Amsterdam Business School
Auditor provided tax services and tax avoidance
Name: Indre Jancaite Student number: 11083999
Thesis supervisor: Mr M. Schabus Date: 15 August 2016
Word count: 8173,0
MSc Accountancy & Control, specialization [Accountancy] Faculty of Economics and Business, University of Amsterdam
Statement of Originality
This document is written by student [Indre Jancaite] who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
ABSTRACT
Taxes present an important cost to the firm and shareholders, and it is generally expected, that shareholders prefer to implement tax avoidance activities. Recent years have witnessed increased attention to various determinants of tax avoidance and one of them is auditor provided tax services. Using a sample from specific databases with detailed firm characteristics and tax information to examine the relationship between auditor provided tax services and tax avoidance measures: effective tax rate, cash effective tax rate and book tax difference, it is found, that tax NAS has a positive relationship only with cash effective tax rate, but not significant. Therefore, this study shows, that there is no significant relationship between tax NAS and analyzed tax avoidance measures.
TABLE OF CONTENT
ABSTRACT ... 3
1. INTRODUCTION ... 5
1.1. BACKGROUND ... 5
1.2. RESEARCH QUESTION ... 6
1.3. MOTIVATION AND CONTRIBUTION ... 7
1.4. STRUCTURE ... 8
2. LITERATURE REVIEW AND HYPOTHESES ... 9
2.1. WHAT IS TAX AVOIDANCE? ... 9
2.2. MOTIVES FOR, AND DETERMINANTS OF, TAX AVOIDANCE ... 10
2.3. AUDITOR- PROVIDED TAX SERVICES ... 13
3. METHOD ... 16
3.1. SAMPLE SELECTION ... 16
3.2. MEASURES OF TAX AVOIDANCE ... 17
3.3. NAS AND CONTROL VARIABLES... 18
3.4. EMPIRICAL MODEL ... 19 4. RESULTS ... 20 4.1. DESCRIPTIVE STATISTICS ... 20 4.2. MULTIVARIATE ANALYSIS... 21 5. CONCLUSION ... 24 6. REFERENCES ... 26 APPENDIX 1 ... 30 APPENDIX 2 ... 31 APPENDIX 3 ... 32 APPENDIX 4 ... 33
1. INTRODUCTION
1.1.
BACKGROUND
Traditionally and the most broadly, tax avoidance is defined as the reduction of explicit taxes (Hanlon and Heitzman, 2010). This definition follows from the research of Dyreng et al. (2008), where tax avoidance is described as reflecting transactions that have any possible effect on the firm’s explicit tax liability. It is relevant to mention that tax avoidance does not necessarily infer that firms are involved in improper activity, rather it reflects both tax reductions that are relatively in compliance with the law, as well as those that imply “wrongdoing” (Dyreng et al., 2008). However, as Hanlon et al. (2010) and Weisbach (2003) stated, there are no universally accepted definitions, or constructions for, tax avoidance or tax aggressiveness, the terms may have different meanings for different research angles. Allingham and Sandmo (1972) provided the first thorough theoretical and empirical analysis on why, and to what extent, an individual would intentionally evade taxes. With reference to the most recent reviews of tax research from Hanlon et al. (2010) there is a growing strand of literature examining various determinants of tax avoidance, starting with ownership and control (see Slemrod, 2004; Chen and Chu, 2005; Crocker and Slemrod, 2005; Desai, 2007.) up to multiple firm- level characteristics (Gupta and Newberry, 1997; Phillips, 2003; Chen et al., 2010).
Recent years have witnessed increased attention to determinants concerned with auditor independence issues, especially post- Sarbanes- Oxley Act (SOX). Even though SOX prohibited various auditor-provided services, due to the possibility of impaired auditor independence, tax services approved by audit committee remain viable (U.S. House of Representatives 2002). This decision has led to discussion on the effect of tax NAS to auditor independence and other relevant factors. The benefits of having the same audit firm to provide tax NAS include enhanced audit effectiveness and financial reporting quality, as well as internal control quality (Simunic, 1984; Kinney et al., 2004; Robinson, 2008; de Simone et al. 2015). Although, the most active researches incorporate the drawbacks of tax NAS, mostly related with impaired auditor independence and negative impact on financial reporting quality (Krishan et al. 2011). In addition, lately, literature has developed research related directly to tax avoidance; Krishan et al. (2011) in brief analysis, found that tax avoidance is not significantly associated with tax fees, while Hogan et al. (2015) identified a substantial long-term negative relationship between Tax NAS and long term tax- avoidance.
1.2.
RESEARCH QUESTION
A recent overview of tax research by Hanlon et al. (2010) states, that there is a widespread interest in, and concern over, various determinants of tax avoidance. Additionally, with reference to Krishan et al. (2011), despite the increased interest in tax avoidance, limited empirical evidence exists of the association between tax NAS and tax avoidance behaviour of firms. Based on these rationales, this paper will investigate the following research question:
What is the relationship between auditor- provided tax services (tax NAS) and tax avoidance?
After SOX, tax NAS has become a widely examined question in accounting literature. Proponents of a prohibition on tax NAS discuss the issue that proposed tax strategies are becoming increasingly complex, as audit firms start to market them to earn additional fees. (Dhaliwal et al. 2013). Critics of prohibition argue that audit committees do not oversee clearly NAS approval because they could be “fooled” by auditors’ thorough and explicit documentation in authenticating the tax NAS they provide. Thus, basically they lose the capability to govern tax NAS, which may culminate in abusive tax strategies. (PCAOB, 2004). However, there also exists controversy, fuelled by the belief that tax NAS benefits outweigh the drawbacks and so could not be completely banned. This controversy incentivized several studies into the effect of NAS on tax avoidance. This paper hypothesizes, that “Tax NAS fees are negatively associated with tax avoidance”.
Using sample of 1936 firms and 5686 observations over a period between 2010 and 2015 evidence is provided to support this hypothesis.
1.3.
MOTIVATION AND CONTRIBUTION
This paper will contribute to the emerging body of literature related to the determinants of tax avoidance. The main motivation for this paper is the lack of tax research into the various determinants of tax avoidance, especially in regard to the provision of non-audit services by audit firms. This is of importance, as the effect of tax NAS on tax avoidance within firms has consequences, not only for possibly impaired auditor independence and current debate on the prohibition of tax NAS, but also for thinking about regulation in tax planning as a whole. To be more specific, current tax research literature contains a number of signals for further research related to diverse determinants of tax avoidance (Hanlon et al. 2010). As prior literature mainly sought to analyse determinants of tax avoidance related with ownership and control, as well some firm level characteristics, there is lack of research related to auditor provided non-audit services, particularly tax NAS. Krishnan et al. (2011) stated:
“Although corporate tax avoidance is frequently discussed in the popular press and alleged by regulators, there is limited empirical evidence of the linkage between auditor-provided tax services and tax-avoidance behaviour by firms.”
Several authors briefly analysed the effect of tax NAS on tax avoidance. Krishan et al. (2011) provided adjacent research in their paper into the relationship between tax NAS and tax avoidance but, however, found no significant association. Hogan et al. (2015) investigated the association of tax NAS and long-term corporate tax avoidance and found a significant negative relationship between tax NAS and taxes paid. Therefore, this paper will examine the relationship between effective tax rate, cash effective tax rate and book tax difference measures of tax avoidance.
Furthermore, this paper contributes to current knowledge and the viable debates regarding tax NAS and concerns whether auditor-provided tax services should be banned, such as internal audit and information technology consulting services. One side of debate debate related to tax NAS provides drawbacks and expresses concern about the impact of non-audit fees on auditor independence (Abbott et al. 2003; Davis et al. 2008; Grant et al., 2008;), while the proponents of not limiting non-audit services maintain a positive opinion, stating that synergies are created by combining the audit and tax service functions within one accounting firm (De Simone et al., 2015; Gleason et al., 2011; Kinney et al., 2004; McGuire et al., 2012).
Nonetheless, evidence on the effects of tax NAS are still mixed and additional empirical research in relation to tax avoidance would be useful.
1.4.
STRUCTURE
The remainder of this paper will be structured as follows. Section two provides the discussion of relevant literature and hypotheses development. The research methodology is discussed in section three. Section four provides results and overview and finally section five will present the conclusion and research limitations.
2. LITERATURE REVIEW AND HYPOTHESES
2.1.
WHAT IS TAX AVOIDANCE?
Tax avoidance has become an essential part of tax research and despite its recency, it is being examined actively. Empirical research studies into tax avoidance have been developed, ranging from individual to corporate tax avoidance behaviour. As early studies mainly focused on individuals as tax payers and explicitly examined their non-compliant activities, such as understating taxable income and other governmental factors related with tax enforcement policies (Allingham et al., 1972; Andreoni et al., 1998; Franzoni, 1999; Slemrod et al., 2002; Srinivasan, 1973), modern studies now concentrate on corporate tax planning behaviour. One of the first authors to introduce tax avoidance in a corporate context was Rice (1992), who provided empirical research into the corporate tax-gap.
Without detracting from the growing literature in tax avoidance, it must be noted that there is no universally accepted definition. As Hanlon et al. (2010) noticed, authors use different terms to describe tax-reporting behaviour, including “aggressiveness”, “sheltering”, “evasion”, and “noncompliance”, however, despite all semantics, tax avoidance is a more generic term. Traditionally, accounting literature described tax avoidance as the reduction of explicit taxes. The framework of this definition was proposed in Dyreng et al. (2008) study, where tax avoidance was defined as reflecting transactions that may have any possible effect on the firm’s explicit tax liability. Additionally, the author emphasized, the importance of understanding tax avoidance in a broader way and to not distinguish between activities, which are compliant or non-compliant with the law. Hanlon et al. (2010) and Weisbach (2003) supported this view and stated that these activities may hold different meanings from different angles. As an example, Weisbach (2003) indicated that especially lawyers and economists promptly classify proper and improper activities in tax avoidance, as if one could easily decide what is the legality of a tax structure.
Considerable attention has to be given to agency theory while analysing tax avoidance, especially in the corporate context (see Chen et al., 2005; Crocker et al., 2005; Desai et al., 2006; Slemrod, 2004). According to Slemrod (2004), the agency setting is most important due to separation of ownership and control in the corporate setting. With the reference to the agency theory in tax avoidance, the conflict arises between shareholders and managers as firms’ owners expect management to act on their behalf, which includes maximizing the
profits of the firm. However, this often includes tax-planning activities, opportunities to reduce tax liabilities. Although, Hanlon et al. (2010) points out that tax avoidance is not a reflection of agency problems itself.
2.2.
MOTIVES FOR, AND DETERMINANTS OF, TAX
AVOIDANCE
The concept of tax avoidance is very often linked to tax aggressiveness, which is the concept for tax evasion. However, many researches try not to distinguish between avoidance and evasion, as mostly firms deploy at least technically legal behaviour and even if illegal behaviour was deployed, the legality of tax evasion is often determined after the fact. Therefore, prior literature often deals with the consequences of tax avoidance while analysing the motives, as the benefits and costs have to be measured while considering on the level of aggressiveness of tax planning.
The most traditional motive for tax avoidance is possible economic benefit arising from lower corporate taxes paid to the government. This might be useful, not only for the firm cash flow, but likewise for management and shareholder cash flow. Chen et al. (2005) confirms this view: “Taxes represent a significant cost to the company and a reduction in cash flows available to the firm and shareholders, leading to firms’ and shareholders’ incentives to reduce taxes through tax aggressive activities”. Additionally, Desai et al. (2009) provides a different view, which assumes that the possibility for firms to avoid taxes by making use of tax shelters is a good incentive also. The authors provide evidence that firms with worse economic results are more likely to employ tax avoidance activities to appear more profitable without creating additional value with their business. An interesting perspective is provided by McGill et al. (2004), that in recent years there is an increasing trend to treat tax departments as a profit centre instead of a cost centre as it should be. Considerable attention has been devoted to agency theory and likewise, the relationship between managers and shareholders (see Desai et al., 2009; Chen et al. 2010; Hanlon, 2005; Armstrong et al., 2012). Chen et al. (2010) found significant empirical evidence that shareholders are aware of the issues of the agency problem and are keen on regulatory activities to prevent managers from engaging in tax aggressiveness. Armstrong et al. (2012) suggest solving the agency problem with incentives for managers to avoid taxes and not act opportunistically per se. While Wilson (2009) suggests employing corporate governance to reduce the agency problem.
A wide spread of literature has analysed the relationship between various determinants and tax avoidance. The quantity of research notably increased after the call of Shackelford et al. (2001), wherein the emerging interest in research on tax avoidance was also identified. The studies related to determinants of tax avoidance can be grouped into these sections: Firm characteristics and tax avoidance, ownership structure and tax avoidance, management compensation and tax avoidance, corporate social responsibility, auditing and tax avoidance.
One of the determinants of tax avoidance- firm characteristics, such as size, age, growth, capital intensity, industrial membership, asset tangibility lays a significant role when determining the magnitude of corporate tax avoidance. Gupta et al. (1997) were one of the first to examine the interrelations of a wide range of firm level characteristics, such as firm profitability, leverage, and capital intensity on tax avoidance measured GAAP ETRs. The evidence showed that tax avoidance is associated with the mentioned firm level characteristics. Further studies mostly focused on firm size as a determinant of tax avoidance. However, the results of these different scholars are mixed. Omer et al. (1993), Minnick et al. (2010), Rego (2003) and Zimmerman (1983), found a positive association between firm size and tax avoidance, particularly using an ETR measurement. Wilson (2009) used a non-ETR measurement of tax avoidance- book tax differences and found a positive relationship between corporate tax shelter participants and firm size. Conversely, other scholars derived different results of their empirical research. Gupta et al. (1997), Mills et al. (1998), and Richardson et al. (2007), found a negative or no association between firm size and tax planning activities.
The ownership structure of a firm may also be relevant, in affecting the firm’s level of tax avoidance (Desai et al. 2006; Hanlon et al. 2010), and different types of ownership structures have a diverse impact on tax avoidance. The most important questions to answer while analysing the relationship between tax avoidance and ownership structure is, who owns the firm and in what proportion is the ownership? The company’s ownership identity can vary; from public, international, family owned, hedge fund owned, et cetera, thus, ownership structure is highly important and affects tax avoidance as well as impacting various firm aspects, such as financing or investing. Badertscher et al. (2009) found that privately held firms, majority-owned by private equity, tend to engage in less tax avoidance activities compared with privately held, publicly traded firms, which engage in more non-conforming tax planning. Similarly, Slemrod (2007) and Chen et al. (2010) the most explicitly analyze
association between family firms and tax avoidance, and document significant evidence, that family firms tend to avoid fewer taxes compared to non-family firms. This result is also consistent with the evidence of Desai et al. (2006). The authors show that family firms are more likely to behave like individuals. In addition, hedge fund firms were analysed by Cheng et al. (2012), who concluded that hedge fund ownership employs less tax avoidance than firms prior to the hedge funds involvement.
In addition, a young but growing literature incorporates agency predictions into empirical studies related to tax avoidance. Slemrod (2004), Chen and Chu (2005), and Crocker and Slemrod (2005) were some of the initiators in the examination of corporate tax avoidance within an agency framework. While analysing tax avoidance activities with agency settings, Hanlon et al. (2010) mentions that it is important to remember: “if avoidance activities create value and compensation incentives align the manager’s interest with shareholders, then firms, that use more after- tax performance- based incentives should engage in more tax avoidance”. Armstrong et al. (2012) and Phillips (2003) studied the association between incentives-based compensation and tax avoidance. Phillips (2003) finds no evidence that after-tax CEO performance measures would be associated with significantly lower ETRs, while Armstrong et al. (2012), who specifically examined incentives for tax directors, found a significant negative affect on GAAP effective tax rates. Additionally, Rego et al. (2012) documented empirical research with the discovery of a positive relation between top executives equity risk incentives and tax avoidance measures.
Corporate responsibility may also be important. Specifically studied by the scholars were the relationship between tax avoidance and different dimensions of corporate social responsibility, such as corporate governance, community, diversity, et cetera. Huseynov et al. (2012) describes different dimensions of corporate social responsibility. The authors examine whether tax management fees in combination with different levels of corporate social responsibility affect tax avoidance levels in a firm that uses tax NAS. This study suggests, that there is an association between tax management fees paid to the auditors and tax avoidance, using GAAP and cash ETR as measures.
However, the literature analyses audit related issues in firms connected with tax avoidance, as audit fees and tax fees for non-audit provided tax services is comparably new. The relation
(2013), Hogan et al. (2015) and Krishnan et al. (2011). Dhaliwal et al. (2013) research was based on extensive hand-collected samples of firms concluded, with empirical evidence, that there is a positive relationship between tax NAS fees and tax avoidance, which is driven by tax planning and tax NAS fees. Hogan et al. (2015) examined the association between tax NAS and long-term corporate tax rates and using empirical data and multivariate regression models found a significant negative relationship between firm level tax NAS and taxes paid. Krishnan et al. (2011) analysed the association between earnings management and auditor-provided tax services and additionally auditor-provided analysis between two measures of tax avoidance and tax NAS to show empirical evidence that none of the measures are significantly associated with tax fee. However, some findings are mixed, as not complete measures of tax avoidance were analysed.
2.3.
AUDITOR- PROVIDED TAX SERVICES
Accounting scandals in 2000s have resulted in an increased public solicitude regarding non-compliant businesses and accounting professional practises. Seida (2003) and Slemrod (2004) scrutinized these scandals from a tax avoidance perspective. As an example, with reference to Seida (2003) who specifically examined Enron, large differences between book income and taxable income were found. This increase in concern has led to the US Congress passing the Sarbanes-Oxley Act in 2002. According to Rockness et al. (2005), SOX intention was to impose ethical behaviour on firms and accounting professionals, and to control the incidence of financial reporting issues. SOX prohibited the majorityof non-audit services provided by auditors, for the reason that there is a high possibility of impairment to auditor independence. However, auditor-provided tax services approved by the audit committee were allowed to exist (U.S. House of Representatives, 2002). For this reason, interest in research related to tax NAS has increased progressively.
Markelevich et al. (2005), Maydew et al. (2007) and Omer et al. (2006) have noticed significant changes in audit and tax fees after SOX. New limitations on the provision of non-audit services resulted in a substantial increase in non-audit fees for non-audit services, while fees for non-audit services have declined dramatically. Additionally, Bedard et al. (2010) provided a thorough study of decisions made by firms to disclose tax NAS fees and documented a negative relationship between companies who disclose tax NAS fees before the obligatory mandate of SOX and the ratio of non-audit fees to total fees. Additionally, Maydew et al.
(2007) supported these findings, concluding that tax NAS fees declined significantly after SOX, whilst total fees collected from accounting firms remained stable. These variations in paid taxes may be attributed to a shift in clients between the accounting professionals and thus possibly related to external tax consulting agreements. This shows that additional monitoring of firms’ possible aggressiveness in tax planning is vital (Hogan et al. 2015).
As mentioned before, the most important reason to prohibit the majority of non-audit services was a foresight of a possible impairment of auditor independence. However, many scholars argue that tax NAS may also be the reason for impaired auditor independence. The research related to auditor independence was vital for many years but the interest has dramatically increased after SOX. The proponents of tax NAS existence introduce some reasonable benefits, such as enhanced audit effectiveness and financial reporting quality, as well as internal control quality and less restatements (Simunic, 1984; Elder et al., 2008; Kinney et al. 2004; Robinson, 2008; de Simone et al. 2015). On the other hand, proponents of prohibition of tax NAS claim that auditor independence may be impaired, especially if large amounts of tax NAS are paid to accounting firms (Krishan et al. 2011).
During the pre-SOX period, few scholars had already begun to examine possible linkages between NAS and tax measures. Mills et al. (1998) explored the link between tax NAS and tax measures, and subsequently found a negative relation between tax NAS fees and tax liabilities. As mentioned, interest substantially increased after SOX, and some recent studies specifically analysed issues surrounding tax planning NAS and its relation to tax avoidance. Dhaliwal et al. (2013), Hogan et al. (2015) and Krishnan et al. (2011) examined the relationship between tax planning activities and tax NAS fees. The authors were oriented mainly to the tax avoidance measures of book tax gap and cash effective tax rate, and long-run effective tax rates. Dhaliwal et al. (2013) focused analysis on the influence of the audit committee on tax NAS and outlined significant evidence that the effect of tax NAS fees on tax avoidance is more visible when there is an active audit committee oversight. Hogan et al. (2010) broadly examined the influence of tax NAS on long-term tax avoidance. The results have contributed to the policy discussion on auditor independence. To be precise, the authors found that in the long term a lower tax NAS results in higher taxes paid to accounting firms. Controversially, in Krishnan et al. (2011) comparably brief study about tax avoidance, the primary analysis delved into the association of earnings and tax avoidance. As an additional
aggressiveness and avoidance. Krishnan et al. (2011) stand more as proponents of tax NAS existence, since they did not found any positive correlation between tax NAS and corporate tax avoidance. In conclusion they do not support Maydew et al. (2005) assertion that separating audit and tax service providers could potentially contribute to higher audit quality.
After analysing prior literature and given arguments for and against the effects of tax NAS on tax avoidance, the following hypothesis is stated:
3. METHOD
3.1.
SAMPLE SELECTION
The initial sample will be obtained from Audit Analytics (2010-2015) and the Compustat Fundamentals annual financial statement data (2010-2015). In this study, the most recent data is examined, due to the lack of recent literature related with determinants of tax avoidance. The extraction of relevant data from Compustat resulted in an initial sample of 76.578 observations and from Audit Analytics- 69.266. After the merge of Compustat and Audit Analytics data, the sample had 44.343 observations in total. Subsequently, 28.132 observations were dropped from this sample because of missing or non numerical observations and later 612 duplicates were removed. Finally, due to insufficient data on variables, 9.913 observations were dropped from the sample. Additionally, Total Assets variable was winsorized for tangential values to be removed. This leads to final sample of 5.686 observations, which consists of 1.936 firms. Table 1 summands the sample selection process.
Table 1 Sample Selection
# Observations # Firms
Compustat data set for years 2010-2015 76 578 10 609 Audit Analytics data set for years 2010-2015 69 266 15 023 Merged data set (Compustat, Audit
Analytics) 44 343 8 840
Missing or non numerical observations
(28 132) (6 183) 16 211 4 139
Duplicate observations for same fiscal years (keeping the first observation)
(612) (2)
15 599 4 137
Insufficient data (9 913) (2 201)
Full sample 5 686 1 936
The industry composition of the sample relative to all firms in Compustat for the same sample period is provided in Table 2. It is important to mention, that The Global Industry
Classification Standard is used for this composition. Relative to the merged Compustat and Audit Analytics sample, our sample of firms includes a greater proportion of firms from Industrials, Consumer Discretionary and Information technology industries and a somewhat smaller proportion of firms Energy, Telecommunication and Utilities industries.
Table 2
Industry Composition of Sample
Industry Proprietary sample Merged Compustat and Audit analytics sample Number Frequency (%) Number Frequency (%)
1. Energy 203 4 3 355 8 2. Materials 337 6 2 723 6 3. Industrials 1 120 20 4 649 10 4. Consumer Discretionary 1 150 20 4 906 11 5. Consumer Staples 292 5 1 695 4 6. Healthcare 677 12 5 918 13 7. Financials 426 7 12 641 29 8. Information Technologies 1 377 24 6 615 15 9.Telecommunication Services 60 1 578 1 10. Utilities 44 1 1 263 3 Total 5 686 100 44 343 100
3.2.
MEASURES OF TAX AVOIDANCE
In accordance with prior literature on tax avoidance, we use 3 measures of tax avoidance. The first measure is effective tax rate (ETRi,t):
ETRi,j= Total Tax Expensei,t / Pre- tax Incomei,t
This measure defines tax avoidance through permanent book- tax differences. Chen et al. (2010) provides examples of this tax planning activities: investment in tax havens with lower foreign tax rates (provided that foreign source earnings are classified as permanently reinvested), investment in tax exempt or tax favoured assets, participation in tax shelters, that give rise to losses for tax purposes, but not for book purposes.
CETRi,t= Cash Taxes Paidi,t/ Pre-tax Incomei,t
This measure reflects both permanent and temporary book-taxes differences.
Additionally, tax planning activities are captured with book-tax difference measure. This measure takes as its starting point the gap between financial and taxable income. Desai et al. (2006) defined book tax gap as the difference between income reported to capital markets, using Generally Accepted Accounting Principles and to the tax authorities. As Manzon et al. (2002) and Desai et al (2006) stated tax returns are often confidential and income reported to tax authorities cannot be observed directly and must be inferred using mainly financial accounting data. Following Desai et al. (2006) tax avoidance is estimated by this composition of book tax difference measure:
BTDi,t= (Pre tax incomei,t-(Income taxes federali,t+ Income taxes foreigni,t)/Tax loss carry
forwardi,t))/Total Assetsi,t
3.3.
NAS AND CONTROL VARIABLES
To examine the impact of tax NAS on corporate tax avoidance, two variables are used to capture the potential influence of tax NAS. First variable is the fees paid to auditors for tax NAS (TAXFEE) and the second - auditor brand name (BIG4). Measure for auditor brand name is used with reference to Becker et al. (1998) and DeAngelo (1981) to control for reputation effect in the auditing industry.
The results are controlled for firm characteristics with the reference to the authors (see Dyreng et al., 2008; Frank et al., 2009; Manzon et al., 2002; Mills, 1998). These specific firm characteristics are correlated with tax planning measures in order to ensure the consistency of the results, and ensure, that they are not driven by fundamental firms’ characteristics.
The first set of control variables: firm size (SIZE) and profitability (ROA) are measured by logarithm of total assets and by the return on total assets, respectively. Huang et al. (2015) argues, that the effect of firm size on tax avoidance is controversial. The authors as an example of controversy provide political power and political costs theories. Political power theory argues, that larger firms have more resources to lobby on tax planning strategies, while political cost theory argues, that larger firms are too scrutinized by the government and the public to be able to avoid taxes. Profitability is expected to positively affect tax
avoidance, according to Huang et al. (2015) states, that there is possibility, that more profitable companies take on additional initiatives for tax planning.
In addition, potential growth of the firm is controlled as growing firms may make more investments in tax favoured assets, that are able to generate timing differences in the recognition of expenses (Chen et al. 2010). In order to measure the potential growth of the firm, the sales growth (GSALE) is used, which current sales or revenue of the firm. According to Tang et al. (2011) the growth in revenue will increase tax avoidance due to inconsistent income and expense recognition in accounting regulations and tax laws. On the contrary, Dyreng et al. (2010) find a negative relation between sales growth and tax avoidance.
3.4.
EMPIRICAL MODEL
The empirical models to test hypothesis are:
1a. TAXAVOIDETR= f (TAXFEE, BIG4, SIZE, ROA, GSALE, YEAR)
1b. TAXAVOIDCETR= f (TAXFEE, BIG4, SIZE, ROA, GSALE, YEAR)
1c. TAXAVOIDBTD= f (TAXFEE, BIG4, SIZE, ROA, GSALE, YEAR)
where TAXAVOID(ETR,CETR,BTD) are tax avoidance measures and TAXFEE, BIG4, SIZE,
ROA, GSALE, YEAR – additional control measures as discussed above. All variables are obtained from Compustat or Audit Analytics, unless otherwise noted.
4. RESULTS
4.1.
DESCRIPTIVE STATISTICS
Descriptive statistics for the sample are reported in Table 3. The mean (median) sales and total assets for our sample is $ 1.811,99 ($ 780,96) million and $ 1.867,85 ($983,37) million, respectively. Our sample firms are profitable, reporting both mean and median return on assets of approximately 2-5 percent. Finally, most of our sample firms use services from BIG4 auditors with mean of 0,82 and median- 1 and pay relatively high amounts for tax NAS, where mean is equal $ 0,28 million and median $ 0,08. Thus the sample consists of relatively large firms and therefore have a relatively greater proclivity to engage in tax planning.
Furthermore, it is relevant to analyse descriptive statistics for our dependent variables. The mean value for CETR is 0,18, which is slightly lower than the figures reported in the most recent literature in tax avoidance from Hogan et al. (2015) with reported mean of 0,25 and also with Dyreng et al. (2008) study, where the mean of CETR is 0,27. In addition, the mean of BTD is higher (0,04) than reported in significant Huang et al. (2016) study (0,015) and ETR is likewise higher (0,41) comparing with relevant studies, see Armstrong et al. (2012).
Table 3 Descriptive Statistics Mean Median Standard deviation 25th percentile 75 percentile Tax variables ETR 0,41 0,53 3,61 0,23 0,67 CETR 0,18 0,21 2,06 0,04 0,34 BTD 0,04 0,06 0,29 0,02 0,11 TAXFEE 0,28 0,08 0,62 0,01 0,30
Firm control variables
GSALE 1 811,99 780,96 2 914,31 247,13 2 264,90
ROA 0,02 0,05 0,28 0,01 0,08
BIG4 0,82 1,00 0,38 1,00 1,00
Table 4 presents the correlations among control variables. Consistent with Table 3, tax avoidance measures are positively correlated with most of control variables, only CETR is negatively correlated with BIG4 variable. Visual inspection of the correlations does not suggest the presence of multicollinearity. To further address this, multicollinearity test diagnoses are performed for all of the independent variables for each of estimated models. In accordance with Chen et al. (2012) interaction variables are included in these test, since they are regarded as continues variables.
Table 4
Correlations among variables
ETR CETR BTD TAXFEE GSALES ROA BIG4 SIZE
ETR 1 CETR 0,8006 1 BTD 0,0280 0,0232 1 TAXFEE -0,0050 0,0014 0,0005 1 GSALES 0,0086 0,0111 0,0958 -0,0302 1 ROA 0,0264 0,0223 0,9690 0,0046 0,0870 1 BIG4 0,0005 -0,0128 0,1174 0,0169 0,2308 0,1202 1 SIZE 0,0037 0,0069 0,0946 -0,0420 0,6633 0,0927 0,3156 1
4.2.
MULTIVARIATE ANALYSIS
This study hypothesizes whether tax NAS fees are negatively associated with tax avoidance. The first model specifically tests whether tax avoidance measure ETR is associated with tax NAS and additional control variables. More specific it is expected, that those firms, that grant a relatively high portion of tax NAS to audit firms, exhibit a stronger proclivity for tax avoidance activities. Thus, in order for hypothesis not to be rejected TAXFEES coefficient needs to be significantly greater than zero.
Inconsistent with predictions, the summarized regression results as presented in a Table 5 indicate insignificant negative TAXFEES coefficient of -0,029 (p-value 0,708). Thereby indicating that, compared to other firms, those firms, that grant a relatively higher portion of tax NAS to audit firms, exhibit slightly lower proclivity for tax planning activities. Hence suggesting, that tax NAS is positively associated with ETR.
Table 5 Model 1a
Variables Coef. t-Stat p-value
Intercept 0,428 3,704 0,00021 TAXFEES -0,029 -0,375 0,708 GSALES 0,000 0,559 0,576 ROA 0,336 1,973 0,049 BIG4 -0,033 -0,250 0,803 SIZE -0,00001 -0,245 0,806 Number of observations 5686 Standard error 3,605 R- squared 0,000791 Adjusted R-squared -0,000089
The second model specifically tests whether tax avoidance measure CETR is associated with tax NAS. It is expected, that firms who engage in tax planning activities, pay greater amounts of tax NAS to audit firms. Thus, in order for hypothesis not to be rejected TAXFEES coefficient needs to be significantly greater than zero. The summarized results of the regression, as reported in Table 6, indicate positive TAXFESS coefficient of 0,007 (p-value 0,875), thereby providing evidence that, firms who pay a greater amount for tax NAS, engage in tax planning activities. Thus indicating, that tax NAS is negatively associated with tax CETR.
Table 6 Model 1b
Variables Coef. t-Stat p-value
Intercept 0,235 3,550 0,0004 TAXFEES 0,007 0,158 0,875 GSALES 0,00001 0,639 0,523 ROA 0,169 1,740 0,082 BIG4 -0,104 -1,377 0,169 SIZE 0,000003 0,189 0,850 Number of observations 5686 Standard error 2,065 R- squared 0,00092
The third model specifically tests, whether tax NAS is negatively associated with tax avoidance measure BTD. More specific it is expected, that those firms, that grant a relatively high portion of tax NAS to audit firms, exhibit a stronger proclivity for tax avoidance activities. Thus, in order for hypothesis not to be rejected TAXFEES coefficient needs to be significantly greater than zero.
Inconsistent with predictions, the summarized regression results as presented in a Table 7 indicate insignificant negative TAXFEES coefficient of -0,0017 (p-value 0,2615). Thereby indicating that, compared to other firms, those firms, that grant a relatively higher portion of tax NAS to audit firms, exhibit slightly lower proclivity for tax planning activities. Hence suggesting, that tax NAS is positively associated with BTD.
It is important to mention, that this model is highly significant as R-squared is 0,939, which implies, that the model explains a large part of the variance in the dependent variable.
Table 7 Model 1c
Variables Coef. t-Stat p-value
Intercept 0,0176 7,5888 0,0000 TAXFEES -0,0017 -1,1231 0,2615 GSALES 0,000002 3,4102 0,0007 ROA 1,0016 292,9903 0,0000 BIG4 -0,0006 -0,2410 0,8095 SIZE -0,000001 -1,1035 0,2699 Number of observations 5686 Standard error 0,072 R- squared 0,939 Adjusted R-squared 0,939
5. CONCLUSION
This paper examines whether and to what extent tax NAS is associated with tax avoidance. Based on tax avoidance and tax NAS literature it is predicted, that tax NAS has effect and is negatively associated with tax avoidance. This study uses three tax avoidance measures: effective tax rate, cash effective tax rate and book tax differences.
The findings are consistent with with this projection only with tax avoidance measure- cash effective tax rate. Other tax avoidance measures: effective tax rate and book tax rate are positively associated with tax NAS, which means are inconsistent with the hypothesis of this study. Using a sample of 5686 observations and 1936 firms over a period between 2010 to 2015 strong evidence is provided that tax NAS is positively associated with ETR and ETR and on the contrary tax NAS is negatively associated with CETR. However, the results are not significantly positive or negative, which suggest, that tax NAS has low influence on tax avoidance activities. It is important to mention, that there is a lack of empirical research related with tax NAS and tax avoidance measures and this study contributes to tax planning literature. The results are consistent with Krishnan et al. (2011) study, who also found no significant association between tax NAS and tax avoidance. However, it is inconsistent with Hogan et al. (2015) research, who found a significant negative relationship between tax NAS and taxes paid.
It is relevant to mention, that results of this study suggests, that tax NAS has no or insignificant association with auditor independence, as a result of insignificant association between tax NAS and tax avoidance.
This study answers multiple calls from prior literature for research regarding determinants of tax avoidance and policy discussion on auditor independence. This study extended existing research on the determinants of tax avoidance as per Hanlon et al. (2010) request and additionally literature related with tax avoidance and tax NAS was extended as per Krishan et al. call for more empirical evidence.
Results of this study are subjected to the following limitations. First of all, tax NAS are not publicly available. While this study provide evidence that tax NAS have value, as Hogan et al. (2015) pointed out, firms that are engaging their auditors for tax services
may also be paying for other firms for tax planning. Secondly, as this study includes firms who paid tax NAS during period 2010-2015, there is a possibility, that some of the firms paid fees to non-audit firms, or did not pay fees to any accounting company and relied on internal tax services.
Future research, should focus on possibility to examine the relationship between tax avoidance and fees for tax services paid to non-audit firms or analyze the association between tax planning activities and internal tax services. Additionally, future research might focus on extending the current set of determinants on tax avoidance, as an example to correlate tax avoidance with environmental and social responsibility constructs.
6. REFERENCES
Abbott, L.J., Parker, S., Peters, G., Raghunandan, K. 2003. “An empirical investigation of audit fees, non- audit fees, and audit committees”. Contemporary Accounting Research, Vol. 20, No. 2, 215-234.
Allingham, M., Sandmo, A., 1972. Income tax evasion: a theoretical analysis. Journal of Public Economics 1, 323–338.
Armstrong, Christopher, Jennifer Blouin, and David Larcker, 2012b. The incentives for tax planning, Journal of Accounting and Economics 53 (1-2): 391-411.
Andreoni, James, Brian Erard, and Jonathan Feinstein, 1998. Tax Compliance, Journal of Economic Literature 36 (2): 818- 860.
Badertscher, Brad, Sharon Katz, and Sonja Olhoft Rego, 2011. The Impact of Private Equity Ownership on Portfolio Firms ́ Corporate Tax Planning. Working Paper, University of Notre Dame, Columbia University, University of Iowa.
Becker, C. L., M. L. Defond, J. Jiambalvo, and K. R. Subramanyam. 1998. “The Effect of Audit Quality on Earnings Management.” Contemporary Accounting Research 15 (1): 1–24.
Bedard, J., Falsetta, D., Krishnamoorthy, G., Omer, T. (2010). “Voluntary disclosure of auditor- provided tax service fees”. The Journal of the American Taxation Association, Vol. 32, No. 1, pp. 59-77
Chen, K.P., Chu, C., 2005. Internal control vs. external manipulation: A model of corporate income tax evasion. RAND Journal of Economics 36, 151–164.
Crocker, K., Slemrod, J., 2005. Corporate tax evasion with agency costs. Journal of Public Economics 89, 1593–1610.
Davis, S.M., Hollie, D. 2008. “The impact of non- audit service fee levels on investors perception of auditor independence”. Behavioral Research in Accounting. Vol. 20, No. 1, 31-44.
DeAngelo, L. E. 1981. “Auditor Size and Audit Quality.” Journal of Accounting and Economics 3 (3): 183–199.
Desai, M.A. & Dharmapala, D. (2009). Corporate tax avoidance and firm value. The Review of Economics and Statistics 91, 537–546.
De Simone, L., Ege, M.S., Stomberg, B. 2015. “Internal Control Quality: The Role of Auditor- Provided Tax Services”. The Accounting Reviw, American Accounting Association Vol. 90, No. 4, 1469- 1496.
Dyreng, S., Hanlon, M., Maydew, E., 2008. Long-run corporate tax avoidance. The Accounting Review 83, 61–82.
Elder, R., Harris, D., Zhou, J. (2008). “Tax consulting and reported weaknesses in internal control”. Working Paper, Syracuse University, Syracuse, NY.
Franzoni, Luigi, 1999. Tax Evasion and Tax Compliance, Encyclopedia of Law and Economics 4 (1): 51-94.
Grant, T., Park, N., Wheeler, S. 2008. “Post-SOX market perceptions of non-audit services and auditor independence”. Advances in Accounting, Finance and Economics. Vol. 1, No. 1, 1-16.
Gleason, C., Mills, L.F. 2011. “Do auditor- provided tax services improve the estimate of tax expense?”. Contemporary Accounting Research, Vol. 28, No. 5, 1484-1509. Gupta, S., Newberry, K., 1997. Determinants of the variability in corporate effective tax
rates: evidence from longitudinal study. Journal of Accounting and Public Policy, 1–34.
Hanlon, M., Heitzman, S., 2010. “A review of tax research”. Journal of Accounting and Economics 50, 127- 178.
Hanlon, M. & Shevlin, T. (2005). Book-tax conformity for corporate income: An introduction to the issues. Tax Policy and the Economy 19, 101–134.
Hogan, B., Noga, T. 2015. “Auditor- provided tax services and long-term tax avoidance”. Review of Accounting and Finance, Vol. 14, No. 3, 285-305.
Huang, D., Chang, M. (2016) Do auditor-provided tax services improve the relation between tax-related internal control and book- tax differences?, Asia-Pacific Journal of Accounting & Economics, 23:2, 177-199.
Jensen, M., Meckling, W., 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305–360.
Kinney, W., Jr., Z.-V. Palmrose, and S. Scholz. 2004. Auditor independence, nonaudit services, and restatements: Was the U.S. government right? Journal of Accounting Research 42 (June): 561–588.
Krishnan, G.V., Visvanathan, G. 2011. “Is the an Association between Earnings Management and Auditor-Provided Tax Services”. American Accounting Association, Vol. 33, No. 2 Fall, 111- 135
Leitz, G (2013) Determinants and consequences of corporate tax avoidance. Working Paper, University of Münster
Markelevich, A., Barragato, C.A., Hoitash, R. (year?). “The nature and disclosure of fees paid to auditors: an analysis before and after Sarbanes- Oxley Act”. The CPA Journal, Vol.75, pp. 6-10
Maydew, E.L. and Shackelford, D.A. (2007). “The changing role of auditors in corporate tax planning”, in Auerbach, A., Hones, J.Jr., Slemrod, J. Taxing Corporate Income in the 21st Century, Cambridge University Press.
McGill, G. A. & Outslay, E. (2004). Lost in translation: Detecting tax shelter activity in financial statements. National Tax Journal 57(3), 739–756.
McGuire, S.T., Omer, T.C., Wang, D. (2012). “Tax avoidance: does tax specific industry expertise make a difference”. The Accounting Review, Vol. 87, No. 3,975-1003. Mills, Lillian, Merle Erickson, and Edward Maydew, 1998. Investments in Tax Planning,
Journal of the American Taxation Association 20 (1): 1-20.
Minnick, Kristina, and Tracy Noga, 2010. Do corporate governance characteristics influence tax management?, Journal of Corporate Finance 16 (5): 703-718.
Omer, Thomas, Karen Molloy, and David Ziebart, 1993. An Investigation of the Firm Size-Effective Tax Rate Relation in the 1980s, Journal of Accounting, Auditing & Finance 7 (1): 167-182.
Omer, T., Bedard, J., Falsetta, D. (2006). “Auditor provided tax services: the effects of a changing regulatory environment”. The Accounting Review, Vol. 81, No. 5, pp. 1095-1117.
Phillips, J., 2003. Corporate tax-planning effectiveness: the role of compensation-based incentives. The Accounting Review 78, 847–874.
Rego, Sonja Olhoft, 2003. Tax-Avoidance Activities of U.S. Multinational Corporations, Contemporary Accounting Research 20 (4): 805-833.
Rego, Sonja Olhoft, and Ryan Wilson, 2012. Equity Risk Incentives and Corporate Tax Aggressiveness, Journal of Accounting Research 50 (3): 775-810.
Richardson, Grant, and Roman Lanis, 2007. Determinants of the variability in corporate effective tax rates and tax reform:
Evidence from Australia, Journal of Accounting and Public Policy 26 (6): 689-704. Robinson, D. 2008. Auditor independence and auditor-provided tax service: Evidence
from going-concern audit opinions prior to bankruptcy filings. Auditing: A Journal of Practice & Theory 27 (November): 31–54.
Rockness, H., and J. Rockness. 2005. “Legislated Ethics: From Enron to Sarbanes-Oxley, the Impact on Corporate America.” Journal of Business Ethics 57: 31–54.
Seida, J. 2003. Enron: The Joint Committee on Taxation’s Investigative Report. Testimony before the Senate Committee on Finance, 108th Congress-First Session: 110–117, Washington: US Government Printing Office.
Simunic, D. 1984. Auditing, consulting, and auditor independence. Journal of Accounting Research 22 (Autumn): 679–702.
Slemrod, Joel, and Shlomo Yitzhaki, 2002. Tax Avoidance, Evasion, and Administration, in: ALAN AUERBACH/MARTIN FELDSTEIN, Handbook of Public Economics, Elsevier Science B.V., Amsterdam, Chapter 22.
Slemrod, J., 2004. The economics of corporate tax selfishness. National Tax Journal 57, 877–899.
Srinivasan, T. N., 1973. Tax evasion: a model, Journal of Public Economics 2 (4): 339-346.
Tang, T., and M. Firth. 2011. “Can Book-tax Differences Capture Earnings Management and Tax Management? Empirical Evidence from China.” The International Journal of Accounting 46 (2): 175–204.
U.S. House of Representatives. 2002. The Sarbanes-Oxley Act of 2002. Public Law 107-204 [H.R. 3763]. Washington, D.C.: Government Printing Office.
Weisbach, D., 2003. Corporate tax avoidance. In: Proceedings of the National Tax Association Ninety Sixth Annual Conference, vol. 9.
Wilson, R. J. (2009). An examination of corporate tax shelter participants. The Accounting Review 84(3), 969–999.
Zimmerman, Jerold, 1983. Taxes and firm size, Journal of Accounting and Economics 5 (1): 119-149.
APPENDIX 1
Variable definitions
Variable Definition Source and data codes ETR The GAAP effective tax rate for the year defined as total income-tax
expense scaled by pre-tax income
Compustat TXT/PI
Cash ETR
The cash effective tax rate for the year defined total income taxes paid scaled by pre-tax income
Compustat (TXDI/PI)
BTD
Book tax difference, pre tax book income minus the sum of federal income taxes and foreign income taxes scaled by tax loss carry forward and all scaled by total assets
Compustat
(BI-((TXFED+TXFD)/TLCF)/AT)
TAXFEE Total annual tax fees (in $millions) paid by the firm over total assets Audit Analytics GSALE Total revenue, current sales turnover Compustat ROA Net income (or loss) scaled by
beginning total assets Compustat NI/AT
BIG4 Audit firm hired for specific year. Audit Analytics, 1 if auditor_fkey= 1,2,3 or 4; 0 otherwise
APPENDIX 2
Model 1a, multivariate analysis
Multiple R 0,028118 R Square 0,000791 Adjusted R Square -0,000089 Standard Error 3,605249 Observations 5685 df SS MS F Significance F Regression 5 58,406 11,681 0,899 0,481 Residual 5679 73814,629 12,998 Total 5684 73873,035 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95,0% Upper 95,0% Intercept 0,428 0,115 3,704 0,000 0,201 0,654 0,201 0,654 TAXFEES -0,029 0,077 -0,375 0,708 -0,180 0,122 -0,180 0,122 GSALES 0,000 0,000 0,559 0,576 0,000 0,000 0,000 0,000 ROA 0,336 0,170 1,973 0,049 0,002 0,669 0,002 0,669 BIG4 -0,033 0,132 -0,250 0,803 -0,292 0,226 -0,292 0,226 SIZE 0,000 0,000 -0,245 0,806 0,000 0,000 0,000 0,000
APPENDIX 3
Model 1b, multivariate analysis Multiple R 0,0303 R Square 0,0009 Adjusted R Square 0,0000 Standard Error 2,0645 Observations 5686 df SS MS F Significance F Regression 5 22,289 4,458 1,046 0,389 Residual 5679 24205,190 4,262 Total 5684 24227,478
Coefficients Standard Error t Stat value P- Lower 95% Upper 95% Lower 95,0% Upper 95,0%
Intercept 0,2347 0,0661 3,5499 0,0004 0,1051 0,3643 0,1051 0,3643 TAXFEES 0,0069 0,0440 0,1576 0,8748 -0,0794 0,0933 -0,0794 0,0933 GSALES 0,0000 0,0000 0,6389 0,5229 0,0000 0,0000 0,0000 0,0000 ROA 0,1694 0,0974 1,7399 0,0819 -0,0215 0,3604 -0,0215 0,3604 BIG4 -0,1041 0,0756 -1,3770 0,1686 -0,2523 0,0441 -0,2523 0,0441 SIZE 0,0000 0,0000 0,1890 0,8501 0,0000 0,0000 0,0000 0,0000
APPENDIX 4
Model 1c, multivariate analysis Multiple R 0,9691 R Square 0,9391 Adjusted R Square 0,9391 Standard Error 0,0725 Observations 5686 df SS MS F Significance F Regression 5 460,2527 92,0505 17526,6550 0,0000 Residual 5679 29,8263 0,0053 Total 5684 490,0790 Coefficients Standard
Error t Stat P-value Lower 95%
Upper 95% Lower 95,0% Upper 95,0% Intercept 0,0176 0,0023 7,5888 0,0000 0,0131 0,0222 0,0131 0,0222 TAXFEES -0,0017 0,0015 -1,1231 0,2615 -0,0048 0,0013 -0,0048 0,0013 GSALES 0,0000 0,0000 3,4102 0,0007 0,0000 0,0000 0,0000 0,0000 ROA 1,0016 0,0034 292,9903 0,0000 0,9949 1,0083 0,9949 1,0083 BIG4 -0,0006 0,0027 -0,2410 0,8095 -0,0058 0,0046 -0,0058 0,0046 SIZE 0,0000 0,0000 -1,1035 0,2699 0,0000 0,0000 0,0000 0,0000
