The Effect of Cultural Differences on the Auditors’
Profes-sional Judgement in Disclosing Key Audit Matters
Thesis MSc Accountancy & Controlling
Max Laurens Hofstra S3191281 06-11702455 M.L.Hofstra.1@student.rug.nl De Populierenlaan 1 – 273 9741 HA Groningen 18th of January 2021 Prof. Dr. D. A. de Waard
Abstract: The implementation of ISA 701 required auditors to expand the traditional pass/fail auditor’s report with Key Audit Matters (KAMs) in order to reduce the information gap and create more uni-formity. KAMs are those matters that were of most significance in the audited financial statements based on the auditor’s professional judgement. This study responds to the call from the research field to explore the cultural influence on KAM disclosure. An empirical analysis was conducted on a sample of 914 auditor’s reports fiscal-year end 2018 from 27 different countries all over the world using cultural di-mensions of Hofstede and Gray. The empirical results indicate that there appears to be an effect of culture on the level of KAM disclosure by an auditor. More specifically, power distance, uncertainty avoidance, individualism and secrecy all showed significant relationships with KAM disclosure. Firstly, auditors from a power distant country disclose more KAMs which was in the opposite direction of hy-pothesis 1. Secondly, auditors from an uncertainty avoidant and secretive culture disclose a lower level of KAMs, while auditors from individualistic countries disclose a higher level of KAMs. In essence, culture has an impact on the level of KAM disclosure based on the auditor’s professional judgement. This conclusion is important for regulators and standard setters to enhance rule making and improve the auditor’s report.
Keywords: Auditor’s report, Key Audit Matters, Legitimacy Theory, Hogarth’s Theory, Cross-National differences, National Cultures, Hofstede’s Cultural Dimensions, Gray’s accounting values
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 2
Table of contents
1. Introduction _____________________________________________________________ 3 2. Theoretical Background ____________________________________________________ 7 2.1 Legitimacy Theory _____________________________________________________ 7 2.2 Hogarth’s Theory ______________________________________________________ 8 2.3 National Cultures Framework of Hofstede __________________________________ 9 2.3.1 Power Distance ___________________________________________________ 10 2.3.2 Uncertainty avoidance ______________________________________________ 11 2.3.3 Individualism _____________________________________________________ 12 2.3.4 Secrecy _________________________________________________________ 13 3. Research design _________________________________________________________ 15 3.1 Sample and data collection ______________________________________________ 15 3.2 Dependent variable: Level of KAM disclosure ______________________________ 15 3.3 Independent variables __________________________________________________ 16 3.4 Control variables _____________________________________________________ 17 3.5 Statistical model ______________________________________________________ 18 4. Results ________________________________________________________________ 19 4.1 Descriptive Analysis___________________________________________________ 19 4.2 Correlation Analysis ___________________________________________________ 20 4.3 Regression analysis ___________________________________________________ 23 4.4 Robustness tests ______________________________________________________ 24 5. Conclusion and Discussion ________________________________________________ 26 5.1 Discussion and Conclusion _____________________________________________ 26 5.2 Limitations and future research __________________________________________ 28 5.3 Implications _________________________________________________________ 29 6. References _____________________________________________________________ 31 7. Appendices _____________________________________________________________ 35The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 3
1. Introduction
Since 1890 there exists a debate amongst regulators, standard setters, financial statement users and accountants about the usefulness of the auditor’s report (Chandler & Edwards, 1996). After the financial crisis of 2008 this criticism on the informative value of the auditor’s report was growing even more (Church, Davis & McCracken, 2008). The traditional auditor’s report was seen as insufficiently informative, not transparent and thus not useful, which started people questioning the informative and communicative value of these reports (Asare and Wright, 2012; IAASB, 2015). Similarly, it has also drawn the attention of researchers who emphasized the need for a change in auditor’s reports. Prior research found that users had shown little interest in actually reading auditor’s reports since the format is highly standardized. Auditor’s reports are seen as uninformative since nearly every public company receives an unqualified opinion that is as standard as any other opinion given by audit firms (Gray, Turner, Coram & Mock, 2011; Church et al., 2008). Mock, Bédard, Coram, Davis, Espahbodi & Warne (2013) found that stakeholders wanted more information about the auditor, the audit process and the financial statements. Where Vanstraelen, Schelleman, Meuwissen & Hoffmann (2012) found that users were interested in additional disclosure, like the specific areas of risks found during the audit. This has resulted in a certain gap between the information that users want from the audit and the financial statements and the information that was made available through the auditor’s report and the audited financial statements (Gold & Heilmann, 2019). Standard setters are call-ing this phenomenon the “information gap” which is closely related to the long-standcall-ing ex-pectation gap between the exex-pectations of an audit by the user and what an actual audit contains (IAASB, 2011; IAASB, 2012). Multiple initiatives by the International Auditing and Assurance Standards Board (IAASB), the UK Financial Reporting Council (FRC) and the Public Com-pany Accounting Oversight Board (PCAOB) have been the result of all the debates and papers that tried to improve the communicative value in the auditor’s report. The result was the adop-tion by the European Union of Regulaadop-tion (EU) No 537/2014 of the European Parliament and the Council in April 2014. This new rule affected all the accounting periods that started after the 17th of June 2016. The new Regulation emphasizes the importance of several measures like the auditors’ independence, greater transparency within the auditor’s report and requirements of the audit committee. The most important change was the expansion of Key Audit Matters (hereafter: KAMs) within the auditor’s report (Pinto & Morais, 2019). Alongside new EU Reg-ulation, the International Standard on Auditing (ISA) 701 was introduced which affected al financial statements ending on or after the 15th of December 2016 (IFAC, 2017). ISA 701.8
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 4 states that KAMs are those matters that, in the professional judgement of the auditor, were of most significance in the financial statements audit of the current book year. These new regula-tions had the objective to transform the auditor’s report from the traditional pass/fail model into an auditor’s report that was more informative and uniform in order to reduce the information gap (Gold & Heilmann, 2019). But instead of reducing the criticism concerning the auditor’s report, it raised new questions about the informativeness of these KAMs. (Lennox, Schmidt & Thompson, 2019).
Sirois, Bédard & Bera (2017) argued that the new information puts focus on critical matters but might lessen the focus on other parts of the financial statements. While others state that the KAMs would only increase the value of an auditor’s report if it includes actual relevant information (Vanstraelen et al., 2012). This indicates that the debate on the informative value of the auditor’s report lasts for more than one century now. The value relevance has been criti-cized in a published article by Accountant.nl (2019). While Most of the recent literature found that the informational relevance of the auditor’s report has improved since the implementation of KAMs (Edilson Divino Alves Júnior & Galdi, 2020; Kohler, Ratzinger-Sakel & Theis, 2020). Sirois et al., (2018) for example, found that KAMs have a direct impact on drawing the attention of professional investors. Other studies found that KAMs enhanced the quality of fi-nancial reports (Gold, Heilmann, Pott & Rematzki, 2020). This shows that KAMs indeed im-prove the new auditor’s report and the more important it is to get a better understanding on what influences the auditors’ professional judgement in disclosing KAMs. As we know from ISA 701.8, the level of KAM disclosure is determined based on the auditors’ professional judge-ment. Consequently, Patel, Harrison & McKinnon (2002) argue that culture has an influence on the professional judgement of auditors. Therefore, it is interesting to investigate if culture also influences the level of KAMs disclosed in the auditor’s report. This study aims to try to address new determinants, rather than examining the informative value of KAMs. This leads us to the following research question:
What is the effect of cultural differences on the auditors’ professional judgement in disclosing Key Audit Matters?
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 5 This research tries to respond to the demand from prior literature to investigate the effect of cultural differences on the auditors’ professional judgement in disclosing KAMs (Pinto & Morais, 2019). Some unexplainable differences were found in literature between countries and the level of KAM disclosure suggesting that it could be a function of institutional or cultural factors (Nolder & Riley, 2014). In the past, many have studied the informative value of the enhanced auditor’s report to the users of the financial statements (Edilson Divino Alves Júnior & Galdi, 2020; Kohler et al., 2020; Moroney, Phang & Xiao, 2020; Sirois et al., 2017). But until now, research on the factors that influence the judgement of auditors in disclosing KAMs, rather than the value of the auditor’s report, hasn’t been popular. Therefore, this study tries to extent the existing literature by providing more evidence on factors that influence the auditors’ judgement to disclose KAMs,
Suttipun (2019) tried to find which factors influenced KAM reporting at companies listed on the Stock Exchange of Thailand (SET) from 2016-2018. This paper found a significant positive relationship between the size and complexity of the firm, while a negative relationship was found between profitability and the level of KAM reporting. Ferreira & Morais (2020) showed a positive relationship between the complexity of the audited firm and Big-4 or Non-Big-4 and the level of KAMs disclosed at Brazilian listed companies (BM&FBovespa) in 2016. Furthermore, they found a negative relationship between the number of KAMs and the auditors’ fee and opinion.
In addition, Pinto & Morais (2019) tried to investigate the effect of reputation loss, pre-cision of accounting standards, litigation risk, auditor-client relationship and the effect of su-pervisors and regulators activities on the number of KAMs reported in the main European coun-tries (France, the Netherlands and the UK). In line with other papers, they also found a positive relationship between the complexity of the firm and the level of KAM reporting. In contrast to Ferreira & Morais (2020), did Pinto & Morais (2019) found a positive relationship between audit fees and level of KAMs reported. They also expanded the literature by adding a positive relationship between precise accounting standards and the level of KAMs reported.
Kitiwong & Srijunpetch (2019) conducted an interesting research about the effect of cultural differences between Asian countries on the level of KAMs reported. They defined cul-ture by the cultural dimensions of Hofstede (2001) from which they only investigated the effect of uncertainty avoidance and masculinity in Malaysia, Singapore and Thailand. They found no significant relationship.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 6 This research contributes to the literature by expanding it with new factors that may influence the level of KAM reporting since the available research failed in finding a significant relationship between the level of KAM reporting and cultural differences. Furthermore, the sample of this study consists of countries from different continents while prior literature only focused on samples of companies from one country or one continent (e.g. Europe, Asia & Bra-zil). Especially in the research field of culture, it is interesting to look at differences between countries from different continents.
Finally, this research can give regulators, financial managers and standard setters a bet-ter understanding of cultural factors that may have an influence on the auditors’ judgement and the level of KAM disclosure.
The remainder of this thesis proceeds as follows. The next section presents the theoret-ical background. Thereafter, follows an elaboration on the research design. In the fourth section, the results will be presented. The last section will discuss and conclude the findings of this research together with the limitations and implications.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 7
2. Theoretical Background
The next section will describe the theoretical framework. Firstly, the legitimacy theory will be discussed. The next paragraph will elaborate on Hogarth’s’ theory. Lastly, the cultural framework of Hofstede will be described and forms the basis of the hypothesis used in this research.
2.1 Legitimacy Theory
This paper firstly tries to explain the level of KAM disclosure by using the legitimacy theory. The concept of this theory can help explain how auditors use KAM reporting to reduce the legitimacy gap between the actions and the expectations of society. The theory builds upon the relationship between the company and the society at large and is based on the concept of companies continually trying to ensure that they act within the boundaries and norms given by society, in order to be seen as “legitimate” by these society groups. If a company wants to survive, it is crucial to gain a status that is seen as legitimate. This specific relationship between the company and the society is based on a “social contract” (Deegan & Samkin, 2009).
According to this contract, companies need to act within the boundaries and norms which are expected from the society at large in order to get permission to continue their opera-tions. This is a challenging task for companies since the requirements and the attitudes of the society at large are continually subject to change (An, Davey & Eggleton, 2011; Deegan, 2006). Therefore, companies need to adapt their operations to these changing expectations and norms. However, it is very difficult for a company to attain coherence between the needs of the organ-ization and the expectations and norms of the society at large because of the legitimacy gap. This gap is the result of the display between how a society perceives the actions of a company and if they are in line with their own perceptions on how they should have acted. The legitimacy gap will threaten a company’s reputation of being seen as legitimate by society (An et al., 2011). Within the context of the audit, the requirements of the society are changing towards a need for more transparency of the audit process and more specific information about the audited companies (Mock et al, 2013; Vanstraelen et al., 2012). ISA 701 was introduced to provide more information about the audit procedures to enhance the transparency of the auditor’s report (Gold & Heilmann, 2019). The legitimacy theory argues that auditors need to adapt their work-ing activities by implementwork-ing such new standards which help them to respond to the new ex-pectations. As mentioned before, the most prominent change from the new standard was the introduction of KAMs, which contain specific information about the key risks of a company.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 8 Moreover, auditors will report more KAMs if they believe that it is demanded by society to disclose more information and be more transparent (De Villiers & Van Staden, 2006). Thus, auditors will use KAM disclosure in order to respond to a changing need for more information by society to sustain their legitimacy status.
2.2 Hogarth’s Theory
Another theory that might explain the level of KAM disclosure, is the theory of Hogarth. More specifically, this theory can help in getting a better understanding of the underlying pro-cess of the disclosure of a KAM in the auditor’s report. The theory of Hogarth (1980) is con-cerned with the information assimilation of professional judgement and making decisions. It consists of four stages. The first stage contains the acquisition of information, while the second stage focusses on the processing of information. The third stage contains the output of the afore-mentioned stages. And lastly, the fourth stage is concerned with the feedback given on the output. The process that underlies the decision to disclose a certain level of KAMs falls within the third stage and is affected by auditors themselves, their task environment and the resulting actions (Pinto & Morais, 2019; Sierra-García, Gambetta, García-Benau & Orta-Pérez, 2019).
The decision of an auditor to disclose a KAM depends on the conflict whether a risk is significant or not based on professional judgement. According to Hogarth’s theory, an auditor can choose between two strategies to solve such judgement conflict. Firstly, an auditor can choose for a confrontation strategy which consists of compensatory options expressed in the expected utility framework. It simply means that there is a trade-off between the loss of repu-tation when the auditor doesn’t disclose a KAM and the loss of a client when the auditor does disclose a KAM. Thus, an auditor will disclose a KAM if it has a high expected utility which depends on the risk aversion. Secondly, an auditor can choose for an avoidance strategy. In this case, the auditor won’t disclose a KAM because of a feeling of less responsibility associated with the consequences of not reporting a KAM. (Pinto & Morais, 2019; Ferreira and Morais 2019)
In conclusion, the decision of an auditor to disclose a certain level of KAMs is influ-enced by the perceived effect of the trade-off between the expected cost of losing a client and on the other hand, the expected costs of loss of reputation. This will lead to an auditor avoiding or confronting clients with the disclosure of a KAM which depends on the expected conse-quences (Einhorn & Hogarth, 1981).
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 9
2.3 National Cultures Framework of Hofstede
Prior literature has defined culture in many different ways, for example House, Hanges, Javidan, Dorfman & Gupta (2004) defined culture as the shared beliefs, motives, values, inter-pretations and identities of events which resulted from experiences of members from commu-nities and are passed through different generations. While Hofstede (2001, pp.9) described cul-ture as “the collective programming of the mind that distinguishes the members of one group or category of people from another”. Culture can be seen as an environmental factor of a com-pany that is very powerful in affecting an accounting system and how individuals process in-formation and exists at different levels, namely national, geographical and organizational cul-ture (Tsui & Gul, 1996).
In this paper the focus will be on national culture, which includes the standards, behav-ioural patterns and principles of a nation (Ghemawat & Reiche, 2011; Leung, Bhagat, Buchan, Erez & Gibson, 2005). Tsui & Gul (1996) stated that culture affects the behaviour of an auditor and that it leads to differences in their professional judgement. KAMs are defined, according to ISA 701, as the product of the auditor’s professional judgement in deciding which issues of a financial statement are of most significance according to his/her judgement (Pinto & Morais, 2019). Therefore, it can be argued that culture helps in explaining the differences between coun-tries in disclosing KAMs by auditors.
The National Cultures Framework of Hofstede (Hofstede, 2001) is a framework that can help explaining such cultural differences and is the first theoretical model that enabled research-ers to quantify culture. The framework consists of five dimensions: Power distance (PD), Mas-culinity (MAS), Individualism (IDV), Long-term orientation (LTO) and uncertainty avoidance (UA). It is believed amongst researchers that PD, UA and IDV are the main cultural drivers of an auditors’ professional judgement and the most used in other literature (Patel et al., 2002; Tsui & Windsor, 2001). Furthermore, these three dimensions bring the clearest implications for regulators and standard setters which suggests that culture affects auditor’s judgement. Moreo-ver, Gray (1988) extended this framework by introducing four accounting values which are related to the societal values (dimensions) of Hofstede. Gray argues that sub-cultural account-ing values and societal values both influence the behaviour of an accountant. This study will only use the secrecy (SEC) dimension since it relates to UA, PD and IDV. Prior literature found a clear relationship between secrecy and the dimensions of Hofstede while there was no clear
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 10 relationship found between the cultural dimensions of Gray and Hofstede. This makes the other dimensions of Gray more difficult to measure (Hope, Kang, Thomas and Yoo, 2008)1.
To conclude, this paper will expect that the level of KAM disclosure differs cross-na-tionally based on PD, UA, IDV and SEC. The effect of culture on the level of KAM disclosure will be addressed below within the different dimensions.
2.3.1 Power Distance
Hofstede (1980) defined PD as the extent to which members of an organization with less power expect and accept a power distribution that is unequal. It can also be described as the degree of hierarchy, where power should be concentrated within the highest levels of a society or organization (Jang, Shen, Allen & Zhang, 2017; House et al., 2004). In countries with a high PD, managers and supervisors have centralized power and are willing to share less of this decision-making power. On the other hand, junior accountants for example are more reluctant in challenging their supervisors and expressing their disagreement with them. People are afraid of superiors and the influence they have on their own jobs (Dickson, Den Hartog & Mitchelson, 2003; Newbury & Yakova, 2006; Taylor & Curtis, 2003). This is in line with Hof-stede’s (1980) definition of PD, where people in cultures with high PD accept and abide their roles within such hierarchy.
Another important aspect of a power distant country, is that upper-level management accounting can be seen as a tool to present desired results (Hofstede, 2001, pp.383). People lower in the afore mentioned hierarchy are unlikely to question acts of people with authority and let alone report these questionable acts (Schultz, Johnson, Morris & Dyrnes, 1993). Thus, within the audit context, the level of PD can affect the interaction between the auditor and their client, since auditors are unlikely to question the acts of their clients in countries with high PD. Prior research found that in these cases, auditors are more likely to acquiesce to pressures from their powerful client. More precisely, it can be concluded that auditors from high PD countries are less likely to question financial results presented by their clients (McKinnon, 1984; Patel, et al., 2002; Yamamura, Frakes, Sanders & Ahn 1996). Moreover, Hogarth’s Theory tells us that auditors from a power distant culture will disclose less KAMs, since these people won’t
1Gray’s (1988) framework consists of three more variables, namely: professionalism, uniformity and conservatism.
Profes-sionalism relates to the preference of an auditor for exercising his professional judgement as opposed to complying with reg-ulatory requirements. Uniformity indicates an auditor’s preference for uniform accounting rules rather than flexible practices which depend on the perceived circumstances. Lastly, conservatism relates to a preference for a more careful approach of considering uncertainty instead of taking a more optimistic or laissez-faire approach. Prior research hasn’t determined a rela-tionship between these three variables of Gray (1988) and Hofstede’s framework (Mir, Chatterjee & Rahaman, 2009). There-fore, in this study only secrecy is included while professionalism, uniformity and conservatism are excluded.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 11 choose to confront their client. So, when company or industry events suggest that there is a risk of material misstatement that is significant, auditors are more reluctant to disclose such risk as a KAM, because otherwise these auditors will have to question and consult the acts of their clients (Hughes, Sander, Higgs & Cullinan, 2009).
Lastly, prior literature states that people in countries with high PD are more cautious in sharing important information to others and informational openness is discouraged in these countries (Jaggi & Low 2000; Luo & Tang, 2016). For the audit, this means that auditors are less likely to share sensitive information in the form of KAMs.
Based on the argumentation stated above, it can be expected that an auditor from a power distant country is more likely to disclose a lower level of KAMs since they are less willing to consult with superiors or share information with others. Contrary, auditors from less power distant countries are likely to disclose a higher level of KAMs based on the above find-ings. Accordingly, the following hypothesis was made:
H1: The level of Power Distance (PD) is negatively associated with the level of Key Audit Matter (KAM) disclosure.
2.3.2 Uncertainty avoidance
The dimension of uncertainty avoidance reflects the extent to which members of a so-ciety feel either uncomfortable or comfortable in unknown or uncertain situations (Hofstede, 2001, pp.161). Cultures with a high UA are built upon rule-based systems in order to reduce ambiguity. People need more time to seek information in their desire for more clarity and struc-ture. While cultures with low UA make use of principle-based systems during their work pro-cess. They have a higher tolerance for uncertainty while being more confident in their own abilities. The different level of UA can result in auditors reacting differently on certain problems during the audit process. When there is a high level of UA, auditors have a higher work effort during the audit engagement to reduce the risk to a level that is acceptable and obtain sufficient evidence to support their own judgment. While auditors with a low level of UA are relying more on their own professional judgement instead of certain rules (Cowperthwaite, 2010). Moreover, Cohen, Pant & Sharp (1995) argue that it is sometimes better to ignore certain rules while it is in the best interest of a society. Therefore, auditors with a low level of UA are better able to determine whether a risk of material misstatement is significant or not.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 12 Gray (1988) argues that in cultures with high UA there is a strong preference for con-straining the disclosure of information in order to retain security and avoid certain conflicts. According to this theory, auditors with a culture of high UA are more likely to be secretive which means that people are less willing to report certain confidential information. Prior re-search confirmed this theory of Gray, where high UA leads to a lower willingness to disclose specific information (Hope, 2003; Orij, 2010).
Following the analysis above, it can be concluded that auditors from an uncertainty avoidant country are more secretive in disclosing confidential information while making use of rule-based systems. Contrary auditors from countries that are not uncertainty avoidant, are re-lying more on their own professional judgement and confront their clients. Therefore, it can be expected that auditors from a country with a high level of UA will disclose a lower level of KAMs, while auditors from low UA countries disclose a higher level of KAMs. This could be explained with the following hypothesis:
H2: The level of Uncertainty Avoidance (UA) is negatively associated with the level of Key Audit Matter (KAM) disclosure.
2.3.3 Individualism
According to Hofstede (2001, pp.244), Individualism can be described as the degree of integration among society members and implies the preference for a tightly or loosely social fabric. IDV shows how such members relate themselves to society. Cultures with a high level of IDV are seen as individualistic, while cultures with a low level of IDV are seen as collec-tivistic. Within individualistic cultures people are acting in the best interest of their own needs. These people are more focused on their own tasks while emphasizing their own needs at the expense of the needs of others such as managers and clients. This indicates that there is a higher level of independency towards their manager or client. While on the other hand people from collectivistic cultures are seen as members of the “in-group” and act in the best interests of the group’s needs. Collectivistic cultures consist of teams (the “in-group”) that are interdependent with team members searching for loyalty from other members. The work approach from col-lectivistic people can be described as relationship-orientated, while individualistic people are more task-orientated. (Cowperthwaite, 2009; Hughes et al., 2009). Prior literature also argued that people from collectivist cultures show a preference for secrecy to avoid competition or conflict. While in individualistic cultures a competitive environment is encouraged which leads to a preference for secrecy. (Gray, 1988; Jaggi & Low, 2000)
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 13 As regarding the audit, individualistic auditors are less dependent on their client than collectivistic auditors since they have a task-orientated approach. This will enhance the com-municative environment between the auditor and their client which will result in a higher ten-dency of asking critical questions to the client. Contrary, collectivistic auditors are uncomfort-able in questioning the client because of their relationship-orientated approach. (Yamamura et., 1996)
To conclude, individualistic auditors disclose a higher level of KAMs because they are less dependent. These auditors are more focussed on their self-interests at the expense of their client’s interest. Collectivistic auditors disclose a lower level of KAMs since they are more dependent and value the relationship between the client. Therefore, the following hypothesis was made:
H3: The level of Individualism (IDV) is positively associated with the level of Key Au-dit Matter (KAM) disclosure.
2.3.4 Secrecy
Gray (1988) described secrecy as the preference for people to disclose confidential and restricted business information to only those people that are seen as insiders of the company. On the opposite side, people can have a more publicly accountable approach and be more trans-parent. Therefore, it can be seen as an interplay between secrecy versus transparency, where people can be secretive or transparent. In this interplay, secretive people prefer to disclose lim-ited information to stakeholders, while transparent people prefer to disclose transparent and honest information. Secrecy is closely related to PD, UA and IDV, where people with a high level of PD and UA tend to be more secretive while individualistic people prefer to be more transparent (Hope et al., 2008). This relationship will be made visible in the research design section.
As mentioned before, within a culture with a strong preference for UA people want to avoid conflict and competition and therefore tend to be more secretive in disclosing infor-mation. The same applies to cultures with a high level of PD, where people restrict certain information to sustain the power inequalities within the society (Hope et al., 2008). On the other hand, individualistic cultures prefer to be less secretive, while collectivistic cultures prefer to be more secretive. Collectivistic people are more concerned for insiders of the firm or closely involved with stakeholders (Gray, 1988).
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 14 Within the audit context, secrecy is an important value dimension since auditors deter-mine the quantity of information that will be disclosed to external parties (Jaggi, 1975). Apply-ing prior literature, it can be argued that secretive auditors are more conservative and apply a cautious work approach while disclosing information like KAMs to external parties (Gray, 1988). Therefore, it can be concluded that a higher level of secrecy will lead to a lower level of KAM disclosure. Contrary, auditors from transparent countries disclose a higher level of KAMs. This hypothesis can be drawn upon the relationship between secrecy and the three di-mensions of Hofstede. The last hypothesis was formulated as follows:
H4: The level of Secrecy (SEC) is negatively associated with the level of Key Audit Matter (KAM) disclosure.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 15
3. Research design
This section describes the design method used to investigate the research question. Firstly, the sample and collection of data will be explained. Then, the dependent variable will be elaborated on. Thenceforward, the independent and control variables will be explained. And lastly, the statistic model will be presented and explained.
3.1 Sample and data collection
For this study, data from listed companies on KAMS was manually gathered from 27 different countries for fiscal-year 2018. The sample is based on a convenience sampling ap-proach to make sure that a proper cross-country analysis could be done on the selected countries that were considered to be diverse in terms of the cultural dimensions of Hofstede (Hooghi-emstra, Hermes & Emanuals, 2015). This study used a dataset on KAMs from the Rijksuniver-siteit Groningen which was hand-collected by multiple students. Fiscal-year 2018 was the most recent year in the dataset that contains information on all variables for the countries used in this study. An overview of the selected countries and their scores on the cultural dimensions used in this study are presented in appendix 1. The sample only consists of companies that present an annual report with an auditor’s report. Data from stock-listed companies on KAMs was hand-collected from the auditor’s reports stated in annual reports. The companies were selected from the main national indices of the specific country since it is assumed that these companies and their auditors are an appropriate reflection of the overall sentiment of a country based on their size.
The initial dataset compromised of 1002 observations from 27 different countries. Com-panies were excluded from the dataset if the audit firm of a company was headquartered in a different country than the country of origin of the stock-market, when no auditor’s report was available or when other data of KAMs was missing. This has resulted in a final sample of 914 observations.
3.2 Dependent variable: Level of KAM disclosure
Focussing on the aim of this research to determine the effect of cultural differences on the auditors’ professional judgement in disclosing KAMs, this study used the number of KAMs to determine the level of KAM disclosure by the auditor. Pinto & Morais (2019) claim that this is a relevant measure to identify KAMs because of several reasons. One of them is that the importance of KAMs will be diluted because more KAMs disclosed will increase the complex-ity (Sirois et al., 2017). Another reason which shows the relevancy is that KAMs make other
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 16 related disclosures in the annual report more salient because the users’ focus is attracted by KAMs (Orquin & Loose, 2013). But when the number of these elements increases, it can neg-atively affect the effectiveness of signalling from the auditor (Li, Qi, Tian & Zhang, 2017). Therefore, to quantify the level of KAM disclosure, the number of KAMs was found to be an appropriate measurement tool by prior literature.
3.3 Independent variables
The cultural value of a country was characterized using Hofstede’s (2001) and Gray’s (1998) framework. This framework is a widely used, accepted and a well-defined measurement tool amongst different areas of research to characterize the culture of a country. Stahl and Tung (2015) described these dimensions as “a means for characterizing and differentiating the way
people in different societies think and behave”. The cultural dimensions of Hofstede (2001) are
based on a study done within a business organization (Hooghiemstra, et al., 2015). This makes Hofstede’s framework an appropriate tool for this research to define culture, since this study is focussed on the top businesses of a country. One side note that must be addressed is that this framework can’t characterize the whole culture of a country and it only helps to get a better understanding of the cultural values of that country (Chan & Cheung, 2012). The scores for Hofstede’s dimensions (2001) were collected from his website (Hofstede, Hofstede & Minkov, 2010)2.
As mentioned before, the cultural dimensions of Hofstede are globally the most used indicators for culture. However, the GLOBE values are a widely used alternative measure for culture, which is based on the cultural values and practices of a country (Miska, Szöcs, & Schiffinger, 2018). This paper used the GLOBE values in robustness tests to check for potential underlying false assumptions. Firstly, to use this alternative measure, the cultural dimensions PD, UA and IDV were transformed into the GLOBE values. This was done, to make sure that the GLOBE values and Hofstede’s dimensions were equal and comparable to each other. Con-sequently, the cultural dimension of PD was replaced by GPD, UA was replaced by GUA while IDV was replaced by GIDV. Moreover, the GLOBE study distinguishes two dimensions of individualism/collectivism: societal in-group collectivism and societal institutional collectiv-ism. According to House et al., (2014) the dimension of societal in-group collectivism is most equal to the cultural dimension individualism of Hofstede. The GLOBE values were collected
2 Information on the cultural scores of Hofstede’s framework was retrieved from his website:
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 17 from the dataset of the GLOBE project3. These values were measured on a scale of 1, which indicates a low score, and 7, indicating a high score. Furthermore, some observations were deleted from the dataset since the auditor was headquartered in a country that was not included in the GLOBE project. This compromises the following set of countries: Austria, Belgium, Norway and Romania.
Lastly, the scores for secrecy from Gray’s (1998) framework were measured using the equation below, which is formulated by Hofstede’s dimensions. This equation represents the preference of an auditor for confidentiality which might restrict the reporting of information in KAMs. Hope et al. (2008) found a relationship between PD, UA and IDV from Hofstede (2001) and SEC from Gray (1998). Moreover, they used the following equation to measure the SEC variable. The cultural scores on each dimension for all countries are presented in appendix 1.
SEC= UA +PD – IDV (Orij, 2010).
3.4 Control variables
This paper uses several variables to control for the effects these variables have on the level of KAM disclosure found in prior research:
Ferreira and Morais (2019) found a positive relationship between the auditor being a Big-4 firm and the number of KAMs disclosed. It is argued that Big-4 firms provide a better audit quality (Francis & Yu, 2009). Because of their size, they have better financial and human resources and industry-specific knowledge which gives them access to more advanced technol-ogy to increase their audit process (Simunic & Stein, 1987). In addition, it is believed that these Big-4 firms possess higher wealth values so there is more at stake for these audit firms in terms of litigation and sanctions. Therefore, KAM disclosure can be seen as a tool to communicate relevant issues and insure themselves against the consequences of litigation and sanctions (Fer-reira and Morais, 2019). As a result, this paper used BIG4 as a control variable.
This study used AUDITROTATION to control for the change of auditor during a fiscal-year. Velte (2018) argues that learning effects are regarded as a result of longer audit tenure. This results in the new appointed auditor being more conservative to reduce their liability and therefore increase the number of KAMs (Sierra-García et al., 2019). Contrary, Brown and Knechel (2016) state that there can be a better fit with the client and newly appointed auditor based on compatibility. Sierra-García et al. (2019) state that it could also be possible that
3 Information on the cultural GLOBE values were retrieved from the following website:
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 18 auditors will disclose less KAMs based on their lower experience. Nevertheless, prior research (Velte, 2018; Pinto & Morais, 2019) found a negative relationship between audit rotation and the level of KAM disclosure.
Prior literature found a positive relationship between client SIZE and the level of KAM disclosure (Pinto & Morais, 2019; Sierra-García et al., 2019; Suttipun, 2018; Velte, 2018). Pinto and Morais (2019) claim that larger firms show more complexity which require more audit work and possesses a greater risk of liability for the auditor. It was thus expected that auditors will disclose more KAMs when auditing a larger client. SIZE is measured by the natural loga-rithm of a firms’ total assets. To account for currency differences, the total assets were con-verted to Euro’s using the conversion rate of 31 December 2018. These conversion rates for each currency can be found in appendix 2.
YEAREND is a control variable included in this study to control for the busy season and its impact on the disclosure of KAMs (Bédard et al., 2019). There is a possibility that during the busy season auditors may be overloaded with work as fiscal-year end at 31 December is most common (Hay, Knechel & Wong, 2006). This could potentially lead to auditors disclosing more KAMs to lower their liability (Pinto & Morais 2019). A positive relationship was ex-pected.
3.5 Statistical model
The hypotheses of this study were tested with a multiple regression analyses in order to get an understanding of the effect of culture on the auditors’ professional judgement in disclos-ing KAMs. A multiple regression analyses is typically used in prior studies to examine rela-tionships between characteristics and level of KAM disclosure (Suttipun, 2020). A regression method was employed with archival data. The following model was tested:
!"# = &'+ &')* + &'+" + &',*- + &'./0 + &'1,24
+ &'4/"5/6* + &'"+*,7587"7,89 + &'.,:/ + ;<
in this equation KAM stands for the level of KAM disclosure, &' is constant, &< the independent and control variables and ;< the error term. Table 1 presents an overview of all variables and their definition.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 19
Table 1
Variable Definitions
Variables Description
Dependent variable
KAM The number of Key Audit Matters disclosed by the
auditor in the auditor’s report at fiscal-year 2018.
Independent variables
PD The level of power distance contains a score
be-tween 0 and 100. Where a higher (lower) level of power distance indicates a relatively more unequal (equal) distribution of power within a culture.
UA The level of uncertainty avoidance contains a score
between 0 and 100. Where a higher level of tainty avoidance indicates a relatively more uncer-tainty avoidant culture.
IDV The level of individualism contains a score
be-tween 0 and 100. Where a higher (lower) level of individualism indicates a relatively more individu-alistic (collectivistic) culture.
SEC The level of secrecy is based on the equation
UA+PD-IDV and contains a score between -100 and 200. Where a higher (lower) level of secrecy indicates a relatively more secretive (transparent) culture.
Control variables
BIG4 A dummy variable where 1 indicates a firm being
audited a Big 4 audit firm (Deloitte, Ernst & Young, KPMG, PwC) and 0 otherwise
YEAREND A dummy variable where 1 indicates a fiscal-year
of the audited company ending on December 31 and 0 otherwise. However, some countries have a different fiscal-year ending and adjustments to YEAREND were made
AUDITROTATION A dummy variable where 1 indicates that the
au-dited firm has switched from audit company and 0 otherwise.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 20
4. Results
In the next section, the results of the regression analyses are presented. The first part contains the descriptive analysis, where information on all variables are discussed. This will be followed up by an overview of the different analyses and robustness tests that were done.
4.1 Descriptive Analysis
The descriptive statistics of the independent and dependent variables are presented in table 2. Panel A provides information on the dependent variable KAMs per country. The mean value of KAMs for the complete dataset is 2.83 (Std. Dev.=1,482), which implies that on aver-age 2.83 key audit matters were reported in the auditor’s report. The minimum of reported KAMs was 0 while the maximum was 8. Contrary, ISA 701 became active after fiscal year 2016 for European organizations (IFA, 2017), and therefore it was expected that the minimum of European countries was higher than 0. Panel B represents the descriptive statistics of the independent and control variables. Regarding the cultural dimensions of Hofstede, all scores were above or around 50 which indicates that the dataset contained diverse countries that were well balanced. Based on the information from panel B, it can be stated that the sample contained on average countries that are more power distant, uncertainty avoidant, individualistic and less secretive (different scale). The average firm of this dataset has a 33.4658 natural logarithm of assets. Furthermore, of the audited firms 93% was audited by a BIG4 firm, 89% had a fiscal year ending on 31 December and only 6% has switched from auditor.
4.2 Correlation Analysis
Table 3, which can be found in the appendix, presents the correlation matrix where the presence of multicollinearity was tested. It is difficult to make unique estimates of the regres-sion coefficients if there are multicollinearity issues. Such an issue indicates that the predictor variables in the regression model can be predicted by other predictor variables from the model. A possible multicollinearity problem can be tested through a Pearson’s correlation analysis where multicollinearity is present at a correlation higher than .8 or lower than -.8 (Field, 2013) between two variables. The results indicate that there is one possible issue with multicollinear-ity. Secrecy has a strong significant multicollinearity issue with power distance (.864). How-ever, this relationship is logical and consistent with prior studies since secrecy was explained by power distance, uncertainty avoidance and individualism. This also explains the high corre-lation value of secrecy and individualism which falls just under the critical .8 value to call it a multicollinearity issue.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 21
Table 2
Descriptive Statistics
Panel A: Dependent variable (KAM) per country
Country N Mean Median Min Max Std. Dev.
Australia 53 3.53 3.00 1.00 7.00 1.265 Austria 20 2.30 2.00 1.00 5.00 .865 Belgium 18 2.89 2.50 1.00 8.00 1.745 Brazil 62 2.81 3.00 .00 7.00 1.513 China 11 2.18 2.00 1.00 3.00 .603 France 37 3.46 4.00 .00 7.00 1.386 Germany 29 3.31 3.00 2.00 6.00 1.004 Great Britain 47 4.34 4.00 .00 7.00 1.903 Hong Kong 73 2.56 3.00 1.00 7.00 1.105 Italy 34 2.85 3.00 .00 6.00 1.395 Mexico 20 2.25 2.5 .00 7.00 1.773 Malaysia 29 2.48 2.00 1.00 6.00 1.430 Netherlands 30 4.10 4.00 2.00 8.00 1.561 Norway 23 2.04 2.00 .00 4.00 1.022 New Zealand 44 2.02 2.00 1.00 4.00 .821 Philippines 29 3.59 4.00 1.00 7.00 1.763 Poland 26 3.15 3.00 1.00 6.00 1.347 Romania 32 2.09 2.00 .00 7.00 1.489 Russia 39 2.26 2.00 .00 5.00 1.229 Singapore 26 2.50 2.00 1.00 7.00 1.556 South Africa 34 2.76 3.00 1.00 5.00 1.208 South Korea 42 1.98 2.00 .00 5.00 .897 Spain 33 3.36 3.00 1.00 7.00 1.454 Sweden 26 3.04 3.00 .00 7.00 1.509 Switzerland 20 3.35 3.00 1.00 7.00 1.531 Thailand 49 2.14 2.00 .00 6.00 1.155 Turkey 28 2.57 3.00 .00 5.00 1.034 Total 914 2.83 3.00 .00 8.00 1.482
Panel B: Independent variables
Mean Median Min Max Std. Dev.
PD 58.21 64.00 11.00 100.00 22.060 UA 61.04 64.00 8.00 95.00 23.377 IDV 51.57 51.00 18.00 90.00 24.439 SEC 67.68 72.00 -19.00 150.00 51.822 BIG4 .93 1.00 .00 1.00 .248 YEAREND .89 1.00 .00 1.00 .311 AUDITROTATION .06 .00 .00 1.00 .238 SIZE 33.4658 33.2424 19.96 41.68 2.98988
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 22 The correlation matrix presents some significant relationships between the cultural pre-dictor variables with the level of KAM disclosure. As for the control variables, only SIZE showed a significant relationship with KAM. Firstly, there exists a negative relationship be-tween power distance and the level of KAM disclosure which is strongly significant (-.119). This implies that auditors in a power distant country will disclose a lower level of KAMs. Sec-ondly, uncertainty avoidance showed a significant relationship with KAM that is negative (-.077). This indicates that auditors from a more uncertainty avoidant country will disclose less KAMs. Thirdly, a positive relationship between individualism and the level of KAM disclosure existed (.276) that is strongly significant. This relationship implies that auditors from a more individualistic country will disclose more KAMs. Lastly, secrecy showed a negative relation-ship with the level of KAM disclosure that is strongly significant (-.215). This relationrelation-ship indicates that an auditor from a country that is more secretive will disclose fewer KAMs. Sig-nificant relationships between cultural dimensions can be explained by the definition of culture by Gray (1988) where culture is a system of shared values. Another explanation is that secrecy is compiled by the three dimensions of Hofstede. All correlations support the directions as hy-pothesized.
In addition, to assure that there weren’t any other issues of multicollinearity, the Vari-ance Inflation Factors (VIFs) of all variables were calculated. The VIFs are presented in table 4, where a VIF higher than 5 implies an issue of multicollinearity (Stine, 1995). There is one problem of multicollinearity in the regression model which contains the variable secrecy and is in line with the results of the correlation matrix. This variable is formulated by the three dimen-sions of Hofstede and therefore resulted in a multicollinearity problem. Consequently, secrecy was excluded from the model containing all variables. The low VIF values of the other variables showed that there were no other problems of multicollinearity.
Table 4
Variation Inflation Factors (VIFs)
Tolerance VIF
PD .447 2.237
UA .844 1.185
IDV .465 2.151
SEC n/a n/a
BIG4 .936 1.068
YEAREND .915 1.093
AUDITROTATION .982 1.019
SIZE .933 1.072
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 23
4.3 Regression analysis
In order to test the impact of culture on the level of key audit matter disclosure, a re-gression analyses was conducted. Table 5, outlined in the appendix, present the results of this regression analysis. The hypotheses were tested in respectively model 1, 2, 3, 4, 5, and 6 in order to verify whether a certain hypothesis must be rejected or accepted.
In the first model, a test was done to verify the effect of the control variables. The ad-justed R-square of this model was .157, this means that 15.7% of the variation in the level of KAM disclosure can be explained by the control variables YEAREND, AUDITROTATION, BIG4 and SIZE. In model 1, there was a significant negative relationship found between the fiscal-year end of a company and the level of KAM disclosure (β = -.243, p < .10). This indi-cates that when the fiscal-year of a company ends on 31 December, they will disclose less KAMs than when the fiscal-year ends on a different date. Moreover, there exists also a signifi-cant relationship between AUDITROTATION and KAM (β = .313, p < .10), which indicates the auditor will disclose more KAMs when the audited company has switched from auditor. Then lastly, model 1 showed a significant relationship between SIZE and KAM (β = .200, p < .01) indicating that an auditor will disclose a higher level of KAMs when the audited company becomes a larger cooperation.
Model 2 tested for hypothesis 1 which argues that a higher level of PD will lead to a lower level of KAM disclosure, thus a negative relationship was expected. The explanatory power of this model including PD increased by .5 % compared to model 1. This model indeed showed a negative relationship between PD and KAM which was also significant (β = -.005, p < .05), consequently hypothesis 1 was accepted. The model also presents relationships of two control variables BIG4 and SIZE and KAMs. The relationship between BIG4 and KAM was significantly negative (β = -.349, p < .10), were an auditor from a BIG4 firm will disclose less KAMs.
Model 3 tested for hypothesis 2, which proposes that a higher level of UA will lead to a lower level of KAM disclosure. Thus, a negative relationship was expected where an auditor from an uncertainty avoidant country will disclose a lower level of KAMs. The explanatory power of this model stayed equal at 15.7% compared to model 1. As proposed in hypothesis 2, model 3 showed a negative but not significant relationship between UA and KAM (β = -.001). Therefore, hypothesis 2 was rejected. Model 3 also contained significant relationships between the control variables AUDITROTATION and SIZE and KAM.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 24 The fourth model 4 tested for hypothesis 3 which implies that auditors from a more individualistic country will provide a higher level of KAMs. Therefore, a positive relationship was expected between IDV and KAM. The model showed an increase in explanatory power from 15.7% in model 1 to 21% in model 4. The relationship between IDV and KAM in this model was positively significant (β = .014, p < .01). Thus, the findings from this model lends support to hypothesis 3. The model also presented significant relationships between the control variables BIG4 and SIZE and KAM.
The fifth model was used to test hypothesis 4, which proposes that auditors from a more secretive country will disclose less KAMs. This hypothesis assumes a negative relationship between SEC and KAM. When including the independent variable SEC in the regression model, the explanatory power was 2.2% higher. Model 5 showed a negatively significant rela-tionship between SEC and KAM (β = -.005, p < .01), so hypothesis 4 was accepted. This model also showed significant relationships between all control variables, except for YEAREND, and KAMs.
Lastly, the sixth model added all independent and control variables used in this research, except for SEC because of multicollinearity issues. The main purpose of model 6 was to identify the explanatory power of a model with all the variables from this research. This model showed the highest explanatory power (22.6%) of all variables on the level of KAM disclosure com-pared to the other models. Moreover, this model only showed significant relationships between the control variables BIG4 and SIZE and KAM. Contrary, model 6 showed a positively rela-tionship between PD and KAM compared to model 2, which is also statistically more signifi-cant. Model 6 also showed a new significant relationship between UA and KAM, which was in line with hypothesis 2 and was therefore accepted. Lastly, the β coefficient from IDV increased from .014 to .023. To conclude, model 6 provided support for hypothesis 2 and 3.
4.4 Robustness tests
In order to test if the regression model was based on true underlying assumption or by false assumption, robustness tests were examined. In this way, it was verified if the results were misleading or not. Table 5, presented in the appendix, provides the results of these robustness tests, where regressors were added or removed from the original regression model.
Model 7-10 present the first robustness tests, where culture was measured with the GLOBE values, which is an alternative measure for culture. As mentioned before, the cultural dimensions of Hofstede were transformed into the GLOBE values. Consequently, PD was re-placed by GPD, UA was rere-placed by GUA while IDV was rere-placed by GIDV.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 25 The explanatory power of the all models including the GLOBE values were almost equal to the models with Hofstede’s dimensions, the highest difference was 2.1%. In accordance with model 2, model 7 also showed a significant negative relationship between power distance and the level of KAM disclosure (β = -.576, p < .01). Contrary, model 8 showed a negative signif-icant relationship between uncertainty avoidance and level of KAM disclosure (β = -.436, p < .01) where no significant relationship was found in model 3. Furthermore, model 4 showed a significant relationship between individualism and KAM disclosure that was not existent any-more in model 9. Furtherany-more, model 10, with all GLOBE values, also showed three significant relationships (β = -.615; -.476; .522, p < 0.01) compared to model 6. These results were in line with the hypotheses from this paper. While model 6 showed a significant positive relationship between PD and KAM, this model showed a negative relationship in line with hypothesis 1.
With the cultural dimensions of Hofstede being transformed into GLOBE values, the cultural value SEC could also be transformed into a GLOBE value. This has resulted in a new measure for secrecy called GSEC. The results in model 11 showed a significantly negative relationship between GSEC and KAM (β = -.511, p < .01) which was in line with hypothesis 4 and model 5. The explanatory power of this model has increased from 17.9% to 20.6%.
The last robustness test consists of an alternative measure for secrecy. In this paper, SEC was measured through the link between PD, UA and IDV which was tested in model 5. How-ever, Orij (2010) argues that SEC also can be measured with the following formula: SECalt = PD + UA - IDV – MAS and was tested in model 12. MAS stands for masculinity and is another cultural dimension of Hofstede. MAS is defined as a clear distinction between the gender roles where men are focused on assertiveness and material incentives, and woman on the quality of life (Hofstede, 2010). The explanatory power of model 12 was equal to that of model 5 (17.9%). Furthermore, model 12 showed a negative significant relationship (β = .004, p < .01) which was in line with model 5 and hypothesis 4.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 26
5. Conclusion and Discussion
In this section, the research question will be answered. Firstly, the results of the previous section are discussed. Subsequently, this study’s limitations and suggestions for future research are provided. Finally, the implications of this study are given.
5.1 Discussion and Conclusion
The aim of this study was to investigate whether culture has an influence on the profes-sional judgement of an auditor in disclosing KAMs. The International Federation of Account-ants released new auditing standards to increase the decision-usefulness of the auditor’s report that became effective after fiscal-year 2016. ISA 701.8 was one of these new standards imple-mented to improve the traditional pass/fail model. This new standard introduced the usage of key audit matters. The aim of the KAMs was to reduce the information gap and create more uniformity around the world. KAMs are those matters that were of most significance in the financial statements based on the professional judgement of the auditor. Moreover, Tsui & Gul (1996) argued that culture has an effect on an auditor’s professional judgement. Thus, it was expected beforehand that the level of KAM disclosure would differ across nations since culture affects the behaviour and professional judgement of an auditor.
To empirically test this expectation, a regression model was conducted while potential relations were explored. The regression model included three of Hofstede’s cultural dimensions and one of Gray’s dimensions, namely PD, UA, IDV and SEC. The analysis was done on a data sample consisting of 914 auditor’s reports from listed companies in fiscal-year 2018 from 27 countries all over the world. The purpose was to explore new factors that may had an influence on the auditor’s professional judgement in reporting KAMs. Most previous studies only iden-tified the potential consequences of KAM disclosure. Therefore, this study contributes to the growing literature on the entry of ISA 701.8 by expanding it with factors influencing KAM disclosure. Furthermore, the dataset consists of 27 countries with diverse cultures. This makes it a unique study on culture in the existing literature. But most important, this study can give regulators, financial managers and standard setters a better understanding of cultural factors that influence the auditors’ judgement on KAM reporting.
The regression results from the separate models support all hypotheses except for the second one. The results show that there are significant links between PD, IDV and SEC and the level of KAM disclosure. Thus, there is support to state that power distant and secretive coun-tries lead to a low level of KAM disclosure, while individualistic councoun-tries will disclose a high
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 27 level of KAMs. However, the results did not indicate that uncertainty avoidant auditors disclose a lower level of KAMs. The robustness tests show additional support for the relationship of PD and SEC and KAM. Based on these empirical results, it can be concluded that there appears to be a link between culture and the level of KAM disclosure.
The results from the complete model showed significant relationships for the dimen-sions of Hofstede, indicating that hypothesis 1-3 could be accepted. However, the relationship between PD and KAM was in the opposite direction as hypothesized in H1. This was in line with the results found in prior research (Hope, 2003; Ho, Wang & Vitell, 2012; Zarzeski, 1996). All of these researchers found a positive relationship between PD and disclosure while putting all of Hofstede’s dimension in one model. This indicates that countries with a higher level of PD disclose more information. These findings were also hypothesized in an opposite direction. Ho et al., (2003) argued that these findings can be explained by the level of regulation and law enforcement. Furthermore, Zarzeski (1996) claims that these confounding effects are the result of the high correlation between PD and IDV. This research also found a high correlation be-tween PD and IDV (-.714) and two of the highest VIF scores. However, there wasn’t a problem found beforehand with multicollinearity. But looking at the separate regression model contain-ing PD, there indeed was support for hypothesis 1 indicatcontain-ing a negative relationship. Thus, Zarzeski’s claim that it is a problem of collinearity sounds very likely, combining the regression and correlation results found in this paper with the results from prior research. Another expla-nation can be found in the used measurement tool of Hofstede’s dimension to determine a coun-try’s culture. While using the GLOBE values there exists support for all hypotheses in the pro-posed direction including PD. Based on the above findings from the complete model, it can be concluded that there exists a relationship between culture and the auditor’s professional judge-ment in disclosing KAMs.
As for the control variables, significant relationships were found in different regression models between the level of KAM disclosure and the control variables. Contrary, there existed negative relationships between BIG4 and YEAREND and KAM, while positive relationships were expected.
To conclude, this study provides evidence that indicates that culture has an effect on the auditor’s professional judgement in disclosing KAMs. However, some limitations and sugges-tions for future research need to be considered together with this conclusion. But most im-portantly, these conclusions have some important implications for regulators and standard set-ters.
The Effect of Cultural Differences on the Auditors’ Professional Judgement in Disclosing Key Audit Matters 28
5.2 Limitations and future research
Regardless of several efforts, some limitations need to be considered while interpreting the outcomes of this study. Consequently, this leads to suggestions for further research.
The first important group of limitations that needs to be addressed, focuses on the da-taset of this study. Moreover, the sample of this study only consists of public firms which makes it unclear if the results are generalizable to non-listed firms. These types of firms may be subject to lesser regulations or other stakeholder-groups with different incentives. This makes it inter-esting for further research to see if culture still can be an important determinant in different contexts as for non-listed firms. Additionally, some data was excluded from the dataset, for example when the annual report wasn’t available in the English language. This could potentially lead to small differences in results and therefore needs to be addressed, although it only consists of a small number of such reports. To fully understand this potential effect, further research on this subject needs to be done. Another limitation of this dataset is that the focus of this study was only on fiscal-year 2018. This makes it a relatively new standard for some auditors, since the new standard ISA 701.8 became effective after fiscal-year 2016 (IFAC, 2017). Therefore, the implementation of this standard could still lead to trouble in properly exercising it. This also makes it an interesting topic for future research to look at the possibility of a learning effect. Furthermore, some argue that after a longer period a possible learning effect could exist (Chloe, 2002), which makes it interesting to conduct a comparative study in the future and see if this study’s cultural differences still hold in the future.
Secondly, the use of Hofstede’s and Gray’s cultural dimensions in this study also gen-erates some limitations, despite the popular use of the cultural frameworks of Gray (1988), Hofstede (2001) and House et al., (2004) to describe the culture of a country. For instance, these frameworks are based on a sample of all kinds of professions but doesn’t focus only audit prac-titioners. Therefore, it would be interesting for future research to dive more into the cultural differences in the audit profession in order to get a better understanding of the effect of culture in the professional behaviour of auditors. This could be done, for example, with a definition of culture that is based on a more specific sample of audit practitioners only (Bik & Hooghiemstra, 2017). Another limitation of such frameworks like that of Hofstede (2001) to describe culture, is that it doesn’t account for within-country differences of culture. Hofstede assumes that a country has a uniform culture, while in practice culture differs within a country and more spe-cifically within a profession (McSweeny, 2002). Additionally, if countries have the same cul-tural score on a dimension, it is assumed that auditors behave the same in exercising the new KAM standard for example. However, this is an unrealistic assumption of such framework,
