Management Control Systems,

Performance and Uncertainty

A quantitative study on the role of MCS in
uncertain contexts

MASTER THESIS WITHIN: Business Administration

NUMBER OF CREDITS: 30 credits
PROGRAMME OF STUDY: Civilekonom
AUTHOR: Julia Lewinson Skörd & Tova Racov
JÖNKÖPING May 2023
Master Thesis in Business Administration
Title: Management Control Systems, Performance and Uncertainty: A quantitative
study on the role of MCS in uncertain context
Authors: Julia Lewinson Skörd & Tova Racov
Tutor: Miguel Gil
Date: 2023-05-22

Key terms: Contingency theory, Institutional theory, Levers of control, Management control
systems, Organizational learning, Performance, Uncertainty

Abstract
Background: Management Control Systems are shown to be valuable tools for managing
different contexts and enhance organizational performance. To succeed with intended
strategies, organizations must balance different levers of controls (LOC). The right balance is
argued to be dependent on the external environment and contingent factors that surrounds the
organization. To understand the role of MCS, one must explore different contexts. Hence, this
paper broadens the knowledge about the role of MCS by studying the unexplored context of
post-Covid-19. When facing uncertainty, MCS are claimed to be especially important for
organizational performance – something that tested in this paper.

Purpose: The purpose of this paper is to contribute to a more comprehensive understanding

of the relationship between MCS, organizational performance, and the external environment.
Also, this paper tests if uncertainty has a moderating effect on these relationships.

Method: This study is based within the positivistic paradigm and follows a deductive
research approach. It addresses a quantitative methodology where data is gathered from 164
Swedish managers through a structured questionnaire. The empirical data is analysed using
multiple linear regressions with- and without moderating effect.

Conclusion: The results partly support the alleged positive relationship between the use of
MCS and organizational performance. Three out of four LOC are found to be positive
correlated with organizational performance. The study did not find support for the claim that
uncertainty has a moderating effect on these relationships. This occurrence can be explained
by the phenomenon of organizational learning.

1
Acknowledgements

We would like to dedicate this section to thank everyone involved in the process of writing
this thesis.

Firstly, we want to thank all the managers that participated in our study for taking their
valuable time to contribute with their experience of using management control systems.
Without you, this study would not be possible.

Secondly, a big thank to the fellow seminar students who has given us valuable feedback and
interesting discussions during the seminars. Additionally, we also want to thank Toni Duras
for helping us with the statistical analyses.

Finally, we want to express our deepest gratitude to our supervisor Miguel Gil. Thank you for
your engagement and all the feedback, knowledge, and guidance you have provided us with
during these weeks. We are beyond grateful for your support!

Jönköping International Business School

22nd of May 2023

Julia Lewinson Skörd Tova Racov

2
Table of Contents
1. Introduction.................................................................................................. 6
1.1 Background ................................................................................................................ 6
1.1 Theoretical Problem ................................................................................................... 8
1.2 Research Purpose and Question ............................................................................... 11
1.3 Disposition................................................................................................................ 12

2. Literature Review ...................................................................................... 13

2.1 Literature Review Method........................................................................................ 13
2.2 Contingency Theory ................................................................................................. 15
2.3 Institutional Theory .................................................................................................. 17
2.4 Organizational Learning .......................................................................................... 18
2.5 Management Control Systems.................................................................................. 20
2.6 Simons Levers of control ......................................................................................... 21
2.6.1 Beliefs Systems ..................................................................................................... 23
2.6.2 Boundary Systems................................................................................................. 23
2.6.3 Diagnostic Control Systems ................................................................................. 24
2.6.4 Interactive Control Systems.................................................................................. 25
2.7 Research on Simons’ LOC ....................................................................................... 26
2.8 Hypotheses Development ......................................................................................... 28
2.8.1 MCS and Organizational Performance................................................................. 28
2.8.2 MCS, Organizational Performance, and Uncertainty ........................................... 31
2.9 Research Models ...................................................................................................... 35
2.9.1 Research Model 1 ................................................................................................. 35
2.9.2 Research Model 2 ................................................................................................. 36

3. Method ........................................................................................................ 37
3.1 Introduction .............................................................................................................. 37
3.2 Research Philosophy ................................................................................................ 37
3.3 Research Approach................................................................................................... 38
3.4 Research Methodology ............................................................................................. 39
3.5 Choice of Theory ...................................................................................................... 40
3.6 Research Design ....................................................................................................... 41
3.6.1 Method of Data Collection ................................................................................... 41
3.6.2 Sampling ............................................................................................................... 42
3.6.3 Pilot study ............................................................................................................. 43
3.7 Survey Design .......................................................................................................... 43
3.7.1 Control Questions................................................................................................. 44
3.7.2 LOC, Uncertainty and Performance .................................................................... 46

3
 3.8 Regressions ............................................................................................................... 47
3.8.1 Control Variables ................................................................................................. 47
3.8.2 Independent Variables .......................................................................................... 47
3.8.3 Dependent Variables ............................................................................................ 48
3.8.4 Moderating Variable ............................................................................................ 48
3.8.5 Regression Models ............................................................................................... 49
3.8.6 Model 1 –Direct Effect ......................................................................................... 49
3.8.7 Model 2 – Moderating Effect................................................................................ 50
3.9 Data Analysis ........................................................................................................... 50
3.10 Research Quality ...................................................................................................... 51
3.10.1 Reliability ......................................................................................................... 51
3.10.2 Validity ............................................................................................................. 52
3.11 Ethical Considerations .............................................................................................. 53

4. Empirical Findings .................................................................................... 54

4.1 Reliability Test ......................................................................................................... 54
4.2 Descriptive Statistics ................................................................................................ 55
4.2.1 Control Variables ................................................................................................. 55
4.2.2 Independent Variables .......................................................................................... 56
4.2.3 Dependent Variable .............................................................................................. 57
4.2.4 Moderating Variable ............................................................................................ 57
4.3 Pearson Correlation .................................................................................................. 57
4.4 Regression Analysis ................................................................................................. 60
4.4.1 Model Fit .............................................................................................................. 61
4.4.2 Hypothesis Testing Model 1 ................................................................................. 63
4.4.3 Hypothesis Testing Model 2 ................................................................................. 65

5. Discussion ................................................................................................... 67
5.1 Discussion Model 1 .................................................................................................. 67
5.1.1 Hypothesis 1 ......................................................................................................... 67
5.1.2 Hypothesis 2 ......................................................................................................... 69
5.1.3 Hypothesis 3 ......................................................................................................... 70
5.1.4 Hypothesis 4 ......................................................................................................... 71
5.2 Summary Model 1 .................................................................................................... 73
5.3 Discussion Model 2 .................................................................................................. 74
5.3.1 Hypothesis 5 ......................................................................................................... 74
5.3.2 Hypothesis 6 ......................................................................................................... 76
5.3.3 Hypothesis 7 ......................................................................................................... 78
5.3.4 Hypothesis 8 ......................................................................................................... 79
5.4 Summary Model 2 .................................................................................................... 81

6. Conclusion .................................................................................................. 82
6.1 Thesis Summary and Main Conclusions .................................................................. 82
6.2 Theoretical Contributions ......................................................................................... 84

4
 6.3 Empirical & Methodological Contributions ............................................................. 85
6.4 Practical Contributions ............................................................................................. 86
6.5 Societal and Ethical Issues and Consequences......................................................... 87
6.6 Limitations................................................................................................................ 88
6.7 Future Research ........................................................................................................ 88

7. Reference List ............................................................................................. 90

8. Appendix ................................................................................................... 107

8.1 Appendix 1 – Literature Review Table .................................................................. 107
8.2 Appendix 2 – Survey Questions English ................................................................ 109
8.3 Appendix 3 – Survey Questions Swedish .............................................................. 114
8.4 Appendix 4 – Coding ............................................................................................. 119
8.5 Appendix 5 – Summary Variables ......................................................................... 120

Figures
Figure 1 - Literature Review .................................................................................................... 14

Models
Model 1 - LOC and Organizational Performance .................................................................... 35
Model 2 - LOC, Organizational Performance and Uncertainty ............................................... 36

Tables
Table 1 - Reliability Test .......................................................................................................... 54
Table 2 - Descriptive Statistics Control Variables ................................................................... 55
Table 3 - Descriptive Statistics Independent Variables............................................................ 56
Table 4 - Descriptive Statistics Dependent Variable................................................................ 57
Table 5 - Descriptive Statistics Moderating Variable .............................................................. 57
Table 6 - Pearson Correlation ................................................................................................... 59
Table 7 - Model Fit ................................................................................................................... 61
Table 8 - Hypotheses Testing Model 1 .................................................................................... 63
Table 9 - Hypotheses Testing Model 2 .................................................................................... 65

5
1. Introduction
______________________________________________________________________
The first chapter of this paper provides a background on the topic of management control
systems. This is followed by a discussion of the theoretical problem and thereafter the purpose
and research questions of this study are presented. The chapter ends with a description of the
disposition of this paper.

1.1 Background

Control is a word that frequently appears in management literature, and the function refers to
the tools managers use to ensure a well-performing organization (Malmi & Brown, 2008).
These tools are fundamental in all organizations as they encompass managers' approaches to
monitor, influence, and evaluate their employees. One way to operate control is through
management control systems (MCS). MCS facilitates and includes all the tools and systems
managers use to define, communicate, and reinforce behaviors in desirable ways (Simons,
1995). They enable the efficient utilization of resources (Simons, 1995) and facilitate
effective monitoring and reporting processes within organizations (Jokipii, 2010). Control
systems may be formally designed and implemented through rules, standard operating
procedures, and budgeting systems with either a feedback or forward-looking nature.
Financially oriented controls often focus on output and results, while controls of non-financial
characters, such as rules, standard operating procedures, and ongoing monitoring of activities,
often have a forward-looking approach (Langfield-Smith, 1997). Managers can create and
communicate different control systems through formal documents and statements of purpose
to give direction to search for new opportunities and enforce wanted and unwanted behavior
(Simons, 1995).

MCS plays a crucial role in assisting managers in formulating strategies, defining operational
actions needed for strategy implementation, establishing clear expectations, prioritizing
operational enhancements, and setting performance targets that can impact current and future
performance (Simons, 2014). Controls can enhance desirable behaviors, motivate, and guide
employees to create value for the organization. It can be used as a constraining force to avoid
taking risks and maintain a balance to run the organization efficiently (Baird et al., 2019;
Widener, 2007). Additionally, establishing a culture of active dialogue and communication

6
within the organization is essential for effectively recognizing and capitalizing on new
strategic opportunities (Davila, 2000). Cross-functional integration and collaborative
discussions can improve organizational performance. Simons (1995) proposes that one
achieves control when the tension between creative innovation and expected goal
achievements are converted into profitable growth. Hence, a high degree of learning and
control will contribute to the effectiveness of the organizations.

Organizations operate in business environments that are constantly changing and are not
always easy to predict (Otley & Soin, 2016). Organizations may encounter uncertainty in the
context of their operations, which can swiftly and dramatically change. This can make it
difficult to forecast economic fluctuations, competitor behavior, and the timing of potential
technological advancements. Additionally, organizations operate within systems of
stakeholders who influence one another in the business environment (Otley, 2016). In this
dynamic landscape, MCS is utilized by managers to drive organizational performance and
achieve desired outcomes (Chenhall, 2003). Organizations need to adapt to their unique
environmental conditions, which often involves adopting different control systems based on
the specific context. The selection of appropriate control techniques is contingent upon the
circumstances surrounding the organization (Otley, 1999).

The outbreak of Covid-19 and the environmental consequences that followed can be seen as a
big crisis. During the pandemic, several countries were shut down to decrease the spread of
infection, which affected the business environment worldwide. Sweden adopted another
strategy to decrease the spread of infection and implemented recommendations rather than
specific rules (Adler et al., 2022). Nevertheless, recommendations, including social
distancing, avoidance of using public transportation, and remote work, were communicated
(FHM, 2020). Unexpected events and disasters can challenge an organization's performance,
making achieving goals and sustaining competitiveness difficult (Aksoy & Saglam, 2020).
Recently, scholars have turned their attention towards the Covid-19 pandemic's ongoing crisis
(e.g., Delfino & Van Der Kolk, 2021; Gomez-Conde et al., 2022; Kober & Thambar, 2022;
Passetti et al., 2021), discovering that MCS can serve as a valuable approach to handling
uncertainties. For instance, Passetti et al. (2021) have found that MCS can be helpful by
coordinating actions and processes essential for crisis management.

7
From April 2022, the Public Health Authority (FHM, 2022) no longer classified Covid-19 as
a pandemic. However, uncertainties still remain, and the aftermath of the pandemic together
with the Ukraine war (Regeringskansliet, 2023) has forced organizations to deal with, e.g.,
large disruptions in the supply chains and higher costs due to inflation (European
Commission, n.d). The period of uncertainty that has followed the pandemic until this day is
referred to as post-Covid-19. Hence, the post-Covid context is continuously characterized by
high uncertainty that organizations must manage. Irrespective of the circumstances, one can
assume that for-profit organizations constantly strive to achieve optimal outcomes. However,
to achieve high performance, it is essential that they possess the appropriate tools. This
becomes even more crucial in times of turbulence and uncertainty, as witnessed in the post-
Covid-19 era. Therefore, research on the role of MCS assumes significance to help
organizations navigate through such challenging times and emerge successfully.

1.1 Theoretical Problem

Since Robert Anthony released his seminal work (1965) describing the concept of
“management control”, many studies have emerged addressing the topic (Ferreira & Otley,
2009). Different controls are implemented to protect against threats (Merchant & Otley, 2006)
and steer organizations in their environment to achieve their short- and long-term goals (Otley
& Soin, 2014). Together these form different types of Management Control Systems (MCS),
which are defined as “formal, information-based routines” (Simons, 1995). According to
Simons, there are four such Levers of Control (LOC): belief systems, boundary systems,
diagnostic control systems, and interactive control systems.

Simons’ LOC framework (1995) is useful for studying how managers deal with strategic
change and renewal (Arjaliés & Mundy, 2013; Barros & da Costa, 2021; Laguir et al., 2022;
Tuomlea, 2005). Throughout the years, researchers have found different positive influences
on organizational performance resulting from the effective use of levers of controls (e.g.,
Langfield-Smith, 1997; Simons, 1995; Tuomela, 2005; Widener, 2007). Control systems and
performance measurements are used by managers to make strategic decisions, reach
organizational goals, and set the direction of the company (Simons, 2014). The four levers of
control constituting a MCS need to be balanced for organizations to implement strategies
effectively (Simons 1995).

8
Since the 1970’s, contingency theory has been used to explain organizations’ variating use of
MCS (Davila, 1980). Otley and Soin (2014), argue that the existence of change in the external
and internal environment in which organizations operate implies that there will be a constant
change in the design and use of control systems as organizations adapt. Fundamental
differences in how organizations use control systems are found due to the different
environments and circumstances surrounding organizations (Simons, 1995). It is not
implicated that each lever of control is used simultaneously, and they could be used to a
differing degree (Barros & Ferrerira, 2022; Kruis et al, 2016). Depending on the
appropriateness and contingent fit of MCS in a specific context, there could be a link to
enhanced performance of the organization (Chenhall, 2003). There is no single best solution
for the design of MCS, and Fisher (1998) denotes that the success of MCS will contribute to
higher performance for organizations that properly fit the contingent factors by adapting to
environmental changes. Evidence supporting the assumptions derived from contingency
theory can be found in current research. Different constellations of MCS are claimed to be
valuable in different contexts, e.g., for strategy implementation (Ditillo, 2004; Simons, 1994),
in collaborative settings (Malmi & Brown, 2008), for exploitative and explorative
organizations (Bedford, 2015; Bedford et al., 2016; Chenhall & Moers, 2015), and innovation
and new product development (Davila, 2000).

To understand the role of MCS, it is essential to study the contemporary dimensions of MCS
within the context in which organizations operate (Chenhall, 2003). It is also important to
gain sufficient knowledge to succeed with strategies and reach organizational goals in
different contexts. Lately, researchers have focused on the recent crisis associated with the
Covid-19 pandemic (e.g., Delfino & Van Der Kolk, 2021; Gomez-Conde et al., 2022; Kober
& Thambar, 2022; Passetti et al., 2021) and found MCS to be valuable tools for dealing with
uncertainties. This paper contributes to these findings by exploring the Swedish post-Covid-
19 context. It is an unexplored context characterized by new sorts of uncertainty and
uncertainties resulting from the previous crisis. After the Covid-19 pandemic, a decrease in
economic growth and enhanced inflation in countries in the European Union have been
noticed (Hordofa et al., 2022). Post-Covid-19 is also characterized by uncertainty regarding
varying customer demands and inconsistency, price volatility, and supply chain delays (Raj et
al., 2022). Widener (2007) found that as uncertainty increases, the information deficit leads to
a higher reliance on the mechanisms that provide and process necessary information.
Therefore, interactive- and diagnostic controls are shown to be more frequently used and

9
essential in uncertain settings (Bisbe & Otley, 2004; Janke et al., 2014; Simons, 1991;
Widener, 2007). Further it is also essential to clearly communicate the organization’s core
values and direction (Heinicke et al., 2016). However, boundary systems are needed to
maintain control and encourage reasonable creativity (Simons, 1995). Hence, MCS are
arguably crucial for the performance of organizations as uncertainty increases, which is tested
in this paper.

Based on the assumptions from Contingency theory, it is necessary to study different contexts
and address contemporary issues to fully understand the role of MCS. As crises significantly
threaten organizations’ survival (Delfino & Van Der Kolk, 2021), it is important to gain
knowledge about how to cope with such contexts. This paper addresses this and studies the
relationship between MCS and performance in Swedish organizations in an unexplored post-
Covid-19 context. As most researchers have approached a contingent view discussing the
importance of contingent factors, this paper contributes by exploring other perspectives as
well. Post-Covid-19 is a context that follows a long period of uncertainty, which makes it
interesting to study the phenomenon of organizational learning and impact of institutional
forces.

This paper also contributes methodologically. Researchers have mainly focused on examining
the use and tensions between two levers of control, e.g., interactive- and diagnostic control
systems (Abernethy & Brownell, 1999; Henri, 2006; Müller-Stevens, 2020). Other studies
have been carried out as “case studies” examining data on only one- or a few companies
(Barros & Ferreria, 2022; Pešalj et al., 2018). Such studies can be criticized for lacking
generalizability and inability to provide a body of gathered knowledge (Langfield-Smith,
1997). By studying a larger number of organizations’ use of all four levers of controls, this
paper fills this gap. To conclude, this research contributes by broadening the understanding of
the relationship between MCS, organizational performance, and the context in which
organizations operate.

10
1.2 Research Purpose and Question

As the benefit of a constellation of MCS is argued to depend on external circumstances, it is

important to study different contexts to understand the role of MCS for performance. Thus,
this paper aims to contribute to a more comprehensive understanding of the relationship
between MCS, organizational performance, and the external environment. Given the argued
impact of the external environment, this paper tests if uncertainty moderates this relationship.
This paper aims to answer the following two research questions:

 What relationships exist between the use of MCS and organizational performance in
uncertain contexts?
 Does uncertainty impact the relationship between MCS and organizational
performance?

11
1.3 Disposition

Chapter 1
The first chapter of this paper provides a background on management control systems. This is
followed by a discussion of the theoretical problem, and thereafter, the purpose and research
questions of this study are presented.

Chapter 2
The second chapter presents the theoretical foundation for this paper. The LOC framework is
addressed for studying MCS, and institutional- and contingency theory explains the link
between management control and the environment surrounding the organization. Further the
phenomenon of organizational learning is discussed. Finally, the chapter ends with a
discussion of the relationship between LOC, performance, and uncertainty.

Chapter 3
The third chapter consists of the methodology covering the study's underlying research
philosophy and approach. It presents the research design, including the data collection
method, sampling, research- and survey design, and regression models. The chapter also
includes an analysis of the research quality regarding the validity, reliability, and ethical
considerations. It finishes by outlining the data-gathering process.

Chapter 4
The fourth chapter presents the empirical findings generated from the quantitative study. It
includes a reliability test, descriptive statistics tables, and a Pearson correlation matrix.
Finally, the chapter ends with multiple regression analyses of the two models of this thesis.

Chapter 5
The fifth chapter discusses the study's empirical findings together with prior research. Each
hypothesis is discussed separately before overall discussions of each model are presented.

Chapter 6
The sixth chapter presents the conclusions of this paper. It also discusses this study's
theoretical-, empirical-, methodological-, and practical contributions. Finally, the thesis ends
with an analysis of the limitations and suggests future research.

12
2. Literature Review
______________________________________________________________________
The second chapter presents the theoretical foundation for this paper, where the main
theoretical framework used for discussing management control is the LOC framework.
Institutional- and contingency theories explain the link between management control,
organizations’ institutional environment, and contingent factors. After this, a discussion of
organizational learning and the relationship between MCS and organizational performance
follows. Lastly, literature on the relationship between the use of MCS and performance in
uncertainty is presented.
______________________________________________________________________

2.1 Literature Review Method

The literature review is performed through structured- and unstructured reviews (Li et al.,
2008). In this thesis, the unstructured review consists of articles from previous courses, from
our supervisor, and our previous research on the topic. Eight articles are retrieved from
previous course literature on the topic of MCS. Further, 20 articles are found from the authors'
previous research. In total, the unstructured review consists of 28 articles that are collected
without any identifiable structure (Sint et al., 2009).

The structured review consists of articles and books found through different search engines
using keywords. Keywords help identify relevant sources of information and for narrowing
the search (Jesson et al., 2011). The search engines used are Google Scholar, Scopus,
Emerald, and the school library Primo. The keywords used Contingency theory, Institutional
theory, Levers of control, Management control systems, Organizational learning,
Performance, Uncertainty. These keywords are used separately and in different combinations
to exclude topics that are not necessarily relevant (Brocke et al., 2009), often by including the
key term "Management control system.". As the topic of management control is
comprehensive, we chose the most highly cited articles for our initial sample. Also, a
"snowballing" method is used, meaning that some references are found by reading other
references included in this paper. Finally, articles written by well-known authors on the topic
were searched for, for instance, "Robert Simons" and "David Otley." From the structured
review, a sample of 67 articles and books were used in the paper.

13
The PRISMA 2020 model is used as a guideline for structuring the literature review, e.g., by
specifying the inclusion and exclusion criteria (Page et al., 2021). The initial sample consisted
of 95 articles and books, including 28 from the unstructured review and 67 from the
structured review. After this, duplicates and articles that did not fit our purpose and low-
quality articles were removed. Additionally, in the process of writing the thesis, some more
references were found and added to the literature review resulting in a final sample of 61
references. The quality check was performed by using the ABS list (Appendix 1). To ensure
that the references are trustworthy, one should only select literature that is peer-reviewed or
articles from journals that are highly rated (Jesson et al., 2011). Most of the articles included
in the ABS list has a rating of 3 or 4, and only a few has a lower rating of 1 or 2, which means
that the articles used are published in good journals. 79% of the references included in the
literature review are rated in the ABS list. The remaining 21% are either highly cited books,
chapters published by well-known publishers, or written by well-known researchers on the
topic.

Figure 1 - Literature Review

14
2.2 Contingency Theory

Since Robert Anthony released his seminal work (1965) on management control systems,
many studies have been done on the topic (Ferreira & Otley, 2009). MCS literature is varied
and rich including several different frameworks. The use of MCS has been studied in many
different settings and research has shown what works and what does not work in these settings
(Merchant & Otley, 2006). In the original form, MCS is seen as a sort of internal activity
within an organization that aims to coordinate the managers’ actions (Otley & Soin, 2014).
Otley and Soin define management control as “the process of steering organizations through
the environments in which they operate, to achieve both short-term and long-term goals''.

Contingency theory within management accounting began to develop in the 1970s with a
focus on trying to explain the varieties within the practice of that time (Otley, 2016). At that
time, contingency-based research studies mainly discussed the deployment and the use of
budgets as it was the dominant technique used. Otley (2016) further states how the topic has
broadened in its scope and rather than focusing solely on management accounting techniques
within the contingency discussion, it seems sensible to include aspects of MCS. It is also
sensible to argue that all research within the topics of management accounting must take a
‘contingency’ approach as universal solutions to problems in organizational control generally
do not exist (Otley, 2016).

When defining the core of the theory, studies rely on a few assumptions that have been
explicitly stated. It tells us that there is no best way to organize and that there is no universally
appropriate accounting system that can be applied in all organizations, during all
circumstances. Further, organizations need to adapt and devote organizational structures
appropriate to the environmental conditions and work performed to manage the organization
effectively, in which the organizations will find themselves (Otley, 1980). This means,
according to Otley, (1980) that the theory must identify specific aspects of an accounting
system to be associated with assured circumstances and demonstrate an appropriate matching.

Something that has caused difficulty in this regard is the definition of what constitutes an
appropriate matching. Otley (2016) states that the simplicity of existence has been taken as
indicating such a matching. This will in turn assume the achievement of long-run equilibrium.
Another track within more sophisticated studies has examined the firm performance as a point
of reference to whether an appropriate matching has been found (Otley, 2016). The way of

15
using performance as an indicator of perfect matching is potentially problematic in some
ways. MCS is argued to have a small impact on performance, and it is also argued that
performance is an independent contingent variable, in its own right (Otley, 2016). On the
contrary, do other researchers show how the use of MSC enhances performance for
organizations that achieve a proper fit with the contingent factors by adapting to e.g.,
innovation modes and the environment (Bedford, 2015; Felicio et al, 2021).

Otley (2016) aligns the contingency theory with the environmental factors as well.
Organizations operate in environments that are constantly changing and are not always easy to
predict (Otley & Soin, 2014). Sometimes the context is uncertain and could change very
radically and quickly. It can be hard to predict the actions of competitors, technical
innovations that may occur and on what time scale (Otley, 2016). Organizations do neither
operate alone as they are surrounded by stakeholders such as all affecting each other in the
business environment. The overall environment has further grown to be both global and
increasingly unpredictable which strengthens the contingent approach. Fisher (1998) also
identifies external environment uncertainty as a contingent factor. Hence, the contingency
theory suggests that organizations must adapt to environmental conditions and adopt different
systems depending on the environment. The choice of appropriate control techniques will
depend upon the circumstances surrounding the specific organization (Otley, 1999).

Fisher (1998) further identifies interdependencies and firm industry such as size,
diversification, and structure as contingent factors. Performance measures regarding the use of
budgets have been discussed prior in the literature and Otley (1980) illustrated examples with
contradictory evidence. Hopwood (1972) concluded that a more flexible style of budgeting
was likely to lead to more effective organizational performance while Otley, 1978 argued for
better organizational performance using a more rigid style of budgeting. These opposite
results from the two studies display an important situational difference in line with the
contingency approach as the studies was performed in different industries with different
organizational design regarding e.g., interdependencies.

More recent literature from Otley (2016) states that budgetary control has reduced in
significance as a performance measure, rather, non-financial measures have increased in
importance. Simons (1995) constructed a broader framework focusing on a much wider set of
controls. It is further argued that different levers of control are combined and used by the
organization and that the right balance between them is crucial for organizational functioning

16
and performance (Henri, 2006; Mundy, 2010; Simons, 1995). Organizations face different
contingency factors which is the reason for the different use and combinations of the controls
(Kruis et al, 2016). The fundamental body within the contingency theory holds that “it all
depends” on situational factors and that there is not a universally “best” system that every
organization can use (Otley, 1980).

2.3 Institutional Theory

While research on MCS from a contingency perspective argues that organizations must adapt
to the environmental conditions by finding the right combination and use of controls, other
research advocates the institutional perspective where it is assumed that organizations are
passively conforming to normative, coercive, and mimetic pressures (Damayanthi &
Gooneratne, 2017). Institutional theory is a popular theoretical foundation and has been
proven to be a constructive theory for entrepreneurial research (Bruton et al., 2010). Two
important works that can be seen as foundation works of the new institutional theory are the
ones by Meyer and Rowan (1977) and DiMaggio and Powel (1983). However, the ideas of the
institutional theory have existed for many years, and one of the earliest versions is the work
by Selznick in 1957.

In modern society, organizations' formal structures arise in highly institutional contexts

(Meyer & Rowan, 1977). Organizations' structures, as a result of their operating environment,
tend to match this environment through the use of technical interdependencies. This
phenomenon is called "isomorphism" (Meyer & Rowan, 1977) and can be defined as "the
resemblance of a focal organization to other organizations in its environment" (DiMaggio &
Powell, 1991). Organizations that have a similar structure within the same organizational field
become isomorphic with their institutional environment (Suddaby, 2010), meaning that
organizations become more homogenous (Mizruchi & Fein, 1999).

DiMaggio and Powell (1991) propose three sorts of isomorphism: coercive, mimetic, and
normative. Coercive isomorphism arises due to pressures from other organizations that the
organization depends on and the pressure from society regarding cultural expectations
(Mizruchi & Fein, 1999). The second type of isomorphism suggested by DiMaggio & Powell
is a mimetic isomorphism. Mimetic isomorphism arises in situations where the leaders of the
organization are uncertain about how to act. In such uncertain contexts, a typical response is
to mimic a peer that the organization experience as successful (Mizruchi & Fein, 1999).
17
Finally, the third type of isomorphism is normative. Normative isomorphism is associated
with the organization's professionalization, meaning that people who are members of a
profession, e.g., lawyers, are trained similarly, leading to common worldviews (Mizruchi &
Fein, 1991). It is also related to the interaction of members of the same profession, which
enhances the spread of ideas within the profession (DiMaggio & Powell, 1991).

Another claim about the homogeneity of organizations is that when their institutional- and
task environment conflict, the organization decouples their structure from their work activities
(Boxenbaum & Jonsson, 2017). For instance, in highly institutionalized environments, MCS
is found to be decoupled from strategy and the organization's daily operations (Tucker &
Parker, 2015). Decoupling is defined as "separating formal structure from actual practice"
(Meyer & Rowan, 1977). By decoupling, organizations can relieve the tension that arises
from external pressures to change while at the same time keeping flexibility (Jamali, 2010).

Institutional theory is helpful for understanding in what ways and why individuals are
responding in a certain way to management control practices (Barros & Ferreira, 2023). Also,
management controls are formed and implemented in organizations due to tensions between
the internal dynamics within the organization and the external institutional forces. Internal
tension tends to arise as there are different interests within the organization, such as the
interest of managers as well as intra-organizational assumptions, rules, and norms (Gamage &
Gooneratne, 2017). In contrast to the contingent view, the institutional perspective assumes
that organizations must conform to institutionalized expectations. This phenomenon has been
shown to impact the choice of MCS and mechanisms for coordination within the organization.
When choosing MCS, the organization might rather see it as a solution to a legitimacy
problem related to institutional expectations rather than a solution to a control problem
(Tucker & Parker, 2015).

2.4 Organizational Learning

Research within organizational learning draws on and contributes to developments in various

fields, e.g., organizational behavior and theory, economics, information systems, and strategic
management. Argote and Miron-Spector (2011) convey that the core of most definitions is
that organizational learning is a change in the organization's knowledge that appears as the
organization obtains experience. Organizational learning occurs in a context that includes the
organization and the environment surrounding the organization. At a practical level, the
18
ability to learn and adapt is argued to be critical to the performance and long-term success of
the organization (Argote & Miron-Spector, 2011). Within strategic management, a
fundamental premise is an alignment between the organization and its environment that
maintains the competitiveness and survival of the firm in the long run (Hambrick, 1983). To
adjust to the changing environment and to make appropriate strategic choices, organizations
must become aware, make sense of the environment (Daft & Weick, 1984), and draw the
proper lessons, i.e., make the best strategic choices (Child, 1997).

To assess viable strategies, organizations need to learn as it enables them to build an

understanding and interpretation of their environment. Fiol and Lyles (1985) describe four
contextual factors that affect the probability that learning will occur. These have a circular
relationship with learning as they create and reinforce learning and are created by earning.
First, it is asserted that culture influences learning and predicts actions. This is supported by
Miles and Snow (1978), who contend that an organization's strategic posture is intimately
related to its culture. Hence, belief systems and norms partially determine strategy and the
direction of organizational change. Second, it is asserted that strategy and strategic posture
partially influence learning capacity. This influences learning by setting boundaries for
decision-making and a context for how the world is perceived and interpreted. It further
assists by defining the goals and objectives. The third factor is structure, which allows both
innovativeness and new insights. It frequently functions as a result of learning and is essential
in establishing these processes. Organizations can further be designed to promote learning and
reflective action-taking. The last factor is the environment, and the paper describes how
learning involves the creation and manipulation of tension between both change and stability.
The level of stress and the degree of uncertainty about past successes determine the
effectiveness of the conditions of learning discussed, and they also influence how the
environment is perceived and interpreted (Daft & Weick, 1984).

Research further argues that organizations must invest in developing institutionalized learning
mechanisms to revise and elaborate their knowledge. This should be done by facilitating
information gathering and elaboration or by intensifying processes of information
dissemination, storage, and retrieval (Lipshitz et al., 2002). Organizations should further use
mechanisms to support information interpretation and exchange views and attitudes to transfer
individual knowledge and create new organizational knowledge (Nonaka, 1994). Hence,
establishing cross-functional teams, allowing employees to interpret information, and sharing

19
different views, attitudes, and data is argued to make the environment more predictable
(Edmondson, 1996).

2.5 Management Control Systems

There are a lot of different definitions and interpretations of what constitutes an MCS (Malmi
& Brown, 2008). Chenhall (2003) defines MCS broadly and explains it as the systematic use
of a Management Accounting System (MAS) and the use of personal and clan controls.
Chenhall defines MAS as "the systematic use of practices such as budgeting or product
costing to achieve some goal." Further, MCS are typically viewed as tools for strategy
implementation (Simons, 1995). Studies show that control systems are used as tools
contributing to the successful implementation of intended strategies and for the emergence of
new strategies (Henri, 2006).

Malmi and Brown (2008) join in on the broad definition of MCS and state that management
controls consist of the systems and devices that managers use in organizations to control that
the behaviors of their employees correspond to the organization's strategy and objective. Two
examples of MCS suggested by Malmi and Brown are the balanced scorecard and budgeting.
Merchant and Van der Stede (2007) provide a narrower definition of MCS. The authors
describe that MCS is formed to deal with employees' behavior. Management controls are
essential for guarding organizations against unwanted actions of their employees or the
possibility that they fail to succeed with their assignments. Further, the authors argue that
MCS would be unnecessary if one could trust that employees always act in the organization's
interest. (Merchant & Van der Stede, 2007).

In today's literature, there are many different views on how management control functions.
Some authors argue that MCS operates as a package (Bedford & Malmi, 2015; Chenhall,
2003; Malmi & Brown, 2008), while others look at the control components separately
(Simons, 1995). Malmi and Brown (2008) claim that management controls should be seen as
a package of systems rather than one single system. That is because different systems are
introduced at different points in time and by different interest groups and should, therefore,
not be seen holistically as one single system. In line with Malmi and Browns' approach to
viewing MCS as a package, Bedford, and Malmi (2015), discusses management controls as a
package of interrelated mechanisms and examine how these combinations are related in
different contexts. However, one cannot assume that there is always a one-to-one relationship

20
between MCS and a context since the controls constituting a package could also combine
differently with other controls in another context. (Malmi & Brown, 2015).

Although the contention is that controls are designed and used for adapting to changing
environments, the impact of organizational change on MCS as a package has yet to be studied
(Evans & Tucker, 2015). There are some challenges with studying MCS as a package. MCS
as a package is difficult to study empirically due to its complexity and large size (Malmi &
Brown, 2008). Malmi and Brown argue that developing survey instruments that can capture
the underlying phenomena meaningfully is challenging. Contingent factors do not influence
the design of an MCS package, rather, it is influenced by functional issues (Sandelin, 2008).
In particular, MCS as a package instead reveals how conflicting functional demands are being
balanced.

2.6 Simons Levers of control

Organizations achieve control in many ways, ranging from social- and cultural controls to
direct surveillance of employees (Simons, 1995). In his framework from 1995, Simons
presents four levers of control (LOC) that an organization must analyze to implement its
strategies successfully. Top management must understand each of the four vital strategic
variables: core values, risks to be avoided, strategic uncertainties, and critical performance
variables. Each construct is controlled through different systems, so-called levers of control.

The four levers of controls forming the LOC framework described by Simons (1995) are
belief systems, boundary systems, diagnostic control systems, and interactive control systems.
These controls create opposite forces in implementing strategies, a phenomenon referred to as
“the yin and yang.” These controls are either expressed in positive- or negative terms and are
found to be important when managing evolutionary or revolutionary change. The belief
systems and interactive control systems are positive forces, the “yang,” while the two other
levers of control, boundary systems, and diagnostic control systems, are constraining and thus
negative “yin.” Further, the belief systems and boundary systems are described by Simons as
“design attributes of MCS.” In contrast, the diagnostic- and interactive control systems are
instead seen as “attention patterns for performance measurement systems.” (Simons, 1995).

Belief systems enhance opportunity seeking, while boundary systems set rules of the
competition, and combined, these frame the organization’s strategic domain. A similar
relationship exists between the diagnostic- and interactive control systems where the
21
diagnostic controls’ primary focus is the implementation of intended strategies. On the
contrary, interactive control systems focus on opportunity-seeking. Together, these two levers
of control guide the formulation and implementation of the organizational strategy. (Simons,
1995).

Simons further argues that it is a crucial decision for top management to select and use these
different levers of control properly. The author further argues for a combined use of the four
levers as a core feature of the framework. When used together, it is possible to properly
control the business strategy and effectively balance innovation and efficiency (Simons,
1995). The power of all four levers of control cannot be found in using each of them
separately but in how they complement one another when used together (Simons, 2014). How
managers choose to use these four levers of control reflects their values, impacts the
probability that the organization will achieve its goals, and reveal management’s opinions of
subordinates. It also influences the ability to adapt and prosper in the long term. (Simons,
1995).

As discussed, there must be a balance among the four levers of control to implement new
business strategies effectively (Simons, 1995). This is one of the biggest challenges of
management. While managers must allow sufficient innovation and adapt to changing
circumstances in the market, they must also manage the ongoing operations of their business.
MCS must include both strategies that are deliberately intended and incremental strategies
that arise in different parts of the organization (Simons, 1994). Managers use MCS for various
purposes. For instance, they are used to communicate organizational goals or plans, to keep
employees informed, or to monitor performance (Simons, 1995). Further, the use of MCS has
been found valuable for strategic renewal (Simons, 1994), improving commitment (Simons,
1995), and in different stages of a business life cycle (Simons, 2014).

For an organization to achieve its financial- and non-financial goals, the managers must rely
on the initiatives and efforts of their employees (Simons, 2014). Employees must be aware of
the organization’s strategy and goals and their role in achieving these goals. When
organizations grow, the importance of communication and control increases, and at the same
time, they become more challenging to work with it. Therefore, an effective balance and use
of management control systems and performance measures are critical to success. (Simons,
2014).

22
2.6.1 Beliefs Systems

The belief system is the first lever of control described by Simon’s LOC framework (1995).
Managers use belief systems to communicate and systematically reinforce to spread the
organization’s purpose and value and to clarify the direction of the organization. These
controls are value-laden and inspirational forces used by top managers as formal controls.
Belief systems are directed to the subordinates to adopt and are related to the organization’s
business strategy. The communication is done through documents such as mission statements,
credos, and purpose statements.

Belief systems are aimed to help determine what problems to tackle and what solutions to
search for when problems arise in the organization’s implementation of a business strategy.
They must be broad so that members of organizations can commit to the communicated
values and purposes on their terms. It is equally important that they are inspirational for
managers, salespeople, and, for instance, a production worker. (Simons, 1995). To learn new
behaviors, e.g., when the manager is newly appointed or when facing strategic change, belief
systems can provide a new frame of reference for changes and new emerging agendas.
(Simons, 1994). However, belief systems are too vague to be used as formal standards by
which one could measure performance. These positive and inspirational controls are
transformed into purposive activity through the combination of another lever of control,
boundary systems. (Simons, 1995).

2.6.2 Boundary Systems

The second lever of control, boundary systems, are negative forces used to avoid risks within
organizations (Simons, 1994). In contrast to belief systems, boundary systems are negative
forces that establish limits for opportunity-seeking based on the defined business risk
(Simons, 1995). Boundary systems are necessary for enhancing creativity and opportunity-
seeking in a controlled way. Without boundary systems, the positivity and inspiration
communicated through belief systems risk dissipating the firm's resources and energy through
unfocused search behavior. The dilemma described by Simons (1995) is that despite this risk,
it is equally problematic to dictate and specify what subordinates should do. Therefore,
boundary systems are minimum standards dictating what the subordinates should not do.
Individuals in the organization are thus allowed to be creative within these boundaries.
(Simons, 1995). Although these boundary systems are essentially negative and proscriptive,

23
they lead to creativity and flexibility, allowing managers to delegate decisions to subordinates
(Simons, 1995).

One ordinary boundary system is the codes of conduct, including, for instance, proscribed
behaviors regarding conflicts of interest and actions that risk compromising organizations'
trade secrets or other confidential information (Simons, 1995). There is a tendency for
boundary systems to be developed solely in response to discrete incidents or crises. Thus,
boundary systems are often concrete in their prescriptions. However, boundary systems are
generally effective when communicated clearly, and most organizational participants will not
contravene these. (Simons, 1995). Although boundary systems are necessary for, in a
controlled way, allowing creativity, they can hinder adaptation to changes in the market,
environmental conditions, or product and technological developments if they are not set
correctly. (Simons, 1995).

2.6.3 Diagnostic Control Systems

According to Simons (1994), diagnostic control systems are the third formal lever of control
used by top managers to monitor the different organizational outcomes. They are referred to
as “the backbone of traditional management control” and are designed to ensure that
organizations achieve their goals. It is a system built on feedback and includes the correction
of deviations from standards that are predetermined. Three features distinguish these levers of
controls related to the ability to measure outputs, predetermined standards, and to correct
deviations from standards. (Simons, 1995). One can use any formal information system
diagnostically if it fulfills four criteria. These criteria are that one must; (1) be able to set the
goal in advance, (2) measure outputs, (3) calculate or compute the variances of performance,
and lastly, (4) be able to use this variance information as feedback-systems or align the
performance with the goals of the organization. (Simons, 2014).

Diagnostic control systems are negative forces used to ensure that the organization achieves
its goals for the intended strategy (Simons, 1995). It is negative because it constrains
innovations and the search for opportunities. Managers can use diagnostic controls as a tool to
evaluate results and achieve organizational goals. They are used to critically measure the
achievement of certain performance variables necessary to succeed with the organizational
strategy. Two examples of diagnostic control systems are the organizations’ budgets and

24
business plans. (Simons, 1994). The budget could be used as a diagnostic control to convey a
sense of responsibility to the employees and to evaluate their performance. (Simons, 1995).

The diagnostic control systems can only be effective if managers devote sufficient attention to
setting organizational goals, designing incentives, aligning performance measures, reviewing,
and following up on important exceptions (Simons, 2014). However, there are some risks
with using diagnostic control systems, including, for instance, the risk of measuring wrong
variables. The old saying “what gets measured, gets managed” accurately describes this risk
since attention is limited, and therefore, one must pay attention to the right things. (Simons,
2014).

2.6.4 Interactive Control Systems

According to Simons (1995), interactive control systems contrast with diagnostic control
systems. While the diagnostic control systems are negative and constraining, the interactive
control systems stimulate learning and advocate opportunity-seeking. Managers use these
control systems to involve themselves regularly in their subordinates' decisions (Simons,
2014). The idea of interactive control systems is to force communication within the
organization. According to Simons (1995), managers use these systems, for example, to focus
on innovation like new technology and establish new programs. Also, they are used to
implement new strategies, starting at the bottom. However, diagnostic- and interactive control
systems are not different in how they are technically designed. It is the way managers use
these controls that differs (Simons, 2014). Thus, the design of diagnostic and interactive
control systems might look identical.

Diagnostic control systems are insufficient when organizations face competitive markets and
external changes (Simons, 1995). In such conditions, Simons argues that there is a need for
interactive control systems for the organization to survive. Further, Simons explains that four
characteristics define the interactive control system. First, the system's information is essential
and used in management. The second characteristic is that the system regularly requires
attention from the organization's operational managers. Thirdly, the information provided by
the system is discussed in face-to-face meetings, including with top managers and their
subordinates. Finally, interactive control systems are characterized as "catalysts for the
continual challenge and debate of underlying data, assumptions, and action plans." (Simons,
1995).

25
Interactive control systems are "the hot buttons" for managers as they pay much attention to
the information it provides to create ongoing dialogues with their subordinates (Simons,
2014). The interactive dialogue occurs at different organizational levels when new
information must be analyzed. It forces organizational learning and highlights when there is a
need for changing routines, value propositions, or even the whole business strategy. Simons
(1994) further argues that the interactive control system itself is not a unique set of controls.
Instead, any control system can be used interactively by managers. For instance, profit
planning systems, balanced scorecards, or project monitoring systems can be used
interactively. (Simons, 2014).

2.7 Research on Simons’ LOC

Since Simons released his LOC framework in 1995, much research has been made on the role
of different levers of controls (e.g., Bisbe & Otley 2004; Kruis et al., 2016; Tessier & Otley,
2012; and their combinations and interdependencies (e.g., Henri, 2006; Tuomela, 2005;
Widener, 2007). Tessier and Otley (2012) build on Simons' Levers of the Control framework
and look at it holistically. The authors' comprehensive framework clarifies the definition of
positive and negative controls by focusing on the dual role of these controls rather than solely
their quality of them. Further, Tessier and Otley distinguish the managerial intentions from
the employees' perceptions and argue that the quality of control is not the same as the
perception of it. The presentation is a bridge between managers' intentions and the perceptions
of their employees in a way where it influences the way employees perceive the controls.

A negative force can be associated with something inferior, but Tessier and Otley (2012)
support the statement of Simons (1995) as they explain boundary systems in organizations
like brakes on the car. A car cannot operate at high speed without brakes, which applies to
organizational boundaries. Top managers need formal systems to establish explicit limits and
rules for everyone to follow themselves. The authors define boundaries in a way that they can
serve as strategic ways to delineate managers' search for innovative ideas. These systems can
either stand alone as information systems or be incorporated as a part of planning guidelines,
created, and communicated by business managers through different business conduct codes,
strategic planning systems, and operating directives (Simons, 1994). According to Tessier and
Otley (2012), boundaries contribute as social and technical controls of different types, which
specifically will be used to manage compliance. Tessier and Otley further argue ambiguity

26
regarding the relationship between belief and boundary control systems as they operate at
different levels of analysis. Belief systems can be used to encourage performance and manage
compliance, while boundary systems are predominantly interested in compliance. (Tessier and
Otley, 2012).

In their review of the use of Simons' LOC framework, Martyn et al., (2016) found that the
framework has proven to help examine controls, both for qualitative- and quantitative studies.
Although it is based on practices from 25 years ago, the framework is gaining popularity and
is still suggested as helpful in explaining how organizations use MCS (Martyn et al., 2016).
Several researchers have addressed the importance of balancing different levers of control.
For instance, Kruis et al. (2016) examine the concept of balance and find that balance can be
achieved through various combinations of levers. The levers have specific features and ways
of displaying their attendance in practice, and it is in their interdependencies, which warrant
consideration, that relevant aspects reside (Barros & Ferreira, 2022). Even though Simons
(1995) implies that all levers of control are needed to achieve a balance, this balance does not
necessarily mean that the use is of equal weight (Kruis et al., 2016). Kruis et al., conclude that
the balance can be achieved by combining different controls, although these combinations are
limited. To find this balance, managers must be able to choose how they use levers of control
(Mundy, 2010). The manager's choice depends on individual, organizational, and external
factors.

When examining the balance between Simons' LOC, Barros, and Ferreira's (2022) findings
suggest that the presence of control systems that are diagnostic or interactive is stronger than
the belief- and boundary systems. Managers do not have to make a trade-off between
innovation and control, rather it is about finding the right balance. Further, the authors argue
that an organization can be innovative and achieve high performance while having a high
presence of all levers of control. (Barros & Ferreira, 2022). Rezania et al. (2016) findings
support Simons' assumption that all lever of control is essential for different purposes. The
interactive- and diagnostic control systems were found to be significantly crucial for
performance. Also, one should put less emphasis on belief systems as it does not help
managers to adapt as required to changes. Instead, belief systems are being balanced using
other levers of control. (Rezania et al., 2016).

27
2.8 Hypotheses Development

2.8.1 MCS and Organizational Performance

The use of MCS and its positive implications for performance is commonly discussed in the
literature (e.g., Baird et al., 2019; Bedford, 2015; Cardinal et al., 2004; Davila, 2000; Felicio
et al., 2021). Findings show that control systems are crucial for performance, and the balance
between them is even more critical for organizational functioning (Cardinal et al., 2004;
Henri, 2006; Mundy, 2010). The researcher further discussed how organizational crises,
performance problems, slower decision-making, and wasted resources were precipitated by
periods of control system imbalance.

The concept of balance is the core of Simons’ framework, and it is argued that all four levers
are crucial for an effective control system (Simons, 1995). Using the constraining and
enabling levers aims to facilitate the two objectives of monitoring performance to enhance
efficiency and effectiveness and encourage creativity crucial for performance. Hence, it is
argued that the constraining levers, diagnostic and boundary, facilitate control, and the
enabling levers, beliefs and interactive, facilitate experimentation and creativity (Mundy,
2010). Research made by Simons (1994) further shows how top managers can use MCS for
renewal and strategic change and that MCS is necessary for managing both evolutionary and
revolutionary changes.

Effectively deployed belief systems have a positive influence on the other three levers of
control as they provide guidance and help motivate individuals to improve performance and
search for opportunities (Widener, 2007). In an empirical analysis of the LOC framework,
Widener (2007) found several relationships between different levers of controls, both inter-
dependent and complementary. Although these inter-dependent and complementary
relationships may exist, it does not necessarily mean that both conditional impact
performance. Bedford (2015) argues that interactive- and diagnostic controls are often
separated temporally and spatially. In exploitative firms, boundary and diagnostic control
systems are independently related to organizational performance. It has been found that
enabling controls, such as belief and interactive control systems, positively and indirectly
influence performance as it provides new structures (Baird et al., 2019). However, the authors
also found that constraining controls impact performance indirectly.

28
The use of belief systems may be promoted through mission statements to create
organizational structures and routines and communicate core values to inspire and motivate
employees to search, explore, create, and expend (Widener, 2007). Mission statements give
the management a unity of direction to communicate and create a sense of shared expectations
among all levels and generations of employees (Widener, 2007). Bart et al. (2001) concluded
that the ultimate mission statement must have the proper rationale, constrain sound content,
have organizational alignment, and bring about good behavioral change in the desired
direction. Research shows that the development of mission statements has a positive
relationship with organizational performance and contributes positively to it (Bart et al.,
2001). Hence, we hypothesize:

H1: There is a positive correlation between the use of belief systems and organizational
performance.

In line with the contingent view, Felício et al. (2021) conclude that MCS must be adapted to
the external environment and organizations' characteristics to achieve different levels of
performance. Govindarajan (1988) found evidence supporting that when matching the
organizational structure and control systems with the organization's environment, a higher
performance can be reached. However, there is also a risk of changing structures excessively,
as findings suggest that constant change in structure and organizations' routines may lead to
unwarranted disruption that negatively influences organizations' performance (Baird et al.,
2019). However, in dynamic environments, the use of boundary systems may be needed to
provide structure and control and stop employees from engaging in high-risk behaviors (Baird
et al, 2019; Widener, 2007). Having a well-defined code of conduct that communicates
appropriate and off-limits behaviors as well as risks that should be avoided will balance the
inspirational belief systems (Simons, 1995). The balance is needed to enhance efficient
organizations meaning that boundary systems are equally important to affect organizational
performance positively. This leads to the second hypothesis:

H2: There is a positive correlation between the use of boundary systems and organizational
performance.

To reach a balance, the use of diagnostic control systems is known to be a critical success
factor (Simons, 1995). Diagnostic systems motivate employees, monitor their performance,

29
and provide appropriate rewards to ensure they work toward the organization's goals (Baird et
al., 2019). According to Simons (1995), it is further used to provide information to "gain
coherence over the multiple functional strategies that coalesce into realized strategies." The
diagnostic use of budgets can be used to achieve effective performance outcomes by
providing motivation, resources, and information to ensure that organizational strategies and
goals are achieved (Bisbe & Otley, 2004; Davila, 2000; Simons, 1995; Widener, 2007). This
leads to the third hypothesis:

H3: There is a positive correlation between the use of diagnostic control systems and
organizational performance.

The literature on management control further argues for the importance of focusing on
financial and non-financial measures when it comes to performance (Felicio et al., 2021).
Davila (2000) argues that managers work with the implicit assumption that good performance
in non-financials will drive good financial performance for the organization. Dossi and Patelli
(2010) examined the contribution of non-financial indicators to the dialogue within
organizations, which, according to the literature, is expected to impact international
organizations' performance significantly. Their findings strengthened prior literature as
performance measurement systems should be conceived more than just as measurement
diagnostic tools or problem detectors. Instead, they should be used to foster dialogue between
different units and corporate offices and to initiate collaborative discussions about strategic
performance results. The authors provided evidence that non-financial indicators support this
dialogue as it captures key performance drivers and complements the information content of
traditional financial metrics. Davila (2000) supports these findings arguing that cross-
functional integration and collaborative discussions are significantly related to performance.
Collaborative discussions provide managers to learn of new strategic opportunities. As the
strategy develops, objectives and critical success factors must be redefined and conveyed
throughout the organization, meaning that interactive control systems become efficient in
supporting structure (Chenhall & Morris, 1995). Henri (2006) also provides evidence of a
direct relationship between interactive use and capabilities leading to strategic choice and
organizational performance, which supports the fourth hypothesis:

H4: There is a positive correlation between the use of interactive control systems and
organizational performance.

30
2.8.2 MCS, Organizational Performance, and Uncertainty
Langfield-Smith (2007) presents that several studies shed light on the importance of using
MCS for influencing performance and strategic outcomes. Different levers of control are
combined and used by the organization to maintain the right balance between ongoing
operations and innovation to adapt to changing markets (Simons, 1995). The different
combinations are because organizations are experiencing different contingency challenges and
must therefore use levers of control differently (Kruis et al., 2016). To reach the most
effective combinations, the balance of controls must be consistent with the strategic- and
contextual challenges surrounding the organization.

In his LOC framework, Simons (1995) discussed the environmental variable “strategic
uncertainty.” Simons defines strategic uncertainty as “the emerging threats and opportunities
that could invalidate the assumptions upon which the current business strategy is based.” In
strategic uncertainty and risks, the LOC framework argues to drive the choice and use of
control systems, which in turn, influence the organization through organizational learning and
the efficient use of management attention (Widener, 2007). When studying the first lockdown
period during the Covid-19 pandemic, Passetti et al. (2021) found that MCS was used to
maintain organizational control. Enabling and constraining controls were combined to cope
with the problems caused by the Covid-19 pandemic efficiently. These findings add empirical
evidence to an emerging stream of management control literature that discusses the
concomitant and different use of multiple control mechanisms to deal with unplanned
situations (Bedford & Malmi, 2015; Rikhardsson et al., 2021; Van der Kolk et al., 2015),
investigating management control during the unfolding of a pandemic crisis.

Widener (2007) found that all four levers of control are shown to positively impact
performance. Although diagnostic- and interactive control systems have received the most
attention in research related to performance (Tessier & Otley, 2012), belief systems have been
shown to positively influence all other control systems (Widener, 2007). For instance, the
diagnostic- and interactive controls are supported through belief systems as they inspire the
search for new opportunities in line with the organizations’ core values (Tuomela, 2005). In
uncertainty, communicating core values and company direction might be fundamental
(Heinicke et al., 2016). Although it is crucial to emphasize all levers of control, emphasis on
the belief systems is the key. Belief systems set the direction for the other levers of control.
(Heinicke et al., 2016). As belief systems are suggested to have a positive influence on all

31
other control systems, and one could expect that there is an increased need for communicating
the company direction in times of uncertainty, it leads to the following hypothesis:

H5: Uncertainty strengthens the positive correlation between the use of belief systems and
organizational performance.

While Henri (2006) looks at levers of control in different levels of uncertainty, Widener
(2007) develops these findings, investigates the relationships among different levers of
control, and studies how strategic risk and uncertainty impact the use and importance of
diagnostic- and interactive controls. The interdependence among control systems is well-
accepted (Otley, 1999). However, Widener (2007) contributes with evidence from her study
of the LOC framework that these interdependencies are complementary. The idea that control
features are complementary was demonstrated by Milgrom and Roberts in 1995.
Complementarity activities mean that doing more of one activity increases the return of doing
more of other activities. Implying this on control, the authors argue that if organizations
emphasize one of the controls, they will also increase the benefits received from using other
control components more frequently. (Milgrom & Roberts, 1995). For instance, a structure is
necessary for interactive controls to be adequate (Chenhall & Morris, 1995; Widener, 2007).

Widener (2007) found in her study that boundary and diagnostic control systems provide a
necessary structure for belief and interactive control systems. Further, interactive control
systems influence both the use of boundary and diagnostic systems (Widener, 2007). Strategic
performance measurement systems (PMS) can be used diagnostically and interactively, and
their use might also have implications for belief and boundary control (Tuomela, 2005). A
similar conclusion is made by Simons (1995) as he found that belief and boundary systems
are ineffective to use separately. To avoid the risk of unfocused exploratory search behavior
within the organization, boundary systems must be placed to maintain control and encourage
reasonable creativity (Simons, 1995). Delfino and Van der Kolk (2021) found in their study
that constraining controls over employees increased during the Covid-19 pandemic. In line
with Delfino and Van der Kolk’s findings, it could be expected that in an uncertain context
like post-Covid-19 where organizations cannot control the external environment, there is
significant importance in maintaining control within the organization. Also, as belief systems
are claimed to be valuable in uncertain contexts and are complementary to boundary systems,

32
increasing the emphasis on belief systems is expected to increase the benefits received from
using boundary systems. This leads to the following hypothesis:

H6: Uncertainty strengthens the positive correlation between the use of boundary systems and
organizational performance.

Simultaneous to the relationship between belief- and boundary systems, Widener (2007)
discuss that diagnostic control systems are necessary for providing a structure for interactive
control systems. Even though the diagnostic control systems work against deploying
capabilities, these can lead to higher organizational performance through monitoring
achievements of goals, controlling risk-taking, and setting boundaries (Henri, 2006).
Diagnostic control systems can be used for promoting mutual involvement and handling
uncertainties in the external environment and thus lead to higher performance in terms of
higher competitiveness (Laguir et al., 2022). Therefore, it can be expected that using
diagnostic control systems has a significant impact on performance in uncertainty. We
hypothesize:

H7: Uncertainty strengthens the positive correlation between the use of diagnostic control
systems and organizational performance.

As reviewed, it is vital to control risk-taking in organizations. However, when firms face high
environmental uncertainty and value flexibility, the organization benefits from dynamic
tension by balancing diagnostic and interactive performance measurements (Henri, 2006). In
uncertainty, including constant change and high competition, firms need to be creative and
innovative. Thus, interactive use and flexible values are particularly significant in such
contexts. In line with this, empirical research shows that when a firm faces various types of
risk and uncertainty regarding, for instance, competition, market, and environmental issues,
the interactive control systems are effective (Bisbe & Otley, 2004; Simons, 1991; Widener,
2007). However, Henri (2006) concludes that this dynamic tension might not fit organizations
that are experiencing a low amount of uncertainty. Instead, in a stable environment, the
dynamic tension between interactive and diagnostic use could be more useful as organizations
are well-known and have established internal processes. (Henri, 2006). Thus, it can be
concluded that interactive control systems are especially important when organizations face
different sorts of uncertainty. They are used to scan the environment in which the organization

33
operates, for organizations to strategically position themselves (Widener, 2007), and are
needed for finding creative solutions in uncertain contexts (Henri, 2006). Since it is argued
that it is essential to be creative and flexible in uncertainty, for instance, when facing high
competition, using interactive control systems might be even more critical for achieving high
performance in uncertain settings. This leads to the following hypothesis:

H8: Uncertainty strengthens the positive correlation between the use of interactive control
systems and organizational performance.

34
2.9 Research Models
2.9.1 Research Model 1

The first model of this paper illustrates the expected relationship between the four LOC and
organizational performance. As stated in hypotheses 1 to 4 that are deducted from the
theoretical framework presented above, each variable is expected to correlate with
organizational performance positively. However, as researchers argue (Kruis et al., 2016;
Mundy, 2010; Rezania, 2016; Simons, 1995), all levers of control must be combined and
balanced for organizations to succeed with their strategies. Also, Widener (2007) suggests
that interdependencies and complementary relationships exist among the LOC, meaning that
using one control might increase the effect of using another control. Therefore, all LOC are
included in Model 1 when studying their respective relation to organizational performance.
Model 1 displays hypotheses 1, 2, 3 and 4.

Model 1 - LOC and Organizational Performance

35
2.9.2 Research Model 2

The second research model is an extended version of Model 1 which includes a moderating
effect. Like Model 1, this model illustrates the combination of all LOC as researchers assume
that a balanced use of LOC is essential in uncertain environments (e.g., Henri, 2006; Simons,
1995; Widener, 2007). Model 2 extends Model 1 by displaying hypotheses 5, 6, 7 and 8
where this study tests whether uncertainty moderates the expected positive relationship
between each LOC and organizational performance.

Model 2 - LOC, Organizational Performance and Uncertainty

36
3. Method
______________________________________________________________________
The third chapter consists of the methodology covering the study's underlying research
philosophy and approach. It explains the research design, including the data collection
method, sampling, research- and survey design. Further, the multiple regression models are
explained, including the different variables. The chapter also includes an analysis of the
research quality regarding the validity, reliability, and ethical considerations and concludes
by outlining the data analysis process.

3.1 Introduction

This study aims to examine the use of MCS in the post covid context in relation to
organizational performance. The research follows a contingency-based approach and is
conducted using Simons's levers of control framework (Simons, 1995). Depending on the
environment and circumstances surrounding the organization, it is argued to be fundamental
differences in how organizations use control systems. However, it is also argued that the use
of MCS will enhance organizational performance. Each lever of control can be used to a
different degree (Barros & Ferrerira, 2022; Kruis et al., 2016), which allows researchers to
explore them separately. It is further assumed that a balance between the levers of control in
uncertain environments will contribute to the overall organizational performance (Henri,
2006; Simons, 1995; Widener, 2007). The idea of the study conducting a theory and
hypothesis-testing approach and use of scientific methods, which generally coincide with a
quantitative research strategy (Antwi & Hamza, 2015), will be discussed in the following
sections.

3.2 Research Philosophy

As research became more severe, the scientific method was developed to investigate things
that could be tested (Antwi & Hamza, 2015). This research approach underpins the positivism
paradigm, and quantitative research mainly follows ontological and epistemological views
related to the positivistic paradigm (Choy, 2014). A positivistic approach aims to use
quantification to enhance precision in describing parameters and the relationships among
them, which is why thinkers adopt scientific methods and systematize the knowledge
generation process (Antwi & Hamza, 2015). Within the positivistic approach, the quantitative
researcher tends to follow an ontological view assuming that knowledge is objective and

37
quantifiable. It further follows an epistemological view arguing that the goal of science is to
develop the most objective methods possible to get the closest approximation of reality
(Antwi & Hamza, 2015). It is suitable to embrace a positivistic approach and conduct
objective research for this study. Hence, the aim is to get the closest approximation of reality
and to yield pure data and facts uninfluenced by human interpretation or bias (Saunders et al.,
2019). Prior studies within the field have objectively analyzed MCS accordingly to Simons's
levers of control framework (Henri, 2006; Müller-Stewens et al., 2020; Widener, 2007),
which are used as benchmarks within this study.

This study is partly based on the contingency approach, arguing that no universally
appropriate control systems can be applied during all circumstances. It is sensible to argue
that all research on the topic must take a 'contingency' approach as organizations are complex
and work in uncertain environments (Otley, 2016). Hence, using an objective approach to
conducting quantitative research will generate a greater number of answers which will
contribute by taking more perspectives into account. It further facilitates the aim of observing
patterns and being able to draw more general conclusions (Bryman & Bell, 2015). Gathering
quantitative data also allows the performance of comparisons as it is possible to conduct
statistical tests of the data. This study aims to contribute to the developed framework of
Simons (1995) by examining the use and possible interrelations within MCS in relation to
performance. Accordingly, the most appropriate way to investigate organizations' use of MCS
and its relationship to performance is by utilizing a positivist strategy with quantitative,
objective data in numerical form (Bell et al., 2022).

3.3 Research Approach

Concepts are essential for organizing, signaling research interests, and providing discipline
about the study's aims (Bell et al., 2022). The relationship between theory and the research
could be described as involving a choice between theories that can be used as a driver of the
research process in all its phases. This can be depicted as the choice between either inductive
or deductive approaches. For this study, a deductive approach is applied. When using a
deductive approach, the researcher deduces hypotheses based on what is known about the
topic (Bell et al., 2022). Based on their hypotheses, researchers use deductive reasoning to
deduce the observable consequences that are expected to occur with the study's new empirical
data (Antwi & Hamza, 2015). According to Bell et al. (2022), a deductive approach fits well
with the positivist paradigm, which this research applies.
38
3.4 Research Methodology

Within research methodology, there are two general approaches: qualitative and quantitative.
The Quantitative methodology is a strategy that emphasizes the quantification of collecting
and analyzing data (Bryman & Bell, 2015). On the contrary, the qualitative methodology
focuses on words rather than quantity. A quantitative method was chosen for this study since
it provides results that can be summarized and generalized easily (Bryman & Bell, 2007). The
advantage of using a quantitative methodology is that it limits the influence of the researcher's
subjective interpretations (Bryman & Bell, 2015). Such methods allow interpretations of
numbers before conclusions are drawn (Watson, 2015). The idea of quantitative research is
that it operates under the assumption of objectivity (Antwi & Hamza, 2015). Also, as noted by
Bell et al. (2022), using a quantitative method allows the researcher to collect a lot of data
even during a short period and receive generalized results.

This paper aims to study organizations' use of MCS in relation to the external environment
and its impact on organizational outcomes. The idea is to study these relationships
deductively and positively to draw generalized conclusions from the new empirical data
provided in this research. The purpose is not to look deeply into several factors. Instead, it is
to emphasize a few factors and study how they are related. Therefore, a quantitative method
provides a good fit for this paper.

Although the quantitative methodology best suits this paper's purpose, some things could be
improved with this choice. First, quantitative research is criticized for using a "narrow-angle
lens, " focusing only on a few factors (Antwi & Hamza, 2015). For instance, it is criticized for
"failing to distinguish people and social institutes from the world of nature" (Bell et al., 2022).
When using a quantitative methodology, it is traditionally assumed that behavior and
cognition can easily be predicted and explained (Antwi & Hamza, 2015). Thus, a possible
limitation of using a quantitative method is that it solely provides a narrow view of the
concept and use of MCS without considering the unpredicted behavior of people and
organizations. With qualitative methods, things can be studied in their natural settings where
one can make interpretations and where the researcher is the instrument of data rather than
standardized instruments (Antwi & Hamza, 2015).

39
3.5 Choice of Theory

Simons LOC (1995) is chosen as the main theoretical framework for this study. The
framework has been frequently used in literature over the years (Tessier & Otley, 2012). Its
popularity indicate that it is useful when it comes to understanding the role- and use of MCS
(Curtis et al., 2017). Many studies use it to understand the role of MCS in specific contexts
and the influence of the external environment (Martyn et al., 2016). It is suitable to use the
LOC framework for studying the specific context of post-Covid-19 because it allows us to
make comparisons with previous research and thus, broaden the understanding of the role of
each LOC in different contexts and external environments. Also, because this paper aims to
provide generalized conclusions about the role of MCS, it is important to address controls that
can easily be measured. Therefore, the LOC framework is suitable as it only focuses on
formal, measurable controls (Simons, 1995).

Additionally, the contingency theory is chosen as the leading theory to support the research of
MCS in the context of uncertainty. External environmental uncertainty, existing in e.g., the
post-Covid-19 context, is identified as a contingent factor (Fisher, 1998). Contingency
theorists propose that organizations should adjust to these contingent factors and employ
various systems suitable for the environment. As such, the use of suitable control methods is
contingent on the specific circumstances of the organization (Otley, 1999). Moreover, as the
post-Covid-19 context and the uncertainty that follow affect all organizations in Sweden, the
institutional theory was chosen to discuss potential institutionalized behavior of organizations
coping with uncertainty. Also, organizations operating in the post-Covid-19 context have been
surrounded by uncertainty for a long time, both during and after the pandemic. Hence, post-
Covid-19 constitutes an interesting context for studying if the idea of organizational learning
(Argote & Miron-Spector, 2011) is supported.

Finally, the use of MCS, its relationship to organizational performance in this post-covid-19
context, and the concept of contingency are tested in regression with uncertainty as a
moderating variable. The choice of using a moderating variable relies on the notion that
uncertainty would positively strengthen the relationship between the four levers of control and
organizational performance. It is argued that the use of MCS is of significance in times of
uncertainty to reach organizational performance (Delfino & Van der Kolk; 2021; Heinicke et

40
al., 2016; Henri, 2006; Laguir et al., 2022; Widener, 2007), which is why a moderating
variable of uncertainty is used in regression.

3.6 Research Design

3.6.1 Method of Data Collection

The data for the quantitative method is collected through a survey. This method requires a
structured approach, e.g., the survey is designed in advance based on what results the study
aims to find. Further, a self-completion questionnaire is used, meaning that the respondent
answers the questions by completing the survey independently (Bell et al., 2022). The
questionnaire is completed using an online survey called "Google Forms." Since no
interviewer is present, a self-completion survey reduces the risk of biased answers among the
respondents. This is because the survey looks the same for all respondents, and the researcher
cannot influence the answers through, e.g., biased formulations of questions. However, with
self-completion questionnaires, there is a greater risk of missing data as respondents might
actively decide not to answer questions. (Bell et al., 2022). To avoid this, all questions in the
survey were made mandatory to answer.

The data was primarily collected by sending out the survey directly to managers' email
addresses that were found on different company websites. Different business registers found
through UC Select were used to find all the registered companies in Swedish municipalities.
Also, the survey was published on LinkedIn and Facebook and sent directly to managers in
our social networks. The reason for aiming for answers from Swedish managers is that there
needs to be more research on MCS in the context of Sweden, especially in uncertain
environments. Further, the approach to sending out the survey directly to managers was
addressed as it has been a successful strategy used by researchers on the topic previously
(e.g., Bedford & Malmi, 2015; Henri, 2006; Widener, 2007). The reason for publishing the
survey on social media platforms was primarily to reach out to managers in our networks.
Finally, as an attempt to increase the response rate, the participants got the opportunity to fill
in their email addresses using a separate link presented at the beginning of the survey to take
part in the results. This approach was inspired by Widener's (2007) data collection for her
article, as she promised all respondents to provide a summary of the results from her study.

41
3.6.2 Sampling

The survey was conducted over 14 days in February 2023 with a sample consisting of
managers of organizations located in Sweden. Research within MCS mainly conducts
empirical studies on managers as they are considered knowledgeable in perceived uncertainty,
performance, and use of control systems within the organization (Henri, 2006; Mundy, 2010;
Widener, 2007). The sample selection is further based on the philosophy of collecting a
sufficient amount of data. It is also assumed that all kinds of managers with a comprehensive
insight into the organization have the knowledge to respond to the survey, which argues for
not limiting the sample. It is considered that research on the topic conducted in Sweden is
very limited. Hence, the study aims at organizations' use of control systems in general and is
not limited to a specific industry or company size.

The sampling was conducted to gather data that can be analyzed statistically to generalize and
draw conclusions based on the organization's use of MCS. The study aims to use the practice
of surveying one individual per organization to find out about the whole organization (Bell et
al., 2022). This method is covered under non-probability sampling, argued to have limitations,
and is being criticized for not being a good representative. On the other hand, it is argued to
be helpful when randomization is challenging due to a large population. Hence, it appears to
be an appropriate choice when considering the limited amount of time, resources, and
connections available for the study (Etikan et al., 2016). One of the techniques used was
convenience sampling, as the population we first targeted was easily accessible, available, and
willing to participate in the study. To complement this method, a purposive sampling
technique was further used to increase the population by identifying and selecting individuals
with desired characteristics and knowledge (Etikan et al., 2016).

The survey was first sent to contacts in the authors' networks. Further, the survey was
distributed through Facebook and LinkedIn social media platforms. On these platforms, the
survey was promoted, asking contacts with a manager position to respond. It further
encouraged people to share and spread the survey to reach even more managers. Additionally,
the survey was sent to different managers, mainly CEOs and CFOs but also other types of
managers through email. Email addresses were conducted through company websites all over
Sweden to gather a sufficient sample. In total, the survey was sent to 738 managers using this
method. However, some of these were no longer up to date. These companies were found

42
either from their recognition or through the UC company register. The emails also contained a
promotion to spread the survey to other managers suitable to participate in the study, implying
that a snowball sampling was conducted as well (Bell et al., 2022).

As illustrated in the descriptives Table 1 in Chapter 4, the sample consists of 164 participants.
The minimum size of the sample depends on the researcher's purpose for using multiple linear
regression. (Knofczynski & Mundfrom, 2008). Estimating the significance and the hypothesis
tests often improves with a larger sample size (Eberly, 2007). Tabachnick and Fidell (2013)
give guidance to calculate the sample size using a formula. The number of independent
variables used in the multiple linear regression model is considered: N>50+8m (where m =
the number of independent variables). In this case, the number of independent variables is
four meaning that the minimum sample size is 82. This is achieved with a doubled sample
size of 164.

3.6.3 Pilot study

A pilot test was conducted before the survey was distributed on a full scale. It was sent to
family, friends, and the supervisor with comprehensive knowledge to fill out the survey and
give suggestions, identify voice problems, or respond with feedback if specific questions were
unclear and needed adjustments (Bell et al, 2022). The pilot test also served as a technical
check to measure the average time it took to respond to it and correct minor errors. Finally,
after getting feedback from the pilot study participants, some adjustments to improve the
survey were made before the final document could be distributed on a full scale.

The adjustments made to improve the survey were minor. Some additional industries were
added to make the respondents feel more included. The question about years in the current
position was also reformulated to improve and clarify. To withdraw the risk of respondents
missing filling out questions, they were all made mandatory to increase the response rate.
Lastly, a message at the end of the survey thanking the respondents for their participation was
added.

3.7 Survey Design

The survey is a self-completion questionnaire comprising 37 mandatory questions divided

into eight sections, including one introductory section. The survey starts with control
questions, followed by questions about managers' experienced external uncertainty, their use

43
of management control systems, and the organizations' economic performance. The survey
was initially constructed in English (Appendix 2) and was then translated to Swedish
(Appendix 3) to simplify for the respondents as it was directed to managers in Sweden. Both
the Swedish and English versions are used for collecting data. However, only a small fraction
of the sampling is made through the English survey.

For all questions, standard instruments from reliable articles are used to ensure valid questions
and measurements (Bornton & Greenhalgh, 2004). However, some questions are somewhat
modified to fit the purpose of this study. Most questions in the survey are closed questions.
The motivation for choosing closed questions is because it simplifies both the researchers
when processing answers and the respondents to complete (Bryman & Bell, 2015). Further,
using a Likert scale enhanced the comparability of answers, and the questions were coded
automatically as pre-codes from 1 – 7 were placed to the side of the scale.

The survey starts with an introductory text introducing the respondents to the research topic.
This is followed by some clarifications regarding expected benefits and risks for the
respondents as well as information about how data will be treated with discretion and
anonymity. At the end of this section, the respondents must answer whether they agree to
participate in the study. This is followed by questions related to the control-, dependent-,
independent-, and moderating variables.

3.7.1 Control Questions

The survey consists of 7 control questions working as a power of explanation to ensure the
responses' accuracy and reliability by identifying and eliminating unreliable data. It provides a
picture of who the respondent is when it comes to age, gender, education level, industry, work
position, work experience, and company size. The control questions about gender, education
level, and company size were close-ended, while the questions about the managers' age, work
experience, company industry, and work position were open questions. Open questions were
chosen when one could expect a large variety among the respondents (Reja et al., 2003). Also,
it allows the respondent to answer on their own terms (Bryman & Bell, 2015).

Age: The respondents were asked about their age in an open question as studies show that age
can affect the perception of ethical conduct exercised in, e.g., boundary systems (Deshpande,
1997). Demographic variables, as well as demographic similarity may influence the

44
leadership-subordinate relationship and subsequent performance, which is why the control
variable age was tested (Bernerth et al., 2007).

Gender: Respondents were further asked about their gender as this additionally can impact the
management style and their perception of ethical conduct exercised in, e.g., boundary systems
(Davidson & Ferrario, 1992; Deshpande, 1997; McColl-Kennedy & Anderson, 2005). The
question was close-ended with the alternatives of being a man, woman, or "other." They were
coded with 0 = Women and 1 = Man, as no one responded to be "other."

Educational level: To discover patterns when it comes to the degree of education and use of
control systems, the respondents were asked about the level of completed education
(Deshpande, 1997). The question was close-ended and coded with 1= High School, 2 =
University, and 3= Postgraduate.

Industry: Further, questions are asked about the organization's industry (Alla Bolag, n.d.).
This question was included to determine possible patterns regarding perceived uncertainty,
use of control systems, and performance depending on the industry (Henri, 2007). The
question was asked as multiple alternative questions with the possibility to answer open-
ended if no alternative appeared suitable. The control variable was later removed from the
data analysis due to the large number of different answers.

Work position: Respondents were asked about their work position to control their manager
position. As the survey was directed to all types of managers, it was interesting to chart what
kind of managers the respondents were. The question was close-ended and coded with 3= Top
managers, 2=Divisional manager, 1= Another manager, 0=Other.

Work experience: The survey included one open control question asking the respondents
about years in their current work position. This was included to examine differences in the
data regarding their experience at the company and in the same position. The same control
question is used in similar research used as benchmarks (e.g., Foster & Davila, 2007; Henri,
2006).

Company size: The respondents were lastly asked about the organization's size with
alternatives categorized following the EU standards (European Union, 2022). This control

45
question is used in prior studies used as benchmarks, and it is found to be relevant to examine
wheatear there is a difference in the use of control systems depending on the size of the
company (e.g., Foster & Davila, 2007; Henri, 2006). The variables are coded from 1-5
depending on the number of employees. 1 = 1 – 10; 2 = 11 – 49; 3 = 50 – 249; 4 = 250 – 500;
5 = More than 500.

3.7.2 LOC, Uncertainty and Performance

The control questions are followed by questions relating to different types of uncertainty,
using Simons’ LOC and organizational performance. All questions, except the control
questions, are close-ended, where the respondent either evaluate assumptions or grade on a
Likert scale from 1 to 7 (Joshi et al., 2015). This scale is used because it has been frequently
and successfully adopted by researchers in the field of MCS (Widener, 2007; Bedford &
Malmi, 2015; Henri, 2007) and thus is a reliable instrument (Hadi & Chatterjee, 2012). Using
existing questions is beneficial as it allows us to compare research within the field (Bryman &
Bell, 2015). Also, by using existing questions, one can be sure that the questions have been
tested successfully before.

First, the respondents get to answer questions about experienced environmental uncertainty,
e.g., economy, competitors, and technology (Bedford & Malmi, 2015). For these questions,
number 1 on the Likert scale indicates that the uncertainty factor mentioned in the question is
very predictable, while number 7 states it is very unpredictable. These questions are part of
the survey as the aim is to study uncertainty as a moderating effect on the relationship
between the use of LOC and organizational performance.

This is followed by questions related to each of the four LOC. The purpose of including
questions related to Simons’ LOC framework (1995) is to provide a general overview of how
managers use MCS. The questions related to the belief- and boundary systems are retrieved
from the study by Widener (2007). Further, the diagnostic- and interactive control systems are
retrieved from Henri’s article (2006). All questions related to the LOC are stated as
assumptions and are measured on a Likert scale from 1 to 7 (1 = Strongly disagree, 7 =
strongly agree). Lastly, the questions about organizational performance are retrieved from
Carmeli (2008) and Delaney and Huselid (1996). Similar to the questions about LOC and
uncertainty, these are measured on a Likert scale from 1 to 7. However, for these questions,
the idea is to compare the organization to competitors (1 = Much worse than our competitors,

46
7 = Much better than our competitors). The questions further measure the performance
financially, the performance in human resources, and the performance within service
development and quality to take several perspectives into account. This was then added
together to get the overall organizational performance.

3.8 Regressions

This paper bases the analysis on two different multiple linear regression models with
belonging hypotheses. This section describes the control variables, independent variables,
dependent variable, and moderating variable. Finally, the two regression models of this paper
are presented.

3.8.1 Control Variables

Control variables have a central role in research. They are often used to receive more accurate
and purified estimates of observed relationships. A control variable is a variable that is
extraneous and is not related to the hypotheses or theories that the researcher aims to test.
Instead, the role of a control variable is to produce distortion within the observed
relationships. (Spector & Branni, 2011).

The control variables included in this paper are gender, age, education, work experience, and
company size. In the first research model, uncertainty is also included as a control variable,
while it acts as a moderating variable in the second research model. These specific control
variables are chosen as they are believed to provide a good overview of the sample, both
about the individual respondents (gender, age, education, and work experience) and also about
what type of organizations are represented by the respondents (company size). Also, even
though these control variables are not related to the hypotheses or research questions of this
paper, their inclusion of them in the analysis could strengthen and purify the observed
relationships (Spector & Branni, 2011).

3.8.2 Independent Variables

The independent variables reflect the factors that are suggested to influence the dependent
variable (Gamst et al., 2008). The independent variables are the “predictors of behaviors,
attitudes, and characteristics” (Wetcher-Hendricks, 2011). The four independent variables x of
this study is the four LOC: belief systems, boundary systems, interactive control systems, and
diagnostic control systems. As stated in the hypotheses, using each of these independent

47
variables is assumed to have a significantly positive impact on the dependent variable y
performance. As four independent variables are used in this study, one can make a more
precise prediction (Sahay 2016).

3.8.3 Dependent Variables

The dependent variable refers to the “behaviors, attitudes, and characteristics predicted by the
independent variables” (Wetcher-Hendricks, 2011). In this study, the dependent variable is
organizational performance. The dependent variable y, organizational performance, can be
predicted using the independent variables. In this paper, four independent variables are used
in the model, enabling the researchers to use more information provided in the model (Sahay,
2016).

3.8.4 Moderating Variable

A moderating variable can be described as “a subset of a class of variables termed, in the

social sciences, test or specification variables” (Sharma et al., 1981). A relationship is
moderated when the relationship between the variables x and y varies due to a third variable
(Robison et al., 2013). It has an impact on either or both the form or the strength between the
dependent and the independent variable. (Sharma et al., 1981). In this study, the moderating
variable is uncertainty. This paper defines uncertainty as “the emerging threats and
opportunities that could invalidate the assumptions upon which the current business strategy
is based” (Simons, 1995).

This variable uncertainty is suitable as a moderation effect as several researchers have found
that the use of MCS is especially important in a different context characterized by uncertainty
(Bisbe & Otley, 2004; Janke et al., 2014; Simons, 1991; Widener, 2007). Further, research
implies that relationships exist between the use of LOC and organizational performance (e.g.,
Baird et al., 2019; Bedford, 2015; Cardinal et al., 2004; Davila, 2000; Felicio et al., 2021).
These two directions of research on MCS thus make it interesting and suitable to test for the
relationship between LOC and organizational performance with a moderation effect
uncertainty. Uncertainty is also chosen as a moderating variable to be able to test the
assumptions of the contingent view.

48
3.8.5 Regression Models

The multiple linear regression model describes the relationship between the dependent
variable y and multiple independent variables x1, x2, ... xk, where y is the response variable,
and the x variables are the predictors (Hadi & Chatterjee, 2012). It contains one dependent
variable and at least two independent variables (Sahay, 2016). The multiple linear regression
model provides an approximation of the true relationship between these variables. Linear
regression models are primarily used as empirical models to estimate unknown and somewhat
complex relationships (Montgomery et al., 2012). If y increases as a response to the increase
of x, there is a positive relationship between the variables (Hadi & Chatterjee, 2012). The
multiple linear regression model in its original form is as follows:
𝑦 = 𝛽0 + 𝛽1 𝑥1 + 𝛽2 𝑥2 … + 𝛽𝑘 𝑥𝑘 + 𝜀

𝛽0 equals the mean response when all predictors are held 0, which is the intercept. Each 𝛽𝑘
represents the slope associated with 𝛽𝑘 and is commonly referred to as partial regression
coefficients (Eberly, 2007). Furthermore, it indicates “the average change in y for each unit
change in 𝑥1when 𝑥𝑘 is constant. .𝜀 stand for “error” and explains all the variability in y that
the linear relations of the independent variables cannot explain (Sahay, 2016).

3.8.6 Model 1 –Direct Effect

This paper's first multiple linear regression model describes the relationship between the
dependent variable y performance and the independent variables x1, x2, x3, and x4, namely:
belief systems, boundary systems, interactive control systems, and diagnostic control systems.
Model 1 represents hypotheses 1 to 4, assuming that positive relationships exist between the
use of all individual LOC and organizational performance. The regression models include all
LOC since they work together and must be balanced to succeed with strategy (Kruis et al.,
2016; Simons, 1995). They are complementary and thus relate to each other (Widener, 2007).
Therefore, Model 1 looks as follows:

𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 = 𝛽0 + 𝛽1 𝐵𝑒𝑙𝑖𝑒𝑓 + 𝛽2 𝐵𝑜𝑢𝑛𝑑𝑎𝑟𝑦 + 𝛽3 𝐷𝑖𝑎𝑔𝑛𝑜𝑠𝑡𝑖𝑐 + 𝛽4 𝐼𝑛𝑡𝑒𝑟𝑎𝑐𝑡𝑖𝑣𝑒 + 𝜀

49
3.8.7 Model 2 – Moderating Effect

The second multiple linear regression model of this paper describes the relationship between
the dependent variable y performance and the multiple independent variables x1, x2, x3, and
x4, namely: belief systems, boundary systems, interactive control systems, and diagnostic
control systems with the moderating variable uncertainty (Unc). Model 2 represents
hypotheses 5 to 8, assuming that the positive relationship between the use of LOC and
organizational performance is strengthened by uncertainty. Thus, Model 2 is as follows:

𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 = 𝛽0 + 𝛽1 𝐵𝑒𝑙𝑖𝑒𝑓𝑥𝑈𝑛𝑐 + 𝛽2 𝐵𝑜𝑢𝑛𝑑𝑎𝑟𝑦𝑥𝑈𝑛𝑐 + 𝛽3 𝐷𝑖𝑎𝑔𝑛𝑜𝑠𝑡𝑖𝑐𝑥𝑈𝑛𝑐

+ 𝛽4 𝐼𝑛𝑡𝑒𝑟𝑎𝑐𝑡𝑖𝑣𝑒𝑥𝑈𝑛𝑐 + 𝜀

3.9 Data Analysis

To test the models and hypotheses that were developed in chapter 2, some statistical tests
were performed. The data can be analyzed using some database, e.g., statistical packages such
as the Statistical Package for the Social Science (SPSS) used in this study. Before performing
the tests in SPSS, the data was transferred to Excel, where it was reviewed and coded. Firstly,
the incomplete and unqualified answers were removed from the dataset before being
translated into codes (Vogt et al., 2014).

Coding can be described as “the analytical task of assigning codes to non-numeric data”
(Geisler, 2018). The researcher must decide how many details the coding scheme should
consider. When doing so, one must consider the research questions and hypotheses to ensure
the coding fits the purpose. A detailed coding scheme will likely provide a better overview of
coding errors and reliability, but it could also lead to a sparse data set and statistical problems.
(Weingart et el., 2004).

This paper manually performed the coding using Excel, a traditional tool for coding data
(Geisler, 2018). As a Likert scale from 1 to 7 was used for all questions, except for the control
questions in the questionaire, codes were assigned directly as the respondent filled in the
survey. Thus, no coding was needed for this set of questions. Regarding the control questions,
the data was coded into relatively large categories. The reason for that is that this research
aims to make general conclusions rather than looking specifically at individual variables.
Thus, the control variables’ primary purpose is to provide an overview of the sample and as a
tool for checking the generalizability.
50
For the questions about age and work experience, no coding was done as the questions were
on an ordinal level of measurement and provided numbers that could be used for the analysis
(Wetcher-Hendricks, 2011). The closed questions regarding gender, education, and company
size were presented on a nominal level of measurement, meaning that the questions permitted
only a few inferences and consisted of no numerical value (Wetcher-Hendricks, 2011). As
only a few alternatives were provided, categories were automatically created, then manually
translated into codes. For instance, for “women” and “men” the numbers 0, respectively 1
were used. The remaining question about work position was categorized into four categories,
where 3 is the code for top managers, and 0 is the code for other managers. All explanations
of the coding can be found in Appendix 4 and are also discussed in the descriptive statistics
section 4.2.

After finishing the coding process, the data was exported to SPSS for further analysis. To get
an overview, the empirical data is summarized into descriptive statistics. These descriptive
statistics tables present the minimum-, maximum-, and mean values as well as a standard
deviation for the dependent, independent, moderating, and control variables. The empirical
data is then statistically analyzed in SPSS (Watson, 2015) using linear regressions (Hadi 6
Chatterjee, 2012) and Pearson correlation. To ensure that the data analysis was correctly
performed, both researchers of this paper and the supervisor analyzed the same data
separately, enabling comparison between the outcomes.

3.10 Research Quality

For research to be valid, it must avoid being misleading and should be trustworthy (Roberts et
al., 2006). Reliability and validity are two ways of demonstrating the trustworthiness of
research findings (Roberts et al., 2006). Except from considering these two concepts, the
research quality is also ensured through consultation with this paper's supervisor regarding the
survey questions. Further, both authors of this paper and the supervisor have analyzed the
empirical findings to ensure that the results show the same and guarantee that the analysis is
correctly performed.

3.10.1 Reliability

Reliability can be used to describe to what extent a particular test, e.g., a survey, will produce
results that are similar in different circumstances under the assumption that all else is equal
(Roberts & Priests, 2006). One reliability index is Cronbach’s alpha coefficient (Tavakol &

51
Dennick, 2011), developed by Cronbach in 1951. Cronbach’s alpha measures the internal
consistency of a test, which means that it measures to what extent the question in the test
measures the same concept. Cronbach’s alpha is expressed in a number between 0 and 1
(Tavakol & Dennick, 2011), where a reliability of 0.9 indicates that 90% of the variability in
the values is true. The Cronbach’s alpha on the empirical findings of this paper is presented in
Table 1. The threshold for an acceptable value of Cronbach’s alpha is often suggested to be
0.6 or 0.7 (Taber, 2018); therefore, one can conclude that all variables measured in the survey
were above the recommended threshold. The only borderline case is the variable
“uncertainty,” which has a Cronbach’s alpha of 0.627 (Table 1).

Further, to assure high reliability, the questions were, although with minor modifications,
retrieved from previous research on the topic in highly cited articles and published in journals
rated in the ABS-list (e.g., Bedford & Malmi, 2015; Henri, 2006; Widener, 2007). By using
instruments that have been repeatedly tested by researchers, the stability and thus the
reliability increase (Heale & Twycross, 2015). Finally, the regression equations have been
formulated and analyzed in consultation with Toni Duras, statistics teacher at Jönköping
International Business School, as well as the supervisor of this thesis, Miguel Gil.

3.10.2 Validity

The other research terminology discussed is validity. Validity describes whether the research
measures what was intended to be measured (Roberts & Priests, 2006). There are two types of
measures, internal- and external validity. Internal validity addresses the reasons why the
research had its outcomes (Roberts & Priests, 2006), while external validity refers to “the
possibility of generalizing the conclusions to situations that prompted the research” (Schram,
2005).

This study ensures internal validity by using questions from previous, well-known research in
the survey (Roberts & Priests, 2006). Further, criterion-related validity is considered as the
instrument used, the Likert scale (Vogt et al., 2014), is commonly used for research on MCS
(e.g., Bedford & Malmi, 2015; Henri, 2006; Widener, 2007). External validity is considered
in the sampling for this research. As external validity refers to the ability to generalize the
empirical findings and apply them to other populations, one must consider factors such as the
representation of participants, the measurements, and the settings (Findley et al., 2021). As
illustrated in Table 2 in the result section of this paper, there is a broad representation of

52
participants regarding, e.g., work experience and company size. The control variable industry
included in the survey was removed due to difficulties in categorizing as it was an open
question resulting in many different answers. Although the gender distribution for managers is
not precisely representative of Sweden (SCB, 2018), the survey questions focus on
organizations as a whole rather than individuals. Therefore, it should not affect the validity
excessively.

3.11 Ethical Considerations

In May 2018, the EU general data protection regulation (GDPR) was imposed. A binding
regulation on processing personal data protects individuals' rights when obtaining, possessing,
and destroying personal data (Sveriges Riksdag, 2022). Several ethical rules need to be
considered when conducting research. These are usually categorized into four main areas:
harm to participants, lack of informed consent, invasion of privacy, and involved deception
(Bell et al., 2022). Ethical issues likely originate from interactions between the participants
and researchers (Bell et al., 2022), which in this study arise when the respondents fill out the
survey. Hence, conducting a survey consists of the central point for ethical considerations.
Considering the ethical rules, we first ensured that all the participants got the information
about conducting the survey voluntarily, and they also had to confirm their voluntary
participation. Sufficient information about the survey, this study, and the research was further
presented in the beginning to ensure that the participants got all the necessary information to
be able to contribute.

When designing the survey, neutrality in the questions was considered to ensure that the
participants were not influenced by their answers (Bell et al., 2022). Closed questions present
different fixed alternatives and therefore face a risk of spontaneity loss. Also, closed questions
force the respondent to choose an alternative even though they might not agree with any of
them (Bell et al., 2022). Therefore, open questions were carefully considered. Any personal
information was not collected in the survey in line with GDPR to ensure the anonymity of the
participants. This paper's authors applied to processing personal data in thesis work by filling
out a form provided by Jönköping International Business School. The supervisor of this paper
certified that this thesis's design meets the data requirements by signing this form. The
answers cannot be traced back to any respondent, yet the data will be erased when the
research is concluded. Finally, the data is handled carefully, and individual answers are only
available to the authors of this paper, the supervisor, and the examiner.
53
4. Empirical Findings
___________________________________________________________________________
The fourth chapter presents the empirical findings generated from the quantitative study. It
starts with a reliability test of all different variables measured. This is followed by descriptive
statistics table to get an overview of the sample and a Pearson correlation matrix. Finally, the
chapter ends with multiple regression analyses of the two models of this paper, including
model fit- and hypothesis tests.

4.1 Reliability Test

To test the reliability of this study, the tool Cronbach’s alpha is used. Cronbach’s alpha
measures the internal consistency of a test, which means that it measures to what extent the
question in the test measures the same concept (Tavakol & Dennick, 2011). Table 1 below
presents Cronbach’s alpha of all the variables measured. As previously described, each
variable is represented by a set of questions measured in the survey.

Table 1 - Reliability Test

As mentioned in Chapter 3, the lowest acceptable value of Cronbach’s alpha is around 0,6 or
0,7 (Taber, 2018). The reliability of each of the variable’s measures is therefore acceptable.
As presented in Table 1, Cronbach’s alpha is high for all independent variables and lower for
the dependent- and moderating variables. Thus, the consistency and reliability of this test are
good.

54
4.2 Descriptive Statistics

Descriptive statistics consists of graphical- and numerical techniques that aim to present and
analyze the data from research. It helps present a sample description, generalize the results,
and make meaning of the results. (Fisher & Marshall, 2009). It provides valuable information
about the distribution and tendency through minimum, maximum, mean, and standard
deviation. When reflecting on the standard deviation as high or low relative to the mean value,
the coefficient of variation is used (Anderson et al., 2020).

After removing non-qualified answer, the total number of observations remaining are 164
answers. Five answers did not qualify due to either lack of consent or because the participant
lacked a manager position. This section is divided into four parts: control variables,
independent variables, dependent variables, and finally, the moderating variable.

4.2.1 Control Variables

Table 2 - Descriptive Statistics Control Variables

The control variable gender ranges from 0 to 1, where 0 represents women, and 1 represents
men. The mean gender equals 0.71, indicating that 71% of the participants are men. Further,
there is a substantial difference between the participants' ages, with the youngest manager
being 22 and the oldest 67 years. Hence, the standard deviation to the mean value is low,
indicating that data are clustered close to the mean (Anderson et al., 2020). The opposite
applies to their work experience, where the mean work experience is 9.12 years with a high
standard deviation of 8.64. This indicates that the respondent's work experience is widely
scattered to the mean. For the control variable work position, the range goes from 0 to 2,
where 2 represents top managers, 1 represents divisional managers, and 0 represents other
managers. The mean value is 1,42 with a low standard deviation meaning that most of the
sample consists of top managers such as CEOs and divisional managers, e.g., CFOs.

55
For the control variable education, one is equivalent to completing a high school education,
two is completing a university education, and finally, three is for completing other post-
graduate education. Education has a mean value of 1.71 with a low standard deviation of 0.64,
meaning that most of the sample have post-graduate education. As mentioned in Section 3,
the control variable size is categorized following the EU standards. For example, one
indicates a small company with 1 to 10 employees, while 5 indicates a larger company with
more than 500 employees. Looking at Table 2, one can conclude that the mean value
regarding company size is 2.77 with a relatively low standard deviation of 1.08, meaning that
a large part of the sample worked at organizations having 50 – 250 employees.

4.2.2 Independent Variables

Table 3 - Descriptive Statistics Independent Variables

The four independent variables consist of Simons lever controls: belief systems, boundary
systems, interactive control systems, and diagnostic control systems. Belief systems and
boundary systems range from a minimum of 1 to a maximum of 7, meaning that the sample
varies regarding the use of these control systems under this measure. It further shows a mean
value of 5.19 with a low standard deviation of 1.18 for belief systems and a mean value of
5.36 with a low standard deviation of 1.27 for boundary systems. Each belief and boundary
systems variable consists of 4 merged questions indicating that data are clustered close to the
mean under this measurement.

Regarding diagnostic control systems, no respondent strongly disagreed with the full extent of
the use of this measure. The range goes from a minimum of 2 to a maximum of 7, with a
mean value of 5.79 and a low standard deviation of 1.20. The use of interactive control
systems is also high, with a mean value of 5.60 and a low standard deviation of 0.85,
indicating that the data are clustered close to the mean, making it more reliable (Pallant,
2020).

56
4.2.3 Dependent Variable

Table 4 - Descriptive Statistics Dependent Variable

The dependent variable of this study is organizational performance. The variable is measured
through a set of 7 questions referring to different sorts of organizational performance, e.g.,
revenue growth, innovation rate, and relationships within the organization. Merging the
answers regarding perceived organizational performance, the variable ranges between a
minimum of 3.29 to a maximum of 6.86. The mean value of the variable organizational
performance is 5.20 with a low standard deviation of 0.68, indicating that the data are
clustered close to the mean (Anderson et al., 2020).

4.2.4 Moderating Variable

Table 5 - Descriptive Statistics Moderating Variable

The variables are calculated through 6 questions regarding perceived environmental

uncertainty regarding customers, suppliers, competitors, technology, economy, and
regulations. The moderating variable uncertainty ranges between the minimum value of 2 to
the maximum value of 6, meaning that the sample varies in uncertainty. It further means that
no respondent under this measure perceived a highly predictable environment to the full
extent nor a highly unpredictable environment to the full extent. Merging the answers to each
of the questions regarding uncertainty, the mean value equals 3.99 with a low standard
deviation of 0.87, meaning that the responses regarding perceived uncertainty are clustered
close to the mean.

4.3 Pearson Correlation

A Pearson correlation is conducted to test the degree of correlation between the variables. It is
known for being one of the most frequently used procedures in statistics and can be applied in
various circumstances. It can be used to determine the strength and direction of a linear

57
relationship as it shows whether there is a statistically significant positive or negative
relationship between two variables (Pallant, 2020). Hence, it also measures the degree of
statistical significance attached to a correlation. (Armstrong, 2019).

The range of correlation lies between a perfectly negative correlation (-1) and a perfectly
positive correlation (1). If the correlation is perfectly positive, it indicates that as one of the
variables increases, the other variable reacts by increasing by the same amount (Bryman &
Bell, 2015; Pallant, 2020). It also means that no other variables are related to these two
variables. As illustrated in Table 6, a high positive correlation exists between experience (i.e.,
years in the current manager position) and age (.532**). Other interesting findings show a
positive correlation between belief systems and all the other levers, meaning the more use of
belief systems, the more use of boundary systems (.343**), diagnostic control systems
(.498**), and interactive control systems (.512**). Belief systems are further positively
correlated with organizational performance (.377**)—the more use of belief systems, the
better performance of the organization. A positive correlation is also shown between
interactive controls and boundary systems (-251**) and diagnostic- and interactive control
systems (.569**). Diagnostic control systems (.416**) and interactive control systems
(.471**) are separately correlated positively with organizational performance. This means that
the more managers use diagnostic and interactive control systems, the better organizational
performance is achieved.

If the correlation between two variables is below 1, one variable increases, and the other
decreases. It also means that one of these variables correlates with at least one other variable
(Bryman & Bell, 2015; Pallant, 2020). As illustrated in Table 6, a negative correlation is
shown between company size and experience (i.e., years in the current manager position) (-
.208**). The larger the company (i.e., the more employees), the fewer years the manager
works in the current position. This indicates that managers in larger organizations tend not to
remain in the same position for too long. On the contrary, managers working in smaller
organizations stay in the same position for more years.

58
 Table 6 - Pearson Correlation
59
4.4 Regression Analysis

This paper analyzes the empirical data using multiple linear regression analysis. It is a
statistical technique for estimating relationships between dependent and multiple independent
variables (Uyanik & Güler, 2013). It is also helpful as it allows the researcher to explain and
predict the dependent variable (Orme & Orme, 2009). When the relationship between two or
more variables is known, one can use it to predict outcomes. In multiple regression, one can
predict the dependent variable Y using the independent variable X. (Sahay, 2016). In this
case, the dependent variable Y is predicted using independent variables, i.e., belief systems,
boundary systems, diagnostic control systems, and interactive control systems. Multiple linear
regression assumes that there is a linear relationship between the variables. (Sahay, 2016).
The multiple linear regression analysis is performed on both models. The first model
addresses the four first hypotheses without the moderation effect, while the second model
addresses hypotheses 5 to 8 that consider the moderation effect uncertainty.

60
4.4.1 Model Fit

Table 7 - Model Fit

A model must have a good fit to achieve a valid research model (Kenny, 2015). A good fit
refers to a reasonably consistent model with the sample data (Schermelleh-Engel et al., 2003).
One could look at the F-value to see if the models are statistically significant. The F-test is
performed to test the overall significance of the two regressions (Sahay, 2016). Table 7 shows
that the F-value of Model 1 and Model 2 are 6.425 and 5.012 and are, therefore, both
statistically significant at p < 0.001.

After looking at the F-value, the adjusted R2 is addressed. When measuring the model
adequacy, the adjusted R2 is preferred. The adjusted R2 helps estimate the variance explained
in the sample population (Akossou & Palm, 2013). While the R2 estimates the model fit

61
optimistically, the adjusted R2 corrects the optimistic overestimation. It is thus always equal
to or less than R2. If the value of the adjusted R2 is less than or equal to 0, the regression
model lacks predictive value. (IBM, 2023). A value of 1 indicates a perfect model fit, while a
value of 0 indicates that there is no linear fit (Harel, 2009). The adjusted for both models are
0.268 and 0.270.

Table 7 illustrates which dependent and independent variables are statistically significant and
which are not significant. The significance, or p-value, of a variable in a regression, implies
whether a relationship exists between the regressor variables and the response variable
(Montgomery et al., 2012). If the significance value is less than the significance level, the
result is judged to be statistically significant (IBM, 2023). The choice of significant level in
multiple regression analysis depends on the desired confidence level. A significance level of
0.05 means there is less than a 5% chance that the statistically significant results will occur
from chance (IBM, 2023). A lower significance level of 0.01 could be appropriate if the
research requires a high confidence interval. This research is somewhat exploratory, and the
cost of making a type I error is relatively low. Hence a significance level of 0.1 is used. As
illustrated in Table 7, only one control variable is statistically significant for both models.
Work position positively correlates with the dependent variable in Model 1 (0.176*) and
Model 2 (0.161*).

Looking at the significance of the standardized betas for the independent variables in Model
1, one can see that belief systems, interactive control systems, and diagnostic control systems
are somewhat significant. A high beta indicates a strong effect of the independent and
dependent variables, i.e., organizational performance (Chan, 2004). The only independent
variable that is not significant is boundary systems, indicating that there is no significant
effect between the use of this variable and the dependent variable. Belief systems, diagnostic
control systems, and interactive control systems are all positively correlated to the dependent
variable (performance). In Model 2, there is no significant independent variable, and the
moderating variable uncertainty neither strengthens nor weakens the relationship between the
independent and the dependent variables.

Table 7 also presents the standard error. The standard error is used to estimate how much a
test statistic’s value varies from one sample to another and shows the uncertainty of the test
statistics (IBM, 2023). The standard error is below 0.1 for all variables in both models, except

62
for the control variable gender, where it is 0.108 (Model 1) and 0.109 (Model 2), meaning that
there is some unexplained variability in the sample (Sahay, 2016).

A threat to estimating relationships sought by regression techniques is multicollinearity

(Farrar & Glauber, 1967). Multicollinearities occur when a multiple regression model
contains two or more predictor variables that are highly correlated (Daoud, 2017). Performing
a diagnostic test to assess multicollinearity is further essential to avoid type I error.
Multicollinearity can be measured through the tool VIF. If the VIF value is above 5, there is a
high correlation, while a VIF value of 1 means no correlation. (Daoud, 2017). All VIF values
are below 5, indicating no threat of high correlation in the sample. Since the VIF values only
indicate zero or a meager degree of correlation, as the highest observable VIF value for
models 1 and 2 is 1.837, respectively 1.969, one can conclude that the research design is
good.

4.4.2 Hypothesis Testing Model 1

Table 8 - Hypotheses Testing Model 1

Table 8 provides an overview of the regression analysis of Model 1, including the four first
hypotheses of this paper. As discussed in the previous section, 4.4.1, it was found that three of
the four independent variables are significant in Model 1. The independent variable belief
systems have a significance level of p<0.10 (0.159†). This indicates a weak positive effect on
the dependent variable (organizational performance) when increasing the variable belief
systems. However, the empirical findings support the first hypothesis:

H1: There is a positive correlation between the use of belief systems and organizational
performance.

63
The second independent variable, boundary systems, did not significantly affect the dependent
variable (0.007). This means that increasing the predictor variable boundary systems will have
no significant effect on the outcome variable, not positive or negative. Thus, the second
hypothesis is not supported:

H2: There is a positive correlation between the use of boundary systems and organizational
performance.

The third independent variable analyzed in the regression analysis is diagnostic control
systems. As presented in Table 8, this variable shows a standardized coefficient of 0.176*. It
is, therefore, significant on a significance level of p < 0.05. The standardized coefficient is
positive and indicates that an increase in this predictor variable will lead to an increase in the
outcome variable. In other words, diagnostic control systems are positively correlated to
organizational performance. Therefore, we support the third hypothesis:

H3: There is a positive correlation between the use of diagnostic control systems and
organizational performance.

Simultaneously with the first and third independent variables in Model 1, the fourth
independent variable also shows statistical significance. The fourth independent variable
presented in Model 1 is interactive control systems. This variable has a standardized
coefficient of 0.264** and is statistically significant. It shows an even more substantial effect
on the dependent variable and is thus the variable that has shown the most substantial effect
on performance. This results in support of the fourth hypothesis:

H4: There is a positive correlation between the use of interactive control systems and
organizational performance.

To summarize Model 1 and the four first hypotheses of this paper, hypotheses 1, 3, and 4 are
supported, while hypothesis 2 is not supported.

64
4.4.3 Hypothesis Testing Model 2

Table 9 - Hypotheses Testing Model 2

Model 2 represents the following four hypotheses, hypotheses 5 to 8. In Model 2, the

independent variables are impacted by a moderation effect, namely uncertainty. Thus, the
values essential to study for Model 2 are the independent variables with a moderation effect.
The first independent variable with a moderation effect is belief systems (belief systems x
uncertainty). As illustrated in Table 9, this variable is insignificant (-0.145), meaning that
uncertainty does not moderate belief systems' relationship with performance. This means that
uncertainty does not strengthen or weaken the relationship between using belief systems and
organizational performance. Therefore, we reject the fifth hypothesis:

H5: Uncertainty strengthens the positive correlation between the use of belief systems and
organizational performance.

No significant moderating effect could be found when adding uncertainty to the relationship
between boundary systems and organizational performance. The standardized coefficient (B)
equals 0.084; therefore, one can conclude that uncertainty does not impact this relationship
positively, as expected, or negatively. Thus, the sixth hypothesis of this paper is not
supported:

H6: Uncertainty strengthens the positive correlation between the use of boundary systems and
organizational performance.

Simultaneously to the boundary system, uncertainty does not affect the relationship between
the diagnostic control system and organizational performance. The standardized coefficient

65
(B) equals -0.006 and shows that uncertainty does not moderate this relationship. We must
therefore reject the following hypothesis.

H7: Uncertainty strengthens the positive correlation between the use of diagnostic control
systems and organizational performance.

The last hypothesis, with the use of interactive control systems and the moderating effect of
uncertainty, was argued to be positively correlated with organizational performance. The
standardized coefficient (B) equals -0.059 without significant effect. Thus, uncertainty does
not moderate this relationship, and the last hypotheses must be rejected.

H8: Uncertainty strengthens the positive correlation between the use of interactive control
systems and organizational performance.

To summarize Model 2, uncertainty did not moderate the relationship between each LOC and
organizational performance. Therefore, all four hypotheses related to Model 2 and the
moderation effect are rejected, as uncertainty does not strengthen this relationship.

66
5. Discussion
______________________________________________________________________
The fifth chapter discusses the study's empirical findings together with earlier research and
theory presented in the literature review and from the methodology chapter. Firstly, a short
summary of Model 1 is presented. This is followed by discussion and analysis of each of the
four first hypotheses represented in Model. Thereafter, the same procedure follows for Model
2 including an overall summary and deeper discussion of the four last hypotheses.

5.1 Discussion Model 1

The relationship between MCS and performance is commonly discussed in the literature on
management accounting (e.g., Baird et al., 2019; Bedford, 2015; Cardinal et al., 2004; Davila,
2000; Felicio et al., 2021). However, researchers shed light on the need for and importance of
unpacking the complexity of the use of MCS in different contexts to understand their role and
learn the proper fit (Chenhall, 2003). This paper examines the use of MCS in relation to firm
performance in a Swedish post-Covid-19 environment, which is a new and unexplored
context. Going back to normality after an extensive environmental crisis can be challenging
and characterized by various sorts of uncertainties. Implemented strategies and the use of
controls contribute by enhancing desirable behaviors and creating value for the organization.
Based on earlier research and ideas of MCS, the use of Simons' four levers of control is
argued to support organizational performance (Baird et al., 2019; Davila, 2000; Dossi &
Patelli, 2010; Henri, 2006; Widener, 2007). However, contingency theorists argue that to
enhance performance, the organization must find a proper contingent fit to the external
environment (Otley, 1980). This paper addresses performance as an indicator of a good match
between MCS and the external environment (Otley, 2016). Hence, finding the right balanced
use of controls adapted to the external environment is expected to contribute to the
performance of organizations, which is tested and supported in this paper. Further, the
assumptions from institutional theory, e.g., the phenomena of isomorphism, can be noticed in
the sample.

5.1.1 Hypothesis 1

Prior research has found strong links between belief systems and organizational performance.
Widener (2007) found evidence for belief systems as a guide to help motivate individuals to

67
improve their performance and search for opportunities. Bart et al. (2001) support their paper
by showing a positive relationship between the development of mission statements and
organizational performance. Within this study, hypothesis 1 (H1) argued for a positive
correlation between using belief systems and organizational performance. The result shows to
be positively significant in regression, and the hypothesis can be accepted. As organizations
emphasize belief systems, their performance improves. Thus, this paper supports the
assumptions made by, e.g., Baird et al. (2001), and belief systems are indeed crucial for
organizational performance in a post-Covid-19 context. Relying on the assumptions of
contingency theory, this could be explained as organizations in Sweden have managed to find
a good fit for contingent factors (Otley, 1980). For instance, it could be by adapting to their
operating environment (Bedford, 2015; Felicio et al., 2021).

Belief systems are recognized for being communicated through mission statements, credos,
and purpose statements containing, e.g., core values to inspire the workforce and spread
purpose, value, and clarify the organization's direction (Simons, 1995). The descriptive
statistics showed that managers in the sample perceived a moderately high use of belief
systems with a low standard deviation, meaning that the data is closely clustered (Anderson et
al., 2020). This means that most of the respondents in the sample perceived a similar use of
belief systems which can be explained by the institutional structures (Meyer & Rowan, 1977).
The result can be explained by mimetic isomorphism, as managers tend to follow and mimic a
successful peer when uncertain about how to act. Managers tend to follow what other
managers do as they might be trained similarly to lead an organization which can be
supported by normative isomorphism. The result can further be explained by coercive
isomorphism, as organizations adapt to other organizations' pressure and societal and cultural
expectations (Mizruchi & Fein, 1999). Prior research suggests that mission statements may
legitimately be viewed as statements for communication of means with the external audience,
such as essential stakeholders (Morphew & Hartley, 2006). This implies that organizations
use belief systems to communicate core values, purpose, and the direction of the organization
not only to the workforce but also to the stakeholders as a result of institutional pressure
(Meyer & Rowan, 1977). Having a workforce that is inspired by the core values to search,
explore, create, and expand reflects how they behave around customers, suppliers, and other
stakeholders to improve organizational performance. Hence, belief systems are social controls
used to encourage performance and manage organizational compliance.

68
Hypothesized and supported by Widener (2007), the more a firm emphasizes the belief
systems to communicate the intended strategy and inspire employees to search for
opportunities, the more the firm will emphasize the other three levers in the LOC framework.
The empirical findings of this paper support these assumptions. When analyzing the empirical
data, a positive correlation could be found between using belief systems and the three other
LOC. Placing more emphasis on belief systems increases the emphasis on boundary systems,
diagnostic control systems, and interactive control systems. To conclude, Belief systems
positively influence all other LOC and have a direct positive relationship with organizational
performance in a post-Covid-19 context. This additionally supports the need for balance
between the levers to achieve performance.

5.1.2 Hypothesis 2

Boundary systems work as an opposing and constraining force to avoid organizational risks.
According to prior literature, it is necessary to enhance creativity and opportunity-seeking in a
controlled way (Simons, 1995). Creativity is allowed but within specific boundaries to
enhance efficiency. Having a boundary system to keep up the balance with the inspirational
belief system may be needed to provide structural control and stop employees from engaging
in high-risk behaviors'' (Baird et al., 2019; Widener, 2007). It is also important as the constant
change in structures and routines may lead to unwarranted disruptions that negatively
influence organizations' performance (Baird et al., 2019). Hence, it is hypothesized that
boundary systems would enhance organizational performance.

The empirical result showed no significant relationship between the use of boundary systems
and organizational performance, leading to the rejection of hypothesis 2 (H2). Hence,
boundary systems do not directly improve organizational performance in a post-Covid-19
context. There could be several possible explanations for this occurrence. As discussed,
contingency theorists argue that using MSC enhances performance for organizations that
properly fit the contingent factors (Bedford, 2015; Felicio et al., 2021). Based on the
contingent view, it could be argued that organizations in Sweden have yet to manage to find
the right balanced use of boundary systems in the post-Covid-19 context. However, the more
credible explanation is presumably that boundary systems themselves are design attributes
rather than formal controls, as suggested by Simons (1995). Prior literature is focused on the
importance of balance between the use of all levers to enhance efficient organizations
(Cardinal et al., 2004; Henri, 2006; Mundy, 2010; Simons, 1995). This may imply that

69
boundary systems are equally needed with the other control systems to enhance organizational
performance, but not directly on their own.

The importance of a balance between the use of all four levers is further supported in the
correlation table, where it is shown that boundary systems are positively significant with all
the other three levers. The highest significant correlation is between belief systems and
boundary systems, indicating that the more managers used belief systems, the more they used
boundary systems, which supports prior literature (Baird et al., 2019; Widener, 2007). The
empirical results show a similar but a notch higher mean value of using boundary systems
compared to belief systems. The standard deviation is accordingly low, conforming to the
institutional pressure of managers acting similarly (DiMaggio & Powell, 1991; Mizruchi &
Fein, 1991). Using a boundary system may be stamped by a national culture where Sweden is
grouped with the Nordic (Hofstede, 1983). Research further finds Sweden as a country that
does not pursue prescriptive codes to a high degree, and the intensity of codification is
distinctly lower compared to Anglo countries (Singh et al., 2005). This may explain the
moderate use of boundary systems for most respondents supported by institutional pressure.

5.1.3 Hypothesis 3

Although researchers shed light on the importance of finding a balance between all four LOC
(Henri, 2006; Mundy, 2010; Simons, 1995), Kruis et al. (2016) argued that this balance does
not mean equal weight. For instance, Barros and Ferreira (2022) found a more substantial
presence of diagnostic- and interactive control systems than belief- and boundary systems. As
presented in the descriptive statistics, the empirical data of this paper support these findings.
The mean values of the diagnostic- and interactive control systems are higher than the belief-
and boundary systems. This indicates that the diagnostic- and interactive control systems are
used to a greater extent than belief- and boundary systems, and therefore, this paper supports
the findings made by Kruis et al. (2016) and Barros and Ferreira (2022).

The empirical data showed a positive relationship between the use of diagnostic control
systems and organizational performance in a Swedish post-Covid-19 context. Although the
diagnostic control systems are seen as constraining (Simons, 1995), they are essential both for
motivating and monitoring employees and have, therefore, a significant impact on
organizational performance (Bisbe & Otley, 2004; Davila, 2000; Simons, 1995; Widener,
2007), which have been noticed after the pandemic. The empirical findings, therefore, support

70
previous research and the third hypothesis suggesting a positive correlation between the use of
diagnostic control systems and organizational performance. The contingent view would
explain these findings as a result of a proper fit for contingent factors and the external
environment (Otley, 1980). Simultaneously as with belief systems, it could be the case that
Swedish organizations have found the right balanced use of MCS in aligned with their
contingent environment. The positive impact on performance indicates a good contingent fit
(Cardinal et al., 2004; Henri, 2006; Mundy, 2010; Simons, 1995).

As reviewed, the institutional theory assumes that organizations passively conform to

normative, coercive, and mimetic pressures (Damavanthi & Gooneratne, 2017). The
phenomena of isomorphism could explain the different amount of emphasis placed on the
diagnostic- and interactive control systems compared to the belief- and boundary systems. As
previously discussed, the standard deviations of all the independent variables, LOC, are
relatively low, indicating that the data is clustered close to the mean and the respondents’
answers are not widely spread on the Likert scale. As suggested by institutional theorists,
organizations become isomorphic or homogenous with their institutional environment
(DiMaggio & Powell, 1999; Mizruchi & Fein, 1999; Suddaby, 2010). Because there is a
homogeneity observed in the sample, the assumptions regarding isomorphism and
homogeneity belonging to institutional theory seem to be supported.

The empirical findings indicate that the organizations have generally found a proper
contingent fit. Hence, the positive relationship between diagnostic control systems and
organizational performance is in line with contingency theory, claiming that organizations
have found a good balance in using these controls that fits in this specific post-Covid-19
context.

5.1.4 Hypothesis 4

Interactive control systems are argued to have great significance on performance (Chenhall &
Morris, 1995; Davila, 2000; Henri, 2006). Therefore, fostering a dialogue within the
organization is crucial for, e.g., recognizing and implementing new strategic opportunities
(Davila, 2000).

This paper's fourth hypothesis suggests a positive correlation between the use of interactive
control systems and organizational performance. When analyzing the empirical data of this

71
paper, one can conclude that the data support this hypothesis. Among all four LOC, the
interactive control systems have the strongest significance for organizational performance.
These empirical findings are in line with the conclusions made by Rezania et al. (2016). When
analyzing the LOC framework provided by Simons (1995), the authors found that each LOC
is necessary for different purposes. However, one can conclude that diagnostic- and
interactive control systems are especially important for organizational performance (Rezania
et al., 2016). In line with the contingent view (Otley, 1980), the high use of interactive belief
systems compared to the other LOC provides an excellent fit for contingency factors. In a
post-Covid-19 context, interactive control systems are the most important LOC for
organizational performance.

As the empirical data showed indications of homogeneity among the respondents, the
assumptions of the institutional theory seem applicable. The standard deviation is shown to be
lower for interactive control systems compared to the other independent variable, LOC. The
conclusion could therefore be made that the homogeneity is high when it comes to the use of
interactive control systems. The Pearson correlation table showed a slight positive correlation
between the use of interactive control systems and work position (0.165*). This indicates a
relationship between the use of interactive control systems and the manager's position. This an
be interpreted as support for the assumptions about normative isomorphism. Normative
isomorphism assumes that people within the same profession are trained similarly,
contributing to the homogeneity of organizations (DiMaggio & Powell, 1991; Mizruchi &
Fein, 1991). Although this paper does not focus on comparing different industries or
professions, one could apply this assumption too, for instance, top managers assuming that
their networking with other top managers might contribute to the spread of ideas within the
profession, leading to a homogenous use of interactive control systems (Dimaggio & Powell,
1991).

Interactive control systems are further significant at a high level which additionally supports
the arguments about constraining and enabling levers and the importance of a balance
between them (Baird et al., 2019). Interactive control systems belong together with belief
systems to the enabling levers that need to be controlled and limited to a certain degree to
keep a finite balance (Mundy, 2010). Also, diagnostic- and interactive control systems are
shown to be one of the most frequently used levers of control (Barros & Ferreira, 2022).
Milgrom & Roberts (1995) argue that these two control features are complementary, which

72
could be one explanation for the numbers in the descriptive statistics indicating a similar use
of the interactive- and diagnostic control systems. Also, as both diagnostic- and interactive
control systems are found to be positively correlated to organizational performance, the
assumption made by Milgrom and Roberts (1995) regarding complementarity seems to be
true. A structure is needed for interactive control systems to be efficient (Chenhall & Morris,
1995; Widener, 2007). The structure could be in the form of diagnostic control systems.

5.2 Summary Model 1

As hypothesized, a positive relationship between MCS and organizational performance is

supported by this paper’s empirical data. More specifically, three of the four LOC, belief
systems, diagnostic control systems, and interactive control systems all impact organizations’
performance positively. Based on the contingency theory, the enhanced performance can be
seen as an indicator of a proper contingent fit (Otley, 2016). Hence, Swedish organizations
have managed to find a good balanced use of their MCS, which dependents on their external
environment. The homogenous sample could be explained by the phenomena of isomorphism.
The normative isomorphism can perhaps explain the similar use of e.g., interactive control
systems. Because all respondents have similar work positions, they tend to be similarily
trained which contribute to homogeneity of organizations (DiMaggio & Powell, 1991). As the
post-Covid-19 context contain several new situations, one could also argue that organizations
learned and mimic each other to deal with the uncertainty. Organizations tend to mimic each
other when uncertain how to deal with new situation (Mizruchi & Fein, 1999).

As discussed, boundary systems are unexpectedly not positively correlated with

organizational performance. This could either be explained by the contingent view, arguing
that organizations have not yet managed to find a good contingent fit to the post-covid-19
(Otley, 2016) context or it could be explained by the Swedish culture and norms that are
generally not based on prescriptive codes (Singh et al., 2005). Finally, it could also be
explained by the argument that boundary systems are design attributes rather than formal
controls (Simons, 1995) and thus, cannot impact performance directly.

73
5.3 Discussion Model 2

The empirical findings show that there is a positive relationship between organizational
performance and three out of four LOC: belief systems, diagnostic control systems, and
interactive control systems. In accordance with the contingent view, this indicates that the
organizations have found an excellent fit for their external environment (Bedford, 2015;
Felicio et al., 2021; Otley, 1980). When adding uncertainty, stronger positive relationships
were expected since using control systems in uncertain environments is argued to be of
particular importance for organizational performance (Heinicke et al., 2016; Henri, 2006;
Laguir et al., 2022; Tuomela, 2005; Widener, 2007). Prior research has examined how MCS is
used in uncertain contexts (i.e., during the Covid-19 crisis) and found that it can be mobilized
to identify and embrace opportunities but also how MCS could assist comprehensively and
effectively in crisis management (Kober et al., 2021; Passeti et al., 2021). However, the
empirical findings in this study showed that uncertainty had no moderation effect and, thus,
no impact on these alleged positive relationships. The findings in the post-Covid-19 context
could be interpreted as a result of the phenomena of organizational learning. During the
pandemic, organizations might have learned how to deal with different sorts of uncertainty
and thus gathered knowledge about how to balance their use of MCS depending on the
context, which could explain why uncertainty does not impact the relationship between LOC
and organizational performance in a post-Covid-19 context, or general, periods following a
crisis.

5.3.1 Hypothesis 5

Belief systems are positive forces that enhance opportunity-seeking (Simons, 1995). These
systems are essential for tackling problems and implementing new business strategies
(Simons, 1994). In uncertainty, it is especially important to communicate the core values and
direction of the company (Tuomela, 2005). Because the post-Covid-19 pandemic is
characterized by high uncertainty, e.g., economic instability and challenges within the supply
chain, it was hypothesized that belief systems could be a more powerful tool for achieving
organizational performance than in periods of low uncertainty. However, the empirical
findings did not show support for this argument. Nevertheless, uncertainty has been shown to
have no moderating effect on the relationship between belief systems and organizational
performance.

74
As reviewed, some researchers argue that communicating core values is the key to adapting to
changing circumstances (Simons, 1994), while others suggest that belief systems could
disturb the process of adapting (Rezania et al., 2016). This paper does not fully accept or
reject either of these arguments. Instead, the controversy might be explained by the contingent
view. Contingency theory argues that to manage the organization effectively; organizations
must adapt and devote their structures to their external environment (Otley, 1980).
Organizations can achieve a proper contingent fit and enhance organizational performance by
adapting MCS to their environment (Bedford, 2015; Felicio et al., 2021). Thus, despite their
different views, both Simons (1994) and Rezania et al. (2016) could be right. It is possible
that belief systems do not have the same relationship to organizational performance in all
contexts. Therefore, it might not only be essential to balance the use of MCS (Simons, 1994)
but also to find a proper fit with the environment (Bedford, 2015; Felicio et al., 2021; Otley,
1980).

In the context studied in this paper, post-Covid-19, the empirical findings showed that using
belief systems positively impacts organizational performance. It also showed that during the
past year, all organizations had experienced uncertainty regarding unpredictable factors, e.g.,
the economy, suppliers, and regulations. However, uncertainty did not impact this relationship
positively or negatively. These findings could be interpreted as contradictions to the
contingency theory. As discussed, contingency theorists argue that organizations must adapt
and devote their structures to environmental conditions to manage the organization effectively
(Otley, 1980). Moreover, since the post-Covid-19 context is characterized by new sorts of
uncertainties such as inflation, supply chain problems, and changed consumer behavior
(European Commission, n.d.; Gupta & Mukherjee, 2022), one could argue that uncertainty
would impact the relationship between belief systems and organizational performance if
relying on the assumptions of the contingency theory. This implies that organizations only
sometimes adapt their use of MCS to changing environments in order to find an excellent
contingent fit and enhance performance. Hence, the contingent view is not fully supported in
this paper.

On the other hand, regardless of whether organizations experience high or low uncertainty,
belief systems were found to affect performance positively. This could be interpreted as the
organizations represented in this study have found a good balance of their use of belief
systems with other LOC and a contingent fit. Although this study is performed in an uncertain

75
context, where one could expect belief systems to be essential for organizational performance,
the uncertainty may not be that surprising for the organizations. Moreover, as uncertainty did
not significantly impact the relationship between belief systems and organizational
performance, one could argue that it is because organizations have already adapted their use
of MCS to contingent factors during the pandemic and learned how to cope with uncertainty.

Further support for this explanation is the phenomena of organizational learning. Like
contingency theory, research on organizational learning argues that organizations must
become aware of their environment to make appropriate strategic choices. The ability to learn
and adapt is critical for the success of organizations (Argote & Miron-Spector, 2011; Child,
1997). Learning is defined as "a change in the organization's knowledge that occurs as the
organization acquires experience (Argote & Miron-Spector, 2011). As discussed, uncertainty
did not affect the relationship between using belief systems and organizational performance.
Relating this to organizational learning, one could argue that organizations learned how to
deal with uncertainty during the pandemic, which prepared them to deal with the post-Covid-
19 challenges. In this post-Covid-19 context, organizations have already learned and adapted
their use of belief systems to uncertainty, leading to no change in the observed positive
relationship between belief systems and organizational performance.

5.3.2 Hypothesis 6

In times of uncertainty, boundary systems are argued to prevent the risk of unfocused
exploratory search behavior. Thus, boundary systems must be placed to maintain control and
encourage reasonable creativity (Simons, 1995). When the external environment is turbulent
in times of uncertainty, keeping up the balance within the organization is further essential.
This means that using belief systems that inspire the workforce to search and explore need
boundaries to avoid taking too high risks (Simons, 1995). When the environment surrounding
the organization is unexpected and challenging to control, maintaining internal control using
boundary systems is of higher importance. Hence, it was hypothesized that uncertainty would
strengthen the correlation between the use of boundary systems and organizational
performance.

The empirical data showed no significant relationship between the use of boundary systems
and organizational performance with uncertainty as a moderating effect, leading to rejecting

76
hypotheses 6. Research within MCS is characterized by a contingent view that organizations
must adapt their control systems to the external environment (Otley, 1980). However,
organizations have been forced to deal with uncertainty for an extended period starting with
the outbreak of Covid-19. The uncertainty of dealing with the pandemic required
organizations to strategic change and boundaries communicating nonacceptable behavior was
implemented from one day to another supported by the contingent view (Adler et al., 2022;
Otley,1980). Because no relationship can be found between boundary systems and
organizational performance and the fact that uncertainty did not change this, the assumptions
of contingency theory might not be fully supported. As a result, organizations do not seem to
have adapted their use of boundary systems to their environment, as no significant
relationship with performance could be found. Also, as uncertainty did not affect these
findings, boundary systems are quite static rather than dynamic.

As mentioned, the post-Covid-19 period has been characterized by other uncertainties and
economic turbulence (European Commission, n.d; Gupta & Mukherjee, 2022). This type of
external uncertainty may not require boundaries to the same degree as in other contexts
studied to impact performance, which could explain the contradicting findings. Another
explanation of the result may be connected to Sweden's cultural norms and values. During the
pandemic, Sweden was characterized for running a coronavirus strategy based on the
understanding that an appeal to citizens' self-restraint and sense of responsibility would be
sufficient since Swedes have high confidence in public institutions and governmental agencies
(Adler et al., 2022). Swedish organizations tend to have a relatively high degree of trust in
their external stakeholders, such as suppliers, customers, and regulators. This trust is based on
a long tradition of collaboration and consensus-building in Swedish society, which has led to
a relatively stable and predictable business environment. (Singh et al., 2005). The authors
further find Sweden as a country that does not pursue prescriptive codes to a high degree, and
the intensity of codification is distinctly lower compared to Anglo countries. When dealing
with uncertainty, some societies are more accepting of it, while others are more inclined to try
to avoid it. Societies more inclined to avoid uncertainty are likelier to have more laws and
formal rules to guard against unpredictability (Hofstede, 1983). Given the context that
Swedish organizations have high confidence in public institutions and governmental agencies.
They may not feel the need to use boundary systems to deal with uncertainty and enhance
performance, as they trust that their external stakeholders will act predictably and reliably.

77
5.3.3 Hypothesis 7

Diagnostic control systems are designed for implementing intended strategies. A balanced use
of diagnostic control systems is essential for the success of an organization. (Simons, 1995).
They help motivate the workforce, monitoring strategies, and achieve goals (Bisbe & Otley,
2004; Davila, 2000; Simons, 1995; Widener, 2007). In contexts characterized by high
uncertainty, researchers highlight the importance of interactive control system (Bisbe &
Otley, 2004; Simons, 1991; Widener, 2007). In addition, diagnostic controls are necessary for
these interactive controls to be effective (Widener, 2007). This led to the hypothesis that
uncertainty strengthens the positive correlation between diagnostic control systems and
organizational performance.

Similar to the previous two hypotheses, uncertainty showed no moderation effect on the
relationship between diagnostic control systems and organizational performance. As
presented in the discussion of Model 1, diagnostic control systems showed a positive relation
to performance. This relationship looked the same when adding uncertainty to the equation.

Even though it was hypothesized that the positive relationship between diagnostic control
systems and organizational performance is strengthened by uncertainty, the contradictory
findings could be logically explained. Diagnostic control systems are argued to be necessary
for managing uncertainties in the organization’s external environment and could lead to
competitive advantages and improved performance (Laguir et al., 2022). Even if it contains
new sorts of challenges, the context of post-Covid-19 studied in this paper occurred after a
crisis that was also characterized by different types of uncertainty. As discussed, the non-
impact of uncertainty on the relationship between LOC and organizational performance
indicates that organizations already have considered uncertainty in their balanced use of MCS.
This supports organizational learning, where organizations become aware of their
environment, learn lessons, and make the best strategic choices (Child, 1997; Daft & Weick,
1984).

It also supports the assumptions of the contingent view implying that a positive effect on
performance can be noticed when finding an excellent contingent fit (Cardinal et al., 2004;
Henri, 2006; Mundy, 2010; Simons, 1995). The findings supported a positive relationship
between diagnostic control systems and organizational performance. It indicates that

78
organizations have already learned how to balance their use of diagnostic controls following
their environment and contingent factors. However, as previously discussed, the post-Covid-
19 environment is not entirely similar to the one during the pandemic. New challenges
occurred, and organizations are trying to find the new normal after a protracted crisis. Going
back to the contingency theory, one could argue that the findings of this paper partly reject its
assumptions.

5.3.4 Hypothesis 8

The use of interactive control systems is argued to be of particular significance when

organizations face various types of risk and uncertainty (Bisbe & Otley, 2004; Simons, 1991;
Widener, 2007). Sometimes, organizations are exposed to constant change and high
competition, creating a need for creativity and innovation. When organizations face different
sorts of uncertainty, interactive control systems can be used to scan the environment in which
the organization operates and support them in strategic positioning (Widener, 2007). It is
further argued that in stable environments, the dynamic tension between interactive and
diagnostic use is less useful as organizations are well-known and have established internal
processes (Henri, 2006). Thus, it was hypothesized that uncertainty would strengthen the
relationship between interactive control systems and performance.

The result from the empirical data rejected the hypothesis, as no significant relation was found
between the use of interactive control systems and organizational performance in the context
of uncertainty. Hence, it is shown that uncertainty neither strengthens nor weakens the
relationship between interactive control systems and performance. This is a surprising result
when discussing from a contingent view suggesting that organizations must adapt their
control systems and devote organizational structures appropriate to the environmental
conditions and work performed to manage the organization effectively (Otley, 1980). Higher
performance will then be reached for organizations that properly fit the contingent factors by
adapting to the environment using MCS (Bedford, 2015; Felicio et al., 2021).

Research has shown cross-functional planning meetings as a critical feature of crisis

management (Kober & Thamber, 2021). However, the short-term/operational logic of MCS is
strengthened while the long-term/strategic logic is weakened (Makrygiannakis & Jack, 2016).
This is also supported by Müller (1985), who highlights how the pressure of a crisis inclines
managers to adopt a firefighting mentality, focusing on immediate actions to achieve

79
immediate results. Hence, the cross-functional discussions used in organizations to deal with
uncertainty in this context may focus on trying to solve short-term and urgent problems. This
may imply that the use of interactive control systems in the uncertain context characterized
by, e.g., delays in supply chains, changed consumer behavior, inflation, and other economic
turbulence (European Commission, n.d; Gupta & Mukherjee, 2022), does not strengthen
organizational performance more than under normal circumstances.

These findings could suggest that contingency theory cannot always explain the use of MCS.
It could be argued that if interactive control systems were that critical for organizational
performance during uncertainty as argued (Henri, 2006; Widener, 2007) and if the
assumptions of contingency theory were accurate, the findings would show an adaptation of
the use of these controls to fit the contingent factors and a positive impact on performance.
Therefore, the assumptions about MCS being flexible and adaptable to the environment may
not be accurate.

Another explanation could be found by looking at the context before the one studied. The
business environment has been characterized by a high degree of uncertainty starting from the
beginning of the Covid-19 pandemic (Passetti, 2021). Organizations have been operating in
uncertain environments regarding the economy, suppliers, customers, regulations, etc.
(European Commission, n.d). Hence, through a contingent view, they have been forced to
adapt to the changed conditions and a new environment to keep up with competitors and reach
a new operational and strategic logic (Otley, 1980). Therefore, it can be argued that
organizations have learned to navigate and learned to survive in an uncertain environment.
Experience from a prior period of uncertainty has increased the knowledge in organizations.
The ability to learn and adapt is argued to be critical in the performance and long-term
success of the organization (Argote & Miron-Spector, 2011). It is further argued that
organizations must develop institutionalized learning mechanisms to revise and elaborate their
knowledge (Lipshitz et al., 2002). The establishment of cross-functional teams to interpret
information and share different views may be used to deal with uncertainty (Edmondson,
1996). It is then possible that organizations already have these mechanisms in place from the
period of uncertainty during covid-19. Hence, organizations may not need to implement new
strategies and control systems. This can be a possible explanation of the result indicating that
uncertainty does not affect the relation between interactive control systems and organizational
performance.

80
5.4 Summary Model 2

The findings of this paper supported none of the hypotheses in Model 2. This means that
uncertainty do not moderate the alleged positive relationships between the use of MCS and
organizational performance. The uncertainty incused by the post-covid-19 context neither
strengthens nor weakens the relationship and, as a result, has no impact. The phenomenon of
organizational learning could explain these findings in the post-Covid-19 context. During the
pandemic, organizations may have learned how to handle various types of uncertainty and
how to balance their use of management control systems (MCS) depending on the context.
This may explain why uncertainty did not influence the connection between LOC and
organizational performance in the post-Covid-19 context or during periods following a crisis.

The implications of these findings indicate that contingency theory may not always be able to
account for the employment of MCS. As the result shows, none of the levers of controls were
indeed crucial for organizational performance during this period of uncertainty. If the
assumptions of contingency theory were correct, the results would have shown the adaptation
of the use of these controls to fit the contingent factors and a positive effect on performance.
Thus, the assumptions regarding the flexibility and adaptability of MCS to the environment
may not be entirely accurate.

81
6. Conclusion
______________________________________________________________________
The sixth chapter presents the final conclusions of this paper. It also discusses the theoretical-
, empirical-, methodological-, and practical contributions of this study. The thesis ends with
an analysis of the limitations and provides suggestions for future research.

6.1 Thesis Summary and Main Conclusions

In line with previous research on MCS and organizational performance (e.g., Baird et al.,
2019; Bedford, 2015; Cardinal et al., 2004; Davila, 2000; Felicio et al., 2021), this thesis
found evidence supporting the alleged positive relationship between MCS and organizational
performance. This paper aims to contribute to a more comprehensive understanding of the
relationship between the use of MCS and organizational performance in relation to
environmental context. Different constellations of MCS are valuable in different contexts
(Ditillo, 2004; Simons, 1994). Thus, it is important to study different contexts to understand
the role of MCS and its relation to performance (Chenhall, 2003). This paper studies the
unexplored and uncertain context of post-Covid-19 in Sweden. Based on data collected from
164 managers of Swedish organizations, one can conclude that emphasizing belief systems,
diagnostic control systems, and interactive control systems enhances organizations’
performance. Moreover, by addressing performance as an indicator of a good match between
MCS and the external environment (Otley, 2016), these findings could be interpreted as
support for the assumptions from the contingency theory, suggesting that organizations must
adapt their use of MCS and find a proper fit to contingent factors to improve performance.

Surprisingly, the findings show that boundary systems alone do not lead to higher
performance. There could be several explanations for this occurrence. The non-existing
relationship between boundary systems and organizational performance could result from a
non-fit between the use of boundary systems and the organizations’ external environment. For
instance, the findings could be related to Swedish norms and values, leading to less reliance
on prescriptive codes. However, as boundary systems are described as design attributes rather
than formal controls (Simons, 1995), a more likely explanation is that boundary systems are
not enough to impact performance directly. The findings of this paper further support the
assumption that controls work together (e.g., Kruis et al., 2016; Mundy, 2010; Simons, 1994;
Simons, 1995; Widener). A positive correlation between all LOC could be found, meaning

82
that as managers place more emphasis on one of the LOC, they tend to increase the use of the
other LOC as well. Based on the assumptions from contingency- and institutional theory and
research on the LOC framework, this thesis suggests that there might not be a uniform set of
controls. Managers must be able to choose the right balance of controls to succeed, which is
impacted by internal and external factors (Mundy, 2010). This thesis also shows evidence of
institutional forces influencing organizations’ use of MCS. The institutional phenomena of
isomorphism (Meyer & Rowan, 1977) could explain the relatively homogenous sample as the
emphasis placed on the different LOC is generally relatively similar for the organizations
represented.

As researchers highlight the importance of different MCS during uncertain circumstances

(Henri, 2006; Heinicke et al., 2016; Kruis et al., 2016; Simons, 1995), this thesis studies the
relationship between LOC and organizational performance in a post-Covid-19 context and
incorporates uncertainty as a moderator. Since researchers argue that MCS is positively
related to different sorts of organizational performance (e.g., Baird et al., 2001; Chenhall,
2003; Fisher, 1998; Mundy, 2010; Widener, 2007) and valuable for managing uncertainty and
change (e.g., Henri, 2006; Laguir et al., 2022; Simons, 1995; Tuomela, 2005), LOC was
expected to be even more positively related to organizational performance when experiencing
high uncertainty. However, the empirical data supported no moderating effect when including
uncertainty.

Therefore, it can be concluded that in contexts where organizations have experienced

uncertainty for a long time, using LOC is not more important for organizational performance.
This could be explained by contingency theory as well as organizational learning. As
discussed, the findings support the phenomena of organizational learning. It indicates that
organizations learn and develop their knowledge about how to cope with uncertainty and are,
therefore, prepared for the challenges that follow a crisis. Also, the assumptions of
contingency theory can explain this occurrence. Since a positive relationship could be found
between three out of four LOC and organizational performance, it can be concluded that when
organizations recently have managed to find a good balance in their use of MCS (Simons,
1995) with its context and contingent factors (Otley, 1980), facing new sorts of uncertainty
does not have an impact. On the other hand, the findings also imply that contingency theory
cannot always be used to explain the use of MCS as organizations do not seem to react by

83
adapting their use of MCS to the new sorts of uncertainties that characterize the post-Covid-
19 context.

6.2 Theoretical Contributions

Research on the LOC framework has shown that no correct balance of MCS exist, managers
must combine controls suitable for their specific organization (Kruis et al., 2016; Mundy,
2010; Simons, 1995). Simultaneously, contingency theorists argue that there is no universally
appropriate use of control for organizations to apply during all circumstances (Otley, 2016).
Organizations must adapt to their environment and adopt different controls depending on the
circumstances surrounding the organization (Fisher, 1998; Otley, 1999). This paper
contributes by supporting this through the findings showing a positive relationship between
organizational performance and three out of four LOC: belief systems, diagnostic control
systems and interactive control systems, indicating a proper contingent fit (Otley, 2016) and a
good balance of MCS (Simons, 1995). This paper further contributes to contingency theory by
providing insight that institutionalism and organizational learning also contributes to the
choice of MCS and its relationship with performance. Also, in line with contingency theory
and research on MCS and uncertainty, it was hypothesized that LOC are especially important
for organizational performance when facing uncertainty. Through the findings it was
concluded that uncertainty has no effect on these relationships, indicating that organizations
that have faced uncertainty for a long period of time, like in the post-Covid-19 context, seem
resistant to new sort of uncertainty. Thereby, this paper also contributes with some critique
against contingency theory.

The paper contributes by discussing how the phenomenon of isomorphism may provide an
explanation for the relatively similar emphasis placed on different types of management
control systems (i.e., LOC) in the organizations included in the study. This suggests that
institutional forces, such as normative pressures for conformity (Damayanthi & Gooneratne,
2017) may influence organizations' adoption and use of MCS, resulting in homogeneity in
their practices. This aligns with the fundamental premise of institutional theory, which argues
that organizations are influenced by social norms, cultural values, and external pressures, all
of which can shape their behavior and practices, including their adoption and utilization of
management control systems.

84
This paper found evidence supporting the phenomena of organizational learning. It is crucial
for organizations’ performance and long-term success to have the ability to learn and adapt
(Argote & Miron-Spector, 2011). Research on organizational learning suggest that
organizations learn by facilitating information gathering and intensify the process of
disseminating information (Lipshitz et al., 2002). Organizational learning is observed in the
empirical findings of this paper where the impact of uncertainty on the relationship between
LOC and performance is tested. Because uncertainty did not show an impact on the positive
relationship between organizational performance and belief-, diagnostic, and interactive
control systems, it could be interpreted as an organization have already learnt how to deal
with uncertainty during the pandemic. This indicate that organizations are aware of their
environment and have learnt to make appropriate strategic choices (Daft & Weick, 1984). In
conclusion, this paper contributes with evidence of organizational learning in contexts
characterized with uncertainty for a long period of time, like the Covid-19- and post Covid-19
context.

6.3 Empirical & Methodological Contributions

Limited research studies all four levers of controls individually and their respective
correlation with organizational performance (Abernethy & Brownell, 1999; Henri, 2006;
Müller-Stevens, 2020), especially in Sweden. Additionally, limited research studies this
relation in a post-crisis context characterized by a high degree of uncertainty. This thesis
adopts a broad approach, resulting in several empirical contributions to the study's subject
matter. First, the findings support the positive relationship between MCS and organizational
performance. The findings are consistent with previous research on MCS and organizational
performance (e.g., Baird et al., 2019; Bedford, 2015; Cardinal et al., 2004; Davila, 2000;
Felicio et al., 2021) and contribute by further strengthening the understanding of the
relationship between MCS and organizational performance in the specific context of post-
covid-19 in Sweden.

Secondly, the study found that emphasizing belief systems, diagnostic control systems, and
interactive control systems enhances organizational performance. The non-existent
relationship between boundary systems and organizational performance also contributes as it
challenges the assumption that boundary systems are sufficient for improving organizational
performance. It also suggests that other types of MCS may be necessary for effective
performance (Baird et al., 2019; Widener, 2007; Simons, 1995). Additionally, the study found
85
a positive correlation between different types of MCS, indicating that as managers emphasize
one type of control, they also tend to increase the use of other types of controls. This
complementary relationship supports the idea that controls work together and need to be
balanced by managers to achieve optimal performance (e.g., Kruis et al., 2016; Mundy, 2010;
Simons, 1994; Simons, 1995; Widener). Lastly, the study found no moderating effect of
uncertainty on the relationship between MCS and organizational performance. It challenges
the assumption that MCS is more important for organizational performance during uncertain
circumstances (Henri, 2006; Heinicke et al., 2016; Kruis et al., 2016; Simons, 1995) and
suggests that organizations may have learned and developed their knowledge about coping
with uncertainty (e.g., Child, 1997; Daft & Weick, 1984), which can impact the effectiveness
of MCS in the post-Covid-19 context.

The study further contributes with some methodological approaches to the subject matter.
Several case studies have previously been exercised in the field of MCS (Barros & Ferreria,
2022; Pešalj et al., 2018) or focused on only some of the LOC (Abernethy & Brownell, 1999;
Henri, 2006; Müller-Stevens, 2020). However, this thesis contributes by pursuing a
quantitative approach in a new context, focusing on all four LOC. The thesis further pushes
the idea of quantitative style with a moderating effect inspired by, e.g., Davila (2005) and
Müller-Stewens et al. (2020). Incorporating uncertainty as a moderator adds to understanding
how uncertainty may interact with MCS to influence organizational performance, providing
insights into the role of uncertainty as a moderator in the research area. The findings show
that uncertainty does not moderate this relationship, indicating that uncertainty neither
strengthens nor weakens the relationship between the LOC and organizational performance.
Despite these unexpected findings, this paper rearranges and finds reasonable explanations for
this occurrence instead of excluding it or trying to disregard it, making it a methodological
contribution.

6.4 Practical Contributions

One practical contribution of this thesis is related to the criticism of contingency theory and
the phenomena of organizational learning in times of uncertainty discussed in the theoretical
contributions. Although the findings support the contingency theory, this paper also shows
that institutional forces play a part in MCS's design to fit the external environment properly.
Further, the phenomenon of organizational learning is essential for managing uncertainty.
Therefore, a practical contribution is that the organization can understand the importance of
86
the external environment when deciding on the use of MCS, which practically could lead to
enhanced performance.

This thesis further makes a practical contribution to understanding the role of MCS and what
factors impact Swedish organizations' design and use of controls. In the post-Covid-19
context, using belief, diagnostic, and interactive control systems positively impact
performance. This thesis provides guidelines for achieving higher performance by
highlighting the relationship between MCS, organizational performance, and uncertain
contexts. These findings have practical implications for managers of Swedish organizations
operating in uncertain contexts.

6.5 Societal and Ethical Issues and Consequences

The consequences resulting from crises such as the Covid-19 pandemic has proven how
fragile the modern society is. Both individuals and organizations in the society have been
forced to deal with unexpected events and drastic changes. Organizations must adapt to their
external environment and find a proper use of controls to efficiently reach intended goals and
manage uncertainties (e.g., Fisher, 1998; Otley & Soin, 2014; Simons, 1995). This is not only
important for the survival and success of organizations, but also for society. For instance, the
post-Covid-19 pandemic is characterized by large disruptions in supply chains (European
Commissions, n.d.) that occurred due to difficulties to adapt to the drastic changes. Such
disruptions could lead to devastating consequences e.g., in the healthcare industry. In Europe,
manufacturers of medical supplies were forced to deal with supply shortages of materials
needed for their production (Spieske et al., 2022). Hence, as organizations’ actions impact the
society as a whole, it is important to gain knowledge about how organizations efficiently can
deal with uncertainties and succeed with intended strategies. This paper contributes by
highlighting the benefits of organizations’ use of MCS and the importance of adapting to the
external environment to succeed.

Further, this thesis addresses performance as a measure of a proper contingent fit (Otley,
1980). Although the participants of this study represent for-profit organizations, there are
ethical issues related to how to measure performance. One can argue that performance is not
only about the financial performance, but also other factors such as employee well-being and
environmental impact. This paper addresses these ethical issues by using 7 different measures
87
of performance. These are related to financial performance, product quality, employee
satisfaction, and internal relationship. However, this study can still be criticized for not
addressing all relevant measures of performance.

6.6 Limitations

Given the relatively small sample studied in this thesis, the findings may not reflect how the
use of MCS looks in Swedish organizations in general. Hence, one must address the
conclusions with some caution. The same applies to generalizing the findings outside of this
Swedish context. As discussed in Chapter 5, norms and values that characterize Sweden as a
country could impact the use of MCS and response to uncertainty. Further, in line with
contingency- and institutional theory, this could also be impacted by organizations' external-
and institutional environment. Therefore, the findings might not be representative of countries
other than Sweden.

There are also limitations of this thesis related to the sampling. Managers are chosen as
representatives of the different organizations participating in the study. Because the managers
are asked to rate questions regarding, e.g., their relationship with employees or the
organization's economic performance, there is a risk of dishonesty. The respondents might
want to portray themselves as good managers, which could impact the credibility of this
thesis. Because respondents are held anonymous, there is a risk that several respondents
represent the same organization in the study. This constitutes a limitation of this paper since
the 164 responses analyzed are interpreted as representatives of different organizations.
Finally, there is also a limitation related to the questionnaire. Because no question regarding
what country the managers work in is included, there is a risk that the respondents represent
an organization outside of Sweden, which goes against the purpose of this paper.

6.7 Future Research

Although the importance of MCS for organizational performance is commonly discussed in

current literature (e.g., Baird et al., 2019; Bedford, 2015; Cardinal et al., 2004; Davila, 2000;
Felicio et al., 2021), only limited research on how the context in which organizations operate
impact this relationship exists. By exploring the contingent- and institutional view, it became
clear that the external environment is crucial to study to understand the role of MCS. While
this thesis addresses the role of MCS on organizational performance, another suggestion for
future research is to consider internal factors. For instance, it could be to study internal factors
88
such as the structure and size of the organization (Dragnić, 2014) to broaden the
understanding of how to find the most appropriate balance of MCS for organizational
performance. Also, one could look at organizations' annual reports rather than relying on
managers' perceptions when measuring organizational performance to increase the study's
trustworthiness.

In contrast with contingency theory and the assumption that organizations' use of MCS is
dynamic with the environment (Otley, 1980), this thesis found that organizations do not react
to contingent factors, such as new sorts of uncertainty in the external environment when they
have faced a long period of uncertainty before. Therefore, another suggestion for future
research is to continue studying other contexts characterized by various types of uncertainty
for a long period of time to test if this phenomenon, which this thesis interprets as
organizational learning (Argote & Miron-Spector, 2011; Daft & Weick, 1984; Lipshitz et al.,
2002), can be noticed elsewhere. The paper suggests that the relationship between MCS and
organizational performance may vary across different contexts. Further research could
conduct comparative studies across different countries, industries, and organizational types to
explore MCS and performance and how contextual factors may influence this relationship.
Finally, this paper examines organizational performance by measuring economic
performance, service/product quality and development, and human resource performance. A
suggestion for future research is to decompose performance and examine, e.g., financial
performance or human resource performance individually in relation to the use of MCS.
Lastly, it could be suggested to examine the relationship between the use of MCS and
performance in uncertain contexts by conducting a qualitative research approach to get a
deeper understanding of the interrelationships.

89
7. Reference List
Abernethy, M. A., & Brownell, P. (1999). The role of budgets in organizations facing
strategic change: An exploratory study. Accounting, organizations, and society, 24(3),
189–204. https://doi.org/10.1016/S0361-3682(98)00059-2

Adler, N. J., Sackmann, S. A., Arieli, S., Akter, M. M., Barmeyer, C., Barzantny, C., ... &
Zhang, Z. X. (2022). The Grand Challenge None of Us Chose: Succeeding (and
Failing) Against the Global Pandemic 1. In Advances in global leadership (Vol. 14,
pp. 3-85). Emerald Publishing Limited.

Akossou, A. Y. J., & Palm, R. (2013). Impact of data structure on the estimators R-square and
adjusted R-square in linear regression. Int. J. Math. Comput, 20(3), 84-93.

Anderson, D., Sweeney, D., Williams, T., Camm, J., Cochran, J., Freeman, J., & Shoesmith,
E. (2020). Statistics for Business and Economics, 5th edition.

Anthony, R. N. (1965). Planning and control systems: a framework for analysis. Division of
Research, Graduate School of Business Administration, Harvard University.

Antwi, S. K., & Hamza, K. (2015). Qualitative and quantitative research paradigms in
business research: A philosophical reflection. European journal of business and
management, 7(3), 217-225. https://doi.org/10.1080/095372897235082

Argote, L., & Miron-Spektor, E. (2011). Organizational learning: From experience to

knowledge. Organization science, 22(5), 1123-1137
https://doi.org/10.1287/orsc.1100.0621

Arjaliès, D. L., & Mundy, J. (2013). The use of management control systems to manage CSR
strategy: A levers of control perspective. Management Accounting Research, 24(4),
284-300. https://doi.org/10.1016/j.mar.2013.06.003

Armstrong, R. A. (2019). Should Pearson's correlation coefficient be avoided? Ophthalmic

and Physiological Optics, 39(5), 316-327 https://doi.org/10.1111/opo.12636

90
Baird, K., Su, S., & Munir, R. (2019). Levers of control, management innovation and
organisational performance. Pacific Accounting Review, 31(3), 358-375.
https://doi.org/10.1108/PAR-03-2018-0027

Barros, R. S., & da Costa, A. M. D. S. (2021). Management control systems and innovation:
A levers of control analysis in an innovative company. Journal of Accounting &
Organizational Change. https://doi.org/10.1108/JAOC-09-2020-0137

Barros, R. S., & Ferreira, A. M. D. S. da C. (2022). Management control systems and

innovation: A levers of control analysis in an innovative company. Journal of
Accounting & Organizational Change, 18(4), 571–591. https://doi.org/10.1108/JAOC-
09-2020-0137

Barros, R. S., & Ferreira, A. M. D. S. D. C. (2023). Management control systems and

innovation: a case study grounded in institutional theory. Journal of Management
Control, 1–25.

Bart, C. K., Bontis, N., & Taggar, S. (2001). A model of the impact of mission statements on
firm performance. Management Decision, 39(1), 19–35.
https://doi.org/10.1108/EUM0000000005404

Bedford, D. S. (2015). Management control systems across different modes of innovation:

Implications for firm performance. Management Accounting Research, 28, 12–30.
https://doi.org/10.1016/j.mar.2015.04.003

Bedford, D. S., & Malmi, T. (2015). Configurations of control: An exploratory

analysis. Management Accounting Research, 27, 2–26.
https://doi.org/10.1016/j.mar.2015.04.002

Bedford, D. S., Malmi, T., & Sandelin, M. (2016). Management control effectiveness and
strategy: An empirical analysis of packages and systems. Accounting, Organizations
and Society, 51, 12–28. https://doi.org/10.1016/j.aos.2016.04.002

91
Bell, E., Harley, B., & Bryman, A. (2022). Business research methods. Oxford University
Press.

Bernerth, J. B., Armenakis, A. A., Feild, H. S., Giles, W. F., & Walker, H. J. (2007). Leader–
member social exchange (LMSX): Development and validation of a scale. Journal of
Organizational Behavior: The International Journal of Industrial, Occupational and
Organizational Psychology and Behavior, 28(8), 979-1003
https://doi.org/10.1002/job.443

Bisbe, J., & Otley, D. (2004). The effects of the interactive use of management control
systems on product innovation. Accounting, organizations, and society, 29(8), 709–
737. https://doi.org/10.1016/j.aos.2003.10.010

Brocke, J. V., Simons, A., Niehaves, B., Niehaves, B., Reimer, K., Plattfaut, R., & Cleven, A.
(2009). Reconstructing the giant: On the importance of rigour in documenting the
literature search process. https://doi.org/10.17705/1CAIS.03709

Bruton, G. D., Ahlstrom, D., & Li, H. L. (2010). Institutional theory and entrepreneurship:
where are we now and where do we need to move in the future? Entrepreneurship
theory and practice, 34(3), 421–440 https://doi.org/10.1111/j.1540-6520.2010.00390.x

Boxenbaum, E., & Jonsson, S. (2017). Isomorphism, diffusion, and decoupling: Concept
evolution and theoretical challenges. The Sage handbook of organizational
institutionalism, 2, 77–101.

Bryman, A., & Bell, E. (2015). Business research methods. Oxford University Press.

Cardinal, L. B., Sitkin, S. B., & Long, C. P. (2004). Balancing and rebalancing in the creation
and evolution of organisational control. Organisation science, 15(4), 411-431.
https://doi.org/10.1287/orsc.1040.0084

Carmeli, A. (2008). Top management team behavioral integration and the performance of
service organizations. Group & Organization Management, 33(6), 712–735.
https://doi.org/10.1287/orsc.1040.0084

92
Chenhall, R. H. (2003). Management control systems design within its organizational context:
findings from contingency-based research and directions for the future. Accounting,
organizations, and society, 28(2–3), 127–168. https://doi.org/10.1016/S0361-
3682(01)00027-7

Chenhall, R. H., & Moers, F. (2015). The role of innovation in the evolution of management
accounting and its integration into management control. Accounting, organizations,
and society, 47, 1–13. https://doi.org/10.1016/j.aos.2015.10.002

Chenhall, R. H., & Morris, D. (1995). Organic decision and communication processes and
management accounting systems in entrepreneurial and conservative business
organizations. Omega the international journal of management science 23(5), 485-
497. https://doi.org/10.1016/0305-0483(95)00033-K

Child, J. (1997). Strategic choice in the analysis of action, structure, organizations, and
environment: Retrospect and prospect. Organization studies, 18(1), 43-76.
https://doi.org/10.1177/017084069701800104

Choy, L. T. (2014). The strengths and weaknesses of research methodology: Comparison and
complimentary between qualitative and quantitative approaches. IOSR journal of
humanities and socialscience, 19(4), 99-104
https://doi.org/10.6084/m9.figshare.1233344

Cronbach. L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika,
16(3), 297–334. https://doi.org/10.1007/BF02310555

Daft, R. L., & Weick, K. E. (1984). Toward a model of organizations as interpretation

systems. Academy of management review, 9(2), 284-295
https://doi.org/10.2307/258441

Damayanthi, G., & Gooneratne, T. (2017). Management controls in an apparel group: an

institutional theory perspective. Journal of Applied Accounting Research, 18(2), 223–
241. https://doi.org/10.1108/JAAR-09-2015-0075

93
Daoud, J. I. (2017). Multicollinearity and Regression Analysis. Journal of Physics.
Conference Series, 949(1), 12009–. https://doi.org/10.1088/1742-6596/949/1/012009

Davidson, M., & Ferrario, M. (1992). A comparative study of gender and management
style. Management Development Review. 5(1), 13–.
https://doi.org/10.1108/EUM0000000000661

Davila, T. (2000). An empirical study on the drivers of management control systems’ design
in new product development. Accounting, Organizations and Society, 25(4-5), 383–
409. https://doi.org/10.1016/S0361-3682(99)00034-3

Davila, T. (2005). An exploratory study on the emergence of management control systems:

formalizing human resources in small growing firms. Accounting, Organizations and
Society, 30(3), 223-248 https://doi.org/10.1016/j.aos.2004.05.006

Chan, Y. H. (2004). Biostatistics 201: linear regression analysis. Age (years), 80, 140.

Delaney, J. T., & Huselid, M. A. (1996). The impact of human resource management
practices on perceptions of organizational performance. Academy of Management
journal, 39(4), 949–969. https://doi.org/10.2307/256718

Delfino, G. F., & van der Kolk, B. (2021). Remote working, management control changes and
employee responses during the COVID-19 crisis. Accounting, Auditing &
Accountability Journal, 34(6), 1376–1387. https://doi.org/10.1108/AAAJ-06-2020-
4657

Deshpande, S. P. (1997). Managers’ Perception of Proper Ethical Conduct: The Effect of Sex,
Age, and Level of Education. Journal of Business Ethics, 16(1), 79–85.
https://doi.org/10.1023/A:1017917420433

DiMaggio, P. J., & Powell, W. W. (1991). The new institutionalism in organizational

analysis. University of Chicago Press.

94
Ditillo, A. (2004). Dealing with uncertainty in knowledge-intensive firms: the role of
management control systems as knowledge integration mechanisms. Accounting,
organizations, and society, 29(3–4), 401–421.
https://doi.org/10.1016/j.aos.2003.12.001

Dossi, A., & Patelli, L. (2010). You learn from what you measure: financial and non-financial
performance measures in multinational companies. Long Range Planning, 43(4), 498-
526. https://doi.org/10.1016/j.lrp.2010.01.002

Dragnić, D. (2014). Impact of internal and external factors on the performance of fast-
growing small and medium businesses. Management-Journal of Contemporary
Management Issues, 19(1), 119–159.

Eberly, L. E. (2007). Multiple linear regression. Topics in Biostatistics, 165–187.

Edmondson. (1996). Learning from Mistakes is Easier Said Than Done: Group and
Organizational Influences on the Detection and Correction of Human Error. The
Journal of Applied Behavioral Science, 32(1), 5–28.
https://doi.org/10.1177/0021886396321001

Etikan, I., Musa, S. A., & Alkassim, R. S. (2016). Comparison of convenience sampling and
purposive sampling. American journal of theoretical and applied statistics, 5(1), 1-4.

Evans, M., & Tucker, B. P. (2015). Unpacking the package: Management control in an
environment of organizational change. Qualitative Research in Accounting &
Management, 12(4), 346–376. https://doi.org/10.1108/QRAM-07-2015-0062

European Commission (n.d) Jobs and economy during coronavirus pandemic. Retrieved 23
Mars, 2023, from https://commission.europa.eu/strategy-and-policy/coronavirus-
response/jobs-and-economy-during-coronavirus-pandemic_sv.

Eurostat (2022). Structural business statistics overview. European Union.

https://ec.europa.eu/eurostat/statistics-
explained/index.php?title=Structural_business_statistics_overview

95
Farrar, D. E., & Glauber, R. R. (1967). Multicollinearity in Regression Analysis: The Problem
Revisited. The Review of Economics and Statistics, 49(1), 92–107.
https://doi.org/10.2307/1937887

Felício, T., Samagaio, A., & Rodrigues, R. (2021). Adoption of management control systems
and performance in public sector organizations. Journal of Business Research, 124,
593-602. https://doi.org/10.1016/j.jbusres.2020.10.069

Ferreira, A., & Otley, D. (2009). The design and use of performance management systems:
An extended framework for analysis. Management Accounting Research, 20(4), 263–
282. https://doi.org/10.1016/j.mar.2009.07.003

Findley, M.G., Kikuta, K., & Denly, M. (2021). External Validity. Annual Review of Political
Science, 24(1), 365–393. https://doi.org/10.1146/annurev-polisci-041719-102556

Fiol, C. M., & Lyles, M. A. (1985). Organizational learning. Academy of management

review, 10(4), 803-813. https://doi.org/10.2307/258048

Fisher, M. J., & Marshall, A. P. (2009). Understanding descriptive statistics. Australian

Critical Care, 22(2), 93–97. https://doi.org/10.1016/j.aucc.2008.11.003

Fisher, J. G. (1998). Contingency theory, management control systems and firm outcomes:
past results and future directions. Behavioral research in accounting, 10, 47.

FHM (2022, 29 Mars). Covid-19 ska fortsätta smittspåras i vård och omsorg.
Folkhälsomyndigheten. https://www.folkhalsomyndigheten.se/nyheter-och-
press/nyhetsarkiv/2022/mars/covid-19-ska-fortsatt-smittsparas-i-vard-och-omsorg/

Gamage, S. D. D., & Gooneratne, T. (2017). Management controls in an apparel group: an

institutional theory perspective. Journal of Applied Accounting Research, 18(2), 223–
241. https://doi.org/10.1108/JAAR-09-2015-0075

96
Gamst, G., Meyers, L. S., & Guarino, A. J. (2008). Analysis of variance designs: A
conceptual and computational approach with SPSS and SAS. Cambridge University
Press.

Geisler, C. (2018). Coding for language complexity: The interplay among methodological
commitments, tools, and workflow in writing research. Written Communication, 35(2),
215–249. https://doi.org/10.1177/0741088317748590

Gomez-Conde, J., Lopez-Valeiras, E., Rosa, F. S., & Lunkes, R. J. (2022). The effect of
management control systems in managing the unknown: Does the market appreciate
the breadth of vision? Review of Managerial Science, 1-27.
https://doi.org/10.1007/s11846-022-00601-0

Govindarajan, V. (1988). A contingency approach to strategy implementation at the business-

unit level: integrating administrative mechanisms with strategy. Academy of
management Journal, 31(4), 828-853. https://doi.org/10.2307/256341

Gupta, A. S., & Mukherjee, J. (2022). Long-term changes in consumers' shopping behavior
post-pandemic: an exploratory study. International Journal of Retail & Distribution
Management, (ahead-of-print). https://doi.org/10.1108/IJRDM-04-2022-0111

Hadi, A. S. & Chatterjee, S. (2012). Regression Analysis by Example. (5th ed.). John Wiley
& Sons, Incorporated.

Hambrick, D. C. (1983). Some tests of the effectiveness and functional attributes of Miles and
Snow's strategic types. Academy of Management journal, 26(1), 5-26.
https://doi.org/10.2307/256132

Harel, O. (2009). The estimation of R 2 and adjusted R 2 in incomplete data sets using
multiple imputation. Journal of Applied Statistics, 36(10), 1109-1118.
https://doi.org/10.1080/02664760802553000

Heale, R. & Twycross, A. (2015). Validity and reliability in quantitative studies. Evidence-
Based Nursing, 18(3), 66–67. https://doi.org/10.1136/eb-2015-102129

97
Heinicke, A., Guenther, T. W., & Widener, S. K. (2016). An examination of the relationship
between the extent of a flexible culture and the levers of control system: The key role
of beliefs control. Management Accounting Research, 33, 25-41.
https://doi.org/10.1016/j.mar.2016.03.005

Henri, J. F. (2006). Management control systems and strategy: A resource-based

perspective. Accounting, Organizations and Society, 31(6), 529–558.
https://doi.org/10.1016/j.aos.2005.07.001

Hofstede, G. (1983). National cultures in four dimensions: A research-based theory of cultural

differences among nations. International studies of management & organization, 13(1-
2), 46–74. https://doi.org/10.1080/00208825.1983.11656358

Hopwood, A. G. (1972). An empirical study of the role of accounting data in performance

evaluation. Journal of accounting research, 156-182.

IBM. (2023, 3 January). Significance level. https://www.ibm.com/docs/en/cognos-

analytics/11.1.0?topic=terms-significance-level

IBM. (2023, 13 March). Adjusted r square. https://www.ibm.com/docs/en/cognos-

analytics/11.1.0?topic=terms-adjusted-r-squared

Jamali, D. (2010). MNCs and International Accountability Standards Through an Institutional

Lens: Evidence of Symbolic Conformity or Decoupling. Journal of Business Ethics,
95(4), 617–640. https://doi.org/10.1007/s10551-010-0443-z

Janke, R., Mahlendorf, M. D., & Weber, J. (2014). An exploratory study of the reciprocal
relationship between interactive use of management control systems and perception of
negative external crisis effects. Management Accounting Research, 25(4), 251–270.
https://doi.org/10.1016/j.mar.2014.01.001

Jesson, J., Matheson, L., & Lacey, F. M. (2011). Doing your literature review: Traditional and
systematic techniques.

98
Jokipii, A. (2010). Determinants and consequences of internal control in firms: a contingency
theory based analysis. Journal of Management & Governance, 14, 115-144.
https://doi.org/10.1007/s10997-009-9085-x

Joshi, A., Kale, S., Chandel, S., & Pal, D. K. (2015). Likert scale: Explored and
explained. British journal of applied science & technology, 7(4), 396.

Kenny, D. A. (2015). Measuring model fit. https://davidakenny.net/cm/fit.htm

Knofczynski, G. T., & Mundfrom, D. (2008). Sample sizes when using multiple linear
regression for prediction. Educational and psychological measurement, 68(3), 431–
442. https://doi.org/10.1177/0013164407310131

Kober, R., & Thambar, P. J. (2022). Paradoxical tensions of the COVID-19 pandemic: a
paradox theory perspective on the role of management control systems in helping
organizations survive crises. Accounting, Auditing & Accountability Journal, 35(1),
108-119. https://doi.org/10.1108/AAAJ-08-2020-4851

Kruis, A. M., Speklé, R. F., & Widener, S. K. (2016). The Levers of Control Framework: An
exploratory analysis of balance. Management Accounting Research, 32, 27-44.
https://doi.org/10.1016/j.mar.2015.12.002

Laguir, I., Choi, T. M., Stekelorum, R., Gupta, S., & Kumar, A. (2022). Roles of Mobilized
Controls and Environmental Uncertainty on Supply Chain Resilience: An Empirical
Study from Dynamic-Capabilities-View and Levers-of-Control Perspectives. IEEE
Transactions on Engineering Management.
https://doi.org/10.1109/TEM.2022.3171606

Laguir, I., Gupta, S., Bose, I., Stekelorum, R., & Laguir, L. (2022). Analytics capabilities and
organizational competitiveness: Unveiling the impact of management control systems
and environmental uncertainty. Decision Support Systems, 156, 113744.
https://doi.org/10.1016/j.dss.2022.113744

99
Langfield-Smith, K. (1997). Management control systems and strategy: A critical
review. Accounting, Organizations and Society, 22(2), 207–232.
https://doi.org/10.1016/S0361-3682(95)00040-2

Li, G., Ooi, B. C., Feng, J., Wang, J., & Zhou, L. (2008, June). EASE: an effective 3-in-1
keyword search method for unstructured, semi-structured and structured data.
In Proceedings of the 2008 ACM SIGMOD international conference on Management
of data (pp. 903-914). https://doi.org/10.1016/j.is.2008.08.001

Lipshitz, R., Popper, M., & Friedman, V. J. (2002). A Multifacet Model of Organizational
Learning. Journal of Applied Behavioral Science., 38(1), 78–98.
https://doi.org/10.1177/0021886302381005

Makrygiannakis, G., & Jack, L. (2016). Understanding management accounting change using
strong structuration frameworks. Accounting, Auditing & Accountability Journal.
https://doi.org/10.1108/AAAJ-08-2015-2201

Malmi, T., & Brown, D. A. (2008). Management control systems as a package—

Opportunities, challenges, and research directions. Management Accounting Research,
19(4), 287–300. https://doi.org/10.1016/j.mar.2008.09.003

Martyn, P., Sweeney, B., & Curtis, E. (2016). Strategy and control: 25 years of empirical use
of Simons’ Levers of Control framework. Journal of Accounting & Organizational
Change. https://doi.org/10.1108/JAOC-03-2015-0027

McColl-Kennedy, J. R., & Anderson, R. D. (2005). Subordinate–manager gender combination

and perceived leadership style influence on emotions, self-esteem and organizational
commitment. Journal of Business Research, 58(2), 115-125.
https://doi.org/10.1016/S0148-2963(03)00112-7

Merchant, K. A., & Otley, D. T. (2006). A review of the litera- ture on control and
accountability. In C. Chapman, A. Hop- wood, & M. Shield (Eds.), The handbook of
management accounting research. Elsevier Press. https://doi.org/10.1016/S1751-
3243(06)02013-X

100
Merchant, K. A., & Van der Stede, W. A. (2007). Management control systems: performance
measurement, evaluation, and incentives. Pearson education.

Meyer, J. W., & Rowan, B. (1977). Institutionalized organizations: Formal structure as myth
and ceremony. American journal of sociology, 83(2), 340-363.
https://doi.org/10.1086/226550

Miles, R. E., & Snow, C. C. (1978) Organizational strategy, structure and process. New York:
McGraw-Hill. https://doi.org/10.2307/257544

Milgrom, P., & Roberts, J. (1995). Complementarities and fit strategy, structure, and
organizational change in manufacturing. Journal of accounting and economics, 19(2–
3), 179-208. https://doi.org/10.1016/0165-4101(94)00382-F

Mizruchi, M. S., & Fein, L. C. (1999). The social construction of organizational knowledge:
A study of the uses of coercive, mimetic, and normative isomorphism. Administrative
science quarterly, 44(4), 653-683. https://doi.org/10.2307/2667051

Morphew, C. C., & Hartley, M. (2006). Mission statements: A thematic analysis of rhetoric
across institutional type. The Journal of Higher Education, 77(3), 456-471.
https://doi.org/10.1353/jhe.2006.0025

Mundy, J. (2010). Creating dynamic tensions through a balanced use of management control
systems. Accounting, Organizations, and society, 35(5), 499-523.
https://doi.org/10.1016/j.aos.2009.10.005

Müller, R. (1985). Corporate crisis management. Long Range Planning, 18(5), 38-48.
https://doi.org/10.1016/0024-6301(85)90199-2

Müller-Stewens, B., Widener, S. K., Möller, K., & Steinmann, J. C. (2020). The role of
diagnostic and interactive control uses in innovation. Accounting, Organizations and
Society, 80, 101078. https://doi.org/10.1016/j.aos.2019.101078

101
Nonaka, I. (1994). A dynamic theory of organizational knowledge creation. Organization
science, 5(1), 14-37. https://doi.org/10.1287/orsc.5.1.14

Orme, J. G., & Combs-Orme, T. (2009). Multiple regression with discrete dependent
variables. Oxford University Press.

Otley. D. (1978). Budget Use and Managerial Performance. Journal of Accounting

Research, 16(1), 122–149. https://doi.org/10.2307/

Otley. D. (1980). The contingency theory of management accounting: Achievement and

prognosis. Accounting, Organizations and Society, 5(4), 413–428.
https://doi.org/10.1016/0361-3682(80)90040-9

Otley, D. (1999). Performance management: a framework for management control systems

research. Management accounting research, 10(4), 363-382.
https://doi.org/10.1006/mare.1999.0115

Otley, D. (2016). The contingency theory of management accounting and control: 1980–
2014. Management accounting research, 31, 45-62.
https://doi.org/10.1016/j.mar.2016.02.001

Otley, D., & Soin, K. (2014). Management Control and Uncertainty (Otley & K. Soin, Eds.;
1st ed. 2014.). Palgrave Macmillan UK. https://doi.org/10.1057/

Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., ...
& Moher, D. (2021). The PRISMA 2020 statement: an updated guideline for reporting
systematic reviews. International journal of surgery, 88, 105906.
https://doi.org/10.1016/j.ijsu.2021.105906

Pallant, Julie. 2020. SPSS Survival Manual: A Step by Step Guide to Data Analysis Using
IBM SPSS. 7th edition. Maidenhead: Open University Press.

Passetti, E. E., Battaglia, M., Bianchi, L., & Annesi, N. (2021). Coping with the COVID-19
pandemic: the technical, moral, and facilitating role of management

102
 control. Accounting, Auditing & Accountability Journal, 34(6), 1430-1444.
https://doi.org/10.1108/AAAJ-08-2020-4839

Pešalj, B., Pavlov, A., & Micheli, P. (2018). The use of management control and performance
measurement systems in SMEs: A levers of control perspective. International Journal
of Operations & Production Management, 38(11), 2169–2191.
https://doi.org/10.1108/IJOPM-09-2016-0565

Regeringskansliet (2023, 24 Mars). Rysslands invasion av Ukraina. Regeringen.

https://www.regeringen.se/regeringens-politik/rysslands-invasion-av-ukraina/

Reja, U., Manfreda, K. L., Hlebec, V., & Vehovar, V. (2003). Open-ended vs. close-ended
questions in web questionnaires. Developments in applied statistics, 19(1), 159–177.

Rezania, D., Baker, R., & Burga, R. (2016). Project control: an exploratory study of levers of
control in the context of managing projects. Journal of Accounting & Organizational
Change, 12(4), 614–635. https://doi.org/10.1108/JAOC-10-2015-0084

Roberts, P., & Priest, H. (2006). Reliability and validity in research. Nursing Standard,
20(44), 41–45. https://doi.org/10.7748/ns2006.07.20.44.41.c6560

Sahay, A. (2016). Applied regression and modeling: a computer integrated approach (First
edition.). Business Expert Press.

Sandelin, M. (2008). Operation of management control practices as a package—A case study

on control system variety in a growth firm context. Management Accounting
Research, 19(4), 324–343. https://doi.org/10.1016/j.mar.2008.08.002

Saunders, M., Lewis, P., & Thornhill, A. (2019). Research methods for business students (8.
ed.). Pearson Education.

SCB. (2018). Jämn könsfördelning bland chefer totalt sett, men ojämn könsfördelning bland
chefsyrken. Statistiska Centralbyrån. https://www.scb.se/hitta-statistik/statistik-efter-

103
 amne/arbetsmarknad/sysselsattning-forvarvsarbete-och-arbetstider/yrkesregistret-med-
yrkesstatistik/pong/statistiknyhet/yrkesregistret-med-yrkesstatistik-2016/

Schermelleh-Engel, K., Moosbrugger, H., & Müller, H. (2003). Evaluating the fit of structural
equation models: Tests of significance and descriptive goodness-of-fit
measures. Methods of psychological research online, 8(2), 23-74.

Schram, A. (2005). Artificiality: The tension between internal and external validity in
economic experiments. The Journal of Economic Methodology, 12(2), 225–237.
https://doi.org/10.1080/13501780500086081

Sharma, S., Durand, R. M., & Gur-Arie, O. (1981). Identification and Analysis of Moderator
Variables. Journal of Marketing Research, 18(3), 291–300.
https://doi.org/10.1177/002224378101800303

Simons, R. (1994). How New Top Managers Use Control Systems as Levers of Strategic
Renewal. Strategic Management Journal, 15(3), 169–189.
https://doi.org/10.1002/smj.4250150301

Simons, R. (1995). Levers of control: How managers use innovative control systems to drive
strategic renewal. Harvard Business Press.

Simons, R. (2014). Performance measurement and control systems for implementing strategy
(1. ed., Pearson new international edition). Pearson.

Singh, J., Carasco, E., Svensson, G., Wood, G., & Callaghan, M. (2005). A comparative study
of the contents of corporate codes of ethics in Australia, Canada and Sweden. Journal
of World Business, 40(1), 91-109. https://doi.org/10.1016/j.jwb.2004.10.007

Sint, R., Schaffert, S., Stroka, S., & Ferstl, R. (2009, June). Combining unstructured, fully
structured, and semi-structured information in semantic wikis. In CEUR Workshop
Proceedings (Vol. 464, No. 12, pp. 73-87). Crete, Greece: Heraklion.

104
Spector, P. E., & Brannick, M. T. (2011). Methodological urban legends: The misuse of
statistical control variables. Organizational research methods, 14(2), 287–305.
https://doi.org/10.1177/1094428110369842

Spieske, A., Gebhardt, M., Kopyto, M., & Birkel, H. (2022). Improving resilience of the
healthcare supply chain in a pandemic: Evidence from Europe during the COVID-19
crisis. Journal of Purchasing and Supply Management, 28(5), 100748.
https://doi.org/10.1016/j.pursup.2022.100748

Sveriges riksdag. (2022). Processing of personal data.

https://www.riksdagen.se/en/global/about-the-website/processing-of-personal-data/

Suddaby, R. (2010). Challenges for institutional theory. Journal of management

inquiry, 19(1), 14–20. https://doi.org/10.1177/1056492609347564

Tabachnick B. G., & Fidell, L. S. (2013). Using multivariate statistics (6th ed.) Boston:
Pearson Education

Taber, K. S. (2018). The Use of Cronbach’s Alpha When Developing and Reporting Research
Instruments in Science Education. Research in Science Education (Australasian
Science Education Research Association), 48(6), 1273–1296.
https://doi.org/10.1007/s11165-016-9602-2

Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach’s alpha. International Journal
of Medical Education, 2, 53–55. https://doi.org/10.5116/ijme.4dfb.8dfd

Tessier, S., & Otley, D. (2012). A conceptual development of Simons’ Levers of Control
framework. Management accounting research, 23(3), 171-185.
https://doi.org/10.1016/j.mar.2012.04.003

Tucker, B. P., & Parker, L. D. (2015). Business as usual? An institutional view of the
relationship between management control systems and strategy. Financial
Accountability & Management, 31(2), 113–149. https://doi.org/10.1111/faam.12050

105
Tuomela, T-S. (2005). The interplay of different levers of control: A case study of introducing
a new performance measurement system. Management accounting research, 16(3),
293-320. https://doi.org/10.1016/j.mar.2005.06.003

Uyanık, G. K., & Güler, N. (2013). A study on multiple linear regression analysis. Procedia-
Social and Behavioral Sciences, 106, 234–240.

Vogt, W. P., Gardner, D. C., Haeffele, L. M., & Vogt, E. R. (2014). Selecting the right
analyses for your data: Quantitative, qualitative, and mixed methods. Guilford
Publications.

Watson, R. (2015). Quantitative research. Nursing standard, 29(31).

https://doi.org/10.7748/ns.29.31.44.e8681

Wetcher-Hendricks, D. (2011). Analyzing quantitative data: An introduction for social

researchers. John Wiley & Sons.

Weingart, L., Smith, P., & Olekalns, M. (2004). Quantitative coding of negotiation
behavior. International negotiation, 9(3), 441–456.
https://doi.org/10.1163/1571806053498805

Widener, S. K. (2007). An empirical analysis of the levers of control framework. Accounting,

organizations, and society, 32(7–8), 757–788.
https://doi.org/10.1016/j.aos.2007.01.001

106
8. Appendix
8.1 Appendix 1 – Literature Review Table
Reference Publication Rating on
ABS-list
Anthony (1965) Harvard University Press Missing
Argote & Miron-Spector (2011) Organization science 4*
Baird et al. (2019) Pacific Accounting Review 1
Barros & Ferreira (2022) Journal of Accounting & Organizational Change 2
Barros & Ferreira (2023) Journal of Management Control 2
Bedford (2015) Management Accounting Research 3
Bedford & Malmi (2015) Management Accounting Research 3
Bisbe & Otley (2004) Accounting, organizations, and society 4*
Boxenbaum & Jonsson (2017) The Sage handbook of organizational institutionalism Missing
Bruton et al. (2010) Entrepreneurship theory and practice 4
Cardinal et al. (2004) Organisation science 4*
Chenhall (2003) Accounting, organizations, and society 4*
Chenhall & Morris (1995) Omega the international journal of management science 3
Child (1997) Organization studies 4
Daft & Weick (1984) Academy of management review 4*
Davila (2000) Accounting, Organizations and Society 4*
Damavanthi & Gooneratne (2017) Journal of Applied Accounting Research 2
Delfino & Van Der Kolk (2021) Accounting, Auditing & Accountability Journal 3
DiMaggio & Powell (1991) University of Chicago Press Missing
Dossi & Patelli (2010) Long Range Planning 3
Edmondson (1996) The Journal of Applied Behavioral Science 2
Evans & Tucker (2015) Qualitative Research in Accounting & Management 2
Felicio et al. (2021) Journal of Business Research 3
Ferreira & Otley (2009) Management Accounting Research 3
Fiol & Lyles (1985) Academy of management review 4*
Fisher (1998) Behavioral research in accounting 3
Govindarajan (1988) Academy of management Journal 4*
Heinicke et al. (2016) Management Accounting Research 3
Henri (2006) Accounting, Organizations and Society 4*
Hopwood (1972) Journal of accounting research 4
Jamali (2010) Journal of Business Ethics 3
Kruis et al. (2016) Management Accounting Research 3
Laguir et al. (2022) IEEE Transactions on Engineering Management. Missing

107
 Langfield-Smith (2007) Accounting, Organizations and Society 4*
Lipshitz et al. (2002) Journal of Applied Behavioral Science Missing
Malmi & Brown (2008) Management Accounting Research 3
Martyn et al. (2016) Journal of Accounting & Organizational Change 2
Merchant & Otley (2006) Elsevier Press Missing
Merchant & Van der Stede (2007) Pearson education Missing
Meyer & Rowan (1977) American journal of sociology 4*
Miles & Snow (1978) Organizational strategy, structure, and process Missing
Milgrom & Roberts (1995) Journal of accounting and economics Missing
Mizruchi & Fein (1999) Administrative science quarterly 4*
Mundy (2010) Accounting, Organizations, and society 4*
Nonaka (1994) Organization science Missing
Otley (1978) Journal of Accounting Research 4*
Otley (1980) Accounting, Organizations and Society 4*
Otley (1999) Management accounting research 3
Otley (2016) Management accounting research 3
Otley & Soin (2014) Palgrave Macmillan UK Missing
Passetti et al. (2021) Accounting, Auditing & Accountability Journal 3
Rezania et al. (2016) Journal of Accounting & Organizational Change 2
Sandelin (2008) Management Accounting Research 3
Simons (1994) Strategic Management Journal 4*
Simons (1995) Harvard Business Press Missing
Simons (2014) Pearson Education Missing
Suddaby (2010) Journal of management inquiry 3
Tessier & Otley (2012) Management accounting research 3
Tucker & Parker (2015) Financial Accountability & Management 3
Tuomela (2005) Management accounting research 3
Widener (2007) Accounting, organizations, and society 4*
*The highest possible rating

108
8.2 Appendix 2 – Survey Questions English

Control Questions
I am...
(1) Female
(2) Male
(3) Other

What is your age?

(1) _________

What is the highest degree or level of education you have completed?

(1) Elementary school
(2) High school
(3) University
(4) Postgraduate
(5) Other post-secondary education

What industry does your organization mainly operate in?

(1) Energy and power supply e.g., oil, gas, and coal
(2) Materials e.g., chemistry, wood, paper, and metals
(3) Industry e.g., machines, trucks, vehicles, and parts
(4) Retail and rare items e.g., cars, appliances, and clothes
(5) Daily goods e.g., food, drink, and detergent
(6) Healthcare e.g., medical and biotechnology
(7) Finance and real estate e.g., banks and insurance companies
(8) Information technology and telecom operators e.g., computers, software, and telephone
(9) Construction, design, and interior
(10) Other_________

What is your position at the organization?

(1) Top manager
(2) Division manager
(3) Manager

109
(4) Other_________

For how many years have you held your current position?
(1) _________

How many employees work in the organization?

(1) 1 – 10
(2) 11 – 49
(3) 50 – 249
(4) 250 – 500
(5) More than 500

Perceived environmental uncertainty.

Over the past year, how predictable have important actions or changes in the external
environment been? (1= Highly predictable; 7=Highly Unpredictable)

1.1 Customer (e.g., level of demand, customer requirements)

1.2 Suppliers (e.g., markets for key inputs, quality of resources)

1.3 Competitors (e.g., competitors entering or leaving, tactics/strategies)

1.4 Technological (e.g., R&D advances, process innovation)

1.5 Economic (e.g., costs, interest rates)

1.6 Regulation (e.g., new laws or rules)

110
Belief Systems
Please indicate the extent to which the following items characterize your organization:
(1= Not very characteristic; 7= Very characteristic)

2.1 Our mission statement clearly communicates the firm’s core values to our workforce.

2.2 Top managers communicate core values to our workforce.

2.3 Our workforce is aware of the firm’s core values.

2.4 Our mission statement inspires our workforce.

Boundary Systems
Please rate the extent to which you agree or disagree with the following:
(1= Strongly disagree; 7= Strongly Agree)

3.1 Our organization relies on a code of business conduct to define appropriate behavior for
our workforce.

3.2 Our code of business conduct informs our workforce about behaviors that are off-limits.

3.3 Our organization has a system that communicates to our workforce risks that should be
avoided.

3.4 Our workforce is aware of the firm’s code of business conduct.

111
Diagnostic Control Systems
Please rate the extent to which you agree or disagree with the following. In our organization
we… (1= Strongly disagree; 7= Strongly Agree)

4.1 Track progress towards goals

4.2 Monitor results

4.3 Compare outcomes to expectations

4.4 Review key measures

Interactive Control Systems

Please rate the extent to which you agree or disagree with the following. In our organization
we… (1= Strongly disagree; 7= Strongly Agree)

5.1 Enable discussion in meetings of superiors, subordinates, and peers.

5.2 Enable continual challenge and debate underlying data, assumptions, and action plans.

5.3 Provide a common view

5.4 Enable focus on common issues

5.5 Enable focus on critical success factors

112
Organizational Performance
How would you compare the organizations performance over the past year to that of the
organizations that do the same kind of work?
(1= Much worse than the competitors; 7=Much better than the competitors)

6.1 Growth in sales/revenues

6.1 Return on sales/revenues

6.3 Quality of products/services

6.4 Development of new products/ ´services

6.5 Ability to attract and retain talented employees

6.6 Relations between management and employees

6.7 Relations among employees

113
8.3 Appendix 3 – Survey Questions Swedish

Kontrollfrågor
Jag är en...
(1) Kvinna
(2) Man
(3) Annat

Hur gammal är du?

(1) _________

Vilken är den högsta utbildningsnivå som du avklarat?

(1) Grundskola
(2) Gymnasieutbildning
(3) Högskole-/universitetsutbildning
(4) Forskarutbildning
(5) Annan eftergymnasial utbildning

Vilken bransch verkar organisationen inom?

(1) Energi och kraftförsörjning tex. el, olja, gas och kol
(2) Material tex. kemi, trä, papper och metaller
(3) Industri och tillverkning tex. maskiner, lastbilar, fordon och delar
(4) Detaljhandel och sällanköpsvaror tex. kläder, inredning, bilar och vitvaror
(5) Dagligvaror tex. mat, dryck och tvättmedel
(6) Hälsovård tex. sjukvård och bioteknik
(7) Finans och fastigheter tex. banker, försäkringsbolag och fastigheter
(8) Informationsteknik och teleoperatörer tex. datorer, mjukvara och telefon
(9) Bygg, konstruktion och design
(10) Annat _________

Vilken är din roll i organisationen?

(1) Högsta chef
(2) Avdelningschef
(3) Annan typ av chef

114
(4) Annat _________

Hur många år har du haft din nuvarande roll?

_________

Hur många anställda arbetar i organisationen?

(1) 1 – 10
(2) 11 – 49
(3) 50 – 249
(4) 250 – 500
(5) Mer än 500

Upplevd osäkerhet i organisationens omgivning

Hur förutsägbara har viktiga händelser eller förändringar i organisationens omgivning varit
under det senaste året? (1= mycket förutsägbart; 7=väldigt oförutsägbart)

1.1 Kunder (tex. efterfrågan, kundkrav)

1.2 Leverantörer (tex. marknader för nyckeltillgångar, kvalitet på resurser)

1.3 Konkurrenter (tex. konkurrenter som dyker upp eller försvinner, taktik/strategier)

1.4 Teknik (tex. forsknings- och utvecklingsframsteg, teknisk och processinnovation)

1.5 Ekonomi (tex. kostnader, räntor)

1.6 Regleringar (tex. nya lagar eller regler att förhålla sig till)

115
Styrsystem för att engagera anställda i organisationens vision och kärnvärden
Ange i vilken utsträckning följande påståenden beskriver din organisation:
(1= Håller inte alls med; 7= Håller med fullt ut)

2.1 Vår verksamhetsplan/strategi förmedlar tydligt organisationens kärnvärden till vår

personal

2.2 Toppchefer kommunicerar kärnvärden till vår personal.

2.3 Vår personal är medveten om organisationens kärnvärden

2.4 Vårt verksamhetsplan/strategi inspirerar vår personal.

Återhållande system
Vänligen betygsätt i vilken utsträckning du håller med/inte håller med om följande:
(1= Håller inte med alls; 7= håller med fullt ut)

3.1 Vår organisation förlitar sig på en uppförandekod för att förmedla lämpligt beteende för
vår personal.

3.2 Vår affärsuppförandekod informerar personalen om beteenden som är förbjudna.

3.3 Vår organisation har ett system som kommunicerar ut vilka risker som bör undvikas till
personalen.

3.4 Vår personal är medveten om organisationens uppförandekod.

116
Styrsystem för att övervaka att organisationens mål uppnås
Vänligen betygsätt i vilken utsträckning du håller med/inte håller med om följande. vår
organisation... (1= Håller inte med alls; 7= Håller med fullt ut)

4.1 Följer upp framsteg mot satta mål

4.2 Kontrollerar/mäter resultat

4.3 Jämför resultat med förväntningar

4.4 Utvärderar nyckeltal

Interaktiva styrsystem
Vänligen betygsätt i vilken utsträckning du håller med/inte håller med om följande. Vår
organisation... (1= Håller inte med alls; 7= Håller med fullt ut)

5.1 Möjliggör diskussion på möten med överordnade, underordnade och jämlika.

5.2 Möjliggör kontinuerlig utmaning och diskuterar underliggande data, antaganden och
handlingsplaner.

5.3 Förmedlar ett gemensamt synsätt

5.4 Möjliggöra fokus på vanliga frågor

5.5 Möjliggöra fokus på kritiska framgångsfaktorer

117
Organisationens prestation
Hur skulle du beskriva organisationens prestation under det senaste året jämfört med de
organisationer som utför samma typ av arbete när det kommer till…
(1= Mycket sämre än våra konkurrenter; 7=Mycket bättre än våra konkurrenter)

6.1 Tillväxt i försäljning/intäkter

6.1 Tillväxt i resultat

6.3 Kvalitet på produkter/tjänster

6.4 Utveckling av nya produkter/tjänster

6.5 Förmåga att attrahera och behålla duktiga medarbetare

6.6 Relationer mellan ledning och medarbetare

6.7 Relationer mellan medarbetare

118
8.4 Appendix 4 – Coding

I am...
0 = Female, 1 = Male

What is the highest degree or level of education you have completed?

1 = High School, 2 = University, 3 = Postgraduate

What is your position at the organization?

0 = Other, 1 = Another manager, 2 = Divisional manager, 3 = Top manager

How many employees work in the organization?

1 = 1 – 10, 2 = 11 – 49, 3 = 50 – 249, 4 = 250 – 500, 5 = More than 500

119
8.5 Appendix 5 – Summary Variables

120
121