Amsterdam Business School
Reputation and control tightness
Name: Lavinia Mihaela Cristea Student number: 11422653
Thesis supervisor: dr. ir. S.P. Sander van Triest Date: 24 June 2017
Word count: 13224
MSc Accountancy & Control, specialization Control
1
Acknowledgements
I would like to thank God for the opportunity which made possible to be a student at UvA. I am very grateful to my supervisor, Sander van Triest, for all support, suggestions and pleasant meetings. I want also to thank Ms Helena Klooosterman, who gave me ideas and included me in her interesting project.
I am grateful to my parents and my brother for their encouragement and good words in difficult times. Also, I would like to thank my friend, Shabir Jaan for his unconditional help.
Lavinia Mihaela Cristea Amsterdam, June 2017
Statement of Originality
This document is written by student Lavinia Mihaela Cristea 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.
2
Abstract
In this master thesis, I examine whether a relation exists between reputation and control tightness. The survey method with the use of professional answers provides the support to develop and investigate this effect. I believe that this exploratory study is the first which explicitly examine the relationship between those two variables. MCS was conceived in the form of four control settings (action, result, personnel and cultural), where two sets of controls (action, result) were defined between two extends (tight and loose). The questionnaire was transmitted to professionals, part from various industries in PSFs, all over the world. Multiple regression analysis was applied and the outcome shows a positive and significant correlation among reputation and all control tightness constructs, excluding result control tightness.
Keywords
: Reputation, Management Control System, Control Tightness, High Level of Control Tightness, Low Level of Control Tightness, Professional Service Firms, Intangible, Autonomy, Professionals.3 Contents
1 Introduction ... 4
2 Literature Overview ... 6
2.1 Professional Service Firms ... 6
2.2 Reputation ... 6
2.2.1 Professional Service Firms and Reputation ... 8
2.3 Management Control System ... 8
2.3.1 Management Control Systems tightness ... 12
2.4 Hypothesis development ... 14
3 Methodology ... 17
3.2 Survey Design ... 19
3.2.1 Preliminary Survey Testing ... 19
3.2.2 Independent Variable ... 19 3.2.3 Dependent Variables ... 19 3.2.4 Control Variables ... 20 3.4 Statistical Model ... 25 4 Results ... 26 4.1 Descriptive Statistics ... 26 4.2 Correlation Analysis ... 27 4.3 Regression Analysis ... 29 5 Conclusion ... 32
5.1 Discussion of Main Findings ... 32
5.2 Limitations ... 33
5.3 Implications for Future Research ... 34
References ... 35
Appendices ... 40
Appendix A ... 40
4
1 Introduction
In this thesis, I investigate whether there is a relationship between reputation and the design of the management control systems (MCSs) for professional service firms. Merchant and Van der Stede (2007) define an MCS as all the devices or systems managers use to ensure that the behaviours and decisions of their employees are consistent with the organisation’s objectives and strategies (p.6). An MCS aims to align the interests of the managers and employees with those of the organisation (Malmi & Brown, 2008, p. 289).
In professional service firms (PSFs), this alignment is expected to be more difficult. PSFs are firms which have a primary and most valuable asset in highly educated professionals (Greenwood et al., 2005) and particular organisations with a large activity sector from consulting, audit and accounting, to software development, law, engineering and architecture (Greenwood and Lachman, 1996). Professionals value autonomy (Raelin, 1985) and they have a strong position towards their organisation since they represent the most valuable asset of PSFs (Coff, 1997). Additionally, the nature of the services provided by PSFs implies that performance measurement is difficult (Von Nordenflycht, 2010). This poses challenges in designing the management control systems of PSFs.
The characteristics which complicate MCS design in PSFs suggest that organisational reputation may play a particularly relevant role: since the services that PSFs provide are difficult to assess and typically client-specific, reputation is a key selling point (Hall, 1992). PSFs need to guard their reputation diligently, which suggest high levels of quality control, yet the reputation results from the high-performance levels of the professionals: reputation depends on the human capital employed (Greenwood et al., 2005). These professionals are not expected to accept high control levels. This leads to the research question of this paper: How does reputation impact the design of management controls in PSFs?
The purpose of my master’s thesis is to investigate whether reputation, driven by company’s performance, plays a role in the development of MCS – defined as control tightness - in PSFs. More precise, my research question is:
Is there a relation between the reputation of professional service firms and control tightness? Although the importance of reputation for PSFs has been recognised (Starbuck 1992, Rao et al. 2001), this study is the first to investigate how it impacts the control of PSFs.
5 In the next sections, the relevant literature overview will be examined to provide a theoretical background for the hypothesis development, which completes the section. Section 3 will discuss the empirical analysis with procedures used, sample derivatives, instruments and research method selected, while section 4 discusses the results. Section 5 conclude with a discussion.
6
2 Literature Overview
In this chapter, prior literature concerning PSFs, MCSs and reputation will be discussed. The importance of those three variables is examined in separate subheadings.
2.1 Professional Service Firms
PSFs embody a significant part of the business economy which are “becoming ever more pronounced in economies the world over” (Delong and Nanda 2003, p. ix). PSF term is used to depict firms which have a primary and most valuable asset in highly educated professionals (Greenwood et al., 2005) and particular organisations with a large activity sector from consulting, audit and accounting, to software development, law, engineering and architecture (Greenwood and Lachman, 1996). Moreover, considering their skills and expertise, PSFs are “knowledge engines for business” (Lorsch and Tierney 2002, p. 14). Managing knowledge is different from traditional methods and leverage upon special and particular approaches (Robertson and O’Malley Hammersley, 2000). Previous research has focused on the structures (Pettigrew and Fenton, 2000) and formal knowledge management systems (Hansen et al. 1999, Sarvary 1999) of PSFs.
Despite this, literature does not provide a uniform perspective on what PSFs precisely means. Greenwood et al. (2005) argue that two characteristics (Delong and Nanda 2003, Broschak 2004) for PSFs are very important. The first one that outputs are intangible and not easy to measure by customers, professionalism is a function of complex knowledge. The second characteristic belongs to high education and qualification in order to be prepared for the customised requests, essential for business continuity of PSFs (Greenwood et al. 2005). Particularly, management of knowledge has to be carefully handled and services provided in an innovative and creative way to satisfy different needs (Chang and Birkett, 2004).
Those characteristics suggest that PSFs may be influenced by reputation in controlling, rewarding, hiring and retaining the highly educated workforce.
2.2 Reputation
Reputation is an intangible concept as a result of past actions and future behaviour, research universally shows that a good reputation demonstrably increases corporate worth and provides sustained competitive advantage (Hall, 1992). Reputation is a concept which cannot be controlled directly and not easy to be explained. The literature provides some insights into this omnipresent
7 concept as a total effect of a firm’s or individual’s behaviour (Hutton et al., 2001). Previous research has found that reputation is a relevant notion only for particular stakeholders, and not considered an across-the-board measure valid for all public.
From a corporate management perspective, the company’s image is essential to be perceived higher and authentic because corporate identity delivers clearness for a good evaluation of the benefits associated (Lewellyn, 2002). As a distinction between image and reputation, an image is the immediate mental picture that stakeholders had; reputation evolves in time and over time, as a consequence of previous actions and behavioural manifestations of the company (Balmer and Greyser, 2003). The behaviour has to be constant and consistent with a trustworthy exterior perception. Thus, a principle distinction is deemed to be stability in behaviour characterised by consistency in behaviour over time (Helm, 2011). In short, stability in behaviour establishes trust and credibility with company’s various stakeholders (Fombrun et al., 2000a; Larkin, 2003). Thus, a good reputation is the result of the “alignment of expectations shareholders and corporate behaviour” (Hoffmann 2016, p. 37). In this way, PSFs gain reputation. Consistency in behaviour help companies to win competitive advantage (Nordenflyght, 2010) or outrun services of the rivals (Gompers & Lerner, 2001), mostly based on recommendations from the past customers. Moreover, their choice might influence other potential customers by the power of the word, seeing that reputation is the subjective perception of individuals (Fombrun et al., 2000).
Consistently, good reputation increase organisational performance in an easier way (Roberts and Dowling, 1997) to achieve the desired results among clients, key stakeholders as large customers, opinion leaders from the business sector, suppliers or current and potential employees (Fombrun & Shanley, 1996; Fombrun & Van Riel, 2004). Reputation helps companies to attract, recruit and motivate talented employees (Fombrun, 1996; Fombrun, 2002). Reputation takes good years to build and few seconds to lose (Fombrun et al., 1990) and once is achieved, reputation has to be guarded (Brissette, 2005). Whether reputation is gone, the profit will pursue in the same way (Herbig and Milewicz, 1995).
Reputation provides corporate legitimacy and stabilises its stakeholder relationships (Fombrun, 1996; Deephouse, 2000). Corporate legitimacy is defined as “a generalised perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions” (Suchman, 1995, p. 574). Rindova et al. (1999) found a connection between legitimacy, reputation and trust where legitimacy is developed in companies that promote trust, admiration and respect over other companies, in this way building reputation. Trust is a psychological perception by an outside party, where the vulnerability is accepted and strongly believe that others will act beneficially with good intentions
8 (Rousseau et al., 1998). Creation of this social capital (i.e. reputation) is considered a mutual trust between client and organisation which influence the organisation experience, converting in an individual name property (Starbuck, 1992). Social capital assures and converts favourable contracts into a long term relation, the key being reflected by the technical expertise measured as a collective success (Starbuck, 1992).
2.2.1 Professional Service Firms and Reputation
Reputation concept has been researched in other fields than PSFs (Merton 1968, Hall 1992, Fombrun 1996, Deephouse 2000). In PSFs, reputation has been seen important, with emphasis on the intangibility characteristic (Greenwood et al., 2005) defined as the experience which cannot be touched or possessed, besides a hard process of performance measurement (Lecture: Management Control Research). Backing reputation, professionals are an important piece of this image since services are delivered according to customer demands (Pinnington & Morris, 2002).
Through the intangibility of the benefits provided (Greenwood et al., 2005), reputation could be considered a difficult and fluid concept because of the strategy and diversification relation (Greenwood et al., 2005). Reputation help companies to grow, make higher profits and attract more clients with fewer costs (Greenwood et al., 2005). In PSFs, reputation gives the power to hire the best professional staff, promote image with a cost effective and the possibility to switch for premium returns. Thus, a continuous cycle of benefits is incurred for a good reputation (Greenwood et al., 2005). This could lead to the Merton’s Matthew Effect, where “the richer get richer and the poorer get poorer” (Merton 1968, p. 56). This cycle of benefits answers to this effect for PSFs and employees. Moreover, whether services have a high value delivered and stakeholders are satisfied, the probability of choosing the next service that the company has associated with the specialised service chosen before will be higher.
PSFs represent firms of which the primary and most crucial asset is the professionals, characterised as being high educated and accredited workers, with essential and valuable contribution encrypted and embedded in a complex knowledge (Greenwood et al., 2005), with a direct effect on reputation.
2.3 Management Control System
MCSs are important and essential for organisations because prevent from failures, bad reputation, misalignments between principal and agent, losses and risks. Management control is important from two broad perspectives. First, management control comes into managers’ interest to resolve
9 strategic control issues, from an internal perspective. Second, management control drives the capacity of controlling behaviour and how to influence it in a desirable way (Merchant and Van der Stede, 2007, p.4). By doing so, managers make sure that organisation objectives have a common ground with employees’ decisions and behaviours. In the same time, MCSs limit unwanted behaviour and solve agency theory problem (Malmi & Brown, 2008).
Specifically, a control system in place provide guidance for past, present, future and emphasise the company strategic objectives in order to be maintained. Opportunities have to follow strategic objectives and not involve too many risks or persuade unwanted behaviour that can alter organisation’s integrity (Simons, 1995).
In PSFs, professionals may look for independence, authority, being perceived by managers as a difficulty in managing and targeting in a certain direction (Von Nordenflycht, 2010). In this respect, Adler and Borys (1996) recognise between enabling and coercive bureaucracy. As of how the concept suggests, ‘enabling bureaucracy’ gives employees room for flexibility, creativity, which to some extent provide freedom in working outside the scope of their position. ‘Coercive bureaucracy’ restrict any kind of discrepancy from procedures or deviations from internal regulation, being described as a non-creative work environment, dissatisfaction and incapability for contemporary ideas. Dissatisfaction effect would be avoided whether managers apply an enabling system and not punish deviations, variation from procedures would be encouraged and learning opportunities in a pleasant work environment would definitely retain professionals. Auzair and Langfield-Smith (2005) choose to measure the extent to which the control system is bureaucratic in service organisations. According to their findings, a more bureaucratic control systems is based on a restricted response from the managers to the new opportunities, a tight control where the work of the employees is more precise and detailed in procedures, frequently monitored, with formal and action controls. A less bureaucratic control seems to be more flexible and informal controls applied. Managers communicate the best way for achieving the targets; hence, the openness and involvement is higher with an interpersonal emphasis on employee preferences. Compared to a tight control, a loose control perspective is suggested here and a more detailed MCS tightness concept will be examined in the next subsection.
Simons (1995) argue that the aim of a control system is to manage tensions encountered between organisational goals and employees creativity in consideration of a system under control. According to Anthony and Govindarajan’s (2007), under MCS employees can be influenced by the manager concerns. Consequently, MCS steer the behaviour of employees and help managers to achieve their goals. Because the objectives of both parts must be aligned, MCS help this objective to be reached (Merchant and Van der Stede, 2003). The keys present in MCS definition
10 are: ‘behavior’, ‘strategy’ and ‘organizational goals’. Four types of controls are evidenced and explained, based on an object of control: action controls, result controls, personnel controls and cultural controls.
In my master thesis, I will use Merchant and Van der Stede (2007) framework, distinguished between four types: action controls, result controls, personal controls and cultural controls.
Action controls have a central perspective on employee behaviour in order to make sure that his actions are consistent with company’s benefits. Behaviour and processes must be congruent, evaluation mechanisms and direct observations define action controls (Turner and Makhija, 2006; Stouthuysen et al., 2012). Similarly, in the context of what action controls are about, Turner and Makhija (2006) clearly specified that action controls help organisations to define certain tasks employee has and what, when and how managers have to adapt in order to reach desired result (Snell, 1992). Called as behavioural controls, action controls are holding employees accountable for actions, judged by formal controls embodied in rules and procedures (Merchant, 1982). The most explicit system for management control expressed by process controls (Merchant and Van der Stede, 2007) is surrounded by specific details, such as: clear operating rules and procedures, standard established routines, behavioural restrictions, customised job description, etc. (Turner and Makhija, 2006; Merchant and Van der Stede, 2007; Malmi and Brown, 2008; Stouthuysen et al., 2012).
Result controls are described from a performance measure perspective and hold employees accountable for the result of actions performed. Focused on the outcome of work and not the work itself, managers evaluate and judge results delivered by rewards or punishments. Result controls provide the possibility for employees to compare the results delivered in an unambiguous way (Modell, 1996) with targets, used as a measure of reference for performance (Stouthuysen et al., 2012), divided in financial and non-financial performance measurements (Rockness and Shields, 1984). Therefore, decentralisation of control and empowerment are principal characteristics of result controls, where employees are free to choose in attaining targets without any restrictions (Abernethy and Brownell, 1997; Snell, 1992). In the same time, actions performed directly influence results, an immediate effect which creates worries for employees (Merchant and Van der Stede, 2007).
Personnel controls describe the process starting with attracting, selecting and employment, training and retaining employees (Ouchi, 1979). Strictly speaking, Abernethy & Brownell (1997) argue personnel controls are about a rigorous selection process, making sure that right people join the company. This aspiration is very clear, because: the requirements and expectations are specified, employees have to persuade in doing a good work and implicitly, acting in
11 organisational interest being self-engaged in monitoring activities. Aligning both parties’ interests is important from the beginning, with a slight influence in socialising from cultural controls (Ouchi, 1979; Snell, 1992). In addition, Malmi and Brown (2008) argue that selection process is placed under a cultural control system, explained by a straightforward reason: organisations recruit and give job only to people who share the same values as the company has. From another perspective, organisations are not looking only for the same principles, but also for abilities, a capacity of analysing and response in different situations, knowledge, work under control and stress, skills and future perspectives (Snell, 1992). This suggests that the main key is to choose the right people who are transparent and inspired by their own vision but at the same time compatible with organisation’s objectives (Merchant and Van der Stede, 2007).
Cultural Controls are considered as being the softest part of the organisation system (Lecture: Management Control). Cultural controls are described by the same values, beliefs and norms resembled by the belief system (Simons, 1994), in a social context where the pressure is exerted on the individual. Cultural controls represent the tolerance and the extent of which organisation admit employee implicit beliefs, norms and values from the previous workplace applied in the existed job. The fourth category of control from Merchant and Van der Stede model imply an informal socialisation mechanism encountered in each organisation, foster a particular set of values, beliefs, objectives, perspectives, visions between each representative of an organization. Implicitly, cultural controls share and oversight in an indirect way the behaviour, code of conduct, reflection and speculation mechanism, gathered in the actions performed (Malmi and Brown, 2008). Once individual join the company and participate at meetings, different reunions, or other types of socialisation practices (Turner and Makhija, 2006), he will become an important part of the organisation. Thus, cultural controls force in adoption the employees who diverge with organisational model and promote mutual monitoring in a particular conception (Merchant and Van der Stede, 2007). Core characteristics of the cultural control system, such as norms and shared values are applied in extreme intangible situations and comprehensive personal interactions. Cultural controls are important in terms of knowledge transmission among existing employees to strength company’s promoted values (Løwendahl., 2001) with an implicit communication and reciprocal cooperation for making things happen (Kaše et al., 2009).
According to Merchant and Van der Stede (2007), the principal scope of MCS is to make sure that organisational objectives are achieved. Along with the four types aforementioned, they introduce control tightness concept, where every type of control is applied in a loose or tight degree. Moreover, with a tight control in place, the likelihood that employees will comply with
12 organisational objectives is higher. Hereby, the tolerance, flexibility, room for interpretation is reduced to a certain level in order to satisfy both parties’ wishes.
2.3.1 Management Control Systems tightness
The focus point of this paper is to examine whether control tightness has an influence on reputation. Control tightness is not consistently defined in the literature (Van der Stede, 2011) and relatively few academics choose to define tight or loose control as a whole. Instead, most authors prefer to discuss the components or characteristics of tight control. Merchant and Van der Stede (2007) define control tightness as the ‘degree of certainty that employees will act according to organisation wishes’ (p. 118), being a construction of the control system. Amigoni (1978) utilises the definition of Dearden (1971) and Dalton (1971) who characterised a tight control system as one in which administrative control dominates over several, social and individual control.
Hopwood (1974) continue to further explain both concepts of controls were in tight control systems, the achievement of targets are explained to the employees as firm commitments, the participation in setting objectives is not important and performance is based only on accounting measures. On the other hand, loose control is evidenced by managers which are conscious to the social side of the control. Employee participation is high, targets are negotiated and even when both parties agree upon, targets are only set as a reference point in the context of other available information.
In Merchant and Van der Stede model, control tightness in organisations depends on the answer to three important questions: (1) What are the potential benefits of tight control? (2) What are the costs?, (3) Are any harmful side effects likely? Control tightness is contextualised in this model for action controls, result controls, personnel controls and cultural controls.
Added up the definitions of MCSs, the one end of control term is achieved through more controls, more supervision, formal procedures, coercive controls and strict rules with a high degree of MCS. Respectively, in the opposite part of the term are fewer controls, less supervision, informal procedures, enabling and flexible control with room for creativity, innovation and diversity. In this regard, two contributions to the literature require attention. First is represented by the style of control (Amigoni, 1978) and control tightness (Merchant, 1982). In his framework, Amigoni (1978) define various aspects and management control characteristics, applied to any organisation, with emphasis on how those features influence the control mechanism. The organisational style of control is represented by the managers’ attitude in acting and managing the control process, where two distinctive types are evidenced: tight or loose.
13 From a tight control perspective, the environment is dominated by many controls over individuals’ work, a clear setting of targets where the performance is evaluated on an accounting basis, non-participation of individuals in organisational decisions because this arrangement is considered top management responsibility.
Comparatively, a loose control is described as a low monitoring and oversight process over employees. The individual contributes in organisational goal setting because managers are knowledgeable of the importance given to social and individual part. Targets are set in an attainable way and are discussed to make the employee aware that targets represent a relative goal for future perspective. Loose control implies performance measurement on non-financial measures, considered forward-looking measures since providing guidance to concentrate on actions linked with forthcoming performance (Lecture: Management Control).
In addition, control tightness implies a deep analysis conductive to expected return. Merchant (1982) suggest that a tight control is close to a higher certitude in what procedures employees should respect and in the same time, understand what is necessary to be done since are held accountable for their actions. A tight control mechanism is a way in which organisation acts and performs activities equally expected from employees, to avoid certain discrepancies between principal and agent, as to how agency theory state. Hence, a tight control system is considered more effective and efficient in the monitoring process and individuals’ surveillance because of advance clear expectations and prevention measures, strict discipline sanctions and approach compensation system.
Once again, more controls limit creative attempts of individuals and impede the freedom intensity (Merchant, 1982), where he distinguish the subsequent features of control tightness:
Description of organisational objectives – extensive approach, particular goals that are consistent with desired organisational outcome
Communication of organisational objectives – goals are transmitted in a complete way to cover, understand and at the same time obtain the individual acceptance in attaining organisational aim (greater effectiveness, efficiency, promptitude, regularity)
Monitoring – organisational surveillance and oversight responsibility in a detailed, periodically and at the right time
Rewarding – appreciate the individual effort, compensate or punish actions performed, establishing a rigorous reward-performance connection.
Control tightness has been analysed by Auzair and Langfield-Smith (2005), where target settings are very precisely defined and accurately communicated, periodically monitored on a timely basis.
14 Oppositely, loose control materialises flexibility in performing activities, no frequently monitoring and no overlook of the organizational objective.
The tightness of a control system represents the choice of management. However, too much control may restrict to a certain extent employees’ wishes, limit creativity and boost dissatisfaction among professionals, but at the same time protect organisational objectives from unwanted behaviours which could create dysfunctionalities (Auzair and Langfield-Smith, 2005). Considering this, a tight control may not be considered to be applied in order to make sure goals are attained, as well as loosening controls may perform efficiently for the organizational goal.
2.4 Hypothesis development
Reputation is difficult to gain and easy to lose (Fombrun et al., 1990; Herbig and Milewicz, 1995). For PSFs, reputation results from the performance of their professionals (Greenwood et al., 2005; Pinnington & Morris, 2002). To maintain this reputation, PSFs need to ensure that employees are acting in a consistent manner with their reputation and goal. This can achieve a high level of action control tightness. By clearly describing and prescribing the correct activities for employees, the risk of reputation-damaging actions is reduced. This leads to the following hypothesis:
H1a: High reputation is related to high level of action control tightness
Oppositely, PSFs’ main assets consist of highly qualified people with knowledge power (Scott 2008, Von Nordenflycht 2010). Such professionals want to be autonomous in operational and decision making. Professionals seek for autonomy and informal controls (Von Glinow, 1983; Von Nordenflycht, 2010) and dislike surveillance. They prefer flexibility, creativity, innovation and PSFs have to be aware of not restraining the initiative of employees in certain circumstances (Chang and Birkett, 2004). Covaleski et al. (1998) suggest that the organisation depends on professionals and expects from professionals to act according to company objectives, assessed in a business reputation view. The high quality of the services which lead to the PSF’s high reputation is achieved when the professionals get the freedom to be flexible and creative. Accordingly, I envision:
H1b: High reputation is related to low level of action control tightness
Raelin (1985) suggest that when goals are clearly specified, with no other interpretation and most notably, steeply presented in one single motto, the professional perception, understanding and apprehension is higher. Result controls are necessary to empower and in the same time, to oversight professionals activity (Merchant and Van der Stede, 2007). Moreover, result controls
15 enforce accountability and motivate professional to act, making aware of future consequences. For the desired result to occur, there is a need for an efficient control system in place with a high result control tightness (Merchant and Van der Stede, 2003).Therefore, I anticipate the following: H2a: High reputation is related to high level of result control tightness
Nevertheless, it is difficult to choose a good performance measure. Because of their intangible nature, professional services are hard to be measured and have no similarity or adequate reference in the market. The professionalism applied converge with organisational goal and performance outcome might not be completely captured in the measurement applied. Consequently, the result controls applied do not provide an accurate image and professionals might feel that management ambiguously evaluate their performance. Furthermore, tight control reduces autonomy since professionals want to be autonomous in decision making. Hence, I complete this assertion as following:
H2b: High reputation is related to low level of result control tightness
Selecting the best people is essential for PSFs concerning high-quality services and corporate image maintenance. PSFs most valuable intangible asset is represented by professional workforce (Greenwood et al. 2005, Coff and Kryscynski 2011) and the most important aspect for PSFs is to recruit the right human capital, considered the principal advantage in competition. Therefore, always will be chosen the highly educated workforce in the selection and promoting personnel since this is the principal management concern (Nurmi, 1998). Hiring the highest personnel, save money and at the same time, create organisational value (Lepak and Snell, 1999), improve effectiveness and performance (Abernethy and Brownell, 1997). Therefore I expect:
H3: High reputation is related to high level of personnel control tightness
Personnel and cultural controls set free the image of company’s values expectation and reveal true insights from organisation’s centrum. Cultural controls are playing an essential role in the organization besides result control. As long are apportioned in the right extent, reputation is reached and with a good management in place, maintained. PSFs are looking for employees who share the same corporate culture (Merchant and Van der Stede, 2003) and implicitly, provide support to express their knowledge among colleagues (Løwendahl, 2001).
Managers represent the role-model for professionals (Merchant and Van der Stede, 2003) and are the principals who create cultural aims to be fulfilled. As part of cultural control tightness, monitoring is a key activity which enforces control mechanism and is essential to be tight by reason of reputation, easily influenced by external factors. Due to this reason, first is important to
16 be selected right people with the same culture. Second is to give them what they need. Consequently, in managing social activities, professionals find themselves committed to organisational culture (Ouchi, 1979) and therefore do a god job (Merchant, 1982). This professional product delivered influence reputation. Since organisation wants to ensure that reputation is not affected, cultural controls degree will not be loose.
That being said, I propose H4:
H4: High reputation is related to high level of cultural control tightness
Figure 1: Proposed research model (+) / (-) (+) / (-) (+) (+) Reputation
Action Control Tightness
Result Control Tightness
Personal Control Tightness
Cultural Control Tightness
Control Variables Organizational size Strategy Task complexity Environmental uncertainty
17
3 Methodology
To conduct this research, I decided to join the PSF thesis survey project 2016-2017 at the University of Amsterdam, led by Helena Kloosterman. My decision was based on the following reasons.
First, I gained access to a large number of questionnaires from high qualified respondents, which it would be impossible for me to obtain such substantial number. Second, the project includes a pre-developed survey, clearly structured in questions about overall activity in PSFs which allow me to develop the hypothesis and deeper understand the subject. Third, previous research has focused on a particular PSF industry, whereas the current study aims to provide qualitative empirical evidence based on PSFs sector in general, for more reliable conclusions.
3.1 Sample Selection
The main objective of data collection was to gather a substantial sample of employees working in various PSF industries in different geographical areas. In order to get access to the entire database, students who applied for thesis survey project had to submit at least 10 completed surveys. In order to guarantee the authenticity of the completed surveys, every confirmation was sent by e-mail to the coordinator. The data collection were between December 2016 and January 2017. Every respondent was qualified eligible whether following conditions were cumulatively fulfilled:
The respondent must understand and use the English language at a business level
The respondent had to be able to work relatively independently in a medium/large size company (i.e. more than 50 employees)
The respondent must have relevant experience – at least three years in the current field and by preference less than ten years
The respondent must not work in non-profit organisations, NGOs, government, except medicine/physician practitioners
The respondent is neither the owner of the company or part of the Board nor the person who design the management control system.
These requirements are affected by the MCS in a position that the respondent is subject to MCS and small companies were not included because of their characteristic MCS. The original sampling contained 612 of respondents from which 98 of respondents did not answer to relevant
18 questions (e. g. questions regarding reputation and control tightness) and were not taken into account for the analysis.
Table 1: Sample selection process
Initial Respondents 612
Reputation questions unanswered 2
Control tightness questions unanswered 96
Valid Respondents 514
The highest percentage of the sample consist of accounting profession with 19.3%, next to medicine/physician practices (9.3%) and consulting/management strategic (7.6%). The gender is out fulfilled by the male with a 64.6% and 35.4% by the female. 67.5% is represented by Dutch nationality, followed by German with 7.9% and American with a percentage of 2.9. 86.19% of the participants argued that a professional qualification is necessary for their sector and 89.69% already hold a professional qualification. A large number consist of respondents with master degree of 44.8%, followed by bachelor degree or lower of 41.7%. The descriptive frequencies tables of respondents’ occupation and nationality are described in Appendix B.
Table 2 presents the background information where the mean of professional qualification is 1.17 and job tenure mean is 5.84.
Table 2: Background information
PSF N Min Max Mean Median Std. dev.
Age 514 20 65 26.30 34.00 8.693
Gender* 514 1 2 1.35 1.00 0.478
Professional qualification** 514 1 2 1.17 1.00 0.375
Highest qualification*** 514 1 3 1.72 2.00 0.688
Years in the current organization**** 514 1 11 5.84 5.00 3.188 Years in the current field**** 514 1 11 7.13 7.00 3.128 *Male = 1, Female = 2
**Yes = 1, No = 2
***Bachelor degree or lower = 1, Master Degree = 2, PhD or other professional doctorate degree = 3 ****Less than 1 year = 1, 1 year = 2 …9 years = 10, 10 years or more = 11
19 3.2 Survey Design
The survey intends to provide a general research in terms of MCS in PSFs. However, this study aims to investigate whether reputation affects control tightness in PSFs, at the same time controlled by the variables previously mentioned. The questions considered relevant for this study can be inspected in Appendix A.
3.2.1 Preliminary Survey Testing
Before to start the collection of the data, Helena Kloosterman conducted two separate pre-tests to ensure the quality and the effectiveness of the survey, which influence the layout and the design of the constructs. Therefore, it is needed to explain how the final version of the survey was obtained in two stages.
The first test aims to ensure merely the quality of the survey. In this respect, to assess the adequacy, a sort task was used. The sorting task requires the respondents to match 52 statements that are designed to assess the MCS tightness to the definitions they consider it most concisely featured. Following to the answers of 14 subjects (one accountant, one psychologist, one security consultant, one marketing professional, one architect, one IT consultant, one dentist, one graphic designer, two management consultants, four lawyers), 32 statements were enclosed in the survey. The second test intended to ensure the adequacy of survey quality, as an overall assessment. Twenty subjects (one artist, one electrical engineer, one graphic designer, one advertising executive, one mining engineer, one architect, one dentist, one graphic designer, two marketing professionals, two management consultants, three medical doctors, three lawyers, three chemists) were asked to answer and assess the time performed, clarity, appearance and quality of the content survey. Furthermore, minor changes were applied with regard to terminology and further options to hardly any multiple choice questions.
3.2.2 Independent Variable
Reputation (REP), defined as the intangible characteristic embedded in PSFs and professionals feature, represent the independent variable analysed in this study. Respondents were asked to reply to four statements on a five-point Likert scale - strongly disagree, agree, neutral, agree, and strongly agree - which were conducted towards the respect, good value for the price, quality of services and the outside recognition under which company is perceived.
3.2.3 Dependent Variables
In order to measure the respondents' perception regarding how tight or loose is MCS in the organization, this study uses the following dependent variables: action controls tightness (ACT),
20 result control tightness (RCT), personnel control tightness (PCT), cultural control tightness (CCT). For each construct, the level of agreement or disagreement was evaluated with four questions on a five-point Likert scale. The construction of dependent variables is adapted from prior research.
ACT questions consist of rules and procedures used in PSFs applied by management, under various tasks and situations performed by employees, the extent to which flexibility, monitoring and management support are used.
RCT construct is to some extent very much alike ACT, with the difference that result controls have a primary concern in the achievement of goals and targets, and not rules and procedures. RCT statements focus on the expectations of organisational goal achievement and the extent to which individual may differ from pre-settled targets.
PCT questions consist of personal opinion about the recruitment process. Particularly, the main objective was to find out information about the number of interviews, factors which influence the hiring process, and the steps performed by the company. Respondents were asked to provide information about the abilities, skills and values required for the particular job.
For CCT were asked statements about socialisation process, procedures used and the perception of the respondent about social events. The main focus of CCT construct involved questions regarding the manner in which organisational goals were communicated, core values insufflated, but also particular details about team-building activities, meetings and peer interaction.
3.2.4 Control Variables
The aim of this master thesis is to demonstrate whether reputation (REP) has an impact on MCS or MCS is affected by reputation. In order to obtain a clear picture, organisational size (SIZE), strategy (COST and DIFF), task complexity (COMP), environmental uncertainty (UNC) represent the variables which have to be controlled for.
Organisational size is defined as the total number of employees in a company (Abernethy et al., 2010). Size represent one of the most representative and common factors, with effect on many elements of organisational design, namely control (Sharma, 2002; Rockness and Shields, 1984; Lin et al. 2012). It is well known that large organisations face more complexity, which affects many aspects of business aspects. Furthermore, communication and cultural controls might be affected, causing a less effectiveness and more problematic management of control (Lin et al., 2012, Von Nordenflycht, 2011; Morris and Empson, 1998). Since cultural controls aim to minimise the pressure from peer groups (Merchant and Van der Stede, 2007) and considered
21 appropriate to foster high personal interaction (Modell, 1996), a face to face communication is essential in understanding the complexity of tasks (Kaše et al., 2009). It is widely recognised that a bureaucratic control is adopted when organisations become larger and more formal controls are adopted. Large organisations use more rules and procedures because have to control the behaviour of people and be more in control, while small companies rely to a better extent on informal controls. The respondents were asked about the size of their organisation and to choose between four options – less than 100, more than 100 but less than 500, more than 500 but less than 5000, more than 5000.The strategy is another determinant of MCS. Organisational strategy takes into account the choices of the management for a desirable reflection of organisational goals in order to achieve performance (Lorsch and Tierney, 2002). Cost leadership strategy is adopted by organisations which associate it with a formal MCSs and principal focus on cost control. More specifically, cost leadership strategy emphasises a traditional MCS. Contrary, other organisations emphasise differentiation strategy and do not rely on standardised procedures or result controls. The work environment is more flexible and controls are rather informal (Chenhall, 2003). The same idea is suggested by Auzair and Langfield-Smith (2005) since a more bureaucratic MCS is linked with cost leadership strategy, rather than differentiation strategy. In the survey, respondents were asked to indicate the extent of developing efficiency with low costs, and the extent to which customers are satisfied when the company provide a large variety of services or products.
MCS choices are often influenced by task complexity (Merchant and Van der Stede, 2012; Jaworski et. al, 1993). Task complexity is associated with the difficulty, regularity and diversity of responsibility for a particular job (Jaworski et. al, 1993). In general, the more complex the task the more difficult is to control people behaviour (action controls). Difficulties are encountered when management have to measure how predictable the task will be and to what extent can vary. Because of this complexity, it is uncertain to assess the high quality performed by employee (result controls). This lead to the assumption that task complexity influence MCS, specifically, action and result controls. The survey develops a number of seven questions on a five-point Likert-scale and was regarding the complexity, routinization, difficultness and dependency of many different persons.
Another factor suggested with impact to MCS is environmental uncertainty by the reason of PSFs are influenced by their environment more than other organisations (Sharma, 2002). Three factors of environmental uncertainty also influence the design of MCS: environmental turbulence, the capacity to foresee the most important environmental determinants and strength of competition (Sharma, 2002). The statements included in the survey indicate that respondents were asked details about price competition, to what extent the company compete for manpower and the
22 patience to wait for new customers, on a five-point Likert scale as well as. Besides, the survey addresses questions about the new number of products marketed and predictability of the organisational industry.
3.3 Validity Analysis
In order to assure the reliability and validity of the analysed factors, a Cronbach’s Alpha test was applied and disclosed in each table for each construct. The Cronbach’s Alpha is a system construct measurement of the data’s reliability and internal consistency. A level of 0.7 is considered acceptable and respectable (Field, 2009), while a result below 0.5 is considered unacceptable (George and Mallery, 2003). The purpose of this thesis was to obtain a Cronbach’s Alpha higher 0.5 and not essentially a value of 0.7.
The discriminant validity of constructs was proven using the varimax rotated factor analysis to return the factors which are not orthogonal and total variance to be distributed all over the factors concerning the maximisation of their variance. For a sampling from 100 to 200, Field (2009) argues that the value of loadings should range between 0.364 and 0.512.
Table 3 displayed the analysed factor (REP), where the first factor was found with an eigenvalue higher than 1 and variance explained 61.72% from the total. The acceptable value of Cronbach’s Alfa is supported by the high values of all items which provide support for the construct.
Table 3: Factor analysis – REP
Reputation Component 1
1. Well-respected 0.869
2. Good value for price 0.666
3. Consistent quality and service 0.864
4. Brand name recognition 0.723
Variance explained (%) 61.72%
Cronbach’s Alpha 0.78
For ACT, only one underlying factor shows an eigenvalue higher than 1, with 56.90% of total variance explained and unrelated Cronbach’s Alpha of 0.74.
23 Table 4: Factor analysis – ACT
Action control tightness Component 1
1. Procedures and rules whatever situation 0.786
2. Job task coverage 0.765
3. Rules for everything 0.804
4. Frequently supervision 0.654
Variance explained (%) 56.90%
Cronbach’s Alpha 0.74
RCT construct with a high level of control tightness showed only one factor with the eigenvalue greater than 1, with a good level of Cronbach’s Alpha of 0.8 and 62.76% of the variance explained.
Table 5: Factor analysis – RCT
Result control tightness Component 1
1. Measure for everything 0.792
2. Large number of goals/targets 0.756
3. Constantly checking 0.837
4. Frequently checking for performance 0.782
Variance explained (%) 62.76%
Cronbach’s Alpha 0.8
For PCT construct, the first analysed factor has eigenvalue higher than 1 and explain 57.03% of the total variance. The associated Cronbach’s alpha was 0.74 for this construct, implying an acceptable reliability.
Table 6: Factor analysis – PCT
Personnel control tightness Component 1
1. Extensive HR process 0.831
2. Many steps required 0.853
3. Several interviews required 0.703
4. Extensive evaluation before employment 0.606
Variance explained (%) 57.03%
Cronbach’s Alpha 0.74
Factor analysis for cultural control tightness (Table 7) displayed only one component with the eigenvalue greater than 1. This determinant explained 62.87% of the total variance and a sufficient reliability is showed by the Cronbach’s Alpha.
24 Table 7: Factor analysis – CCT
Cultural control tightness Component 1
1. Regularly host social events 0.841
2. Communication of core value 0.708
3. Team-building events 0.850
4. Other types of events 0.764
Variance explained (%) 62.87%
Cronbach’s Alpha 0.8
For control variables, factor analyses were performed, except for the SIZE construct since only one single question was asked. For the strategy construct, three components have the eigenvalue over 1. The factor analysis performed displays a discrepancy on the strategy construct, cost leadership (COST), differentiation (DIFF) and the third component. Each item discloses a value of 34.78%, 12.15%, 9.53% from total variance explained. The loadings which are bold selected were designed as follows: the COST construct constituted elements 1, 2, 3 and 8, with a value of Cronbach’s Alpha of 0.73. DIFF construct comprised items 5, 6, 7 with a reliable Cronbach’s Alpha of 0.67. The information from questions 4 and 11 was not enclosed in the constructs. Table 8: Factor analysis – Strategy
Organizational Strategy COST DIFF Component3
1. Low cost than competitors 0.738 0.142 -0.182
2. Increased cost efficient 0.764 0.056 0.230 3. Cost coordination 0.766 0.188 0.164
4. Utilization improvement 0.444 0.313 0.243
5. Quick and new services 0.301 0.510 0.263
6. Distinction in services 0.066 0.812 0.143
7. Broad range in services 0.136 0.831 0.019
8. Time improvement 0.532 0.088 0.505
9. Quality provided 0.005 0.076 0.837
10. Customized services 0.169 0.294 0.582
11. After sale support 0.161 0.496 0.416 Variance explained (%) 34.78% 12.15% 9.53%
Cronbach’s Alpha 0.73 0.67
Lastly, the factor analyses were run on the ROUTINE construct. Two components have eigenvalue over 1. Based on the loadings, table 9 showed a variance of 29.77% and 19.39% from the total variance explained. The factor analysis divide distinctly between two components, ROUTINE was here considered as a formative construct because all questions define completely the variable and exclude any factor would provide an imperfect picture. The associated Cronbach’s Alpha for this construct has an acceptable level of 0.59.
25 Table 9: Factor analysis – ROUTINE
ROUTINE Component 1 Component 2
1. Easily performance determinacy 0.511 0.140 2. Easily outcome determinacy 0.646 0.127 3. Problems with no immediate solution 0.187 0.804 4. Time spent with no immediate solution 0.179 0.802
5. Similar problems 0.728 -0.040
6. Routine job 0.730 -0.019
7. Job dependence for completion -0.101 0.542
Variance explained (%) 29.77% 19.39%
Cronbach’s Alpha 0.59
Table 10 showed the factor analyses performed for the UNC construct. Two questions had the eigenvalue greater than 1, which is related with 34.18% and 21.77% from the total variance. Similarly, UNC is distinguished between two components. All components were taken into consideration since defining the construct as a whole. The picture of UNC disclosed a reliable level of Cronbach’s Alpha equal with 0.61.
Table 10: Factor analysis – UNC
Environmental uncertainty Component 1 Component 2 1. Biding for new customers 0.715 -0.455 2. Intensive competition for manpower 0.538 -0.168 3. Intensive price competition 0.676 -0.474 4. Market activities classification 0.575 0.249
5. Industry predictability 0.451 0.696
6. Clients predictability 0.509 0.548
Variance explained (%) 34.18% 21.77%
Cronbach’s Alpha 0.61
3.4 Statistical Model
Section 2 described the hypotheses which are tested using a multiple linear regression considered on ordinary least squares (OLS) method, applying the subsequent model:
𝐶𝑇 = 𝛽0 + 𝛽1REP+ 𝛽2𝑆𝐼𝑍𝐸 + 𝛽4𝐶𝑂𝑆𝑇 + 𝛽5𝐷𝐼𝐹𝐹 + 𝛽6 ROUTINE + 𝛽7𝑈𝑁𝐶 + 𝜀
The dependent variable is represented by CT and will be the variable which will be alternated for each regression. In the following section are described the results of the regression analyses.
26
4 Results
Section four start with an analysis of descriptive statistics of complete variables needed to test the hypothesis. Subsequently, the correlation table is described using Pearson correlations coefficient, which provides the first observation into the connection among variables. Consequently, the effects of multiple regression analyses are described, including a short table with the main findings.
4.1 Descriptive Statistics
Independent, dependent and control variables are introduced in Table 11. The mean of 4.22 indicates a high level of reputation variable (REP).
Table 11: Descriptive statistics
N Min Max Mean Median Std. dev.
Independent variable REP 514 1.50 5.00 4.22 4.25 0.67 Dependent variables ACT 514 1.00 5.00 3.18 3.25 0.85 RCT 514 1.00 5.00 2.89 3.00 0.86 PCT 514 1.00 5.00 3.59 3.75 0.91 CCT 514 1.00 5.00 3.29 3.25 0.81 Control variables SIZE* 514 1.00 4.00 2.92 3.00 1.10 COST 514 1.00 5.00 3.54 3.50 0.76 DIFF 514 1.25 5.00 3.59 3.50 0.74 ROUTINE 514 1.86 5.00 3.30 3.39 0.47 UNC 514 1.00 5.00 3.18 3.17 0.64
*1 = less than 100, 2 = more than 100 but less than 500, 3 = more than 500 but less than 5000, 4 = more than 5000
Following, the mean of SIZE (2.92) is somewhat high and explain the large sample of respondents who are a part of large organisations, besides COST (3.54) and DIFF (3.59) which suggest the cost efficiency and differentiation in services.
27 4.2 Correlation Analysis
This section starts with the Pearson correlation between all variables. There are three significant relationships found from four relationships between the independent variable REP and dependent variables CT, respectively ACT, PCT and CCT. No significant correlation was found between REP and ROUTINE (r = 0.032, p > 0.05).
Table 12 provide significant and positive results among all the dependent variables. The highest correlation (r = 0.439, p < 0.05) was between RCT and ACT, followed by PCT and CCT (r = 0.389, p < 0.05), next to RCT and PCT (r = 0.267, p < 0.05). It is worth to mention again that ACT construct positively correlate with a tight level of result controls (r = 0.439, p < 0.05), result which contradict the opinions that ACT might be replaced by other category of controls.
Moreover, I found significant relations among control variables with no less than three of the CT variables. The strongest coefficient was between RCT and COST (r = 0.219, p < 0.05), besides ACT and COST (r = 0.196, p > 0.05), where a possible explanation for this result might be due to the high mean aforementioned.
This section provided the first understanding of the relations between REP, CT and control variables.
28 Table 12: Pearson Correlations
1 2 3 4 5 6 7 8 9 10 1 REP 2 ACT 0.167** 3 RCT 0.074 0.439** 4 PCT 0.175** 0.220** 0.267** 5 CCT 0.337** 0.146** 0.210** 0.389** 6 SIZE 0.181** 0.097* -0.016 0.153** 0.194** 7 COST 0.193** 0.196** 0.219** 0.092** 0.132** -0.007 8 DIFF 0.281** 0.099* 0.128** 0.164** 0.165** -0.029 0.410** 9 ROUTINE 0.032 0.193** 0.158** 0.018 -0.052 0.064 0.201** 0.103** 10 UNC 0.145** -0.003 0.147** 0.135** 0.082 -0.046 0.286** 0.356** 0.115** ** p-value significant at < 0.05 (2-tailed)
29 4.3 Regression Analysis
OLS multiple regressions results are introduced in this section. Multicollinearity is run by analysing the variance inflation factors (VIF). The results for all regressions range among 1.1 and 1.6, implying that multicollinearity is not an issue because the scores are below 10 (Field, 2009).
Hypothesis H1 concerned the relation between tight action control and reputation. Table 13 presents the regression results for this hypothesis. The significant F-value suggests is appropriate to run the regression analysis for H1. The significance of the model fit (p < 0.05) is indicated by the analysis of variance (ANOVA). The R-square indicates a value of 8.5%, which represents the independent variable explained in ACT. In general, the higher the R-square the better the prediction fits the statistics. In this case, it showed the R-square is not so high by reason of REP not easy to estimate. A positive correlation was found between REP construct and ROUTINE (β = 0.164, p < 0.05), as well as with COST (β = 0.157, p < 0.05), SIZE (β = 0.078, p < 0.05). However, REP is positively correlated with ACT, according to expectations (β = 0.130, p < 0.05). Thus, hypothesis H1a is supported, while hypothesis H1b is not supported.
Table 13: Regression analysis - ACT
H1: ACT Predicted β t-statistic p-value
(constant) 2.907 0.004*** REP + / - 0.130 2.906 0.004*** SIZE 0.078 1.815 0.070* COST 0.157 3.291 0.001*** DIFF 0.011 0.222 0.824 ROUTINE 0.164 3.762 0.000*** UNC -0.101 -2.202 0.028** N: 514 Adj. R2: 0.085 F-statistic: 8.900 p-value: 0.001***
Regression analysis for RCT construct is presented in the following table. Adjusted R-square indicates a relatively strong model fit (p < 0.05) and explains 5.8% of the variance of RCT. The model do not provide support for the REP and RCT because of the non-significant effect (β = 0.029, p > 0.05). A positive correlation was found among RCT and COST (β = 0.146, p < 0.05), as well as RCT and ROUTINE (β = 0.116, p < 0.05). The reason for the non-significant effect might lay from the poor correlation presented in Table 14 between RCT and REP. Hence, H2a is not supported, nor is H2b.
30 Table 14: Regression analysis - RCT
H2: RCT Predicted β t-statistic p-value
(constant) 2.787 0.006*** REP + / - 0.029 0.633 0.527 SIZE -0.026 -0.604 0.546 COST 0.162 3.324 0.001*** DIFF 0.014 0.270 0.787 ROUTINE 0.116 2.637 0.009*** UNC 0.076 1.637 0.102 N: 514 Adj. R2: 0.058 F-statistic: 6.277 p-value: 0.001***
Regression analysis which tests whether REP is positively correlated with PCT is showed in the following table. A strong model fit (p < 0.05) explains the 6.1% of the variance in PCT and according to expectations, a positive and significant relationship (β = 0.107, p < 0.05). Similarly, table 15 provides significant results between PCT and SIZE (β = 0.158, p < 0.05), as well as PCT and DIFF (β = 0.102, p < 0.05). Furthermore, H3 is supported.
Table 15: Regression analysis - PCT
H3: PCT Predicted β t-statistic p-value
(constant) 5.212 0.000*** REP + 0.107 2.332 0.020** SIZE 0.158 3.611 0.000*** COST 0.003 0.069 0.945 DIFF 0.102 2.035 0.042** ROUTINE -0.014 -0.325 0.746 UNC 0.075 1.601 0.110 N: 514 Adj. R2: 0.061 F-statistic: 6.495 p-value: 0.000***
The last hypothesis has the highest percent of total variance explained in CCT, 13.9%, which indicates a very strong fit model. This suggests that in good reputation companies, cultural controls are very tight since the adjusted R-square is the highest from all regressions analysis. Again, the high F-value indicate the suitability of data and prediction. Moreover, REP shows support for this hypothesis (β = 0.280, p < 0.05), besides SIZE and ROUTINE.
31 Table 16: Regression analysis - CCT
H4: CCT Predicted β t-statistic p-value
(constant) 4.404 0.000*** REP + 0.280 6.379 0.000*** SIZE 0.153 3.640 0.000*** COST 0.068 1.465 0.144 DIFF 0.063 1.314 0.189 ROUTINE -0.093 -2.200 0.028** UNC 0.011 0.240 0.811 N: 514 Adj. R2: 0.139 F-statistic: 14.741 p-value: 0.000***
Table 17 presents the summary results from each hypothesis tested using the regression analysis. According to the results, H1a, H3 and H4 are absolutely supported while H1b, H2a and H2b are rejected.
Table 17: Regression analysis – Summary results
REP Predicted β p-value Adj. R2 Support
H1a + Yes ACT 0.130 0.004*** 0.085 H1b - No H2a + No RCT 0.029 0.527 0.058 H2b - No PCT H3 + 0.107 0.020*** 0.061 Yes CCT H4 + 0.280 0.000*** 0.139 Yes
32 5 Conclusion
The importance of reputation has been recognised in the literature, however, few researchers provide support for the role of reputation in PSFs. This study is the first which examines whether a relation exists between reputation and MCS in PSFs. This master thesis provides evidence of a significant and positive relationship between reputation and different types of controls.
Section five discusses the main findings, next to an overview of limitations. Besides these, this section settles with limitations and implications for future research.
5.1 Discussion of Main Findings
Previous research on reputation, management control, PSFs and a dataset observation from a large number of respondents all around the world permit to this exploratory study to investigate and analyse the relation between reputation and control tightness. Principally, the results provide support for the assumption that reputation impact MCS design. At the same time, the evidence challenge to a certain extent prior research.
It is widely recognised that PSFs provide professionals flexibility to accomplish various tasks and allow them to be autonomous over their work. Since they trust professionals’ abilities without being necessary to follow strict procedures, clearly description of activities or strictly supervision, many researchers suggest that those actions are not necessary to be taken (Von Glinow, 1983; Von Nordenflycht, 2010). Furthermore, Covaleski et al. (1998) argue that organisation itself depends on a high quality of professional services for the business reputation. However, the result of this study contradicts those opinions because a positive relation is found between action control tightness and reputation. In other words, the tighter the action controls the higher the reputation is. This finding indicates that in good reputation companies, there are many strict rules, procedures which have to be respected and employees must act in a consistent way with the reputation and company’s goal. Hence, the proposed hypothesis where a low level of control tightness might be related with high reputation is rejected. A first probable answer for why autonomy does not support or at least not anymore is because nowadays those types of controls actually help professionals. Procedures and rules further professionals’ work. In fact, the use of controls contributes to time-saving which they could spend on performing actions for reputation maintenance and not waste time on mundane tasks. In this respect, professionals are actively engaged in organisational objectives and do not oppose resistance to bureaucratization (Brivot, 2011). The second explanation might be that professionals possess enough experience and in the same time, know when is necessary to apply action controls, to provide high-quality services and protect organisational goal (Merchant, 1989). A third reason is pointed by Brivot (2011) where professionals were involved in many scandals, which decrease the public trust about their professional judgement. Because of those past events, the main solution adopted was to create procedures for as many problems as possible and develop rational answers to increase the outside perception and guard their reputation.
