Master thesis
Rijksuniversiteit Groningen
High Quality, Good or Bad?
A global study about the relation between audit quality, corporate tax avoidance and CSR disclosure
PERSONALIA
Master thesis Accountancy & Controlling, track Accountancy
Name Joey A. Huisman
Email j.a.huisman.2@student.rug.nl Faculty Economie en Bedrijfskunde Student number s2811650
First Supervisor Dr. N. (Nazim) Hussain, PhD Second Supervisor Dr. W.G. (Wilmar) de Munnik
Date 23 januari 2017
Word count 8,534
Internship Deloitte, Amsterdam
Counselor Deloitte T. (Thomas) Moeke, Msc RA
ABSTRACT
This study broadens the perspective on the role of the auditor in relation with the behavior and strategic choices of a firm with regard to tax avoidance and Corporate Social Responsibility (CSR). Legitimacy theory and agency theory are used to explain the impact of audit quality on the extent of firm’s tax avoidance and CSR. This study uses the Big4 proxy for audit quality and uses two types of effective tax rates (ETR) to measure tax avoidance. CSR is measured by the CSR disclosure score from Bloomberg. This study finds strong evidence by means of three multiple regression models with data covering a period of ten years. The sample consists of a worldwide list of firms and the results are highly significant and robust. The results indicate that firms with a Big4 auditor engage more in tax avoidance and more in CSR disclosure than firms with a non-Big4 auditor. This finding implies that bigger firms are using their social practices (CSR) in order to camouflage their unethical accounting practices (tax avoidance), by means of an entrenchment strategy.
KEY WORDS
Audit quality, tax avoidance, ETR, CSR disclosure, legitimacy, agency theory, entrenchment strategy
TABLE OF CONTENTS
PREFACE 3
I. INTRODUCTION 4
II. LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT 6
AUDIT QUALITY,TAX AVOIDANCE AND AGENCY THEORY 7
AUDIT QUALITY AND CSR 9
III. RESEARCH METHOD 11
SAMPLE DESIGN 12
RESEARCH DESIGN 13
DESCRIPTIVE ANALYSIS 15
CORRELATION MATRIX 16
IV. ANALYSIS, FINDINGS AND ROBUSTNESS CHECKS 19
MAIN ANALYSIS AND FINDINGS 19
ROBUSTNESS CHECK:BIG6 AUDIT FIRMS 21
ROBUSTNESS CHECK:ESG DATA FROM ASSET 4 22
ROBUSTNESS CHECK: EXCLUDING FINANCIAL FIRMS 22
V CONCLUSION 23
VI REFERENCES 26
3
Preface
The past five months I have worked on this Master thesis. Looking back at the moment I needed to choose a subject for this thesis, I am sure I picked the right one. I found it very interesting to study the topics ‘tax avoidance’ and ‘Corporate Social Responsibility’. Especially because tax avoidance is a hot topic all over the world and, in addition, the Panama Papers made me even more enthusiastic about this topic. I involved the auditor to my study because I am following the Master to become one. I am going to work at a Big4 audit firm and despite I know those firms are the best possible option to learn and develop as much as possible, I really wanted to investigate whether the Big4 also leads to positive outcomes for firm’s practices. On the other hand this study makes me aware of the decision to engage in tax avoidance or not and I believe everybody avoids tax to some extent. We are all seeking for ways to pay as less tax as possible, isn’t it? So, frequently I am wondering: to what extent tax avoidance is ethical and from which point tax avoidance becomes unethical. I guess the answer is not that simple and therefore a lot of research needs to be done before we understand the complex mechanisms that clarify the determinants of tax avoidance and the interference of CSR.
I could never write this thesis without help of others, so I am very grateful to everybody who helped me during my thesis period. Of course I want to thank my supervisor Nazim Hussain for all the help, the feedback and useful suggestions. I really liked working with you and enjoyed the times you asked: how you doing? (referring to Joey from the friends series). Besides that, I want to thank Thomas Moeke as my supervisor at Deloitte. You made me feel welcome at Deloitte and took a lot of time to support me with my thesis and also with some practical issues. In addition I want to thank all other colleagues at Deloitte Regio audit for the good times, everybody was really helpful. Sjoerd Verhoogt, Bastiaan Oosterveld and Liesbeth Keulers were my fellow trainees at Deloitte and I want to thank them for all the good times we had and the help with statistical, writing and layout matters. Special thanks to my girlfriend Petra Grabundzija for your support and effort to improve my thesis. Last but not least, I want to thank my family for all their support and guidance.
Ede, 23th of January
I.
Introduction
In August 2016, the European Commission (EC) accused Apple of avoiding tax by means of an unlawful tax scheme. The tax scheme of Apple is labeled as unlawful because the commission concluded that Ireland granted illegal tax benefits to Apple (Financial Times, 2016a). Consequently the Financial Times stated that Apple’s effective tax rate (ETR) was less than 1% for a business unit in Ireland. This ETR is extremely low for a business unit located in a country with a corporate tax rate of 12.5%. The EC states that Apple’s avoided tax amounts of €13 billion (The Guardian, 2016). The EC’s ruling became worldwide a ‘hot topic’. In response, Apple and Ireland are preparing appeals separately against the ruling in the European court (Financial Times, 2016a). Besides, CEO’s from leading companies in the US and the US Treasury are supporting Apple in her battle against the EC. The US Treasury states: “commission’s actions could threaten to undermine foreign investment, the business climate in Europe, and the important spirit of economic partnership between the US and the EU” (Financial Times, 2016b). Other organizations fear for a ruling from the EC as well. For example, McDonalds and IKEA are in an ongoing investigation regarding tax avoidance. Apple is not the only well-known firm that is being accused for tax avoidance. The head of Google UK and top managers from Starbucks and Amazon needed to appear in front of the Public Accounts Committee (PAC) concerning to the issue of tax avoidance (BBC, 2012). The Dutch government granted a special tax deal for Starbucks for its royalty payments from the UK. In addition, Google and Amazon admitted they used European tax jurisdictions that were in their favor for their UK businesses.
Elaborating on the role of professions in scandals, auditors’ role in broader accounting scandals are often criticized and condemned. For example, Arthur Anderson, formal auditor of Enron was found guilty for fudging the financial accounts and felt apart after the scandal. In another case the auditor of Bernie Madoff was sentenced to prison. Looking at the tax scandals from above, it is interesting to know which role the auditor is playing in tax scandals and what influence can they have on firm’s tax accounts. The recent scandals about the unethical corporate behavior (tax avoidance practices) have intrigued the interest of academics as well as policy makers to investigate and know determinants and consequences of such corporate practices.
Moreover, academics endeavors to find a link between tax avoidance and ethical corporate behavior in the form of CSR. This study examines the auditor as a determinant of unethical corporate behavior (tax avoidance) and ethical corporate behavior (CSR). Auditors are expected
to assess the validity and reasonableness of the recognition and disclosure of tax-related items in the financial statements (Kanagaretnam, Lee, Lim & Lobo, 2016). An aggressive tax strategy that is applied by the firm is likely to increase the litigation risk for the auditors. The reason for this is that shareholders are holding the auditor responsible for tax-related flaws in the financial statements (Donohoe and Knechel 2014). To clarify, shareholders can be exposed to reputational risk when the tax authority investigates the tax practices performed by their executives. So, in this situation, executives act in a sub-optimal manner. This kind of reputational risk can be seen as an agency problem. Shareholders could use the auditor to react on this agency problem between them and their executives, through auditors verifying the executives for its tax aggressiveness. Hoopes, Mescall and Pittman (2012) provide supporting evidence in their study that examines the role of Internal Revenue Service (IRS) audit. They find that when US firms fear for an IRS audit in the coming year (T+1), these firms undertake less aggressive tax positions for that specific year (T=0). Apparently, firms find it unpleasant to experience pressure from an audit on their tax accounts. So, a strong monitoring audit can be seen as an instrument to reduce suboptimal behavior in tax practices. Big4 auditors are capable of conducting such a strong audit because of their high quality audit. Hence, auditors may directly influence the extent of tax avoidance (Kanagaretnam et al., 2016) by delivering a high degree of audit quality.
In addition, I involve the roll of CSR into this study as ethical corporate behavior. CSR is seen as a form of voluntary disclosure that has significant consequences for the issuing firms (Chen, Srinidhi, Tsang & Yu, 2016). Firms can engage in tax avoidance and meanwhile look socially responsible due to CSR disclosure. It is paradoxical to do both, given that paying taxes is a significant corporate duty to society (Preuss, 2010). Auditors have to deal with CSR as well. Chen et al. (2016) found that CSR performance is embedded in the audit risk and is assessed by the auditor. Thus, when firms have a high CSR performance, the audit fee will be lower in comparison with firms with low CSR performance. In addition, Chen et al. (2016) suggests that firms can be advised or urged to make more CSR disclosure by their auditor. Moreover, auditors may act as watchdogs for firm’s ethical behavior. Ethical behavior is one of the functions of CSR (Deigh, Farquhar, Palazzo & Siano, 2016). In my study I use CSR disclosure as a measure for CSR, hence when firms disclose their CSR activities/outcomes, stakeholders and the public can assess firm’s ethical behavior. Besides that, CSR disclosure is non-financial information and can be important for stakeholders and thus be important for the auditor. Auditors perform audits to give insurance about firm’s financial and non-financial information in order to
strengthen the quality of the communication between the firm and its stakeholders, mainly its investors. Hence, auditors find it important that firms issue CSR disclosures. Considering the importance of ethical behavior and communication towards investors/public, auditors seem to have the same interest as stakeholders when it comes to CSR disclosures. Auditors operate in this way as watchdogs for firms when they are urging to issue more in CSR disclosure.
Involving both tax avoidance and CSR is this study will help researchers to better understand the determinants on tax avoidance and CSR. Furthermore, to my knowledge this study is the first study regarding the impact of audit quality on tax avoidance and CSR on a global scale. Therefore it provides a better understanding of auditor’s role in tax avoidance and CSR worldwide. Other studies use auditor-provided tax services (APTS) to proxy for audit quality (e.g. Gleason & Mills, 2011; McGuire, Omer & Wang, 2012). I follow Kanagaretnam et al. (2016) using Big4 as proxy for audit quality. Other than Kanagaretnam et al. (2016), this study range the ETR outcomes between zero and one, instead of range them by top quantile of country-industry combination in binary numbers. The results are significant and robust due to a unique set of independent and dependent variables. Moreover, the results provide a broad global perspective on audit quality covering 64 countries over all continents.
Concluding I claim that the auditor is related to corporate tax avoidance and CSR and the role of the auditor will be more important in the near future. Therefore it is relevant to ask that: What is the impact of audit quality on the extent of firm’s Corporate Tax Avoidance and Corporate Social Responsibility? During this study the above question will be the focus. In the next section the relevant literature will be discussed and both hypotheses will be developed. In the third section the research methodology will be addressed. In the fourth section the analysis, results and the robustness checks will be described. Thereafter, the conclusion will be discussed in section five.
II.
Literature review and hypothesis development
From 1990 till 2012 the effective tax rate of multinational and domestic firms systematically decreased every year (Dyreng, Hanlon, Maydew & Thornock 2014). In addition, firms seems to be able to avoid tax over a period of ten years (Dyreng, Hanlon and Maydew, 2008). Avoiding corporate tax on the long term must concern the governments because tax is a substantially amount for the government incomes. Academics are trying to find evidence for determinants concerning corporate tax avoidance. Some academics attempted to involve and relate CSR to tax avoidance (Sikka, 2010; Lanis & Richardson, 2015; Hoi, Wu & Zhang, 2013;
Watson, 2015; Huseynov & Klamm, 2012; Davis, Guenther, Krull & Williams, 2016; Dowling 2014). Discussing these papers provides a broad image of the existing literature. Evidence from Davies et al. (2016) suggests that influential stakeholders of social responsible firms do not see the payment of corporate taxes as complementing CSR activities. In addition, firms are inclined to make promises of responsible actions and choices, but then actually engage in tax avoidance and evasion (Sikka, 2010). Thus firms appear being social responsible by engaging in CSR activities, but in fact, they are not due to tax avoidance. Regarding Hoi et al. (2013), firms with irresponsible CSR activities, engage more in tax avoidances. Watson (2015) investigates the impact of pretax earnings on the relation between tax avoidance and CSR. He founds that a lower extent of CSR is positively related to tax avoidance when firms do have a low earnings performance. In addition, Lanis & Richardson (2015) find that firms with a high level of CSR performance is negatively related to tax avoidance.
Unfortunately, evidence from scientific research about the relation between tax avoidance and CSR is ambiguously. Due to the ambiguousness it is relevant to examine the impact of one firm characteristic on both phenomena separately. Evidence should provide a better understanding on the role of the auditor on tax avoidance and CSR and its influence on corporate behavior. This study finds evidence by means of examine the impact of firm’s auditor on the two phenomena: tax avoidance and CSR.
Audit Quality, Tax Avoidance and Agency Theory
DeAngelo (1981) claims that auditor size is a factor for audit quality, therefore Big4 audit firms have more incentive to ensure that the financial statements represents valid outcomes. The reason for this is that auditors with more clients than others, have more to lose if the auditor does not report the misstatements they found in the financial statements. The auditor can lose face or quasi-rents from their current clients, therefore auditors are under pressure to deliver an audit of high quality. Quasi-rents are the potential (high) start-up costs for auditing a new client, and thus are saved costs while maintaining their clients. DeAngelo (1981) refers to the ‘collateral aspect’ what results in more audit quality at larger audit firms in comparison with smaller audit firms. In addition, previous studies argue that Big4 auditors are positively related to higher financial reporting quality in public firms around the world (Khurana and Raman 2004; Choi, Kim, Liu & Simunic 2008; Francis and Wang 2008; Kanagaretnam, Lim & Lobo, 2010). The study of Kanagaretnam et al. (2016) provides a strong evidence for a negative relation between audit quality and the likelihood of tax aggressiveness. They used the Big4 proxy for audit quality and tax avoidance was measured as “the difference between the tax on
pre-tax income computed at the home-country statutory corporate tax rate and the taxes actually paid, expressed as a percentage of pre-tax income” (Kanagaretnam, 2016 p. 17). After the tax avoidance was measured, the researchers compared the Big4 dummy with a tax avoidance indicator when it is in the top quantile of each country-industry combination. This study is going a step further with a measure for tax avoidance on a broader scale which will be discussed in detail in the research method.
On the other hand, there are several studies that address a negative role of accountancy firms relating to tax aggressiveness. For example, tax aggressiveness can be explained by the influence of the financial profession, in particular, the Big4 accountancy firms (Sikka & Willmott, 2013). These firms are developing tax schemes in order to avoid tax for bigger firms. Accountancy firms can provide tax services to improve client’s tax management by means of paying less tax. The reason for firms buying tax services can be found in the enterprise culture. The enterprise culture is producing negative effects for both the firms and major accountancy firms due to the willingness to increase their profits through, inter alia, tax avoidance (Sikka 2007). Addison & Mueller (2015) also criticized the role of large accountancy firms in their paper. They examined an investigation by PAC inquiry about tax avoidance. The tax services or tax planning of the Big4 is seen as an arrangement of “self-interested, unethical gain and based on substantial knowledge and information advantages” (Addison & Mueller, 2015 p. 1282). Furthermore, McGuire et al. (2012) found that the external audit firm’s overall expertise is associated with greater tax avoidance. McGuire et al. (2012) suggests: “that overall experts are able to combine their audit and tax expertise to develop tax strategies that benefit clients from both a tax and financial statement perspective” (p. 978). During this study they examine a sample of firms with tax fees from their audit firm (auditor provided tax services). Sikka & Hampton (2005) addressed in their paper that accountancy firms enable firms to avoid tax. With the help of the financial and political resources, the accountancy firms are able to sell tax avoidance schemes. In order to expose tax avoidance schemes, governments tries to curb the tax avoidance by means of inquiries, but due to globalization and technological changes it becomes easier to avoid paying taxes. Moreover, selling tax avoidance schemes by accountancy firms have led to a skewed distribution of income and wealth and decreases the purchasing power in the UK (Sikka, 2015).
The criticisms from abovementioned studies are related to an accountancy firm as a whole and mainly say something about the impact of the tax consultants and their tax services. Contrary to this view, the auditor could respond to certain conflicts between shareholders and executives.
Managers and executives often avoid paying tax by (aggressive) tax planning, whilst corporate taxes have an important role when it comes to funding public goods (Sikka, 2010) and is considered by the public as an irresponsible and illegitimate activity (Lanis & Richardson, 2013). Nevertheless, the firm risks public concerns and possibly reputation damage while executives engage in tax avoidance. When executives are making choices in tax decisions by their own, the shareholders of the firm are exposed to reputational risk. Because the shareholders cannot monitor the executives constantly, executives may be more aggressive than they supposed to be concerning tax avoidance. Some consequences could be: their firm can be negatively in the news, the public concerns can raise and tax authorities (e.g. the EC) can start an investigation to its tax practices. The above described problem can be seen as an agency problem between the owners and executives. So, an agency problem could occur when managers and shareholders discuss the cost and benefits of tax avoidance separately (Desai & Dharmapala, 2006; Armstrong, Blouin & Larcker, 2015). This means, executives do have the ability to act in self-interest and with sub optimal behavior regarding the owners. The auditor can reduce the problem by delivering audit quality through auditing the tax accounts. Studies concerning the auditor and the impact on tax avoidance is scant. This study extent the literature by means of focusing on the auditor’s role solely. Reviewing the above literature, I expect to see a negative relation between audit quality and tax avoidance.
H1: Audit quality is negatively related with tax avoidance.
Audit quality and CSR
CSR and tax avoidance are often been related to each other through legitimacy theory (Fallan, 2015; Lanis & Richardson, 2013; Preuss, 2010). Given that academics putt a lot of effort to study the relation between both phenomena it is relevant to include CSR as a dependent variable, next to tax avoidance. This study justifies the relation between audit quality and CSR with legitimacy theory. “Legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions” (Suchman, 1995 p. 574). Legitimacy consist of two different approaches, one the strategic and one the institutional. The strategic approach assumes a high level of managerial interference in the legitimation process. In contrast, the institutional approach depict legitimacy as a set of constitutive beliefs.
In this study I will focus on the strategic approach of legitimacy taking into account that managers and executives make strategic choices regarding CSR. To support this assumption, Silberhorn & Warren (2007) provide some evidence about the major drivers of CSR for large
British and German firms. Their study suggests that firms see their CSR as a comprehensive strategy for interacting with stakeholders. According Prior, Surroca & Tribo (2008), executives apply the entrenchment strategy when they have an incentive to improve socially image, by CSR activities, caused by unethical accounting practice (manipulating earnings). Moreover, several studies examined the influence of entrenchment strategy as well and found that some poor accounting practices are camouflaged by the use of social practices (Cespa & Cestone, 2007; Martínez-Ferrero & García-Sánchez, 2015). Their findings correspond with the conclusion that firms report CSR to camouflage their unethical accounting practices (e.g. earnings management). Hence, entrenchment strategy could be part of the comprehensive strategy executives are using to interact with stakeholders. In this way, executives can camouflage their tax avoidance via CSR reporting. Thus, firms can improve their social image by reporting more CSR, which is a form of voluntary corporate nonfinancial disclosure. Non-financial information provides broader information than Non-financial disclosures and, thus, CSR reports can be informative to (potential) shareholders (Ramanna, 2013). In addition, positive CSR information can represent a high extent of management’s responsibility to all stakeholders (Hoi et al., 2013; Deng, Kang & Low, 2013).
To elaborate the importance of the auditor in this matter, I suggest that auditors see the need of proper and adequate communication between firms and stakeholders by means of disclosures. Adequate communication between firms and stakeholders are in line with the core business of an auditor, because an auditor gives insurance over financial and non-financial statements in order to adequately inform investors and the society. At the same time, legitimacy plays an important role in CSR disclosures for both the auditor and the firm. Supporting evidence can be found in Power’s (2003) critical research paper. Power discussed the process of auditing as a production of legitimacy. He distinguishes four auditing themes, namely: “the audit process and formal structure; auditing as a business; working papers and image management; new audits” (p. 380). Power addresses that “the legitimacy of auditing is under fire” (2003, p. 379) and therefore some parts of auditing are more focused on legitimacy instead of efficiency. For instance, official guidance is rather used for legitimize the work when encountering conflicts with clients, instead of being utilized for daily work (Humphrey & Moizer, 1990). In addition, the formal structure of the audit planning is basically used for legitimate the authority of the auditor (Humphrey & Moizer, 1990). Furthermore, Power (2003) discussed the extension of auditing practices within new business aspects. This has led to making new things auditable and an expansion of auditing mandates (Power, 2003). For example, audit firms perform more
comprehensive audits than before and increasingly perform more audits on aspects such as CSR. When an auditor conduct an audit on CSR, it produces legitimacy for both the auditor as the auditee (the firm), according Power (2003). The auditor includes the CSR performance into its audit and provide a new legitimacy in their work as an auditor. Other way around, the firm receives an audit on its CSR performance and this results in legitimacy for their good management, because good management practice comes forth from the legitimate part of auditing (Power, 2003). Power concludes its critical paper with “the role of auditing in the production of legitimacy, and the consequences of this, remain documented and under-researched” (p. 392). This study expands the research on the legitimacy and the auditing process by examine the impact of audit quality on CSR.
In addition, auditors use the CSR performance, among other things, to indicate client’s audit risk (Chen, Srinidhi, Tsang & Yu, 2011). They found a positive relation between CSR concerns and audit fees. Furthermore, they find that auditors reduce their audit fees when the CSR performance becomes stronger. This is interpreted as: “that managers’ commitment to CSR reduces the regulatory, litigation and reputation risk of both the client firm and the auditor” (Chen et al., 2011 p. 34). Moreover, auditors improve financial statements and can advise or urge to make more CSR disclosure (Chen et al., 2016). As a result, owners will be less exposed to reputation and public concerns risk by sub optimal behavior of their executives. Concluding, I expect to see a positive relation between audit quality and CSR disclosure. The second hypothesis is as follows:
H2: Audit quality is positively related with CSR disclosure.
III. Research method
This study focusses on a global set of firms because research about tax avoidance in international context is scant (Atwood, Drake, Myers, 2012; Kanagaretnam et al., 2016). The main variables of interest in this study are audit quality, ETR, Cash ETR and CSR disclosure. Audit quality is proxied by Big4 auditor and is indicated as a dummy variable for two groups: ‘no Big4’ and ‘Big4’. The auditor data is collected from the Bloomberg database. For both hypotheses (H1 & H2) I compared the proxy groups concerning two different ETRs and CSR disclosure with audit quality. I collected tax data from annual reports by means of the Compustat database for hypothesis one (H1). For both dummy groups I conducted the effective tax rate from the income statement (ETR) and the cash effective tax rate (CASH_ETR). Using these different ETRs as proxy for tax avoidance is a commonly accepted measure (e.g. Desai
and Dharmapala, 2006; Dyreng, Hanlon & Maydew, 2010; Chen, Chen, Cheng & Shevlin, 2010; McGuire et al., 2012). A firm with a higher ETR in comparison with a firm with a lower ETR, represents a lower extent of tax avoidance and vice versa. The ETR is conducted through the following formula:
(ETR) = (Total Tax Expenses) (PreTax Income − Special Items)
Where the ‘total tax expenses’ is current plus deferred tax expense. The Cash ETR will be conducted through the following formula:
(CASH_ETR) = (Cash Tax Paid )
(PreTAX Income − Special Items)
In addition, I collected the CSR disclosure score for hypothesis two (H2) from the Bloomberg database. I follow Nollet, Filis and Mitrokostas (2016) concerning the measurement of CSR disclosure it represents the extent of firm’s CSR. When a firm has a high CSR disclosure score in comparison with other firms, it means that this firm has a higher extent of CSR and vice versa.
I compared the difference in outcome between the two proxy groups regarding tax avoidance and CSR disclosure in multiple regression models. The results from the regressions provide evidence about the influence of audit quality on the extent of firm’s tax avoidance and CSR.
Sample design
Table 1 describes the sample selection procedure before conducting the analysis. It is a global sample consisting over 100,000 firm-year observations. I started with a wordwide list of firms from the Bloomberg database. This data list consist of 11.283 unique firms worldwide. I collected financial data from the Compustat database and auditor- and CSR data from the Bloomberg database. Finally, the total sample size is 111.081 frim-year observations. Thereafter I dropped the observations with duplicate year values (including missing year values) caused by different industry formats and differences in the last month of the fiscal year. I used the duplicates syntax command in STATA to remove these observations, what resulted in 103.201 unique firm-year observations.
TABLE 1
Total Firm-year observations
Total firm-years from Compustat 111,081
Restricted firm-years after removing duplicates 103,201
Total firm-years used for H1 with ETR 29,449
Total firm-years used for H1 with Cash ETR 37,105
Total firm-years used for H2 53,110
Appendix C reports the number of observations per country. United States (31,356) and Japan (16,274) are the countries with the most firm-year observations. Besides that, Australia, Brazil, Canada, China, France, Germany, Hong Kong, India, South Africa, South Korea, Taiwan and the UK consists of more than 1,000 firm-year observations. Furthermore, appendix D shows the auditor distribution over the countries. The distribution of auditor observations per country is equivalent to the distribution of total observations per country. The total sample consists of 37,442 non-Big4 auditors and 49,795 Big4 auditors.
Research design
I estimate three multiple regression models with firm-, culture- and country control variables in order to test the relation between audit quality and the three dependent variables: effective tax rate (ETR), cash effectivetax rate (CASH_ETR) and CSR_DISCL (CSR disclosure). Table 2 shows the description and the source of all the regressions variables. The regression models provide evidence for the two hypotheses concerning audit quality. The independent variable is audit quality (AUDITQ), and represents the audit quality of firm’s auditor. The Big4 is ranked based on global revenue as of 2014 (The BIG 4 accounting firms, 2014)
As earlier mentioned, I control for specific firm-, culture- and country characteristics. The firm specific characteristics, following Dyreng et al. (2008), are: return on assets (ROA), firmsize (logSIZE) and price-earnings ratio (PE_RATIO). I expect that firms with higher ROA in comparison with other firms, do have lower ETRs and have higher CSR scores. Besides that, I expect to see a positive relation between firm size and CSR, because bigger firms will experience a higher degree of pressure from the public and therefore feel the need to be social responsible. Next to that, I expect to see a negative relation between the firm size and both ETRs, because I predict that bigger firms are more capable and have more resources to engage in tax avoidance than smaller firms. Since the price-earnings ratio is used to determine how
much investors are willing to pay for firm’s shares, I expect to see a higher CSR score for firms with a high price-earnings ratio than firms with a low price-earnings ratio. Furthermore I include board size (BOARD_SIZE) and Tobin’s Q ratio (TOBINQ). I expect to see a positive relation with board size on both ETRs and CSR. I assume that a larger board will discuss issues like tax avoidance and CSR more on account of external pressure (stakeholders and public). In addition, when a firm is undervalued (low Tobin’s Q ratio) I expect that firm’s CSR will be higher in order to create value through social activities.
I collected country culture data to control for country culture differences in the sample concerning power distance (POWER), masculinity versus femininity (MASCULIN) and uncertainty avoidance (UNCERT_AVOID) from Hofstede. Several studies provided evidence that culture has an impact on the tax practices of the firm (Tsakumis, Curatola and Porcano, 2007; Richardson, 2008). Besides controlling for culture differences, I control for country characteristics that can change over time. I use the statutory corporate tax rate (CORP_TR), following Kanagaretnam et al. (2016), and the worldwide governance indicators (WGI_COUNTRY) to control for country characteristics. The worldwide governance indicators are borrowed from The Worldbank Group (2016). I expect to see that firms located in a country with higher governance indicators, engage less in tax avoidance and engage more in CSR. While using the culture scores, governance indicators and the statutory tax rate as control variables, I endeavor to make the results comparable between countries in the world. Regarding the statutory tax rates, I expect to see a positive relation between corporate tax rate and ETR because when the corporate tax rate is higher, the firm needs to pay more taxes. On the other hand, a higher tax rate means less profit for the firm and thus I expect that firms in countries with higher tax rates invest less in CSR. The country that is been used for the culture scores and country scores is based on the country of firm’s headquarter, assuming that the most important decisions concerning CSR and tax are made in firms’ headquarters.
An overview of all variables can be found in table 2. Moreover, I control for year and industry fixed effects in all regression models. ‘Year fixed effects’ represents the fiscal year of the financial statement data and the ‘industry fixed effects’ is measured as the first two digits from the Global Industry Classification (GIC) code. Appendix B reports the industries in the sample.
TABLE 2 Variable definitions
Mnemonics Variable name Description Source
Dependent variables
and one. Effective tax rate is calculated as total tax expenses divided by pretax income (minus special items)
CASH_ETR Cash effective
tax rate
Indicator for tax avoidance determined by effective tax rate. The rates are truncated at zero and one. Effective tax rate (cash) is calculated as income taxes paid divided by pretax income (minus special items)
Bloomberg
CSR_DISCL ESG overall
disclosure score
Indicator for ESG score consists of environmental, social and governance data
Bloomberg
Independent variable
AUDITQ Audit Quality Indicator for audit quality where 1 is Big4
(KPMG, EY, Deloitte, PwC) and 0 otherwise
Bloomberg
Firm specific variables
ROA Return on
Assets
Determined by net income divided by total assets Compustat
logSIZE Firm Size Firm size determined by natural log of total assets Compustat
TOBINQ Tobin’s Q Ratio Calculated by market value divided by total assets Datastream/Co
mpustat
PE_RATIO Price Earnings
Ratio
Calculated as price per share divided by earnings per share
Compustat
BOARD_SIZE Board Size Number of board members per firm per year
collected
Bloomberg
Culture control variables
POWER Culture Power
distance score
Score is based on country of firm’s headquarter Hofstede
MASCULIN Culture
Masculinity score
Score is based on country of firm’s headquarter Hofstede
UNCERT_AVOID Culture
Uncertainty Avoidance score
Score is based on country headquarter Hofstede
Country control variables
WGI_COUNTRY Country
Governance controls score
Combined score from six governance scores (control of corruption, government effectiveness, political stability, regulatory quality, rule of law, voice and accountability). The scores are
combined with the Principal Component analysis. Country is based on country of firm’s
headquarter. Worldbank Group CORP_TR Country Corporate Tax Rate following Ganagaretnam (2016)
Country statutory tax rate collected from the Tax Foundation (Pomerleau, 2014), complemented with data from the Worldwide corporate tax guide 2015 (EY). Country is based on country firm’s headquarter
Tax Foundation/
EY
Robustness variables
BIGN6 Audit Quality
Big6
Dummy variable where 1 is Big6 auditor and 0 otherwise
Bloomberg
CSR_DISCL2 ESG overall
disclosure score
Indicator for ESG score consisting of
environmental, social, governance and economic data
Asset 4
Descriptive analysis
Table 3 describes the sample variables. I winsorized the ETR and CASH_ETR at zero and one in order to exclude negative outcomes and outliers. The average ETRs are, respectively, 28.4
and 26.6 percent and is consistent with prior studies (Dyreng et al., 2008; Gleason & Mills, 2011). The average audit quality is 57 percent, which means that 57 percent of the auditor firm-year observations had a Big4 auditor. The WGI_COUNTRY variable is a combined score of six worldwide governance indicators collected from the World Bank. These scores needed to be combined as an individual score due to collinearity issues, for justification see appendix A that reports the high correlations between all variables. I merged the country governance scores by means of a principal component analysis (PCA). The average board size is 9.2, with 1 board member as smallest outcome and 33 board members as biggest outcome. The mean statutory corporate tax rate (32.9 percent) is consistent with Kanagaretnam et al. (2016).
TABLE 3 Descriptive analysis
N Mean Std Dev Min Max
ETR 49,173 0.284 0.162 0.000 0.829 CASH_ETR 62,267 0.266 0.177 0.002 0.883 CSR_DISCL 67,100 20.659 13.089 0.830 86.780 AUDITQ 87,237 0.571 0.495 0 1 logSIZE 100,018 8.985 2.726 2.627 16.567 ROA 97,164 0.028 0.101 -0.752 0.289 PE_RATIO 96,867 17.347 36.395 -133.241 368.285 BOARD_SIZE 68,149 9.206 3.091 1 33 POWER 98,447 53.129 17.741 11 100 MASCULIN 98,447 63.338 17.744 5 100 UNCERT_AVOID 98,447 56.324 22.089 8 100 WGI_COUNTRY 90,527 -0.008 2.328 -7.236 2.686 CORP_TR 98,748 0.329 0.077 0.000 0.550
All continuous variables are trimmed at 1e and 99e percentile.
Appendix B reports the distribution of the mean ETR, CASH_ETR and CSR_DISCL over the specific sectors in the sample. I observe higher ETRs in the industrials-, consumer discretionary- and consumer staples sector and lower ETRs in the energy-, real estate- and telecommunication sector. The difference in mean between the lowest and the highest sector score is 10 percentage points for ETR and Cash ETR. Besides that, I observe that the industrials- and consumer discretionary sector do consist of the most firm-year observations. The mean CSR disclosure score differs from 17.57 (real estate) to 26.75 (utilities). The industrials sector (12,356) and consumer discretionary sector (9,739) have the highest amount of observations. Correlation matrix
The pairwise correlations between the variables are reported in table 4. The pairwise correlation between audit quality (AUDITQ) and ETR is negative, and significant (r = -0.30). This
correlation is not consistent with H1, because H1 predicts a negative relation between audit quality and tax avoidance. A negative correlation with ETR, suggests that audit quality is positive correlated with tax avoidance. There is also a significant negative correlation between audit quality and Cash ETR (r= -0.22). Both negative correlations are not in line with H1, which predicts that both variables are positively related to each other. LogSIZE (r= 0.25 & 0.1), PE_RATIO (r= 0.07 & 0.08) and BOARD_SIZE (r= 0.03 & 0.04) are all positive correlated respectively to ETR and Cash ETR. The culture control variables (POWER, MASCULIN and UNCERT_AVOID) all have a positive correlation with ETR and Cash ETR. There is a negative correlation between both ETRs with ROA and TOBINQ. The WGI_COUNTRY score comes with a mixed result, the governance indicator has a positive correlation with ETR (r= 0.08) but has a negative correlation with Cash ETR (r= -0.03). The corporate tax rate has, as expected, a positive correlation with both ETRs.
Besides that, I observe a significant positive correlation (r = 0.14; p <.05) between audit quality and CSR disclosure (CSR_DISCL). These outcomes strongly indicates a relation between the independent and dependent variables. Firm size (logSIZE) has a positive correlation with CSR, which infers that firms are more social responsible when they are bigger. The positive correlation between ROA and CSR supports the prediction that firms with more return on assets do engage more in CSR. The correlation between PE_RATIO and CSR is not significant. Board size is, as predicted, positive correlated with CSR and infers that a larger board does see the need of CSR concerning the stakeholders. A firm in a country of strong governance indicators, has a positive correlation with CSR and a firm with a higher corporate tax rate has, as expected, a negative correlation with CSR.
All the correlations are below 0.50 except for the relation between POWER and WGI_COUNTRY. Because the correlations are pairwise and not include control variables, the main inferences will be drawn from the regression models with multiple control variables. The multiple regression models are reported in the next section.
TABLE 4
Pairwise correlation matrix
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 ETR 1.000 2 CASH_ETR 0.357* 1.000 3 CSR_DISCL -0.020* 0.009 1.000 4 AUDITQ -0.303* -0.220* 0.144* 1.000 5 logSIZE 0.251* 0.099* 0.346* -0.233* 1.000 6 ROA -0.007 -0.078* 0.084* 0.009* 0.127* 1.000 7 TOBINQ -0.098* -0.033* -0.106* 0.032* -0.305* 0.138* 1.000 8 PE_RATIO 0.068* 0.081* -0.003 -0.068* 0.014* 0.136* 0.136* 1.000 9 BOARD_SIZE 0.033* 0.041* 0.317* 0.078* 0.385* 0.052* -0.150* 0.017* 1.000 10 POWER 0.041* 0.043* -0.050* -0.312* 0.347* 0.147* 0.021* 0.111* 0.084* 1.000 11 MASCULIN 0.371* 0.225* -0.097* -0.481* 0.196* -0.033* -0.101* 0.016* -0.001 -0.013* 1.000 12 UNCERT_AVOID 0.348* 0.160* 0.172* -0.315* 0.473* -0.019* -0.218* -0.050* 0.056* 0.002 0.416* 1.000 13 WGI_COUNTRY 0.082* -0.028* 0.110* 0.224* -0.186* -0.148* -0.109* -0.119* -0.072* -0.839* 0.137* 0.297* 1.000 14 CORP_TR 0.291* 0.156* -0.182* -0.185* -0.111* -0.117* -0.049* -0.058* -0.029* -0.353* 0.471* 0.213* 0.360* 1.000 * Indicates p<0.05
19
IV.
Analysis, findings and robustness checks
Main analysis and findings
Due to three multiple regression models I endeavor to find evidence for H1 and H2. All the models use panel data and consist of the same control variables. For H1, ETR and Cash ETR are the dependent variables and CSR disclosure is the dependent variable for H2. Table 5 reports the results for H1 and H2. All the regression models are significant and has explanatory power. The ETR regression consists of 29,449 observations from the sample. Furthermore the R-squared is (0.18), so 18 percent of the variation of the ETR is explained by this model. The regression shows a significant negative relation between ETR and audit quality. The coefficient is -0.018 and is highly significant (p <0.01). This outcome indicates that when the audit quality increases, the ETR will decrease. A lower ETR represents more extent of tax avoidance in comparison with a firm with a higher ETR. Thus, based on the results a firm with a higher degree of audit quality, in comparison with firms with lower audit quality, engage more in tax avoidance. Furthermore the regression model gives highly significant outcomes for most of the control variables. Return on investment (ROA) is positively related to ETR with a coefficient of 0.103. The coefficient of 0.344 indicates a positive relation between the statutory corporate tax rate and ETR. Firms in countries with a higher tax rate, have a higher ETR than firms located in countries with lower tax rates. This outcome is logical given that the corporate tax rate is directly connected with the ETR. POWER (-0.000) is negatively related to ETR and both MASCULIN (0.001) and UNDERT_AVOID (0.001) are positively related to ETR, but all have low coefficients. In addition, BOARD_SIZE and WGI_COUNTRY are not significant. The results for CASH_ETR are similar regarding the most important outcome for H1. The relation between CASH_ETR and audit quality is negative with a coefficient of -0.043 and is highly significant (p<0.01). This indicates and supports the same inference as for the ETR regression, namely Big4 auditors have a positive impact on the extent of tax avoidance. The regression model consists of 37,105 observations and the R-squared is 0.09. The R-squared is relatively low, but this regression model explains 9 percent of the variation in Cash ETR. Hence, both outcomes from the ETR and CASH_ETR regressions provides strong evidence for H1, but in the opposite direction of what is predicted. I elaborate this contradiction comprehensively in the conclusion section.
The fourth column of table 5 presents the results for H2. This regression model consists of 53,110 observations from the total sample. The R-squared has a value of 31 percent. So, 31
percent of the variation in CSR disclosure score is explained by this model. According the results, audit quality is positively related to CSR disclosure with a 2.699 coefficient. The result is highly significant (p<0.01). The coefficient indicates that when firm’s auditor has a higher audit quality, it is likely that the firm engages more in CSR disclosure than a firm with a non-Big4 auditor. Besides that, BOARD_SIZE is positively related with CSR disclosure with a coefficient of 0.785 and is highly significant. This relation is positive and indicates that firms with more board members engage more in CSR disclosure than firms with less board members. Besides that, ROA is positively related to CSR disclosure. It is likely that firms who have more resources to spent, due to a high return on investments, are more willing to invest in CSR disclosure. CORP_TR is negative related to CSR disclosure, and indicates that when firms have more commitments to tax authorities, firms are less likely to invest in CSR disclosure. Furthermore firms in countries with high governance scores engage more in CSR disclosure. This indicates that governance mechanisms find investments by firms in CSR important.
TABLE 5
Multiple regression models
ETR CASH_ETR CSR_DISCL
AUDITQ -0.018*** -0.043*** 2.699*** logSIZE 0.006*** 0.001* 2.041*** ROA 0.103*** -0.096*** 2.456*** TOBINQ -0.006*** 0.002** 0.557*** PE_RATIO 0.000*** 0.001*** 0.003** BOARD_SIZE -0.000 0.003*** 0.785*** POWER -0.000* -0.000* -0.108*** MASCULIN 0.001*** 0.001*** -0.111*** UNCERT_AVOID 0.001*** 0.000*** 0.080*** WGI_COUNTRY -0.001 -0.008*** 0.581*** CORP_TR 0.344*** -0.100*** -43.004***
Constant Yes Yes Yes
Year fixed effects Yes Yes Yes
Industry fixed effects Yes Yes Yes
Adjusted R2 0.18 0.09 0.31
Highest VIF 4.09 3.62 4.62
Mean VIF 2.17 1.97 1.99
N 29,449 37,105 53,110
Robustness check: Big6 audit firms
I run some additional regressions in order to test if the main results are robust. Firstly, I use an extended measure for audit quality. I include several robustness checks to test if the results from the regressions are robust. In the main analyses I use Big4 audit firms as proxy. To strengthen the results I will check for robustness with Big6 auditors. These robustness checks represents the biggest six audit firms in terms of revenue (worldwide) and therefore have more audit quality according DeAngelo (1981) than the smaller audit firms. In order to apply the Big6 variable, the traditional Big4 auditors will be complemented with the audit firms: BDO and Grant Thornton (The BIG 4 accounting firms, 2014). Table 6 reports the results of the robustness check. Consistent with the main analysis I observe a negative relation between audit quality and ETR (-0.025) and CASH_ETR (-0.043). Besides that I observe a positive relation between audit quality and CSR_DISCL (2.906). All the outcomes from the variables of interest are highly significant. Thus, the results from the robustness check regarding Big6 auditors are consistent with the results reported earlier. This means that the results from the main analysis are robust for audit quality.
TABLE 6
Robustness check with Big6 auditors
ETR CASH_ETR CSR_DISCL
AUDITQ -0.025*** -0.043*** 2.906*** logSIZE 0.005*** 0.000 2.083*** ROA 0.101*** -0.102*** 2.510*** TOBINQ -0.006*** 0.002** 0.570*** PE_RATIO 0.000*** 0.001*** 0.003** BOARD_SIZE -0.000 0.002*** 0.785*** POWER -0.000** -0.000* -0.107*** MASCULIN 0.001*** 0.001*** -0.111*** UNCERT_AVOID 0.001*** 0.000*** 0.082*** WGI_COUNTRY -0.001 -0.008*** 0.589*** CORP_TR 0.346*** -0.097*** -43.052***
Constant Yes Yes Yes
Year fixed effects Yes Yes Yes
Industry fixed effects Yes Yes Yes
Adjusted R2 0.18 0.08 0.31
Highest VIF 4.11 3.62 4.61
Mean VIF 2.20 1.98 2.00
N 29,449 37,105 53,110
Robustness check: ESG data from Asset 4
Secondly, I check on robustness with CSR disclosure data from Asset 4. These scores are ESG disclosure scores (similar to Bloomberg) but are rated by another database. Next to environmental, social and governance scores, Asset 4 provides an economic scores. I labeled this robustness variable as CSR_DISCL2 and the regression consists of 21,547 observations. Table 7 reports the results of the regression. The direction of the main variable of interest (AUDITQ) is consistent with the main analysis. The coefficient of 20.226 if highly significant and indicates a positive relation between audit quality and the extent of CSR disclosure. Furthermore the outcomes of the control variables are in line with the main analysis, except for TOBINQ and PE_RATIO. Both variable outcomes are not significant. The results from the robustness check are consistent with the findings for H2. Thus, the results for CSR disclosure are robust regarding the ESG data from Asset 4.
TABLE 7 Robustness check on CSR CSR_DISCL2 AUDITQ 20.226*** logSIZE 4.160*** ROA 48.171*** TOBINQ -0.100 PE_RATIO 0.002 BOARD_SIZE 1.617*** POWER -0.284*** MASCULIN -0.281*** UNCERT_AVOID 0.088*** WGI_COUNTRY 1.788*** CORP_TR -39.481*** Constant Yes
Year fixed effects Yes
Industry fixed effects Yes
Adjusted R2 0.22
Highest VIF 2.74
Mean VIF 1.68
N 21,547
* p<0.1; ** p<0.05; *** p<0.01
Robustness check: excluding financial firms
Finally, I excluded the financial firms from the sample in order to see if the results are robust. Financial firms do have a different financial structure in comparison with other kind of firms
and therefore can cause some noise in the sample. After excluding the financial firms (banks, diversified financials, insurance, real estate) the results are consistent with the findings. The N decreased with 3,832 firm-year observations for the ETR regression, 6,435 firm-year observations for the CASH_ETR regression and 9,535 firm-year observations the CSR_DISCL regression. Table 8 reports a negative relation between audit quality and ETR and CASH_ETR. Besides that, it reports a positive relation between audit quality and CSR_DISCL. All results are highly significant and similar to the output from the main analysis.
TABLE 8
Robustness check, financial firms excluded
ETR CASH_ETR CSR_DISCL
AUDITQ -0.013*** -0.041*** 2.759*** logSIZE 0.006*** 0.001** 2.434*** ROA 0.101*** -0.075*** 0.420 TOBINQ -0.008*** 0.000 0.253*** PE_RATIO 0.001*** 0.001*** 0.002* BOARD_SIZE -0.001* 0.002*** 0.862*** POWER -0.000 -0.000 -0.133*** MASCULIN 0.001*** 0.001*** -0.135*** UNCERT_AVOID 0.001*** 0.000*** 0.070*** WGI_COUNTRY 0.002* -0.006*** 0.489*** CORP_TR 0.371*** -0.114*** -42.810***
Constant Yes Yes Yes
Year fixed effects Yes Yes Yes
Industry fixed effects Yes Yes Yes
Adjusted R2 0.19 0.08 0.33 Highest VIF 4.37 3.86 5.15 Mean VIF 2.27 2.06 2.14 N 25,617 30,670 43,575 * p<0.1; ** p<0.05; *** p<0.01
V Conclusion
This study broadens the perspective on audit quality related to tax avoidance and CSR. It is relevant to know what influence the auditors have on the accounting and social practices of their clients, because research on the auditor on tax avoidance and CSR I scant and under-researched. In this study, tax avoidance represents unethical accounting behavior/practice and CSR represents ethical social behavior/practice. The legitimacy theory and agency theory are used to justify the study on both phenomena, in order to give more insight in the complex relation with the auditor. In this study I provide evidence for the impact of audit quality on tax
avoidance and separately on CSR. Two hypotheses are tested, both outcomes were highly significant.
The first hypothesis focused on the relation between audit quality and tax avoidance. The results from the analysis indicate a positive relation between audit quality and the extent of firm’s tax avoidance. Thus, firms with a Big4 auditor engage, on average, more in tax avoidance than firms with a non-Big4 auditor. This outcome is highly significant and is the opposite of what was expected, despite several controls for firm, culture and country specific characteristics. I believe that the tax advisor plays an important role in this outcome. The opposite outcome could be explained by firms who hire a Big4 auditor, also hires a Big4 advisor for their tax practices. The data concerning firms that hire Big4 tax advisors were not available, this is the reason that I was not able to collect these data. So, it could be the case that firms with an auditor of high quality simultaneously also have a high quality tax advisor who is bending the law in order to pay as less tax as possible. This situation can explain why the relation between audit quality and tax avoidance is positive instead of negative. In order to find evidence for this explanation, further research should collect data concerning tax advisor identification such as Big4 or non-Big4. In this way, the analysis can be controlled for a higher quality of tax advisory.
In addition, the impact of audit quality on the extent of firm’s CSR is tested with the second hypothesis. This hypothesis can be accepted due to the results given for H2, so the outcome of the analysis was as expected. The relation between audit quality and the extent of firm’s CSR is positive when it comes to CSR disclosure. This means that firms with a Big4 auditor engage, on average, more in CSR disclosures than firms with a non-Big4 auditor. This outcome infers the importance in CSR for high quality auditors. It implies that auditors find it important that their clients report more to its stakeholders, especially on non-financial information. In this way, stakeholders can make better judgements or estimations of firms’ activities. The sample consists of firms listed at a stock exchange and firms who are not listed. Every countries implements its own rules about firms that are obligatory to disclose CSR information. For example, in the Netherlands, listed firms are obligated to disclose non-financial information and non-listed firms are not. In essence, listed firms hire Big4 auditors and non-listed firms hire non-Big4 audit firms. Hence, this could bias the results on audit quality due to different legislation for disclosing CSR information for listed and non-listed firms. Further research can endeavor to include the country specific legislation regarding CSR disclosure in order to control for this limitation.
Looking at the results, I can conclude that auditors do have an impact on the extent of tax avoidance and CSR disclosure. Big4 auditors lead to more CSR disclosure by their clients than non-Big4 auditors do. At the same time Big4 auditors also lead to more tax avoidance by their clients regarding the results. Both hypotheses are statistically proven, but a causal relation is questionable due to the rejection of H1. Thus, a causal relation between audit quality and tax avoidance is questionable. The impact of big accountancy firms contra small accountancy firms regarding tax advisory is left out of the study. This could bias the results for tax avoidance in my sample. Further research can endeavor to eliminate this bias on Big4 tax advisor’s impact on tax avoidance. It will be valuable to find out whether the results for H1 differ from the findings in this study, when controlling for Big4 tax advisors. Another limitation is the impact of big Asian audit firms which audit quality may be similar to the global Big4. Asian audit firms operate only in Asia and therefore miss the amount of revenue to belong to the global Big4. The Big4 audit firms are operating worldwide, so their revenue is understandably higher than the big Asian audit firms. Thus, it could be the case that some Asian audit firms do have high audit quality, similar to the global Big4. The impact of big Asian audit firms can bias the results on audit quality, if their quality is as high as global Big4 audit firms. The sample consists of 11,902 auditor observations from China, India and Japan and are coded as non-Big4. In this situation the impact of big Asian audit firms can bias the outcomes for the non-Big4 group. Furthermore, the results provide additional evidence for existing research about firms using entrenchment strategy to camouflage their unethical accounting practices (Cespa & Cestone, 2007; Prior et al., 2008; Martínez-Ferrero & García-Sánchez, 2015). In essence, the bigger firms hire Big4 auditors in order to conduct the audit on the financial statements. Hence referring to the results of this study, I infer that bigger firms do engage in more tax avoidance and at the same time engage more in CSR disclosures than smaller firms. Engaging in both tax avoidance and CSR, indicates that those firms are using the CSR disclosure to camouflage their tax avoidance. This implication could function as additional evidence for executives are using the entrenchment strategy to camouflage tax avoidance with strategic CSR disclosures.
Finally, further research can investigate if there are continent/country differences in the tax avoidance outcomes and CSR disclosure outcomes. It is relevant for policy makers to know if there are differences between the impacts of Big4 auditors per country. It could be the case that, for example, Big4 auditors in the US have a different influence than Big4 auditors in European countries or Asian countries.
VI References
Addison, S. & Mueller, F. (2015). The dark side of the professions: the big four and tax avoidance. Accounting, Auditing, & Accountability Journal, 28 (8), 1263-1290
Armstrong, C.S., Blouin, J.L., Larcker, D.F. (2015). Corporate governance, incentives, and tax avoidance. Journal of Accounting and Economics, 60, 1-17
BBC (2012). Starbucks, Google and Amazon grilled over tax avoidance, Available at: http://www.bbc.com/news/business-20288077
Atwood, T.J., Drake, M.S., Myers, L.A., (2010). Book-tax conformity, earnings persistence and the association between earnings and cash flows. Journal of Accounting and Economics 50, 111-125.
Cespa, G., Cestone, G. (2007). Corporate Social Responsibility and Managerial Entrenchment. Journal of Economics & Management Strategy, 16 (3), 741–771
Chen, S., Chen, X., Cheng, Q., Shevlin, T. (2010). Are family firms more tax aggressive than non-family firms. Journal of Financial Economics, 95, 41–61
Chen, L., Srinidhi, B., Tsang, A., & Yu, W. (2011). How do Auditors Respond to Corporate Social Responsibility Performance? WORKING PAPER
Chen, l., Srinidhi, B., Tsang, A., Yu, W. (2016). Audited Financial Reporting and Voluntary Disclosure of Corporate Social Responsibility (CSR) Reports. Journal of Management Accounting Research, 28 (2), 53–76
Choi, J.-H., Kim, J.-B., Liu, X., Simunic, D. (2008). Audit pricing, legal liability regimes, and Big 4 premiums: Theory and cross-country evidence. Contemporary Accounting Research, 25, 55-99.
Davis, A.K., Guenther, D.A., Krull, L.K., & Williams B.M. (2016). Do social responsible firms pay more taxes? The accounting review, 91 (1), 47-68
DeAngelo, L. (1981). Auditor size and audit quality. Journal of Accounting and Economics, 3, 183-99.
Deigh, L., Farquhar, J., Palazzo, M., Siano, A. (2016). Corporate social responsibility: engaging the community. Qualitative Market Research: An International Journal, 19 (2), 225-240
Deng, X., J. Kang, and Low, B.S. (2013). Corporate social responsibility and stakeholder value maximization: Evidence from mergers. Journal of Financial Economics, 110 (1), 87– 109.
Desai, M.A., Dharmapala, D. (2006). Corporate tax avoidance and high-powered incentives. Journal of Financial Economics, 79, 145–179
Donohoe, M.P., Knechel, W.R. (2014). Does corporate tax aggressiveness influence audit pricing? Contemporary Accounting Research, 31 (1), 284-308.
Dowling, G.R. (2014). The curious case of corporate tax avoidance: Is it social irresponsible? J Bus Ethics, 124, 173-184
Dyreng, S.D., Hanlon, M, Maydew, E.L. (2008). Long run corporate tax avoidance. The Accounting Review, 83 (1), 61–82.
Dyreng, S.D., Hanlon, M, Maydew, E.L. (2010). The Effects of Executives on Corporate Tax Avoidance. The Accounting Review, 85 (4), 1163-1189
Dyreng, S.D., Hanlon, M, Maydew, E.L., & Thornock J. R. (2014). Changes in corporate tax rates over the past twenty-five years. Available at SSRN: http://ssrn.com/abstract=2521497 EY (2015). Global Tax guide archive Current and past editions as PDFs, available at: http://www.ey.com/gl/en/services/tax/global-tax-guide-archive
Fallan E. (2015). Corporate Social Responsibility and Tax Aggressiveness: A Test of Legitimacy Theory. Social and Environmental Accountability Journal, 35 (3), 198-200 Financial Times (2016a). Apple hit with €13bn EU tax penalty over illegal Irish aid, Available at: https://www.ft.com/content/b573ac02-6e90-11e6-a0c9-1365ce54b926 Financial Times (2016b). Apple ruling sparks angry reaction in US, available at: https://www.ft.com/content/68892752-6ecf-11e6-a0c9-1365ce54b926
Francis, J., Wang, D. (2008). The joint effect of investor protection and Big 4 audits on earnings quality around the world. Contemporary Accounting Research, 25, 1-39.
Gleason, C.A., Mills, L.F. (2011). Do Auditor-Provided Tax Services Improve the Estimate of Tax Reserves? Contemporary Accounting Research, 28 (5), 1484-1509
Hoi, C.K., Wu, Q., Zhang, H. (2013). Is Corporate Social Responsibility (CSR) associated with tax avoidance? Evidence form irresponsible CSR activities. The accounting review, 88 (6), 2025-2059
Hoopes, J.L., Mescal, D., Pittman, J.A. (2012). Do IRS audits deter corporate tax avoidance? The accounting review, 87 (5), 1603-1639
Humphrey, C., Moizer, P. (1990). From techniques to ideologies: an alternative perspective on the audit function. Critical Perspectives on Accounting, 1, 217–238.
Huseynov, F. & Klamm, B.K. (2012). Tax avoidance, tax management and corporate social responsibility. Journal of Corporate Finance, 18, 804-827
Kanagaretnam, K., Lee, J., Lim, C.Y. Lobo, G.J. (2016). Relation between Auditor Quality and Corporate Tax Aggressiveness: Implications of Cross-Country Institutional Differences. AUDITING: A Journal of Practice & Theory, 35 (4), 105-135.
Kanagaretnam, K., Lim, C.Y., Lobo, G.J. (2010). Auditor reputation and earnings management: International evidence from the banking industry. Journal of Banking and Finance, 34, 2318– 2327.
Khurana, I.K., Raman, K.K. (2004). Litigation risk and the financial reporting credibility of Big 4 versus non-Big 4 audits: Evidence from Anglo-American countries. The Accounting Review, 79, 473–495.
Lanis, R. & Richardson, G. (2013). Corporate social responsibility and tax aggressiveness: a test of legitimacy theory. "Accounting, Auditing & Accountability Journal, 26, 75–100 Lanis, R. & Richardson, G. (2015). Is corporate social responsibility performance associated with tax avoidance? J Bus Ethics, 127, 439-457
Martínez-Ferrero, J., García-Sánchez, I.-M. (2015). Is corporate social responsibility an entrenchment strategy? Evidence in stakeholder protection environments. Review of Managerial Science, 9, 89-114
McGuire, S.T., Omer, T.C., & Wang, D. (2012). Tax avoidance: Does tax-specific industry expertise make a difference? The accounting review, 87 (3), 975-1003
Nollet, J., Filis, G., Mitrokostas, E. (2016). Corporate social responsibility and financial performance: A non-linear and disaggregated approach. Economic Modelling, 52, 400–407 Pomerleau, K. (2014).Corporate Income Tax Rates around the World, 2014, available at: http://taxfoundation.org/article/corporate-income-tax-rates-around-world-2014
Power, M.K. (2003). Auditing and the production of legitimacy. Accounting, Organizations and Society, 28, 379–394
Preuss, L. (2010). Tax avoidance and corporate social responsibility: you can't do both, or can you? Corporate Governance: The international journal of business in society, 10 (4), 365 – 374
Prior, D., Surroca, J., Tribo, J.A. (2008). Are Socially Responsible Managers Really Ethical? Exploring the Relationship between Earnings Management and Corporate Social
Responsibility. Corporate Governance, 16 (3), 160-177
Ramanna, K. (2013). A framework for research on corporate accountability reporting. Accounting Horizons, 27 (2), 409–432
Richardson, G. (2008). The relationship between culture and tax evasion across countries: additional evidence and extensions. Journal of International Accounting, Auditing and Taxation, 17, 67-78.
Sikka, P. (2007). Enterprise culture and accountancy firms: new masters of the universe. Accounting, Auditing, & Accountability Journal, 21 (2), 268-295
Sikka, P. (2010). Smoke and mirrors: corporate social responsibility and tax avoidance. Accounting Forum, 34, 153-168
Sikka, P. (2015). The hand of accounting and accountancy firms in deepening income and wealth inequalities and the economic crisis: Some evidence. Critical Perspectives on Accounting, 30, 46-62
Sikka, P. & Hamilton, M. P. (2005). The role of accountancy firms in tax avoidance: Some evidence and issues. Accounting Forum, 29, 325-343
Sikka, P. & Willmott, H. (2013). The tax avoidance industry: accountancy firms on the make. Critical perspectives on international business, 9 (4), 415 - 443
Silberhorn, D., Warren, R.C. (2007). Defining corporate social responsibility a view from big companies in Germany and the UK. European Business Review, 19 (5), 352-372
Suchman, M.C. (1995). Managing legitimacy: Strategic and institutional approaches. Academy of management review, 20(3), 571-610
The BIG 4 accounting firms (2014). Top 20 Accounting Firms in The World, available at: http://www.big4accountingfirms.org/the-top-accounting-firms-in-the-world/
The Guardian (2016). Apple ordered to pay €13bn after EU rules Ireland broke state aid laws, available at: https://www.theguardian.com/business/2016/aug/30/apple-pay-back-taxes-eu-ruling-ireland-state-aid
Tsakumis, G.T., Curatola, A.P., Porcano, T.M. (2007). The relation between national cultural dimensions and tax evasion. Journal of International Accounting, Auditing and Taxation, 16, 131-147.
The World Bank Group: (2016). The Worldwide Governance Indicators (WGI) Project. Available at: http://info.worldbank.org/governance/wgi/index.aspx#home
Watson, L. (2015). Corporate Social Responsibility, tax avoidance, and earnings performance. JATA, 37 (2), 1-21
VII. Appendixes
Appendix A
Correlation matrix between WGI country variables
CO GO PO RE RU VO CO 1.0000 GO 0.9616 1.0000 PO 0.9030 0.8725 1.0000 RE 0.9558 0.9501 0.8497 1.0000 RU 0.9682 0.9535 0.8560 0.9500 1.0000 VO 0.8337 0.7658 0.7061 0.8227 0.8767 1.0000
CO = control of corruption; GO = government effectiveness; PO = political stability; RE = regulatory quality; RU = rule of law; VO = voice and accountability
