Abstract: Based on an analysis of a sample of US companies, this study investigates the
influence of corporate governance on the relationship between audit fees and audit quality in the US, where audit quality is measured by the absolute value of discretionary accruals. Firstly, this study examines the relationship between audit fees and audit quality before and after SOX over the period between 2000 and 2006. Subsequently, it analyzes the influence of corporate
governance on the relationship between audit fees and audit quality in the years 2007 through 2012. The most important result of this study is that there is no significant influence of corporate governance on the relationship between audit fees and audit quality. The control variables used were firm size, leverage and big n, which means that the client was audited by a big four audit firm. This study provides an important contribution to the literature since there have not been any previous studies on the effect of corporate governance on the relation between audit fees and audit quality.
Keywords: audit quality, audit fees, corporate governance, SOX, earnings management, audit
The influence of corporate governance on the relationship
between audit fees and audit quality
Rishi Gahar 10220518
Master of Science thesis Accountancy & Control Prof. dr. Vincent O'Connell
Table of content
1. Background and prior literature ... 4
2. Literature review ... 7
2.1 Audit quality ... 7
2.2 Audit fees ... 8
2.3 Relation between audit fees and audit quality ... 9
2.4 Corporate governance ... 10
2.5 Formulation of hypotheses ... 11
3. Research methodology ... 14
3.1 Measuring of audit quality ... 14
3.2 Regression models... 15
3.3 Sample selection ... 18
4. Results ... 20
4.1 Descriptive statistics and correlation coefficients ... 20
4.2 Main results ... 22
4.3 Sensitivity analysis ... 24
5. Conclusion ... 26
Reference ... 28
4
1. Background and prior literature
This section provides an outline of the background and prior literature, as well as introducing the research question and the general structure of this thesis.
In recent decades there have been many accounting scandals, for instance at Enron, Parmalat, Worldcom and Ahold. These scandals have led to discussions about audit quality (Cullinan, 2004).
There were concerns by the Securities and Exchange Commission (SEC) about the effect of audit fees on the quality of financial reports. As result, the auditor independence regulations were revised in 2000. The purpose of this adjustment by the SEC was to maintain and improve the perception of auditor independence (Iyer et al., 2003, pp. 128-129). Auditor independence may be threatened when the economic value of a certain client is relatively high. The risk in such cases is that the auditor becomes financially dependent and sensitive to pressure from the client. Studies by Simunic (1984) and Beck et al. (1988a) show that there is a positive relation between sensitivity to client pressure and earnings management in the financial statements.
There are many studies done about the relation between audit fees and audit quality. Several studies (Hay et al., 2006; Chung & Kallapur, 2003) argue that non-audit service fees have a huge effect on the independence of an auditor, because the non-audit service is very profitable for audit firms. Research by Frankel et al. (2002) found a negative relation between non-audit fees and audit quality, and a positive relation between audit fees and audit quality. However, Chung & Kallapur (2003) did not find any significant relation between audit fees and audit quality, nor did they find a significant relation between non-audit fees and audit quality. Ashbaugh et al. (2003) also failed to find a significant relation between audit fees and audit quality. In other words, studies on this subject have yielded contrasting results.
Prior literature about the relation between audit fees and audit quality discusses the situation before the big accounting scandals in recent decades. As a reaction to these scandals, the Sarbanes Oxley Act (SOX) was implemented in 2002. The implementation of SOX meant that audit firms were no longer allowed to provide a combination of audit services and non-audit services to their clients. The rationale behind this regulation is that audit quality may be
improved by reducing incentives for auditors to keep the client happy for the purpose of securing future benefits (SEC, 2000; Craswell et al., 2002).
5 SOX is a regulation in corporate governance, which Larcker & Tayan (2011) define as a collection of control mechanisms that prevent or dissuade potentially self-interested managers from engaging in activities that are not in the best interest of investors and stakeholders. The need for corporate governance derives from a situation where there is a separation between the ownership of a company and its management. Executives have the opportunity to act in their own interests while investors and stakeholders have to bear the costs of these actions. This is a typical scenario of the agency problem (Larcker & Tayan, 2011). This scenario is known as the
Moral hazard. Some parties cannot observe the actions of others while these actions do affect the interests of all parties in the transaction. Moral hazard arises due to the separation of
management and ownership in companies (Scott, 2009). The costs that result from the agency problem are called agency costs. Corporate governance is a type of control or a monitoring system that is put in place in an organization with the aim of reducing agency costs (Larcker & Tayan, 2011). Cohen et al. (2004, p.88) showed that ensuring the quality of the financial statements is an important function of corporate governance.
As prior literature has indicated, there is as of yet no conclusive knowledge about the exact impact of audit fees on audit quality. Furthermore, only a limited number of studies have investigated the impact of audit fees on audit quality before and after SOX. In addition, no previous studies have examined the effect of corporate governance on the relation between audit fee and audit quality. In order to fill this gap in the existing knowledge, this thesis focuses on the following research question:
“What is the influence of corporate governance on the relationship between audit fees and audit quality?”
From an academic point of view, this study will contribute to this research area in several ways. First, this research will provide new insights into the effects of corporate governance on the relation between audit fee and audit quality. Furthermore, this research will contribute to the existing literature on audit quality. From a social point of view, this is also an interesting study because this research area is a much talked about topic in current debates.
The remainder of the study is organized as follows. The second section provides the theoretical framework, while the third section formulates the hypotheses and describes the
6 research methodology. The fourth section describes the results and the fifth and final section provides the conclusions.
7
2. Literature review
This section will discuss the theory with respect to audit quality, audit fees, corporate governance and the hypotheses. The first paragraph defines the term audit quality while the second paragraph discusses audit fees. The third paragraph discusses the relation between audit fees and audit quality and the phenomenon of corporate governance will be discussed in the fourth paragraph. Finally, the hypotheses will be discussed in the fifth paragraph.
2.1 Audit quality
Arruñada and Paz-Ares (1997) describe audit quality as the outcome of two variables. The first variable is professional competence, which refers to an auditor’s ability to detect anomalies in the financial statements. The second variable is independence, i.e., the preparedness of an auditor to report the detected irregularity.
Carson et al. (2013) also defined audit quality based on the competence and the
independence of an auditor. These two aspects of audit quality can be affected by various other factors. Some examples of factors that may affect the competence and independence of an auditor are economic dependent, the size of the audit fees, auditor rotation, auditor tenure, size of the accounting firm, industry specialization of the auditor firm, relationship with the client, litigation risk and external regulations (Carson et al, 2013).
Auditing of financial statements allows stakeholders to verify the validity of financial statements, which reduces agency problems between management and stakeholders (Becker et al., 1998). Becker et al. (1998) show that dubious accounting practices, such as earnings management, were more likely to be detected by high-quality audits. Detection of earnings management may possibly damage management reputation and lead to a decrease in firm value. Audit quality has not only the function of detecting misstatements, but also of assessing whether these misstatements are truthfully reported (Catanach & Walker, 1999).
DeAngelo (1981) makes a distinction in audit quality between the probability that an auditor discovers material errors and takes no action and that an auditor discovers a material error and takes action. Material errors in financial statements quantify the quality of the audit. The probability that an auditor discovers material errors depends on his expertise, customer knowledge and experience. Whether the auditor also takes action against errors, depends on his
8 independence. Discovering errors and taking action against material errors in the financial
statements are important for public confidence in the reliability of the financial statements (DeAngelo, 1981).
The literature provides several definitions of audit quality. The essence of the different definitions is that audit quality derives from the likelihood that an auditor will find a leak in the accounting system of a client and will take action against this error.
2.2 Audit fees
Audit fee is a fee that a company pays to an external auditor in exchange for performing an audit (Arens et al., 2012). Total audit fees are made up of the following components; audit fees, non-audit fees and non-audit-related fees (Arens et al., 2012). Audit fees include all the expenses that are linked to the financial audit work. Audit fees are linked to the financial audit for all domestic and foreign subsidiaries (Pratt & Stice, 1994). Audit-related fees are fees for other assurance
services, excluding audit of the financial statement and non-audit fees are fees that are not linked to an audit (Arens et al.,2012).
Provision of audit services incurs several costs. Simunic’s (1980) economic model of audit pricing claimed that an auditor’s cost function consist of two elements of cost: the audit effort and the expected costs of the business risk (Vieru & Schadewitz, 2010). The study of Pong and Whittington (1994) discussed a similar two-fold division of auditing costs. Their two groups of costs are, firstly, the direct costs in the auditing process and costs resulting from the efforts required to collect enough evidence to form an audit opinion. The second part of the costs consists in possible future losses in the post-audit period; such costs arising from an incorrect opinion about the audited financial statements. In addition, Lennox (1999a) argues that the level of the audit fee is influenced by the size and complexity of the client.
Audit Analytics (2007) found that in the time period between 2003 and 2005, the total audit fees for companies required to comply with SOX rose by 66%. In addition, they found that for companies that are not required to comply with SOX, the audit fees increased by 42%. This can be explained by the fact that the number of providers of auditing services on the market decreased after 2002 and as result the market for auditing services was less competitive.
Whisenant et al. (2003) have established that there is a link between audit and non-audit fees. This may be explained by theory of knowledge spillover, which means that the knowledge
9 that is gained by an audit service may be used for a non-audit service as well. The combination of both services therefore provides the audit firm with a costs advantage (Whisenant et al., 2003).
2.3 Relation between audit fees and audit quality
Many studies have examined the relation between audit fees and audit quality, though they have shown different results.
Hoitash et al. (2007) found that audit fees can influence audit quality in two ways. Firstly, audit fees can influence audit quality because a high audit fee is an incentive to increase the effort of auditors, which will result in a higher audit quality. Secondly, when an auditor receive a high audit fee, the auditor will become more economically bound to his client. When the auditor is economically dependent on his clients, he will be less independent as will wish to prevent losing these clients (Hoitash et al., 2007). As result, the audit quality will decrease.
In addition, Gul et al. (2003) examined the relationship between discretionary accruals and audit fees. Gul et al. (2013, pp.444-445) tested whether there is a positive relation between discretionary accruals and audit fees, and indeed found a positive relation between discretionary accruals and audit fees.
Antle et al. (2006) investigated the joint determination of audit fees, non-audit fees and abnormal accruals. Antle et al. (2006, pp.258) found a significant, positive and robust effect of audit fees on abnormal accruals. This finding means that higher audit fees result in more acceptance of abnormal accruals. They also found that non-audit fees do not increase abnormal accruals, which contrasts with the results of several other studies (Ferguson et al., 2000; Frankel et al., 2002).
In contrast, Ashbaugh et al. (2003) did not find a significant relation between non-audit fees and discretionary accruals. Chung and Kallapur (2003) also investigated the relation between non-audit fees and discretionary accruals, and reached the same conclusion as Ashbaugh et al. (2003).
Lennox (1999b) studied the influence of non-audit services on audit quality, using the research by DeAngelo (1981) as the basis for his explanation of audit quality. As Lennox (1999b) indicates, DeAngelo (1981) argues that audit quality depends on the expertise of the auditor in discovering a misstatement and taking action for this misstatement. The relationship between non-audit services and audit quality can take different forms. Providing non-audit
10 services increases an auditor’s knowledge about a client and could therefore lead to an increase in the likelihood that the auditor discovers a misstatement. The independence of the auditor could be either increased or reduced.
2.4 Corporate governance
The phenomenon of corporate governance has become a very important topic in the recent years. The scandals of the past century, such as those at Enron, Parmalat, Worldcom and Adelphia, have put this topic firmly on the agenda (Cohen et al., 2004). New laws and regulations have been introduced with the intention of avoiding scandals in the future. An example of a new regulation is the SOX Act in the US (Cohen et al., 2004). Larcker & Tayan (2011) define corporate governance as a collection of control mechanisms that prevent or dissuade potentially self-interested managers from engaging in activities that are not in the best interest of investors and stakeholders.
The audit committee element of corporate governance has as its goal to improve and guarantee the quality of the financial statements. The audit committee does this by monitoring the financial reporting process and providing advice with respect to the appointment and termination of a firm’s external auditors (Felo, 2003). The main aim of the audit committee is safeguarding the independence of the external auditor, because this is decisive for the audit quality (Carcello & Neal, 2003). For an audit committee to be effective, it should increase the independence of the auditor, which may lead to an improvement of the quality of the financial statements.
Vera-Munez (2005) describes three functions of monitoring by an audit committee . The first function of the audit committee pertains to the audit process, and involves the selection and termination of the external auditor and the approval of all audit- and non-audit services. The second function of the audit committee focuses on financial reporting, which includes the review of the auditing process conducted by the internal and external auditor. The audit committee also reviews the standard accounting policy with regard to the financial reports. The final function of the audit committee is to install whistleblower provisions to take care of accounting or auditing issues of employees.
A few characteristics of an audit committee are the independence of the audit committee, their financial expertise and audit committee size. With respect to the independence of the audit
11 committee, it is recommended that US listed firms have an audit committee which consist of at least three independent members with a financial background (Blue Ribbon Committee, 1999). Blue Ribbon Committee (1999) argues that members of the audit committee may be considered independent if they have no relationship to the corporation that could impair their independence. The SEC (2003) require a company to disclose whether it has at least one financial expert in its audit committee. A company that does not have a financial expert must disclose this and explain why it has no financial expert in its audit committee. The SEC (2003) also states that a financial expert in an audit committee must have experience with financial statements and the ability to apply Generally Accepted Accounting Principles. This rule complies with the requirements in section 407 of the SOX Act of 2002. The size of the audit committee is another characteristic of an audit committee. Several reports recommend a minimum of three audit committee members (Blue Ribbon Committee, 1999; New York Stock Exchange, 2002; Capital Market Authority, 2006); Kalbers and Fogarty (1993) argue that a big committee achieves a higher organizational status, authority and knowledge base.
2.5 Formulation of hypotheses
A lot of studies have examined the relation between audit quality and audit fees. There were concerns by the SEC about the effect of audit fees on the quality of financial reports because of the scandals of the past century. This was the primary motivation for many of the studies about this subject. Prior research shows that the independence of the auditor can be reduced by
increasing audit fees. This indicates that the auditor will accept more earnings management from the management when the audit fees will increase. As result, the audit quality will be lower (Antle et al., 2006). In addition, Gul et al. (2003) also found a positive relation between
discretionary accruals and audit fees. However, Hoitash et al. (2007) found that the audit quality increases when the auditor receives a large audit fee. They argue that a large audit fee works as an incentive for auditors to increase their efforts.
Prior literature has also examined the effects of the SOX on the audit practice. Cohen et al. (2008) investigated the presence of earnings management during the pre-SOX and post-SOX period. They found a strong presence of accrual-based earnings management in the pre-SOX period, while in the post-SOX period they found a big drop in accruals-based earnings
12 pre-SOX period, but a significant rise in real earnings management in the post-SOX period. This indicates that there is a shift from using accrual based earnings management to real earnings management.
Li et al. (2008) investigated market reactions to events surrounding SOX and found that positive abnormal stock returns are associated with events surrounding SOX. They also found a positive relation between stock return around SOX events and the extent of earnings
management. These results indicate that investors anticipated that the more extensively firms had managed their earnings, the more SOX would limit earnings management and improve
information value of financial statements. The preceding discussion leads to the following hypotheses:
H1: There is a significant relationship between audit fees and audit quality in the pre-SOX period.
H2: There is no significant relationship between audit fees and audit quality in the post-SOX period.
The third hypothesis is directly linked to the audit committee characteristics and the relationship between audit fees and audit quality. The first characteristic of the audit committee is the size of the audit committee. The size of the audit committee is expressed in terms of the number of members who are part of the audit committee. As mentioned earlier in the literature review, the size of the audit committee may influence the audit quality. Several studies (Goh et al., 2009; Felo, 2003) found a positive relation between the size of the audit committee and the quality of the audit. On the other hand, various other studies (Abbott & Parker, 2004 and Xie et al., 2003) did not find a significant relation between audit committee size and audit quality.
The second characteristic of the audit committee is the financial expertise within the audit committee. Several studies show that an audit committee without financial experience is less likely to be strong enough to maintain audit quality (Knapp, 1987; DeZoort et al., 2002; and Turley & Zaman, 2004). Based on the findings of these studies it may be expected that the financial expertise of the members of an audit committee is positively related with audit quality.
13 The third characteristic of the audit committee is the independence of the audit
committee. Prior literature also studied the relation between audit committee independence and audit quality. Audit committees with a majority of independent members decrease earnings management (Klein, 2002; Jenkins, 2003). Based on prior literature, the following hypothesis may be formulated:
H3: Corporate governance has a positive significant influence on the relationship between audit fees and audit quality.
14
3. Research methodology
This section discusses the research methodology and tests the hypotheses presented in section two by means of a regression model. The first paragraph describes how audit quality may be measured, while the second paragraph presents the regression models. The third and final paragraph of this chapter discusses the research data and sample selection process.
3.1 Measuring of audit quality
The dependent variable in this study is audit quality, which has been extensively studied in prior research (Arruñada and Paz-Ares, 1997; Carson et al., 2013; DeAngelo, 1981). An oft-used proxy for audit quality involves discretionary accruals. Accruals are the difference between the income before extraordinary items and the net cash flow from operating activities. Discretionary accruals are accruals that can be influenced by management, whereas non-discretionary accruals are accruals that cannot be influenced by management (Healy, 1985). Total accruals can be divided into two components: the discretionary accruals and the non-discretionary accruals.
In this study, audit quality will be measured using the discretionary accruals measure from the modified Jones Model outlined in Dechow et al. (1995). The absolute value of the abnormal component is a measure for the quality of earnings. The value of discretionary accruals has an asymmetric relation with the quality of earnings (Dechow et. al., 1995). The original Jones Model assumes that there is no managerial discretion over revenues in the estimation period or the event period. The modified version of the Jones Model takes account of the fact that all changes in credit sales in the event period are the result of earnings management. The modified Jones Model (Dechow et. al., 1995) used in this study is as follows:
NDAi,t /Ai,t = (1/Ai,t-1) + (∆REVi,t – ∆RECi,t / Ai,t) + (PPEi,t)
Where:
NDAi,t = Non-discretionary accruals of firm I in year t. Ai,t-1 = Total assets at t-1.
∆REVi,t = Changes in revenues of firm I from year t to year t-1. ∆RECi,t = Changes in receivables of firm I from year t to year t-1. PPEi,t = Gross property plant and equipment of firm I in year t.
15 The following model is used for calculating residuals:
TACi,t = (1/Ai,t-1) + (∆REVi,t) + (PPEi,t) + νt
Where:
TACi,t = Total accruals scaled by lagged total assets.
The following model is used for calculating discretionary accruals:
DAi,t = (TACi,t /Ai,t-1) – (NDAi,t /Ai,t-1)
Where:
DAi,t = Discretionary accruals of firm i in year t.
3.2 Regression models
Two regression models will be used in order to test the hypotheses. Discretionary accruals are used as a proxy for audit quality. The following multivariate regression model is used to examine the first two hypotheses, H1 and H2;
DAi,t = α + β1*AUDIT FEE + β2*LEV + β3*BIG N + β4*SIZE + ε (1)
Where:
DAi,t = Discretionary accruals of firm i in year t.
AUDIT FEE = Audit fee that was paid by the client. LEV = Ratio of total liabilities to total assets in year t.
BIG N = 1 if audit firm is a Big N audit firm and 0 otherwise.
SIZE = Size of the firm based on the natural logarithm of total assets.
For the examination of hypotheses 1 and 2, the variable AUDIT FEE is used. The research model includes control variables. The first control variable is leverage (LEV). This control variable constitutes the ratio of total liabilities to total assets in year t and is also used in the study by Ashbaugh et al. (2003). The second control variable concerns the question whether a client was audited by a big four audit firm (BIG N). Several studies (Palmrose, 1988; Feroz et al., 1991) have shown that the audit quality of big four audit firms is higher than of other firms.
16 In addition, big four audit firms have invested in their reputation and will protect their reputation by providing high-quality audits (Simunic & Stein, 1987; Francis & Wilson, 1988). Thus, we include a dummy variable that was ‘1’ if the client was audited by a big four audit firm and ‘0’ if the client was not was audited by a big four audit firm. The final control variable concerns the firm size (SIZE), expressed as the natural log of total assets. This control variable was also used in several previous studies (e.g., Becker et al., 1998; Carey and Simnett, 2006; Gul et al., 2003).
The third and final hypothesis will be used in another regression model. Discretionary accruals are also used as a proxy for audit quality. The following multivariate regression model is used to examine the final hypothesis, H3;
DAi,t = α + β1*AUDIT FEE + β2*ACE + β3*(AUDIT FEE*ACE)+ β4*LEV +β5*BIG N + β6*SIZE +ε (2)
Where:
DAi,t = Discretionary accruals of firm i in year t.
AUDIT FEE = Audit fee that was paid by the client. LEV = Ratio of total liabilities to total assets in year t.
BIG N = 1 if audit firm is a Big N audit firm and 0 otherwise. SIZE = Size of the firm based on the natural logarithm of total assets
ACE = Audit committee effectiveness. An audit committee is effective when [(ACI = 1) +
(ACX = 1) + (ACS ≥ 3 = 1)]
ACI = Audit committee independence. Dichotomous with 1 if all audit committee members are
independent non-executive directors, 0 otherwise.
ACX = Audit committee financial expertise. Dichotomous with 1 if the audit committee contains
a member with financial expertise, 0 otherwise.
ACS = Size of audit committee.
In order to test hypothesis 3, the variable Audit committee effectiveness (ACE) is used to investigate the influence of corporate governance on the relationship between audit fees and audit quality. The ACE variable is a compound variable that consists of audit committee
17 independence (ACI), audit committee financial expertise (ACX) and size of audit committee (ACS). The three variables in the ACE variable are consistent with the study conducted by Zaman et al. (2011), who also used a fourth variable for the ACE, namely audit committee meetings (ACM).Because the data related to this variable were not available in the databases, it will not be used in this study. Audit committee independence (ACI) is a dichotomous variable with ‘1’ denoting that all members of the audit committee are independent non-executive directors and ‘0’ that they are not independent. Audit committee expertise (ACX) is a dichotomous variable with ‘1’ denoting at least one member of the committee having financial expertise. The size of audit committee (ACS) is a continuous variable. ACS is coded ‘1’ when the number of members in the audit committee is three or more and ‘0’ if the number of members is less than three. The
ACE variable is equal to ‘1’ when the audit committee membership comprises only independent
non-executive directors, when at least one member of the audit committee has financial expertise and the audit committee has at least three members. The control variables used in the second regression model are the same as those used in the first regression model. Figure 1 provides a schematic outline of the research design.
18
3.3 Sample selection
The sample used for this study encompasses the period between 2000 and 2012. The data for this study are collected from publicly available data of financial statements from US firms. The US have been chosen because it allows for a sufficiently broad sample size, which makes possible a powerful regression analysis that meets the requirements of this research. Some studies exclude the financial sector and the insurance companies, because their financial ratios are significantly different from those of companies in the non-financial sector, which may lead to a distorted view (Geiger et al. 2006). Another argument for excluding the financial sector relates to the fact that the financial structure of these companies is not comparable to the structure of other companies (Carey & Simnett, 2006). For these reasons, the financial sector with SICH codes 6000-6999 will be excluded from this research. The information about audit firms will be extracted from
COMPUSTAT, Audit Analytics and Risk Metrics. The sample consisted of 23,489 observations over a time period of 12 years. Incomplete observations were removed, as were 4,836
observations from the financial sector. The outliers in the sample are identified when the Z-score is higher than 3.29 or lower than -3.29 (Tabachnick & Fidell, 2013). The Z-score is a score which has been converted into the number of standard deviations that the test score is being converted above or below the mean. Thus a Z-score of -3.30 means that the test score is 3.30 standard deviations below the mean (Tabachnick & Fidell, 2013). The outliers of the variables were removed. The outliers of the variables were below the 2% of the sample of each panel.
For the examination of the first hypothesis, the time period between 2000 and 2001 (Panel A) will be studied. This hypothesis investigates the relation between audit fees and audit quality before the implementation of SOX. Since data before 2000 were not available, this study investigates the relationship between audit fees and audit quality between 2000 and 2001. The analysis of the second hypothesis will focus on the time period between 2003 and 2006 (Panel B). This hypothesis investigates the relation between audit fees and audit quality after the implementation of SOX. In order to prevent the effects of the financial crisis from influencing the results, the time period chosen is until 2006. In examining the third hypothesis, the time period chosen will be 2007 through 2012 (Panel C). This hypothesis investigates the influence of corporate governance on the relationship between audit fees and audit quality. It investigates the
19 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
time period from 2007 onwards, because corporate governance data of previous years were not available. Figure 2 provides a schematic outline of the time period of investigation.
Figure 2 Time period of investigation Hypothesis 1 (Panel A) Hypothesis 2 (Panel B) Hypothesis 3 (Panel C)
20
4. Results
This section discusses the results of this study. The first paragraph introduces the descriptive statistics and correlation coefficients. The second paragraph presents the main results of this paper and the third paragraph of this chapter discusses the sensitivity analysis.
4.1 Descriptive statistics and correlation coefficients
Table 1 shows an overview of the descriptive statistics of all panels in this research. Panel A and panel B have 5 variables and panel C has 6 variables. The mean of the absolute value of
discretionary accruals in panel A is 0.0348, in panel B 0.0474 and in panel C 0.0571. The minimum and maximum AUDIT FEE lie very far apart in each of the panels: the lowest value of
AUDIT FEE is in the hundreds and the maximum value is in the hundred millions. The variables BIG N and ACE are dummy variables that can only have the value of 0 or 1. In addition, Table 1
also presents the descriptive statistics of the variables SIZE and LEV. Table 1 Descriptive statistics
Descriptive Statistics
N Minimum Maximum Mean
Standard Deviation Panel A DA 6383 -2.20 2.22 0.0348 0.3414 AUDIT FEE 13254 275 48000000 419040.02 1393072.8510 LEV 6476 0.00 2577.00 3.0656 56.3134 SIZE 6468 -4.07 13.11 5.2564 2.8526 BIG N 6492 0 1 0.6200 0.4860 Valid N (listwise) 3623 Panel B DA 15351 -2.83 2.84 0.0474 0.3855 AUDIT FEE 58088 150 201560000 840161.96 3250621.1985 LEV 15635 0.00 9242.00 4.9265 120.5885 SIZE 15604 -4.71 13.53 5.3571 3.0073 BIG N 15663 0 1 0.6629 0.4727 Valid N (listwise) 14241 Panel C DA 25351 -1.61 1.66 0.0571 0.2713 AUDIT FEE 75568 250 162100000 1077702.79 3911877.7333 ACE 8831 0 1 0.7807 0.4138 LEV 26009 0.00 25968.97 5.2963 189.4978 SIZE 25927 -4.20 13.59 5.7727 2.9036 BIG N 26073 0 1 0.6397 0.4801 Valid N (listwise) 5625
21 Table 2 shows correlations between the variables DA, AUDIT FEE, LEV, SIZE, BIG N and
ACE. Table 2 shows that there is a significant correlation between DA and all other variables
except ACE. AUDIT FEE has a weak positive correlation with DA (0.021), which indicates that when AUDIT FEE increases, DA also increases. The same applies to the variables SIZE with DA, and BIG N with DA, which are also positive correlated. In contrast, DA has a weak and negative correlation with LEV (-0.045), which indicates that when LEV increases, DA decreases. The variables SIZE and BIG N have a strong positive significant correlation (0.622). This indicates that when the SIZE increases, it is more likely that the client is audited by a big four audit firm, which indeed makes sense. Finally, there is also a positive significant correlation between
AUDIT FEE and SIZE (0.493). This indicates that when the SIZE increases, AUDIT FEE will
also increase, which also makes sense. Table 2 Correlation coefficients
Correlations
DA AUDIT FEE LEV SIZE BIG N ACE
DA Pearson Correlation 1 Sig. (2-tailed) -
N 50996
AUDIT FEE Pearson Correlation 0.021** 1
Sig. (2-tailed) 0.000 - N 44646 158771
LEV Pearson Correlation -0.045** -0.010* 1
Sig. (2-tailed) 0.000 0.035 - N 50878 45343 51813
SIZE Pearson Correlation 0.034** 0.493** -0.035** 1
Sig. (2-tailed) 0.000 0.000 0.000 - N 50866 45264 51584 51702
BIG N Pearson Correlation 0.016** 0.244** -0.038** 0.622** 1
Sig. (2-tailed) 0.000 0.000 0.000 0.000 - N 50996 45453 51813 51702 51931
ACE Pearson Correlation -0.005 0.039** 0.057** 0.077** 0.011 1
Sig. (2-tailed) 0.734 0.004 0.000 0.000 0.403 - N 5650 5646 5630 5650 5650 8831 **. Correlation is significant at the 0.01 level (2-tailed).
22 The independent variables were also checked for multicollinearity. Multicollinearity means that when independent variables are strongly correlated, this could reduce the reliability of the result. The independent variables were checked on the Variance Inflation Factor (VIF) value and Tolerance level. The VIF values of the independent variables were below 5 and the Tolerance levels were above 0.20, which means that there is no multicollinearity between the independent variables.
4.2 Main results
The variables in this study were first checked to see whether they are normally distributed. The variables AUDIT FEE and LEV are not normally distributed.The variables AUDIT FEE and LEV were transformed with log10. Please see the appendix for the histograms of these variables.
Appendix E presents the Anova table for panel A, for which model 1 was used. The table shows that model 1 is not significant for panel A. The significant value is higher than 0.05. This means that hypothesis 1, which examines whether there is a significant relationship between audit fees and audit quality in the pre-SOX period, cannot be tested.
Appendix F presents the Anova table for Panel B, for which model 1 was also used. As the table shows, model 1 is significant for hypothesis 2, which examines whether there is a significant relationship between audit fees and audit quality in the post-SOX period. The significant value is below the 0.05. The results in appendix G show an adjusted R-square of 0.70%, which means that the total variation of the outcome is explained by the model for 0.70%. Table 3 shows the regression results of hypothesis 2, indicating that LEV is significant, which means that LEV influences DA. Thus if LEV increases by 1, DA will decrease with 0.0852. This indicates that if the leverage ratio increases, the discretionary accruals will also decrease. The other variables AUDIT FEE (-0.0053), SIZE (0.0032) and BIG N (0.0125) are not significant, which means that these variables do not influence DA. Based on Table 3, the results provide support for hypothesis 2, since they show that there is no significant relationship between audit fees and audit quality in the post-SOX period. The results of panel B can be explained by the findings of Cohen et al. (2008), who found a strong presence of accrual-based earnings management in the pre-SOX period, while in the post-SOX period they found a big drop in accruals-based earnings management. This could suggest that the enforcement of the SOX act
23 2002 has led to less earnings management by US firms and that the audit fees are not associated with the drop in accruals-based earnings management in the post-SOX period.
Table 3 Regression results Panel B
Regression Results Unstandardized Coefficients Standardized Coefficients t Sig. Model 1 B Std. Error Beta (Constant) 0.0239 0.0385 0.6194 0.5357 AUDIT FEE -0.0053 0.0085 -0.0106 -0.6298 0.5288 LEV -0.0852 0.0090 -0.0810 -9.4637 0.0000 SIZE 0.0032 0.0023 0.0244 1.4205 0.1555 BIG N 0.0125 0.0089 0.0158 1.3965 0.1626
Appendix H presents the Anova table for Panel C, for which model 2 was used. As appendix H shows, model 2 is significant for hypothesis 3. Hypothesis 3 examined whether corporate
governance has a positive influence on the relationship between audit fees and audit quality. The significance value is below the 0.05. The results in appendix I show an adjusted R-square of 1.60%, meaning that the total variation in absolute discretionary accruals is explained by the variation of the six independent variables; AUDIT FEE, ACE, AUDIT FEE*ACE, LEV, SIZE and
BIG N for 1.60%. Model 2 has a higher adjusted R-square compared with model 1. Table 4
shows the regression analysis of hypothesis 3, which reveals that all variables are significant except ACE and the moderator AUDIT FEE*ACE. This means that audit committee effectiveness does not have an influence on DA and on the relationship between audit fees and audit quality. In this model, AUDIT FEE has a significant influence on DA (0.051). So if AUDIT FEE increases by 1, DA will increase with 0.051. This indicates that higher audit fees are associated with more
discretionary accruals. SIZE, BIG N and LEV are also significant. If SIZE goes up with 1, DA decreases by 0.009. This indicates that large firms are associated with less discretionary accruals. Firms that are audited by big 4 audit firms are also associated with less discretionary accruals (-0.032). This is consistent with several studies, e.g. Palmrose (1988) and Feroz et al. (1991). Finally, firms with high leverage ratios are associated with less discretionary accruals (-0.048). Based on table 4, hypothesis 3 can be rejected, since it shows that corporate governance has no positive significant influence on the relationship between audit fees and audit quality.
24 Table 4 Regression results Panel C
Regression Results Unstandardized Coefficients Standardized Coefficients t Sig. Model B Std. Error Beta 1 (Constant) -0.176 0.036 -4.879 0.000 AUDIT FEE 0.051 0.007 0.157 6.937 0.000 ACE 0.000 0.005 0.000 0.024 0.981 LEV -0.048 0.010 -0.074 -4.878 0.000 SIZE -0.009 0.002 -0.102 -4.286 0.000 BIG N -0.033 0.008 -0.057 -4.074 0.000 2 (Constant) -0.169 0.038 -4.430 0.000 AUDIT FEE 0.051 0.007 0.157 6.907 0.000 ACE -0.008 0.013 -0.021 -0.592 0.554 AUDIT FEE*ACE 0.007 0.011 0.023 0.648 0.517 LEV -0.048 0.010 -0.074 -4.880 0.000 SIZE -0.009 0.002 -0.103 -4.310 0.000 BIG N -0.032 0.008 -0.057 -4.069 0.000 4.3 Sensitivity analysis
In order to provide additional support for the findings, hypothesis 3 was tested without the variable SIZE. This variable was excluded because it have a significant strong positive
correlation with BIG N. The following multivariate regression model was used to examine the sensitivity analysis of H3;
DAi,t = α + β1*AUDIT FEE + β2*ACE + β3*(AUDIT FEE*ACE)+ β4*LEV + β5*BIG N +ε
Where:
DAi,t = Discretionary accruals of firm i in year t.
AUDIT FEE = Audit fee paid by the client.
LEV = Ratio of total liabilities to total assets in year t.
BIG N = 1 if audit firm is a Big N audit firm and 0 otherwise.
ACE = Audit committee effectiveness. An audit committee is effective when [(ACI = 1) +
25
ACI = Audit committee independence. Dichotomous with 1 if all audit committee members are
independent non-executive directors, 0 if they are not.
ACX = Audit committee financial expertise. Dichotomous with 1 if the audit committee contains
a member with financial expertise, 0 if not.
ACS = Size of audit committee.
Appendix J presents the Anova table for the sensitivity analysis of hypothesis 3. It reveals that the sensitivity analysis of hypothesis 3 is significant. The significant value is below 0.05. The results in appendix K show an adjusted R-square of 1.30%, meaning that the total variation of the outcome is explained by the model for 1.30%. The adjusted R-square of the sensitivity analysis of hypothesis 3 is lower than the adjusted R-square of the normal regression of
hypothesis 3. Table 5 shows the regression of the sensitivity analysis of hypothesis 3. It indicates that all variables are significant except ACE on the moderator AUDIT FEE*ACE. In other words, the findings of the sensitivity analysis are the same as the primary results and hypothesis 3 is rejected. In conclusion, the sensitivity analysis of hypothesis 3 is consistent with the primary results.
Table 5 Regression results sensitivity analyses
Regression Results Unstandardized Coefficients Standardized Coefficients t Sig. Model B Std. Error Beta
1 (Constant) -0.0404 0.0278 -1.4506 0.1470 AUDIT FEE 0.0261 0.0047 0.0808 5.5728 0.0000 LEV -0.0595 0.0097 -0.0886 -6.1551 0.0000 ACE -0.0009 0.0050 -0.0023 -0.1761 0.8602 BIG N -0.0378 0.0079 -0.0664 -4.7891 0.0000 2 (Constant) -0.0337 0.0307 -1.0991 0.2718 AUDIT FEE 0.0258 0.0047 0.0799 5.4835 0.0000 LEV -0.0596 0.0097 -0.0888 -6.1654 0.0000 ACE -0.0073 0.0133 -0.0196 -0.5501 0.5823 AUDIT FEE*ACE 0.0059 0.0114 0.0187 0.5221 0.6017 BIG N -0.0378 0.0079 -0.0663 -4.7867 0.0000
26
5. Conclusion
This study has investigated the influence of corporate governance on the relationship between audit fees and audit quality in the US. Carson et al. (2013) defined audit quality based on the competence and the independence of an auditor. DeAngelo (1981) made a distinction in audit quality between the probability that an auditor discovers material errors and takes no action, and the situation where an auditor discovers a material error and takes action. The proxy used for audit quality in this study was discretionary accruals.
The first part of this study investigated the relationship between audit fees and audit quality in the pre-SOX period and the post-SOX period. Audit fees were defined as fees that a company pays to an external auditor in exchange for performing an audit (Arens et al., 2012). This study hypothesized that audit fees have a significant relationship with audit quality in the pre-SOX period, and no significant relationship in the post-SOX period, as was formulated in hypothesis 1 and 2. With respect to hypothesis 1, the regression model was not significant, which means that hypothesis 1 was not tested. However, the results did support the second hypothesis. In other words, in the post-SOX period, 2003 through 2006, no significant relationship was found between audit fees and audit quality. The final hypothesis, H3, examined whether
corporate governance has a positive influence on the relationship between audit fees and audit quality. Larcker & Tayan (2011) defined corporate governance as a collection of control mechanisms that prevent or dissuade potentially self-interested managers from engaging in activities that are not in the best interest of investors and stakeholders. The results provided no support for hypothesis 3. Audit committee effectiveness was not found to influence DA and the relationship between audit fees and audit quality. In this model, AUDIT FEE does have a significant influence on DA. For additional support of the findings of hypothesis 3, hypothesis 3 was tested without the variables SIZE and BIG N. These sensitively analyses provided no support for hypothesis 3 either and were consistent with the primary results.
This study has various limitations. Firstly, the sample size for hypothesis 1 was relatively limited. In future studies, a bigger sample size could be used and perhaps a cross-country
analysis may be worthwhile as well. In addition, several variables were chosen to examine audit quality in this study. In future research, it may be useful to opt for a different set of variables. This study used discretionary accruals as a proxy for audit quality; further research could use a
27 different proxy for audit quality. For instance, future studies could focus on extending prior research on going concern opinions, frauds or restatements (Geiger & Raghunadan, 2002; Carcello & Nagy, 2004; Stanley & Dezoort, 2007).
28
Reference
Abbot, L.J., Parker S. & Peters, G.F., (2004). Audit committee characteristics and restatements.
Auditing: A Journal of Practice & Theory, 23, (1), 69-87.
Antle, R., Gordon E., Narayanamoorthy G. & Zhou L., (2006). The joint determination of audit fees, non-audit fees and abnormal accruals. Review of Quantitative Finance and Accounting,
27, 235-266.
Arens, A.A., Elder R.J. & Beasley M.S., (2011). Auditing and Assurance services an integrated approach. Harlow: Pearson.
Arruñada, B., (1997). Mandatory rotation of company auditors: A critical examination.
International Review of Law and Economics, 17, (1), 31-61.
Ashbaugh H., Lafond, R. & Mayhew B. W., (2003). Do non-audit services compromise auditor independence? further evidence. The Accounting Review, 78, (3), 611-639.
Audit Analytics, (2007).“A comparison of audit fee changes experienced by SOX section 404 filers and non-filers. Compliance Week, Available at:
http://www.auditanalytics.com/doc/report-af-20070223.pdf
Barnier, M. (2010). Groenboek: Beleid inzake controle van financiële overzichten: lessen uit de crisis. Brussel: Europese Commissie.
Beck, P. J., Frecka, T. J. & Solomon, I., (1988a). A model of the market for MAS and audit services: knowledge spillovers and auditor-auditee bonding. Journal of Accounting Literature,
7, 50-64.
Becker, C. L., Defond, M. L., Jiambalvo, J., & Subramanyam K. R., (1998). The Effect of Audit Quality on Earnings Management. The Effect of Earnings Management. Contemporary
Accounting Research, 15, (1), 1-24.
Blue Ribbon Committee (1999). Report and recommendations of the Blue Ribbon Committee on improving the effectiveness of corporate audit committees. The Business Lawyer, 54, (3), 1067-1095.
Capital Market Authority (2006). The code of corporate governance in the Kingdom of Saudi Arabia. Retrieved November 30, 2011
Carcello, J.V. & Neal T.L., (2003). Audit committee characteristics and auditor dismissals reporting following “new” going-concern reports. The Accounting Review, 78, (1), 95-117.
29 Carcello, J.V., & Nagy A.L., (2004). Audit firm tenure and fraudulent financial reporting.
Auditing, 23, 55-69.
Carey, P., & Simnett R., (2006). Audit partner tenure and audit quality. The accounting review.
81, (3), 653-676.
Carson, E., Fargher, N., Marshall, A., Lennox, C., Raghunandan, K., & Willekens, M., (2013). Auditor reporting on going-concern uncertainty: a research synthesis. A Journal of Practice &
Theory American Accounting Association, 32, 353–384.
Catanach, A., & Walker, P., (1999). The international debate over mandatory auditor rotation: A conceptual research framework. Journal of International Accounting, Auditing & Taxation, 8, 43-66.
Chung, H. & Kallapur, S., (2003). Client importance, non-audit services, and abnormal accruals.
The Accounting Review, 78, 931-955.
Cohen, D.A., Dey, A. & Lys T.Z., (2008). Real and accrual-based earnings management in the pre- and post-Sarbanes-Oxley periods. The Accounting Review, 83, (3), 757-787.
Cohen, J., Krishnamoorthy, G. & Wright, A., (2004). The Corporate Governance Mosaic and Financial Reporting Quality. Journal of Accounting Literature, 23, 87-152.
Craswell, A., Stokes D.J. & Laughton J., (2002). Auditor independence and fee dependence.
Journal of Accounting and Economics, 33, 253-275.
Cullilan, C., (2004). Enron as a symptom of audit process breakdown: can the Sarbanes-Oxley act cure the disease? Critical Perspectives on Accounting, 15, 853-864.
DeAngelo, L.E., (1981). Auditor size and audit quality. Journal of Accounting and Economic, 3,
(3), 183-199.
Dechow, P., Sloan R. & Sweeney A., (1995). Detecting earnings management. The Accounting
Review, 70, 193-225.
DeZoort, F.T. & Salterio S., (2001). The effects of corporate governance experience and
financial reporting and audit knowledge on audit committee members judgments. Auditing: A
Journal of Practice & Theory, 20, 31–47.
Felo, A.J., Krishnamurthy S. & Solieri S.A., (2003). Audit committee characteristics and the perceived quality of financial reporting: an empirical analysis. Working paper, School of Graduate Professional Studies.
30 Ferguson, M., Seow, G. & Young D., (2000). The effect of non-audit services on audit quality.
Working Paper. The University of Connecticut.
Feroz, E., Park K. & Pastena V., (1991). The financial and market effects of the SEC’s
accounting and auditing enforcement releases. Journal of Accounting Research, 29, 107-142. Francis, J. & Wilson E., (1988). Auditor changes: a joint test of theories relating to agency costs
and auditor differentiation. The Accounting Review, 63, (4), 663-682.
Frankel, R.M., Johnson M.F. & Nelson K.K., (2002). The relation between auditors’ fees for non-audit services and earnings quality. The Accounting Review, 77, 71-105.
Geiger, M.A. & Raghumandan K., (2002). Auditor tenure and audit reporting failures. Auditing:
A journal of practice and theory, 21, 67-78.
Geiger, M., Raghunandan, K. & Rama, D., (2006). Auditor decision making in different litigation environments: The private securities litigation reform act, audit reports and audit firm size. Journal of Accounting and Public Policy, 25, (3), 231-252.
Goh, B.W., (2009). Audit committees, boards of directors, and remediation of material weaknesses in internal control. Contemporary accounting research, 26, (2), 549- 579. Gul, F.A., Chen C.J.P., & Tsui J.S.L., (2003). Discretionary accounting accruals, managers’
incentives, and audit fees. Contemporary Accounting Research, 20, (3), 441-464. Hay, D., Knechel, R. & Li V., (2006). Non-audit services and auditor independence: New
Zealand evidence. Journal of business finance & accounting, 33, (5-6), 715-734. Healy, P.M., (1985). The effect of bonus schemes on accounting decisions. Journal of
Accounting and Economics, 7, 85-107.
Hoitash, R., Markelevich A. & Barragato C.A., (2007). Auditor fees and audit quality.
Management Auditing Journal, 22, (8), 761-786.
Kalbers, L.P. & Fogarty, T.J., (1993). Audit committee effectiveness: an empirical investigation of the contribution of power. Auditing: A Journal of Practice and Theory, 12, 24–49.
Klein, A., (2002). Audit committee, board of director characteristics and earnings management.
Journal of Accounting and Economics, 33, 375–400.
Knapp, M. C., (1987). An empirical study of audit committee support for auditors involved in technical disputes with client management. The Accounting Review, 62, (3), 578.
31 Iyer, G.S., Iyer, V.M. & Mishra B.K., (2003). The impact of non-audit service fee disclosure
requirements on audit fee and non-audit service fee in the United Kingdom: an empirical analysis. Advances in Accounting, 20, 127-140.
Jenkins, N., (2003). Auditor independence, audit committee effectiveness, and earnings management. Working paper, Washington University in St Louis.
Larcker, D. & Tayan B., (2011). Corporate governance matters: a closer look at organizational choices and their consequences. New Jersey: Pearson Education
Lennox, C., (1999a). The relationship between auditor accuracy and auditor size: an evaluation of reputation and deep pockets arguments. Journal of Business, Finance and Accounting, 26, 779-806.
Lennox, C.S., (1999b). Non-audit fees, disclosure and audit quality. The European Accounting
Review, 8, (2), 239-252.
Li, H., Pincus M. & Rego S.O., (2008). Market reaction to events surrounding the Sarbanes-Oxley Act of 2002 and earnings management. Journal of Law and Economics, 51, (1), 111-134.
New York Stock Exchange (2002). Corporate accountability and listing standards committee Report. New York.
Palmrose, Z., (1988). An analysis of auditor litigation and audit service quality. The Accounting
Review, 63, (1), 55-73.
Pong, C.M., & Whittington, G., (1994). The determinants of audit fees: some empirical models.
Journal of Business Finance & Accounting, 21, (8), 1071-1095.
Pratt, J., & Stice, J.D., (1994). The effect of client characteristics on auditor litigation risk
judgments, required audit evidence, and recommended audit fees. The Accounting Review, 69,
(4), 639-656.
Scott, W. R., (2009). Financial Accounting Information. 5th edition. Pearson Education: Canada. Securities and Exchange Commission (SEC) (2000). Final Rule: Revision of the Commission's
Auditor Independence Requirements. Washington, D.C.: Government Printing Office. Securities and Exchange Commission (SEC) (2003). Release Nos. 33-8177; 34-47235
Simunic, D.A., (1980). The pricing of audit services: theory and evidence. Journal of Accounting
32 Simunic, D., (1984). Auditing, consulting, and auditor independence. Journal of Accounting
Research, 22, 679-702.
Simunic, D. & Stein M., (1987). Product differentiation in auditing: auditor choice in the market for unseasoned new issues. The Canadian certified general accountants research foundation. Stanley, J.D. & Dezoort T., (2007). Audit firm tenure and financial restatements. Journal of
Accounting and public policy, 26, 131-159.
Tabachnick, B.G. & Fidell, L.S, (2013). Using multivariate statistics. California State University-Northridg.
Turley, S. and Zaman M., (2004). The corporate governance effects of audit committees.
Journal of Management and Governance, 8, 305–32.
Vera-Munez, S.C., (2005). Corporate governance reforms: redefined expectations of audit committee responsibilities and effectiveness. Journal of Business Ethics, 62, (2), 115-127. Vieru, M. & Schadewitz. H., (2010). Impact of IFRS transition on audit and non-audit fees:
evidence from small and medium-sized listed companies in Finland. The Finnish Journal of
Business Economics, 1, 11–41.
Whisenant, S., Sankaraguruswamy S., & Raghunandan K., (2003). Evidence of Joint
Determination of Audit and Non-Audit Fees. Journal of Accounting Research, 41, (4), 721-744.
Xie, B., Davidson W.N. & DaDalt, P.J., (2003). Earnings management and corporate
governance: the role of the board and the audit committee. Journal of Corporate finance, 9,
(3), 295-316.
Zaman, M., Hudaib, M. & Haniffa, R., (2011).Corporate governance quality, audit fees and non-audit services fees. Journal of Business Finance & Accounting, 38, (1-2), 165–197.
33
Appendix
Appendix A
Histogram AUDIT FEE
Appendix B
34
Appendix C
Histogram LEV
Appendix D
35 Appendix E Anova Panel A Anova Model 1 Sum of Squares df Mean Square F Sig. Regression 0.693 4 0.173 2.164 0.071b Residual 288.463 3600 0.080 Total 289.157 3604 Dependent Variable: DA
b. Predictors: (Constant), BIG N, LEV, SIZE, AUDIT FEE
Appendix F Anova Panel B Anova Model 1 Sum of Squares df Mean Square F Sig. Regression 14.896 4 3.724 27.314 0.000b Residual 1921.421 14093 0.136 Total 1936.317 14097 Dependent Variable: DA
b. Predictors: (Constant), BIG N, LEV, AUDIT FEE, SIZE
Appendix G
Model summary Panel B
Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 0.088a 0.008 0.007 0.369
36 Appendix H Anova Panel C Anova Model Sum of Squares df Mean Square F Sig. 1 Regression 1.909 5 0.382 19.266 0.000b Residual 111.326 5619 0.020 Total 113.234 5624 2 Regression 1.917 6 0.319 16.123 0.000c Residual 111.317 5618 0.020 Total 113.234 5624 Dependent Variable: DA b. Predictors: AUDIT FEE, ACE, LEV, SIZE, BIG N
c. Predictors: AUDIT FEE, ACE, AUDIT FEE*ACE, LEV, SIZE, BIG N
Appendix I
Model summary Panel C
Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 0.130a 0.017 0.016 0.141 2 0.130b 0.017 0.016 0.141 a. Predictors: AUDIT FEE, ACE, LEV, SIZE, BIG N
b. Predictors: AUDIT FEE, ACE, AUDIT FEE*ACE, LEV, SIZE, BIG N
Appendix J
Anova sensitivity analyses
Anova Model Sum of Squares df Mean Square F Sig. 1 Regression 1.533 4 0.383 19.284 0.000b Residual 111.701 5620 0.020 Total 113.234 5624 2 Regression 1.539 5 0.308 15.480 0.000c Residual 111.696 5619 0.020 Total 113.234 5624 Dependent Variable: DA
b. Predictors: (Constant), ACE, AUDIT FEE, LEV, BIG N
37
Appendix K
Model summary sensitivity analyses
Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 0.116a 0.014 0.013 0.141 2 0.117b 0.014 0.013 0.141
a. Predictors: (Constant), ACE, AUDIT FEE, LEV, BIG N
