variables (i.e. strategy, external environment, technology, organizational structure, firm size and culture) on MCS design. As an organization strives to achieve effectiveness, it will seek to attain “fit” between the contingent variables and the MCS (Auzair & Langfield-Smith, 2005). The identification of contextual variables potentially implicated in the design of effective MCS can be traced back to the original structural contingency frameworks developed within organizational theory.

2.1 Agency Theory and Management Control Systems

The essence of the agency theory of Jensen and Meckling (1976) is that a principal (the shareholder) is not capable to entirely control the behavior of the agent (the manager or employee). This theory is based on the assumptions that the principal and agent should have a different risk characteristic, there is a conflict of interest between the principal and agent and there is information asymmetry. The agent has more information than the principal. This can lead to a misalignment of interests between principal and agent.

Management control is the process to mitigate this risk by influencing behavior of employees to increase probability that employees carry out organizational objectives and strategies, i.e. act in the organization’s best interest. Flamholtz et al. define control “as attempts by the organization to increase the probability that individuals will behave in ways that will lead to the attainment of organizational objectives” (1985, p. 35). The task of directing employee’s efforts for the attainment of organizational objectives has always been of the utmost importance. Organizations use a combination of mechanisms, including personal supervision, standard operating procedures, position descriptions, performance measurement and reward systems to control over the behavior of employees. Taken together, these mechanisms constitute the

(9)

organizational comprehensive management control system. As King and Clarkson (2015) propose, the MCS is part of the control solution.

MCSs have been characterized in many different ways. These include, for example, the conceptual framework of Merchant & van der Stede (2012). Their framework consists of a management control system composed of result controls, action controls and personnel and cultural controls. Result controls define performance measures and reward good results (and punish bad results). Employees are accountable for delivered results. Action controls focus on employee behavior to ensure that employees perform actions beneficial to the organization (or do not perform actions harmful to the organization). Personnel and cultural controls are bases on self-control and social control, based on intrinsic motivation and ethics. Other examples of often-used characteristics of MCS design in research are the levers of control of Simons (1994), more or less bureaucratic MCS (Auzair and Langfield-Smith, 2005) and the performance measurement systems framework by Ferreira and Otley (2009). Especially Ferreira and Otley use a very comprehensive approach on MCS (as they call it performance management systems). They conceptualize PMS as the “evolving formal and informal mechanisms, processes, systems and networks used by organizations for conveying the key objectives and goals elicited by management, for assisting the strategic process and ongoing management through analysis, planning, measurement, control, rewarding, and broadly managing performance, and for supporting and facilitating organizational learning and change” (2009, p. 264) resulting in an extended framework. Parts of their performance management systems framework are the key performance measures and the performance evaluation.

2.2 Performance Measurement Systems

Performance measures are used at different levels in organizations to evaluate success in achieving their objectives, key success factors, strategies and plans, and thus satisfying the expectations of different stakeholders. Agency models (e.g. Feltham and Xie, 1994) demonstrate that performance measurement systems should include any performance measure that provides incremental information on the actions the principal wishes to motivate in order to promote congruence between the principal’s objective and that of the agent. A common distinction is between financial (or accounting) measures and nonfinancial (or operational) measures.

Designing a perfectly complete PMS remains challenging, if not impossible, and would require nothing less than the expression of all relevant aspects of performance in quantitative terms (financial and nonfinancial), estimation of the tradeoffs among such dimensions of performance in the setting of targets for performance measures, and consideration of

(10)

interdependencies between different organizational units (Wouters and Wilderom, 2008). Performance measures used at lower levels in the hierarchy become increasingly noisy as interdependencies among units within the firm increase (Abernethy et al., 2004) because the indivisibility of certain resources makes the attribution of performance to individual employees or units within the firm increasingly difficult.

Nevertheless, the area of performance evaluation represents a critical part of control activities (Ferreira and Otley, 2009). Evaluation involves assessing the performance of individuals or groups against the pre-established goals and standards, based upon the information provided by the measurement system and the personal observation of the superior. Rewards are outcomes of behavior which are desirable to a person and which can be either extrinsic or intrinsic. Extrinsic rewards are administered after the evaluation process, usually conducted by the hierarchical superior. Rewards are expressions of approval and recognition by senior management through financial rewards (bonuses and salary increases). This relates to the core of the agency theory that it is necessary to provide an employee incentive to exert efforts in the best interest of the company and that these incentives are not provided by mere the payment of fixed salary (Jensen and Meckling, 1976).

The effectiveness of extrinsic rewards in channeling work behavior is reduced when inequity is perceived. Its effectiveness is further reduced when employees do not accept the evaluation based upon which reward decisions are made (Flamholtz et al., 1985). Acceptance of evaluation depends upon the perception that the evaluation is fair and valid. Performance evaluations can be objective, subjective, of fall in between these two extremes. The use of subjective evaluations has the important advantage of enabling evaluators to correct for identifiable flaws in performance measurement (Gibbs et al., 2004).

2.2.1 Subjective and objective PMS

Effective performance measures provide accurate, informative and timely indications of the individual’s contribution to firm value (or other organizational goals), at low risk to the employee (Gibbs et al., 2004). When quantitative performance measures are effective, objective formula incentives are likely to be used intensively. However, objective performance measures often distort incentives (e.g. because they are incomplete or prone to manipulation) or impose undue risk on the employee (e.g. because they include uncontrollables).

Therefore the performance evaluation procedures of most organizations allow supervisors some level of discretion in evaluating the performance of their subordinates. Discretion enables supervisors to provide a more accurate and complete picture of employees

(11)

performance than would be the case were their evaluations based solely on available objective performance indicators (Gibbs et al. 2004). Subjectivity (qualitative assessments), opposed to objective performance measurement, is the judgment based on personal impressions, feelings, and opinions rather than external facts.

While performance measures can differ along many dimensions, the objective/subjective distinction is one of the most common. Objective criteria measure (financial or non-financial) results, not behaviors (Hoffman et al., 1991). Bommer et al. (1995) have a practical definition of the difference between objective and subjective performance measures. They define objective measures as direct measures of countable outcomes, whereas subjective measures consist of supervisor ratings of employee performance. Baker et al. (1988, 1994) define subjectivity as the extent to which the person responsible for the evaluation has a direct personal influence on the rating. Therefore, subjective performance measurement is based on personal impressions, feelings, and opinions rather than external facts. Subjective measurements are non verifiable, i.e., a third party cannot verify the correctness of the subjective assessment. The measure solely consists of information from within the relation between supervisor and employee, outside information isn’t included. For example, customer satisfaction on itself is a subjective measure, but as a performance measure for incentive use, it is considered as an objective (non-financial) performance measure. It isn’t influenced by the personal impressions, feelings and opinions of the supervisor.

Subjectivity can be useful in mitigating various problems faced in assigning rewards through formulas based on quantitative performance measures. The use of subjectivity allows evaluators to exploit any additional relevant information that arises during the measurement period to the benefit of both the organization and the employee. The organization can benefit through improved incentive alignment and the employee can benefit through reduced risk (Gibbs et al., 2004).

In bonus assignment, subjectivity can arise in several ways, which are often used in combination. First, all or part of a bonus is based on subjective judgment about performance. Second, the weights on some or all quantitative measures are determined subjective. Or, third, a subjective performance threshold or override is used, in which case a subjective determination as to whether to pay a bonus is made based on measured performance and other factors (Gibbs et al., 2004). Subjectivity is also an important element of implicit incentives given (e.g. promotions, job assignments and threat of termination).

(12)

Performance measurement and evaluation processes are an important source of ethical dilemmas in organizations (Maas and Torres-González, 2011). Organizations need to ensure that their procedures and information systems allow managers to make unbiased appraisals. Yet, research in psychology, management, and accounting indicates that many employees feel that subjective evaluations are sometimes inaccurate and unfair. The study of Ittner et al. (2003, p. 725) suggests that psychology-based may be equally or more relevant than economics-based explanations in explaining the firm’s measurement practices.

As discussed in the accounting literature on performance measure selection a wide array of objective performance measures will usually be available, and these measures can be used in any possible combination to provide an overall indication of the employees job performance in a specific period (Gibbs et al., 2004). Following Baker et al. (1988), therefore organization’s evaluation procedures can range from very subjective to very objective.

2.3 Organizational Culture

Following Martin (1992) one variable that offers promise in the study of culture is organizational culture (as cited by Chenhall, 2003, p. 154). It is possible that a strong organizational culture may dominate national culture in het work situation. Organizational culture is a broad concept about which a consensus has yet to be reached. Notions of shared beliefs, values, assumptions, and significant meanings are commonly associated with culture (Henri, 2006, p. 79). This thesis follows Henri (2006) and attempts to capture the underlying value structure that creates meaning in organizational settings. Culture is operationalized as the shared values (what is important) that interact with an organization’s structures and control systems to produce behavioral norms (the way we do things around here). Quinn and Rohrbaugh (1983) developed the Competing Values model. The Competing Values model is used as a mean to define types of organizational cultures and to interpret the characteristics of those cultures (Denison and Spreitzer, 1991).

The Competing Values framework focuses on the competing tensions and conflicts inherent in any human system. Primary emphasis is placed on the conflict between stability and change, and the conflict between internal organization and external environment. The model incorporates two sets of competing values along two axes: (1) the control/flexibility dilemma, which refers to preferences about structure, stability, and change, and (2) the internal/external dilemma, which refers to differences in organizational focus. From these two axes emerge four quadrants reflecting four types of culture, namely rational, hierarchical, development and group (Henri, 2006; Denison and Spreitzer, 1991). Each of the four types of cultural orientation represents one of the four major models in organizational theory.

(13)

The Competing Values framework focuses on the competing tensions and conflicts inherent in any human system: primary emphasis is placed on the conflict between stability and change, and the conflict between the internal organization and the external environment. The first axis reflects the competing demands of change and stability. One end of the axis represents an emphasis on flexibility, change and spontaneity, whereas the other represents a complementary focus on stability, control, and order. The second axis reflects the conflicting demands created by the internal organizations and the external environment. One end of the axis represents a focus on integration and buffering to sustain the existing organization and is person-oriented, while the other represents a focus on competition, adaption, and interaction with the environment.

The Competing Values model thus enables four resulting cultural types of organizational orientations to be posited: group, development, hierarchical and rational. Each of the four types of cultural orientation represents one of the four major models in organizational theory. These four cultural types, along with their different underlying assumptions about motivation, leadership, and effectiveness are described below.

Group Culture: The group culture in the upper left quadrant of Figure 1 has a primary concern with human relations. The purpose tends to be group maintenance. Belonging, trust, teamwork and participation are core values, and primary motivational factors include attachment, cohesiveness and membership (Denison and Spreitzer, 1991, p. 5).

Development Culture: The development culture in the upper right quadrant of Figure 1 also emphasizes flexibility and change, but maintains a primary focus on the external

(14)

environment. This orientation emphasizes growth, resources acquisition, creativity, and adaption to the external environment. Key factors include growth, stimulation, creativity, risk taking, and variety (Denison and Spreitzer, 1991, p. 5).

Rational Culture: The rational culture in the lower right quadrant of Figure 1 emphasizes productivity, performance, goal fulfillment, and achievement. The purpose of organizations with emphasis on the rational culture tends to be the pursuit and attainment of well-defined objectives. Key factors include competition, goal oriented, structure, efficiency and productivity (Denison and Spreitzer, 1991, p. 5).

Hierarchical Culture: The hierarchical culture in the lower left quadrant of Figure 1 emphasizes internal efficiency, uniformity, coordination, and evaluation. The focus is on the logic of internal organization and the emphasis is on stability. Key factors include execution of regulations, security, order, rules and conservative and control (Denison and Spreitzer, 1991, p. 5).

2.3.1 Perception of Organizational Culture

Organizational culture is produced and reproduced through action and interaction. But it is not just lodged in people’s minds. Culture is public, the product of minds, between minds (Dent, 1991). According to Henri (2006, p. 80) and Denison and Spreitzer (1991, p. 6) no organization is likely to adopt only one culture. The difference in knowledge, beliefs and values of individual employees create different meaning systems within an organization. Moreover, cultures in organizations are not independent of their social context. They are interpenetrated by wider systems of thought, interacting with other organizations and social institutions, both importing and exporting values, beliefs and knowledge (Dent, 1991). Therefore, in the concept of an organization as a culture, it is sensible to recognize the possibility and likelihood of distinct subcultures existing among managerial teams, occupational groups, members of different social classes and so on (Dent, 1991, p. 709). Employees create cultures that underlie the quasi-corporate culture by forming groups around the multiple identities within the organization, for example, relating to subject specialism, length of tenure of role or geographic location of one’s home (Lumby, 2012). Some may be dominant-cultures and others counter-cultures.

Quinn (1988) goes a step further and states that in every organization all four cultural types of the Competing Values framework exist (as cited by Henri, 2006, p. 80). No organization is likely to adopt only one culture. Instead, each organization has its distinctive cultures based on a combination of values. The four cultures described above should be thought of as ideal types defined by the Competing Values model. As mentioned before, organizations are unlikely to

(15)

reflect only one culture; rather, one would expect to find combinations of each culture type, with some type being more dominant than others (Denison and Spreitzer, 1991). One underlying assumption of the Competing Values model is the importance of balance. When one cultural orientation is overemphasized, an organization may become dysfunctional and the strengths of the cultural orientation may even become weaknesses. For example, too much flexibility or spontaneity can become chaos; too much order and control can result in rigidity. Denison and Spreitzer (1991) recognizes that the most effective organizational culture isn’t the one that has incorporated the characteristics of all four cultural types, but nonetheless recognize that balance represents the capacity to respond to a wide set of environmental conditions.

Al previous mentioned papers considering the perception of organizational culture (except Henri, 2006), sought to apply the perspective of organizational culture in an empirical setting through qualitative research. The purpose of the undertaken field studies is to explicate a mode of theorizing of organizational culture. The mode of theorizing is interpretive, getting underneath surface of culture in organizational settings. Henri (2006) uses the Competing Values framework in a quantitative approach to study the differences in the perception of organizational culture. His sample consists of top manager of different firms. Therefore, his level of analysis is different than in this study, where the sample consists of employees within one single organization. Landekic et al. (2015) and Demir et al. (2011) also study the difference in the perception of organizational culture within one single organization using the Competing Values framework. Landekic et al. (2015) studies the organizational culture of Croatian Forests Ltd. The research site of Demir et al. (2011) is a Turkish pharmaceutical company. They both found quantitative empirical evidence that all four cultural types of the Competing Values framework exist within one single organization.

As a typology based on general characteristics of organizational cultures is used, this study does not attempt to highlight the unique qualities of an organization’s culture, but rather the perception of subcultures within an organization by individual employees. Recognizing that the perception of organizational culture by individual employees will vary widely within an organization (Landekic et al., 2015 and Demir et al., 2011). This perception of organizational culture is shaped by the difference in knowledge, beliefs, values and social context of individual employees.

(16)

3 Hypotheses Development

Following the work of Henri (2006), Bhimani (2003) and Dent (1991), I intend to examine the extent to which the perception of organizational culture becomes embedded in the use of management control systems. Henri (2006) found consistent evidence that organizational culture has a direct effect on the diversity of measurement of PMS. This perception of organizational culture is shaped by the difference in knowledge, beliefs, values and social context of individual employees. Dent (1991, p. 728) claims that “cultural knowledge in organizations vests organizational activities with symbolic meaning”. And Bhimani (2003) discusses that a management accounting system that is more reflective of the organization culture values of one group, is likely to be seen as being more successful by that group. These different studies indicate that the perception of organizational culture influence the use of performance measurement systems. All organizations contain simultaneously the opposites control and flexibility values, and internal and external values (Quinn, 1998). Since a cultural orientation has a polar opposite (Denison and Spreitzer, 1991), I will refer to the notion of “dominant type” following Henri (2006). I will refer to the notion of “dominant type” in order to capture the specific position of each employee on the control/flexibility and internal/external continuum. Hereafter, control value employees will refer to employees reflecting a control dominant type, while flexibility value employees will refer to those reflecting a flexibility dominant type. Also, internal value employees will refer to employees reflecting a dominant internal focus, while external value employees will refer to employees reflecting a dominant external focus.

The first cultural values that are polar opposites are the control versus flexibility values. The key characteristics of the control values are among others well-defined objectives, goal orientation, structure and efficiency, thus a cybernetic approach. By definition, a cybernetic logic is more compatible with control values than flexibility values. Indeed, predictability, order, goal clarity and formality are more compatible with a cybernetic logic. Control values are associated with stability, enforced roles and bureaucracy (Henri, 2006). The control values have therefore key characteristics that are consistent with the use of objective PMS.

In contrast, flexibility values mirror change, teamwork and cohesion. Key characteristics are trust, spontaneity, change, creativity and openness. According to Gibbs et al. (2004) these are characteristics that relate to the use of subjective bonuses. They find that the use of subjective bonuses is positively related to organizational interdependencies (noisy and fail to encourage cooperation) and the level of trust between the employee and the manager. The consistency

(17)

between the characteristics of flexibility/control values and a subjective/objective use of PMS are presented in figure 2. Therefore I formulate the next hypothesis:

Hypothesis 1: Employees with a perception of a flexibility (control) dominant cultural type tend to use more subjective (objective) performance measures.

The second cultural values that are polar opposites are the internal orientation versus the external orientation. The key characteristics of the internal orientation are among others uniformity, stability, order and control. The key factors of the opposite external orientation are among others external environment, risk taking, development and growth. These are also characteristics of a prospector strategy, which according to Chenhall (2003) require a more informal and open MCS characterized by more subjective long-term controls. Gibbs et al. (2004) find that organizations make greater us of subjectivity in awarding bonuses to mitigate distortions or reduce risk. Next to this Govindarajan (1984) found that a subjective evaluation style is related to uncertainty. The more uncertain the external environment the more open and externally focused the MCS. As mentioned, these characteristics are also key characteristics of the external cultural orientation. Therefore I formulate the next hypothesis:

Hypothesis 2: Employees with a perception of a dominant internal (external) orientation tend to use more objective (subjective) performance measures.

These two hypotheses lead to the theoretical model as presented in figure 2. The theoretical model reflects the influence of the flexibility/control orientation and external/internal orientation as pairs of competing values on the subjective/objective use of PMS.

(18)

4 The Research Method

4.1 Research Site

The research site is a large Dutch insurance company with 3600 employees. As Dent (1991) mentions in his study it is sensible to recognize that there are distinct subcultures existing among managerial teams, occupational groups, and members of different social classes within an organization. To study if the perception of organizational culture defers between employees at the research site, organizational units are recognized. These functional departments are based on the organogram of the insurance company. The recognized organizational units are the functional departments under the CEO (Audit), CFO (BSM, Tax/Legal, Financial Shared Services, Reporting and Control), COS (HR, Communications and General Office), CTO (Strategy & Change and ITC), CRO (Risk) and the Front Office (which has sponsorship of different board members) divided in Life Corporate, Individual Life and Non-Life (P&C).

4.1.1 PCB Cycle1

The performance metrics used in determining the periodic assessment, bonus decisions, and career path of the employees at the research site are organized through the PCB cycle. PCB is an abbreviation for performance and competence evaluation. The PCB cycle is the basis for the dialogue between the employee and the supervisor. Together they discuss how the deployment of talents, development of skills, and ambition of the employee can contribute to the needs of the organization. Every year there are three PCB meetings; a planning meeting, a performance review and evaluation. Agreements are made about performance measures and competences and the progress and ambitions are discussed. Finally, the assessment of performance is made. The difference between performance and competences is considered next. The performance of an individual is application of his or her knowledge, skills and attitudes, and the interplay with the practice setting. The level of performance varies when the clinical scenarios change and the individuals apply skills accordingly. Based on the above, competency is the ‘skill’ and ‘competence’ is an attribute of a person (Khan and Ramachandran, 2012). Competence is an idealized capacity that is located as a psychological or mental property or function and performance is the production of actual utterances. In short, competence involves ‘knowing’ and performance involves ‘doing’.

1_{Source: company policy document}

(19)

Sustainable employability is a shared responsibility of the employee and the employer. The organization beliefs that it is of the utmost importance that talents of the employees are optimally used, so that the employees are motivated and that they work with pleasure and in a healthy condition. Self-knowledge plays a central role. The PCB Cycle consists of three stages.

The first stage of the PCB cycle is the planning meeting. Planning forms the core of the cycle. A good preparation by the employee is necessary. Agreements for the upcoming year are made and the performance measures and competences to work on are determined. The organization demands an active and self-directed attitude of the employee. Therefore the employee prepares for the planning meeting by formulating performance measures and competences based on the organizational objectives and the talents and development needs of the employee. These measures are forwarded to the supervisor before the planning meeting. By linking the objectives of the department with personnel objectives, the employees are considered to work on both with pleasure and effectiveness. The performance measures are divided in result measures and competence measures and are primarily determined by the employees themselves as input for their planning cycles for the upcoming year. The manager assesses these suggested performance measures and agrees with them or makes some remarks or suggestions. After the planning cycle the performance measures are agreed upon and are therefore a joint consultation between the employee and the supervisor.. After the planning part of the PCB cycle they employee has approximately four to six result measures and two to four competence measures. Competence measures require a lot of attention because they require change of behavior and skills.

The organization demands that performance measures are measurable or observable. Performance measures that are measurable are objective measures and could be financial (e.g. cost, profit, return measures) or nonfinancial (e.g. customer satisfaction, number of errors). Performance measures that are observable are more subjective (e.g. development of certain behavior). For example if an employee wants to develop his or her visibility in the organization an measurable objective performance measure could be the number of internships undertaken at

(20)

other departments. On the other hand the development of visibility in the organization could also have an observable subjective performance measure such as the change in introvert behavior. The latter depends much more on the relation between the employee and his or her supervisor. Therefore you could say that there is a distinction between the measurable objective performance measures, which exist outside the relation between the employee and the supervisor, and the observable subjective performance measures, which depends on the relationship between the employee and the supervisor. The employee is free to choose either one of them in the preparation of the planning meeting.

The second formal PCB meeting is the performance review halfway the year. The progress on the performance and competence objectives over the first half-year is discussed and the employee and the supervisor look ahead to the second half of the year. The employee prepares this meeting by filling in the performance form. As a preparation the employee could ask feedback from colleagues or customers. During the meeting the progress on the objectives is discussed with the supervisor and if necessary actions are taken if the employee is behind schedule to meet the objectives. This doesn’t include changing the performance measures or objectives. The performance measures or objectives are set in the planning meeting at the beginning of the PCB Cycles and no changes are made during the year.

The last stage of the PCB cycle is the evaluation. At the end of the year an evaluation meeting is scheduled. The employee fills in the appraisal form as a preparation for the evaluation meeting and therefore assess his or her performance as first. This rating can be supported by feedback of colleagues or clients and by collecting evidence of performance and competence development. At the evaluation meeting the assessment of the employee is discussed with the supervisor and they determine how the outcome of the performance measures and competence development relates to the objectives set in the first stage of the PCB cycle (the planning meeting). Based on this the reward is set. The organization stresses specifically that the performance evaluation isn’t one-way traffic. As an employee you have a lot of influence on the evaluation outcome. In the preparation of the evaluation the employee rates the result on every performance and competence measure. If the supervisor and the employee couldn’t come to an agreement on the evaluation of a performance measure the supervisor has the final call.

The objectives set in the planning meeting are the basis of the evaluation. The realized results and the development on the competences are rated on a five items assessment level (inadequate, adequate, good, very good and excellent). The assessment of the result measures and competence measures determine the variable compensation (bonus) and the periodic reward

(21)

(salary raise). This is done using a formula that calculates the average assessment of the individual performance measures. It also happens that a subjective performance threshold or override is used because the determined objectives on the performance measures are not achieved cause of circumstances outside the influence of the employee. A subjective adjustment to the performance evaluation is made on the objective performance measures.

The organization stresses specifically that the employee has a significant influence on the PCB cycle. The employee prepares every step in the PCB cycle (planning, performance and evaluation). Therefore the employee has a great influence whether subjectivity is used or not. According to Briers and Hirst (1990) this is participative performance evaluation, referring to the extent to which employees contribute to the evaluation of their own performance. The employee decides whether measurable (objective financial or nonfinancial measures) or observable (subjective) performance measures are used. Next to this a subjective adjustment to the performance evaluation of objective measures can be made in the evaluation meeting. Therefore you could say that there is a distinction between an evaluation based on measurable objective performance measures, which exist outside the relation between the employee and the supervisor, and an evaluation based on the observable subjective performance measures, which depends on the relationship between the employee and the supervisor. All or part of an evaluation (bonus) could be based on subjective judgment about performance.

4.2 Sample

The sample consists of individual employees of the large Dutch insurance company. The random sample will contain employees with different functions, departments and levels of education. Because a sample of employees of a single organization is used, there is no need to control for the other contingency variables next to culture (i.e. strategy, external environment, technology, organizational structure and firm size). Every employee in the sample has the same exposure to these variables. Conform Merchant et al. (2011) this provides relatively pure tests of the effects of organizational culture. The one-company setting allows me to control for many potentially relevant factors that vary across organizations and industries to focus on the differences of organizational culture perception.

4.3 Methodology

Considering the research question, the lack of public databases with relevant data and accessibility to an interesting sample, this study consists of survey research. For the process of data collection, I have requested the employees to participate in the survey with an email directly

(22)

addressed to them. The purpose of the survey and the confidentiality of the data are explicitly stressed out in the email. The survey is conducted as an online web-based survey. The advantages of online surveys are increased speed of response, lower cost and less data entry than mail surveys (Crawford, Couper and Lamias, 2001). The survey implementation follows two steps: (1) initial mailing and, (2) follow up.

4.4 Constructs

The questionnaires used in the survey are presented as Appendix A.

4.4.1 Independent Variable – Organizational Culture

Organizational culture is measured using one section of the Institutional Performance Survey (IPS) developed at the National Center for Higher Education Management Systems. The validity of this instrument has been demonstrated and it has been used in an accounting setting (Henri, 2006). Denison and Spreitzer (1991) mention in their paper that the applicability of the model is at several levels of analysis. The model can be applied at both the individual and the organization level.

Respondents are asked to distribute 100 points among the four ideal cultural types along each of the following four dimension of culture: department character, department leader, department cohesion, and, department emphases. For each dimension, respondents must distribute 100 points among four sentences where department A refers to group culture, department B refers to development culture, department C refers to hierarchical culture, and department D refers to rational culture.

The aim of the questionnaire is to capture the specific position of each employee on the control/flexibility and internal/external continuum (i.e. dominant type). The dominant-type score is derived from a cultural-type score and a value score. First, the cultural-type score is compiled for each culture by averaging the ratings obtained on the four dimensions. For each employee, the sum of the four cultural types equals 100. Second, the value score is computed for the control/flexibility and internal/external continuum as follows:

Flexibility-value score = (Group-culture score + Development-culture score) Control-value score = (Hierarchical-culture score + Rational-culture score)

(23)

Internal-value score = (Group-culture score + Hierarchical-culture score) External-value score = (Development-culture score + Rational-culture score)

Third, the dominant-type scores are obtained by subtracting the control-values score from the flexibility-values score and by subtracting the external-values score from the internal-values score. Considering that the flexibility and control value scores and the internal and external value scores are the extremes of a competing values continuum, a different score captures the specific position of each employee on these two continuums. A positive score on the flexibility/control continuum indicates a flexibility dominant type, while a negative score indicates a control dominant type. Subsequently a positive score on the internal/external continuum indicates an internal dominant type, while a negative score indicates an external dominant type. Each mix of control/flexibility and internal/external values will provide a different dominant-type score ranging from -100 to 100. Figure 4 illustrates different combinations of value scores and the dominant-type score resulting.

4.4.2 Dependent Variable – Subjective versus Objective PMS

The use of subjective or objective performance measurement is measured using a construct from the study of Indjejikian and Matejka (2012). The validity of this instrument has been demonstrated. Following Bouwens and Van Lent (2007) I ask for the actual weight placed on a range of performance measures (rather than determining each by Likert scales or to force respondents to rank measures). Performance measures could in fact be (almost) equally important. The instrument asks employees to state the percentage of their latest PCB Cycle that depends on (1) financial measures, (2) nonfinancial measures and (3) subjective evaluations. Whereby the financial and nonfinancial measures are objective, measurable measures outside the relation between the employee and the supervisor. The subjective evaluation consists of non-measurable measures within the relationship between employee and the supervisor.

(24)

The aim of the questionnaire is to capture the specific position of each employee on the subjective/objective continuum (i.e. dominant type). The dominant-type score is derived from a subjective and objective value score. First, Financial Measures is the weight on measurable financial measures, Nonfinancial Measures is the weight on measurable nonfinancial measures and Subjective is the weight on non-measurable subjective performance measures. The study focuses on the use of subjective or objective performance measurement. Secondly, subjective performance measurement is directly measured with the variable Subjective. Objective performance measurement (Objective) is determined by the sum of Financial Measures and Nonfinancial Measures. By definition, Subjective + Objective = 100.

4.4.3 Control Variables

To augment the confidence that my findings are possibly attributable to the perception of cultural differences, I control for the effects of size and years of experience. Variables that have been argued to also affect incentive practices (e.g. Bouwens and Van Lent, 2007). I also collect employee’s characteristics (i.e. gender, age and highest level of qualification) to assist in the testing of possible non-response bias and as controls for the statistical analysis.

As an employees experience grows, there is an increasing likelihood that his or her knowledge base exceeds that of the immediate superior. Size is the natural logarithm of the number of employees that work in the department of the respondent. To study the existence of different perceptions of organizational culture within the organization I also collect the functional department the employee works at.

4.5 Survey respondents

The sample consists of 300 employees, which are randomly selected. Of these, 165 employees filled out the survey. Twenty-five responses were identified as having significant missing values, leaving a final sample of 140 respondents. This represents u usable response rate of 46.7%.

I screened the survey data for possible non-response bias by comparing the first and last 30 responses via t-tests (Armstrong and Overton, 1977). It is common practice to check for non-response bias by comparing “early” and “late” respondents. I used the first 30 respondents as “early” respondents and the last 30 respondents as “late” respondents. The result of the t-test to check for significant differences in variable scores between early and late respondents can be found in Appendix B. The variable nonfinancial measures, is significantly different between “early” (mean = 46.53) and “late” (mean = 36.50) respondents.

(25)

Table 1 reports a recapitulation of the demographics from the sample respondents. Sixty-two percent of the respondents are men and the employees are between the ages of 36 – 45 on average. Most respondents have a HBO degree. The employees have about 6 – 10 years of experience on average and works in a department with 6 – 10 colleagues.

For details of the demographics from the sample respondents see Appendix C.

4.6 Validity and Reliability

To establish content validity, existing and validated constructs used in existent literature have been employed. Moreover, the questionnaire was pre-tested by two employees, an academic and a fellow student. They were asked to complete the questionnaire and to provide comments on its form and content. Some adjustments were made in terms of wording and presentation.

The reliability of each construct and its specific dimensions were assessed with Cronbach Alpha coefficients. The Group culture and Development culture constructs exceeded the recommended cut-off point of 0.70. The Hierarchical culture and Rational culture construct don’t, but the Hierarchical culture (0.58) is nevertheless acceptable (see table 1).

(26)

The Rational culture construct has a low reliability (α = 0.39). The data showed that removing the question about the department leader resulted in a higher reliability (α = 0.45) for the Rational culture construct. Field (2013, p. 709) states that there are many reasons not to use the general guidelines of interpreting Cronbach’s Alpha, not least of which is that they distract you from thinking about what the value means within the context of the research you’re doing. He goes on to say that when dealing with psychological constructs, values below .7 can, realistically, be expected because of the diversity of the constructs being measured. Considering that these constructs of the Competing Values framework are frequently used in studies about organizational culture (for example Henri (2006), Landekic et al. (2015) and Demir et al. (2011)), and that the Gronbach Alpha is relatively close to acceptable, the Rational culture construct is considered usable.

(27)

5 Results

5.1 Descriptive Statistics 5.1.1 Organizational Culture

As a typology based on general characteristics of organizational cultures is used, this study does not attempt to highlight the unique qualities of an organization’s culture, but rather the perception of subcultures within an organization by individual employees. Recognizing that the perception of organizational culture by individual employees will vary widely within an organization (Landekic et al., 2015 and Demir et al., 2011). This perception of organizational culture is shaped by the difference in knowledge, beliefs, values and social context of individual employees. I found quantitative empirical evidence that all four cultural types of the Competing Values framework exist within one single organization. Table 3 shows the different functional departments within the organization and the cultural emphasis of the employees working in that department.

Hierarchical type of culture has, in the evaluation of the representative sample, achieved the highest score, i.e. highest average value according to respondents’ responses (Mean = 28.22). Nevertheless all four cultural types are almost equally represented within the organization. Even within almost all functional departments are the four cultural types present in the perception of the employees. The functional departments have different dominant cultural types but even within these departments the cultural types are fairly equally represented. The more respondents

(28)

from a functional department the more evenly represented are the four cultural types. Therefore I can conclude that all four cultural types of the Competing Values framework exist within one single organization.

Table 4 presents descriptive statistics for each of the scores relevant for the independent variable perception of organizational culture analysis. The dominant-type score on the Flexibility/Control continuum has a negative mean (-6.29), which indicates that the employees in the sample have more emphasis towards control. The dominant type score on the Internal/External continuum has a positive mean (4.10). This indicates that the employees perceive an internal dominant type. These summary scores are close to zero and therefore also indicate that al four cultural types of the competing values are almost perceived equally within the organization by the respondents. The minimum and maximal values of the different cultural types range form 0.00 to 100.00 (and for the rational culture type 57.50). This indicates that on an individual employee level very different perceptions on the extremes of the control/flexibility and internal/external continuum exist within the organization.

To further analyze the differences in culture within the organization I summarize the different functional departments in two distinct sub departments from which you should expect that the characteristics and function within the organization are significantly different; the front office and the back office. The front office of the organization deals with customers and is more commercial. The back office deals with the administrative burden of the insurance company. The front office consists of the functional departments communication, strategy & change, life corporate, individual life and P&C. The back office consists of the functional departments audit, BSM, tax/legal, financial shared services, reporting, control, HR, ITC and risk. Table 5 shows

(29)

these different departments and the cultural emphasis of the employees working in that department.

Employees working in the front office and in the back office both emphasize towards control. They score highest on average on the control value (Mean = 52.36 and 55.34). The employees in the back office emphasize slightly more towards internal values (Mean = 52.55), where employees of the front office don’t have a clear preference on the internal/external continuum. Table 6 show the results of a t-test to examine whether there exists a difference between the mean scores on the cultural emphasis between the front office and back office.

The results of table 6 indicate that there is no significant difference between the means of the dominant type score on the Flexibility/Control and Internal/External continuum for the front office and back office.

All previous scores and tests indicate that there are cultural differences within the organization on an individual employee level but there is no dominant cultural emphasis within the organization as a whole. All four cultural types exist almost equally within the organization. Even within the summary departments front office and back office and the individual functional departments the four cultural types exist almost equally. It appears that there is no significant difference between the departments on its own.

(30)

5.1.2 Subjective versus Objective PMS

As discussed, performance measures can differ along many dimensions. The objective/subjective distinction is one of the most common. Following Bouwens and Van Lent (2007) I ask the employees for the actual weight placed on a range of performance measures (rather than determining each by Likert scales or to force respondents to rank measures). Performance measures could in fact be (almost) equally important. The instrument asks employees to state the percentage of their latest PCB Cycle that depends on (1) financial measures, (2) nonfinancial measures and (3) subjective evaluations. Whereby the financial and nonfinancial measures are objective, measurable measures outside the relation between the employee and the supervisor. The subjective evaluation consists of non-measurable measures within the relationship between employee and the supervisor. Table 7 presents descriptive for each of the scores relevant for the dependent variable PMS use.

The scores indicate that within the sample the nonfinancial measures are on average used the most (41.44%), followed by subjective evaluation (37.97%) and financial measures (20.59%). This indicates that on average 62.03% of the used performance measures are measurable and objective. An average of 37.97% of the performance measures are non-measurable subjective measures within the relationship between the employee and the supervisor. The minimum and maximal values of the different performance measures ranging from 0.00 to 100.00 indicate that on an individual employee level very different types of performance measures are used within the organization.

5.2 Sample reduction

Previous analysis of the cultural types within the organization and tests of significant difference between these cultural types indicate that employees could not be classified unambiguously. To ensure that for further analysis cultural groups had a clearly defined dominant type, employees were categorized as control or flexibility value employees and internal or external value

(31)

employees if their absolute dominant-type score is greater than or equal to 10. Thus, the employees having value scores ranging between 45 and 55 were deleted from the sample for further analysis. While the selection of the cut-off scores is mostly arbitrary, employees deleted were in a grey area and could not be classified unambiguously. After the deletion of 59 cases, the sample consists of 81 employees. Table 8 shows that within this reduced sample, the differences in mean values on the cultural emphasis between the departments is slightly more profound.

The results in table 9 indicate that even after the cut-off of the sample there is no significant difference between the means of the dominant type score on the Flexibility/Control and Internal/External continuum for the front office and back office.

All further analyses are made with the reduced sample without the employees that were in the grey area. All analyses are also executed with the whole sample of 140 employees. The results of these tests are included in Appendix D.

5.3 Correlation

Table 10 shows the Pearson and non-parametric Spearman correlations between all the constructs that were used in this study. As could be expected, the different performance measures are negatively and significantly related to each other. The more an employee uses one specific type of performance measure the less he or she could use other type of measures. Other significant positive relations appear between the experience of an employee (tenure) and age and

(32)

education. Department size has also a significant positive relation with the experience of an employee (tenure). Age is significant positively related with education. The correlations among the other constructs are not sufficiently high to justify concerns with multicollinearity. Therefore these results don’t provide some preliminary evidence for the hypotheses.

The correlations table with the full sample of 140 employees (see Appendix D) also shows a significant positive relation between the dominant type score on the Internal/External continuum and the size of a department (r=0.0197, p<0.05). Furthermore both correlations show a significant negative relation between the dominant type score on the Flexibility/Control continuum and Age (r=-0.227, p=<0.01 / r=-0.206, p<0.05). This indicates that the older the employee the more emphasis towards control.

5.4 Results of ANOVAs

A series of analysis of variance are conducted as a preliminary step to compare the use of the different performance measures financial and nonfinancial (both objective) and subjective between employees reflecting a control and a flexibility dominant type. After the earlier deletion of 59 cases, 41 employees were classified as flexibility value employees and 40 employees were classified as control value employees. Table 11 summarizes the results of ANOVAs and the Kruskal-Wallis tests. The results suggest that there are no significant differences between flexibility value employees and control value employees in the use of the difference performance

(33)

measures. If the 59 cases are included, the results (see table 2 of Appendix D) remain the same except that the significance level of the use of nonfinancial measures and subjective measures gets closer to a significance level of p < 0.05 (respectively the ANOVAs p = 0.063 and p = 0.058). The control variables are also not significantly different between the two groups.

The same analysis are conducted for employees reflecting a internal and a external dominant type. Of the 81 cases, 54 employees were classified as internal value employees and 27 employees were classified as external value employees. Table 12 summarizes the results of ANOVAs and Kruskal-Wallis tests. The results suggest that there are no significant differences between internal value employees and external value employees. If the 59 cases are included, the results (see table 3 of Appendix D) remain almost the same. The main difference between the full sample and the partial sample is that with the full sample the mean gender of internal value employees and external value employees is significantly different. Women tend to have a more internally focus on average where men tend to have a more external focus (p < 0.05). The rest of the main and control variables are not significantly different between the two groups.

(34)

5.5 Hypotheses tests: Linear regression

This paragraph presents the results from the three regression equations, which are summarized in Table 13. Model A represents the first regression in which only the control variables and the dominant-type score of the flexibility and control continuum are added as independent variables. Model B represents the second regression in which the control variables and the dominant-type score of the internal and external continuum are added as independent variables. In Model C, next to the control variables, both dominant-type scores (flexibility/control and internal/external) are added as independent variables. In al three models the use of different performance measures (financial, nonfinancial and subjective) are used as dependent variables.

(35)

The first hypothesis states that flexibility dominant employees place more emphasis on subjective performance measures. The regression shows there is no significant relation between flexibility dominant employees and the use of subjective performance measures (b = -0.032, p ≥ 0.05). The regression also shows that there is no significant relation between control dominant employees and the use of financial and nonfinancial performance measures (respectively b = 0.018, p ≥ 0.05 and b = 0.014, p ≥ 0.05). Therefore this research finds no evidence to support hypothesis 1.

The second hypothesis suggests that employees with a dominant internal focus place more emphasis on the use of objective performance measures (financial and nonfinancial). The results show no support for this statement as the regression shows these variables are not significantly related (respectively b = -0.029, p ≥ 0.05 and b = 0.078, p ≥ 0.05). The regression also doesn’t show evidence for the opposite side of this hypothesis, that employees with a dominant external focus place more emphasis on the use of subjective performance measures (b = -0.049, p ≥ 0.05). Therefore this research finds no evidence to support hypothesis 2.

The only significant relation the regression shows is between the control variable age and the subjective use of performance measures (b = -6.340, p ≤ 0.05). The older an employee gets the less he or she uses subjective performance measures. If the 59 cases are included, the results of the regression (see table 4 of Appendix D) don’t show any significant relations. To further analyze the hypotheses, and show that the results are relative robust, a logistic regression on the main effects is performed. The results of this logistic regression (see Appendix E) show that the aforementioned results of the hypothesis tests also hold with alternative specifications of the variables. All test performed indicate that there is no significant relation between the perception of organizational culture by employees and the use of different performance measures.

(36)

6 Discussion, limitations and conclusions

This thesis draws on prior research to identify if the perception of organizational culture is relevant to the design of MCS, specifically performance measures. The theoretical framework of this thesis combined literature from various perspectives, including areas that have received vast attention in MCS studies, namely, performance measures use and culture as a contingency variable. However, this study offers several new insights into the organizational culture and the use of performance measures. The aim of this study was to provide a better understanding of the relationship between the perception of organizational culture by employees and the use of different performance measures.

Despite insights provided in previous research (e.g. Henri, 2006), the relationship between the organizational culture and the use of PMS, especially on the level of the employee, have been overlooked in current studies. Without neglecting the importance of other contingent factors (e.g. environmental uncertainty, strategy, size, etc.), organizational culture is an omnipresent factor which affects practically all aspects of organizational interactions. This study is one of few in management accounting that examine the perception of organizational culture within one single organization at the level of the individual employee. It draws on previous research by Henri (2006). For the operationalization of MCS, Henri (2006) focuses on the diversity of measurement of performance measurement systems. This study investigates another aspect of the MCS, namely the subjective or objective use of PMS. This is an important characteristic of most incentive contracts for evaluating and rewarding employees. According to Gibbs et al. (2004) theoretical research has suggested various plausible reasons for the use of subjectivity in the assignment of bonuses. Empirical testing of these theories is rare. This study helps to fill this gap.

Considering the exploratory nature of this study, it tests the direct relationship between the variable of organizational culture and MCS. Explicitly the effect of the perception of organizational culture by individual employees on the subjective or objective (financial and nonfinancial measures) use of performance measures is tested. The study results support those of other investigations (e.g. Landekic et al. (2015) and Demir et al. (2011)), which suggest that different employee groups can subscribe to different organization cultural values. The results indicate that all four cultural types of the competing values framework almost equally exist within the studied organization. There is no dominant cultural emphasis within the organization. Even within and between functional departments there isn’t a significant difference in the perception of the organizational culture. This supports the underlying assumption of the Competing Values

The impact of organizational culture on management control systems