Structure of egoism A study on the impact of ownership structure on tax avoidance

(1)

Structure of egoism

A study on the impact of ownership structure on tax avoidance

Roel Wymenga Friesestraatweg 14 9718NG Groningen r.b.wijmenga@student.rug.nl 06-38330001 S2560631 Rijksuniversiteit Groningen Faculty Economie & Bedrijfskunde Master Accountancy & Controlling Track Controlling

03-15-2016

(2)

Structure of egoism

A study on the impact of ownership structure on tax avoidance

Abstract

This paper draws on prior research in the field of corporate social responsibility, agency theory, and stakeholder theory to explore the relationship between corporate structure and corporate tax avoidance. This article seeks empirical evidence that the ownership structure of a corporation has an influence on corporate tax avoidance in multinational corporations. Three hypotheses will be analyzed in this study. Hypothesis 1: The use of ESOP has a positive relationship with the probability of corporate tax avoidance in multinational corporations. Hypothesis 2: The number of shareholders, within a corporation, has a negative relation on the degree of corporate tax avoidance. Hypothesis 3: The more concentrated (dispersed) the ownership structure, the higher (lower) the degree of tax avoidance. The sample in this study consists of 224 American corporations listed in the S&P 500 from 18 different sectors. The data used in this research is provided by the financial database ORBIS and also uses information collected through annual reports and the form 10-K. Effective tax rate (ETR) was used to measure the degree of corporate tax avoidance. The results indicate that the presence of an employee stock ownership plan (ESOP) in a corporation has a significant relationship with corporate tax avoidance. Based on prior research the relationship between ESOP and corporate tax avoidance was expected to be a positive relationship, but the results of this research indicate a negative relationship suggesting that the presence of an ESOP is associated with a lower probability of corporate tax avoidance. The differences in the shareholder group, as in size or concentration, do not have an impact on corporate tax avoidance according to this research.

Keywords

Corporate ownership structure, corporate tax avoidance, employee stock ownership plan (ESOP), shareholder group size, shareholder group concentration, corporate social responsibility (CSR).

(3)

1. Introduction

In 2013, Facebook had 1.23 billion monthly active users worldwide (Sedghi, 2014). The British broadcasting corporation (BBC) reported that the use of Facebook has a negative influence on the users (Facebook use, 2015). Facebook not only has a negative impact on individual users, it also negatively impacts society as a whole because Facebook avoids corporate tax payments (Stewart, 2015). Facebook is not the only corporation that uses subsidiaries and other legal ways to avoid tax payments. A lot of multinationals, including Google, Amazon, and Starbucks, also use subsidiaries in an unconventional manner to avoid tax payments (Barford and Holt, 2013) and, therefore, harm society as a whole, which can be seen as egoistic. Does ownership structure (the presence of an employee stock ownership plan (ESOP), number of shareholders and concentrated or dispersed shareholders group) influence the degree of corporate tax avoidance? In the recent years, the mass media has focused on the tax avoidance behavior of corporations1_{. Scholars are also interested in} corporate tax avoidance and are exploring the link between corporate tax avoidance and corporate social responsibility (CSR). CSR can be defined in different ways, which will be presented in the paragraph below.

According to Tang et al. (2015), CSR reflects the extent to which companies (including multinationals) actively engage in social initiatives in response to a wide range of stakeholder interests. Fisher (2014) sees CSR as the obligation and inclinations, if any, of corporations organized for profit, to voluntarily pursue social ends that conflict with the presumptive shareholder desire to maximize profit. Carroll (1979) defined CSR as the belief that corporations have an obligation to be productive and profitable and meet consumers and society’s needs, but to comply within the law and ethical responsibilities of business. Aupperle et al. (1985) acknowledge the same CSR definition as Carroll’s. According to Huseynov and Klamm (2012), CSR relates the economic goals of business to their social responsibilities, which includes acting ethically, contributing to economic development, and improving the quality of life of stakeholders. Cheng et al. (2014) define CSR as the voluntary integration of social and environmental concerns in a company’s operations and in their interactions with stakeholders. Joireman et al. (2015) see CSR as actions that appear to further some social good, beyond the interests of the corporation that are required by law. According to the CSR definitions cited above, corporations need to consider the opinions and interests of the stakeholders and not just the opinions and interests of the shareholders in their decisions to make a profit. Corporations need to consider the effects of their decisions, products and the way of production, and how they affect the society as a whole. The United Nations Global Compact (UNGC) failed to recognize paying corporate tax as a part of CSR. UNGC is an organization with the largest CSR initiative involving companies, governments, civil society, and labor organizations (Tax – a corporate responsibility priority, 2014). On their website2_{, they claim that their mission is to create a sustainable and inclusive} global economy that delivers lasting benefits to people, communities and markets. They have ten principles 3_{grouped into four dimensions, which are “human rights, labor, environment and} anti-corruption” to accomplish their mission. They do not include corporate tax payments as CSR. Christensen and Murphy (2004) also conclude that most company directors do not regard tax payments as a part of CSR.

1_{http://www.theguardian.com/business/taxavoidance} 2_{https://www.unglobalcompact.org/what-is-gc/mission}

(5)

Avoiding corporate tax payments could be seen as egoistic due to the fact that tax avoidance may harms the society as a whole. Corporate tax is a big part of government income. According to the Dutch government (Financieel Jaarverslag, 2012), in 2012, 11.9% of the government income came from corporate tax. In the U.S. this percentage was 9.9% in 2013 (Fiscal year 2015 Historical Tables, budget of the U.S. government, 2014). One of the major responsibilities of a government is to provide good healthcare, education, and other government tasks like maintenance of the motorway network. These government tasks contribute to a better society with more wealth and social security. The government needs budgets for these obligations. Most governments receive money by levying various taxes among the population such as income tax, corporate tax, and value added tax (VAT). Freedman (2003), Landolf (2006), and Friese et al. (2008) support this point of view. If all corporations use tax avoidance measures, there will be a gap in government budgets because of decreased tax income, which leads to budget spending cuts. Therefore, corporate tax payment should be included in CSR, but do shareholders and employees (through the influence obtained by the use of an ESOP) agree with this statement or do they see corporate tax payments as costs? As mentioned above, scholars see CSR as a mechanism that corporations not only think about their own profit, but also consider the welfare of the society as a whole. Corporate tax payments will help finance the government tasks, which helps society as a whole. Elizabeth Warren (2011), a Harvard law professor and U.S. Senator, also considers tax as public good.

This study focuses on ESOP, magnitude of shareholders and the difference between concentrated versus dispersed shareholder group as part of corporate ownership structure to see if it has an influence on tax avoidance. The agency theory is an essential theory in this study. Agency theory explains the relationship between shareholders and managers (agents) (Goktan, 2014). Agency conflicts arise when the manager’s interests are not aligned with the interests of the shareholders (maximization of profit). Compensation contracts align the interests of the manager (through their salary) with the interest of the shareholders. According to Jones and Felps (2013), managers of U.S. corporations want to maximize the wealth of company shareholders, primarily through maximization of profits. Managers see corporate tax payments as costs that should be reduced to maximize profit. Thus, through compensation contracts, shareholders stimulate managers to reduce corporate tax payments. Unfortunately, not everyone has the same view of tax as public good. According to agency theory, shareholders have great influence on the behavior of managers because shareholders can use compensation contracts to influence the behavior of the manager, but that is not the only influence shareholders have on the managers. Shareholders can also replace the manager by dismissing the manager if the manager does not meet the shareholders’ expectations. Thus, the ownership structure of a corporation influences the probability of corporate tax avoidance. Scholars have investigated the influence of ownership structure on risk profile (Dhillon and Rossetto, 2015), the difference between family owned versus not-family owned corporations with respect to agency conflicts between owners and creditors (Pindado, Requejo and De la Torre, 2015) and the influence of ownership structure on CRS (excluding corporate tax payments) (Kiliç, Kuzay and Uyar, 2015). There are few empirical studies investigating the effect of ownership structure on the corporate tax avoidance of multinational corporations. Chen et al. (2010) have studied this, but limited their research to only the structural ownership differences of family firms versus non-family firms. This study will contribute to the understanding of the influence of ownership structure on corporate tax avoidance of multinational corporations.

(6)

From a financial controller point of view, CSR is often seen as a cost because it takes some investments to make a non CSR corporation into a true CSR corporation (Galaskiewicz, 1997; Navarro, 1988). There is evidence that CSR has a positive effect on corporations. Joireman et al. (2015) found that customers of CSR corporations have positive emotions about the corporation that reduced negative word of mouth and increased positive word of mouth, which in turns promoted the image of the corporation. Cheng et al. (2014) found evidence that corporations with superior CSR performance are better positioned to obtain financing in capital markets and market participants are more willing to allocate scare capital resources to those corporations. Corporations that engage in CRS activities face lower capital constraints. CSR could provide a better image and lower the capital constraints, and thereby help construct better performance of the corporation. If a corporation starts using corporate tax payments as a form of CSR, it could distinguish itself from other CSR corporations, which could result in an even better image and possibly further reduce the constraints of capital. Using corporate tax payments as CSR may be of interest to corporations because it could result in better performance.

This article seeks empirical evidence that the ownership structure of a corporation has an influence on corporate tax avoidance in multinational corporations. The main question of this research is therefore: Does ownership structure influence the degree of corporate tax avoidance?

The remainder of this article is as follows. The next sections of this article address the theoretical background and develop the hypotheses. Then the research method is presented. Next, the findings of this study are discussed and this article closes with a discussion and conclusion.

2. Theoretical background

This section begins with a discussion of the different views of corporate tax and CSR, followed by the influence of corporate structure. It ends by addressing two theories, agency theory and stakeholder theory, related to corporate tax avoidance.

2.1 Differing views of corporate tax and CSR and previous literature

As mentioned in the beginning of the introduction, corporate tax avoidance in combination with CSR is a relatively novel subject. Corporations, politicians, and other stakeholders are still forming opinions about whether taxation should be included in CSR. According to Davis et al. (2013), some view corporate tax as payments to the social welfare of the community. Others argue that high tax rates discourage innovation and investments and harm job creation and therefore prevent the ability of the corporation to contribute to social welfare. Davis et al. (2013) state that there is a difference in how corporations view and report taxes in their corporate accountability reports. Some corporations state that they contribute to the community through ongoing operations of the business and taxes, while other corporations state that taxes are harmful to innovation and economic development. Davis et al. (2013) also report that public corporations are sensitive to how their tax payments are perceived by stakeholders. As an example, in an April 28, 2012 article in the New York Times, Apple stated that it generated almost $5 billion in federal and state taxes. Interestingly, Apple includes in its tax payments the amount of taxes paid by their employees. This shows that Apple tries to create a good tax payment image by pretending to pay more corporate tax than they actually do. Some corporations, like the Coca-Cola Company, do not mention taxes at all in their corporate accountability reports, suggesting that these corporations view taxes as being of little relevance to their stakeholders. There is minimal research in this area, but scholars have started to investigate the corporate tax avoidance practices in combination with CSR.

(7)

Scholars already investigated different aspects of CSR and the influence on corporate tax avoidance. Hoi et al. (2013) investigated the relationship between irresponsible CSR activities and the likelihood of tax-sheltering activities. Their results suggest that corporations with excessive irresponsible CSR activities are more aggressive in avoiding taxes, lending credence to the idea that CSR could be viewed as a facet of corporate culture that affects corporate tax avoidance. In contrast, Lanis and Richardson (2015) examined whether corporate social responsibility performance is associated with corporate tax avoidance. Their results show that corporations that are more socially responsible are likely to exhibit less tax avoidance. Additional analysis indicates that the CSR categories of community relations and diversity represent particularly important elements of CSR performance that reduce tax avoidance. Huseynov and Klamm (2012) focused their research on the effect of three measures of CSR, corporate governance, community, and diversity on corporate tax avoidance that use auditor provided tax services and they found mixed results. Chen et al. (2010) studied the relationship between corporate ownership structure (family firms versus non-family firms) and tax avoidance. Their results show that family owned firms are less likely to be tax aggressive than non-family owned firms. There are other types of ownership structure that should be examined, including ESOP, which will be done in this study.

One of the big four accounting firms (PwC, 2008) has published a document where they state: “Paying tax is considered to be an important part of a company’s economic impact and contribution to society, as taxes fund social investment. A call for a CR (corporate responsibility) agenda on tax is not simply a call for companies to pay more tax.” PwC says that taxes are not a simply legal concept, but there are ethical values that can and should be contemplated when considering and setting a company’s tax strategy. Vodafone, a multinational telecommunications company, published a report about corporate tax and the risk management strategy (Vodafone, 2015). Vodafone believes that it has “an obligation to pay the amount legally due in any territory in accordance with the rules set by governments and in doing so it is not able to determine the ‘fair’ amount of tax to pay. If tax laws are unclear or subjected to interpretation, Vodafone will settle in their favor.” Vodafone wants to develop and foster good working relationships with tax authorities, government bodies, and other third parties related to corporate tax. The risk Vodafone wants to manage in combination with corporate tax payments is the impact on their corporate reputation and brand name. Dyreng et al. (2014) investigated if public pressure influences corporate tax behavior through corporate reputation, among other things. Their findings suggest that activist groups can influence corporate outcomes and suggest real consequences to disclose policies, implying that Vodafone has it right. Aggressive corporate tax strategy could have a negative impact on the corporate reputation and brand name. Hanlon and Slemrod (2007) investigated the impact of tax aggressiveness and the reaction on the stock price of a corporation. They found that if the media reports of the company’s tax aggressiveness, the stock price drops, but not in the same drastic way if the media reports about other accounting mishaps. It still indicates that corporate tax avoidance could hurt the corporate reputation or brand name.

(8)

2.2 The influence of ownership structure

In this study, we focus on ESOP, the number of shareholders, and the concentration or dispersion of shareholders and its impact on corporate tax avoidance in corporations listed in the S&P 500. Prior research have not researched this subject, but have focused on other subjects within ownership structure. La Porta et al. (1998) show us that ownership structure (concentrated versus dispersed ownership) differs by country and that dispersed ownership in large public companies is a simple myth. Also, ownership concentration varies by legal origin (English civil law, French civil law, German civil law, and Scandinavian civil law). According to La Porta et al. (1998), the results of their study support the idea that heavily concentrated ownership results from weak protection of investors in a corporate governance system. The heavily concentration of ownership result in increased protection of investors. Demsetz and Lehn (1985) argue that the structure of corporate ownership varies systematically in ways that are consistent with value maximization. Song et al. (2015) show us that ownership structure (state ownership versus non state ownership) has an influence on innovation. A high level of managerial ownership and high level of ownership concentration has a positive interaction with market orientation on innovation performance. Ersoy and Koy (2015) found that large shareholders and concentrated ownership of corporations would lead to different financial performances. Acero and Alcade (2014) indicate that insider ownership (ESOP) has a relevant influence on the percentage of outside directors. “As the significant shareholders increase their ownership of the company, the percentage of independent directors on the board decreases (Acero and Alcade, 2014, p. 653).” Gordon and Poundt (1993) found that structure of ownership significantly influences voting outcomes, like ESOP and concentrated or dispersed shareholder ownership. These studies show that ownership structure has multiple influences on corporations. The influence of ownership structure on corporate tax avoidance is still not clear, except the difference between family owned and non-family owned corporations.

2.3 Agency theory and tax avoidance

Eisenhardt (1989) originally developed the agency theory to conceptualize three problems, goals asymmetry, risk asymmetry and information asymmetry, which arise out of a principal (e.g. an corporation) delegating work to an agent (e.g. an employee) in a contractual relationship between them. According to Jensen and Meckling (1976), agency theory suggests that managers may be driven by interest, and unless they are restricted from doing so, managers will undertake self-serving activities that may be detrimental to the economic welfare of the principal, leading to an agency problem. Pepper et al. (2015) describes agency theory as inter alia, when boards of directors, acting on behalf of shareholders, must design incentive contracts that make an agent’s compensation contingent on measurable performance outcomes in order to motivate executives (agents) to carry out actions and select effort levels that are in the best interests of shareholders (principals).

According to Hanlon and Slemrod (2007), shareholders want to maximize shareholder value by minimizing corporate tax payments net of the private costs of doing so, but shareholders do not run corporations, managers (CEOs) do. Managers do not always have the same interests as shareholders, and because of information asymmetry, shareholders do not always have the same information that managers have (Ndofor et al., 2015). Shareholders want to have the best rate of return on their investments, but managers might have a different agenda. Managers are accountable for their actions to shareholders, to employees, the government and other stakeholders, which increases the scope of their agenda. Shareholders must ensure that managers have the same interests as they do, and therefore, shareholders need control the managers. Because shareholders appoint managers and determine their compensation, shareholders can greatly influence managers, and therefore, greatly influence corporate tax avoidance decisions.

(9)

2.4 Stakeholder theory and tax avoidance

Paul (2015) indicates that stakeholder theory takes as its basic premise that different stakeholders have different interests at risk, and that the alignment of these interests provide the central mechanism through which value is created by the business. According to Harrison and Wicks (2013, p. 97), “stakeholder theory exists in tension (at least) with shareholder theory (Friedman, 1970), that stakeholder theory provides a vehicle for connecting ethics and strategy (Phillips, 2003), and that corporations that diligently seek to serve the interests of a broad group of stakeholders will create more value over time.” Strand and Freeman (2015) depicts companies and their stakeholders as having shared interests where the role of managers is to focus on creating more value for a greater number of stakeholders by promoting and expanding upon the shared interests between the company and its stakeholders. Sikka (2010) argues that reduction of taxes may benefit shareholders, but the avoidance of taxes may be at the expense of the society as a whole, which embraces stakeholder theory. Therefore, stakeholder theory is based on the positivity of tax payments for the society as a whole and that corporate tax benefits the stakeholders.

Both theories (agency and stakeholder) show the importance of the ownership structure of a corporation. Agency theory shows that shareholders and managers do not always have the same agenda and that shareholders want to influence the managers’ decisions, including on matters such as corporate taxes. Stakeholder theory indicates that focusing on a broader group than shareholders could lead to more wealth, which implies that the shareholder theory impacts corporate tax, because shareholders want to minimize corporate tax and other stakeholders (like the government) do not. Next, this research will examine the hypothesis, to test if ownership structure has an influence on corporate tax avoidance.

3. Hypothesis development

This article studies the relationship between ownership structure and corporate tax avoidance. In this study, ESOP, magnitude of shareholders, and the difference between concentrated versus dispersed ownership structure will be examined in relation to corporate tax avoidance. ESOP will be discussed first.

3.1 Employee stock ownership plan

Before the hypothesis regarding the ESOP is developed, there will first be given an explanation of the ESOP and what it entails. According to the ESOP Association4_{an ESOP is a benefit plan for employees} which makes the employees of a corporation owners of stock in that corporation. The National Center for Employee Ownership5_{(NCEO) explains how an ESOP works. Corporations set up a trust} fund for employees and use this trust fund to buy company stock and contribute these shares directly to the plan. Employees can also choose to invest in company stock or it can be given as a bonus by the corporation. Thus through the use of an ESOP, employees become shareholders of the corporation for which they work.

4_{http://www.esopassociation.org/explore/how-esops-work/what-is} 5_{http://www.esop.org/}

(10)

Katsva and Condrey (2005) made an important finding that employees could differ mentally and could be motivated and stimulated in various ways because of cultural differences. Hofstede (1980, 2001) made a culture model, which is the most outstanding theory about different national cultures (Scheffknecht, 2011; Brewer and Venaik, 2014). Hofstede found that four dimensions distinguish different national cultures. These dimensions are “uncertainty avoidance, power distance, individualism versus collectivism and masculinity versus femininity.” These cultural differences could impact the different motivation and satisfaction of employees. The aspect of cultural differences is omitted in this study.

I argue that the presence of an ESOP ensure that employee focus on the financial aspects of the corporation, such as reducing costs (by avoiding corporate tax payments). I base this assumption on studies that examine the impact of CEOS that have shares of the corporation for which they work. Desai et al. (2006) investigated the influence of stock ownership on CEOs and found that excesses in stock-based incentive compensation can motivate CEOs to use overly aggressive accounting practices. O’Connor et al. (2006) found that stock ownership could also lead to misreport firm financial results. If CEOs are motivated through stock ownership, CEOs are motivated to increase profits because the CEOs will directly benefit from these profits through increasing stock price. If CEOs are motivated to increase profits because of stock ownership, than the same should hold for employees in combination with ESOP. Employees benefit from increasing profits because the share value of the corporation increases, and therefore, the wealth of employees with shares of the corporation increases. Employees are motivated to reduce costs like corporate tax payments, which increases profits. Therefore, if a corporation uses ESOP then employees will not object to corporate tax avoidance because that will increase profits and benefit the employee. The employee will no longer consider the benefits of paying corporate taxes because that will not directly influence the employee, but the society as a whole. An increase in profits (due to corporate tax avoidance) will directly influence the wealth of the employee.

Motivation is very important for the functioning of an employee. There are two ways to motivate an employee: intrinsic and extrinsic motivational factors. Ryan and Deci (2000) define intrinsic motivation factors as desires from within to perform a particular task and extrinsic factors as external influences to an individual and unrelated to the task they are performing. In other words, intrinsic motivation emphasizes rewards derived from the work itself, and extrinsic motivation emphasizes external rewards such as pay. I argue that intrinsic motivational factors are more effective than extrinsic factors, but there are corporations who use extrinsic factors to motivate their employees. One such extrinsic factor is an ESOP. Kim and Ouimet (2014) found that employees and shareholders, according to their relative bargaining power, share productivity gains because of ESOP. They suggest that ESOP is a win -win scheme because average wages, the level of employment, and shareholder value increase. Bergstein and Williams (2013) argue that ESOP is a unique financial tool that contributes to a successful business. ESOP corporations are more productive and more profitable; they also have a higher survival rate than corporations without ESOP. Pendelton and Robinson (2010) found that ESOP, in combination with involvement of employees within the corporation, can have a positive productivity effect over a wide range of values for involvement. ESOP could also lead to employee retention. All these studies imply that employees are motivated by financial incentives when an ESOP is in place, but is this the best way to motivate employees?

(11)

Perry et al. (2010) developed a theory of public service motivation (PSM) which posits that public sector employees are more likely to be motivated by intrinsic rewards compared with private sector employees. Ertas (2015) found that job and pay satisfaction, creativity, personal development, promotion based on merit, and having a good work group was very important to the performance of federal employees. These findings imply that corporations should strive to improve workplace characteristics (like, among other things, job satisfaction, personal development and having a good work group) that are valued by all employees and develop human resource management practices and policies to handle an increasingly mobile workforce. Boxall et al. (2015) argue that greater involvement in the decisions that matter to employees can be very motivating and satisfying for workers. They further suggest that empowerment by facilitating greater skill utilization will lead to higher satisfaction among employees. Their study implies that forms of work organizations and supervision that offer employees greater opportunity for discretion and involvement in decisions that concern them create the conditions for greater learning and in turn contribute to better performance of the employees. Tyagi (1985) and Becherer et al. (1982) investigated the job diagnostic model and found that if an employee (salesperson) has problems with their motivation or satisfaction with their work, corporations should first investigate the five core job dimensions. These dimensions are “skill variety, task identity, task significance, job autonomy and job feedback.” Only after investigation of the five core job dimensions should a corporation investigate other aspects of the work situation, such as the compensation system. Job stability is also very important for an employee (Secară, 2014; Cazes and Tonin, 2010). If a corporation only uses ESOP to motivate employees, it might be beneficial to take another look at this way to motivate the employees. It might be better to motivate employees through an intrinsic model instead of through an extrinsic method like ESOP.

An ESOP, due to the extrinsic nature, motivates employees to focus on the financial aspects of their work. As already mentioned in this paragraph, employee are motivated to reduce costs, like tax payments, which contributes to more wealth for the employee due to the increase in share price. Thus, I argue that an ESOP has a positive relationship with corporate tax avoidance in multinational corporations.

Hypothesis 1: The use of ESOP has a positive relationship with the probability of corporate tax avoidance in multinational corporations.

3.2 The composition of the shareholders group

The second and third hypotheses are about the change in corporate tax avoidance in combination with the number of shareholders and the concentration or dispersion of shareholders within a corporation.

Unfortunately, to the best of my knowledge, there are no studies that review the influence of shareholder group size and the relation to decision making. There are however a number of studies about board size and decision-making. Therefore, I use the board size studies to come to my theoretical foundation because the type of group does not matter, but the group size does influence the decision-making process.

(12)

There are extensive studies testing the performance of a board of directors in combination with their size. Nakano and Nguyen (2012) found that it is difficult to reach consensus in large decision-making groups, suggesting that large boards are associated with lower risk taking. Fama and Jensen (1983) argue that an increase in the number of board members slows down the decision-making processes of a corporation. Lipton and Lorsch (1992) and Jensen (1993) found that as size of the board grows, decision-making processes slow down and this causes communication problems. They interpreted these findings as the probability of the presence of communication and coordination problems in the corporations with bigger boards. As boards becomes bigger the probability of communication and coordination problems increases. Loderer and Peyer (2002) also recognize that bigger boards are ineffective due to coordination issues along with the problems of free riding. Loderer and Peyer (2002) also recognize that bigger boards deteriorate the value of a corporation in the financial market. Upadhyay (2015) shows that larger group sizes reduced extremes in group decision outcomes and that large board size leads to moderate corporate decisions and reduces firm risk, which comes down to the same conclusion as that of Nakano and Nguyen (2012).

Corporate tax avoidance put corporations at risk of damaging reputation, which in terms affect the share prices of the corporation (Hanlon and Slemrod, 2007; Dyreng et al., 2014; Vodafone, 2015). All these studies in the paragraph above show that the bigger the group, the harder it is for that group to make a decision and that the decision that a large group makes are less risky. I argue that this effect is also present for the shareholders. The bigger the group of shareholders, the harder it is for the shareholders to make a decision, and the decisions that they make will have less risk. Thus, using corporate tax avoidance measures could be seen as a risky decision. Therefore, the larger the shareholders group is, the less probable it is that the corporation will use corporate tax avoidance measures, due to the risk involved in using corporate tax avoidance measures.

Shareholders are responsible for hiring and firing a CEO and want to maximize profits and minimize costs like corporate tax. If a CEO of a corporation does something that is not in line with the interests of the shareholders, such as not implementing corporate tax avoidance measure (Jones and Felps, 2013; Hanlon and Slemrod, 2007), the shareholders could fire the CEO and replace him with someone who will follow the interests of the shareholders. To fire a CEO, the shareholders need to make a decision, and the decision-making process is influenced by the size of the group. Lager groups are less likely to make a prompt and important decision. Thus, corporations with a larger group of shareholders are less likely to fire the CEO, and therefore, the shareholders have less influence on the CEO. This means that shareholders cannot put the CEO under pressure to force corporate tax avoidance measures. The agency theory and stakeholder theory shows that CEOs have a bigger agenda than shareholders. CEOs have to take other stakeholders into account (for example employees or the government). Because shareholders in corporation with a larger group of shareholders have less influence, I argue that there is less pressure on the CEO by the shareholders regarding corporate tax avoidance.

Thus, because of the risky nature of corporate tax avoidance measures (Hanlon and Slemrod, 2007) and bigger groups tend to make less risky decisions (Upadhyay, 2015; Nakano and Nguyen, 2012) and the decrease of influence of the shareholders on the CEO within large shareholder groups, I argue that the size of the shareholder group influences tax avoidance.

Hypothesis 2: The number of shareholders, within a corporation, has a negative relation on the degree of corporate tax avoidance.

(13)

Shareholders use their voting rights to make their decisions and therefore the voting rights are the essence of the influence a shareholder has on the CEO of a corporation. The number of voting rights is related to the number of shares the shareholders has. If a shareholder group is concentrated, then there are dominant shareholders (shareholders with at least 10% of the corporations voting rights6₎ and if a shareholder group is dispersed than there is a large group of small shareholders within a corporation. Dahaya et al. (2008) state that dominant shareholders have the ability to divert corporate resources from other shareholders to themselves for personal consumption. From this statement it can be deduced that dominant shareholders can, at the expense of other shareholders, make decisions without full consent of all shareholders. Gomes and Novaes (2005) state that if control is shared among several shareholders, the decisions that they make are the result of bargaining among the shareholders. From this statement it can be deducted that, if the shareholder group is dispersed the decisions which they make are the result of bargaining among shareholders. Therefore the more concentrated the shareholder group is, the easier it is to make decisions, which in terms means that the influence of the shareholders on the CEO is higher when the shareholder group is concentrated. I argue that less pressure on the CEO, due to a dispersed shareholder group, results in a lower degree of corporate tax avoidance, because CEOs have a bigger agenda than shareholders do (agenda shareholders: minimize costs, like corporate tax) and CEOs have to take the agenda of the stakeholders (for example the government) into consideration (stakeholder and agency theory). And therefore, the following hypothesis will also be tested in this study.

Hypothesis 3: The more concentrated (dispersed) the ownership structure, the higher (lower) the degree of tax avoidance.

The research methods of this study are discussed in the next section of this paper.

4. Research design

This section explains how the research was carried out, starting with the description of the sample design. Then the dependent, independent, and the control variables are discussed. This section concludes with a presentation of the statistical model.

4.1 Sample

The sample of this research is based on corporation who are listed in the Standard & Poor’s 500. The data is collected through the financial database ORBIS (actual data from 2014, unfortunately no historical data) for the insights into the ownership structure of the corporations regarding the number of shareholders and how concentrated or dispersed the number of shareholders are. To find out which corporations have an ESOP, this study used information provided by the NCEO7_{. The NCEO} provided a list including all the Standard & Poor’s 900 corporations and information about having an ESOP or not. This study also uses information collected through annual reports and the form 10-K released by the corporations themselves. Some corporations were excluded from the sample due to insufficient data. All observations where the corporation declares negative ETR8_{or losses, the ETR} was set to 0% (Rodriguez and Martinez-Arias, 2014). Based on the research of Fernández-Rodriguez and Martinez-Arias (2014), a maximum of 100% was set for the ETR and all cases that exceeded that were set to 100% ETR. All corporations are grouped by industrial sector based on the

6_{Dahaya et al. 2008 p. 14} 7_{https://www.nceo.org/}

(14)

“Bureau van Dijk major sector indicator” provided by the database ORBIS. The sample in this study consists of 224 American corporations from 18 different sectors.

4.2 Dependent variable

In this study, corporate tax avoidance is the dependent variable. According to Simone et al. (2015), there are four common proxies for identifying tax avoidance, based on financial statement data. These four proxies are GAAP effective tax rate, cash effective tax rate, total book-tax differences, and permanent tax differences. Calculating the cash effective tax rate, total and permanent book-tax differences is very time-consuming and because of the short time span of this research, the corporate tax avoidance is calculated based on the GAAP effective tax rate (ETR). The ETR is calculated by dividing the total tax expense (TTE) by the pre-tax income (PTI). These figures are extracted from the annual reports and the form 10-K reports of the Standard & Poor’s 500 corporations.

ETR = TTE / PTI

According to Lee et al. (2015), ETR include both current and deferred tax expenses and reflect permanent differences between book and taxable incomes with statutory adjustment. A strategy to defer tax payments does not alter the ETR, and the total income tax expense does not necessarily reflect a tax liability. Some accrual adjustments, like changes in the valuation accounts, affect the book income, not the taxable income. To reduce the noise in the proxy for tax avoidance, this study will use a one year ETR from 2014 (ETR1) and a three year average ETR from 2012, 2013, and 2014 (ETR3). Using an average ETR reduces the noise resulting from timing differences within taxation, like accrual accounting (Buijink et al., 2002). The longer the time period, over which the average ETR is calculated, the less noise it will have. In this study, the average ETR is calculated over a time span of three years. The reason why the time span is three years is as follows. The data concerning the ownership structure is based on actual data. Using a long time span to calculate the average ETR results in noise between the relationship with the actual ownership data and the average ETR. There must be a trade-off between the noise with respect to the ETR, where a longer period of time reduces the noise, and the noise with respect to the ownership structure, due to having only actual data and no historic data. I argue that ownership structure fluctuates less over time than the ETR, and therefore, I use a three-year average of ETR.

4.3 Independent variables

In this study, there are three key independent variables: whether or not the presence of an ESOP within a corporation (hypothesis 1), number of shareholders (hypothesis 2) and the degree of concentration or dispersion of the shareholders (hypothesis 3). The data is collected through the database ORBIS and the information provided by the NCEO. There is one9_{way of identifying ESOP} within corporations: through the use of information provided by the NCEO (Chang, 1990). Unfortunately, a caveat has to be made concerning the ESOP data. The NCEO acknowledged that they might have failed to recognize an ESOP in some corporations.

9_{ESOP identification is also possible trough Factiva search}9_{(Kim and Ouimet, 2014). An expensive subscription} is needed to use a Factiva search. Receiving the information from the NCEO is another option that is not free, but the expenses were manageable, especially compared with the cost of a Factiva search. That is why only the information provided by the NCEO was used to identify corporations that have an ESOP.

(15)

To measure the total number of shareholders, the database ORBIS was used. This database provided the current information about the total number of shareholders within a corporation. This was used as measurement for the number of shareholders within a corporation.

To measure the level of concentration or dispersion within the shareholders group, this study uses the percentage of direct ownership of the top ten shareholders (TOP10SHARE) within the corporation. This is based on the basic principle used in the research of Varma et al. (2009). If the top ten shareholders have a high percentage of direct ownership (for example 90%) within the corporation, then the shareholders are concentrated. If the top ten shareholders have a low percentage of direct ownership (for example 10%) within the corporation, than the shareholders are dispersed. The information about the top ten shareholders is provided through the use of the database ORBIS. Unfortunately, not all of the 224 corporations had this information listed on ORBIS, so there are 42 missing values.

4.4 Control variables

In this study, we also examine other mechanisms that could affect the degree of corporate tax avoidance to increase the robustness of our theoretical model. Based on prior research, there are known control variables that scholars consider to have an influence on the corporate tax avoidance practice of corporations, e.g., return on net operating assets, net operating loss carry forward at the beginning of the year, foreign operations, intangible assets, equity in earnings of unconsolidated affiliates, leverage, sales growth, and firm size (Jaafar and Thorntonq, 2015; Badertscher et al., 2013). Due to data availability and the time consuming nature of some control variables, the control variables leverage (LEV), return on equity (ROE), and firm size (ASSETS) are used in this study.

Leverage, the ratio of the total long term debt to total assets, influences corporate tax avoidance because corporations with greater leverage might have less need to avoid tax payments due to the tax benefits of debt financing. ROE is used as a proxy for the current profitability, which influences corporate tax avoidance behavior because corporations that are more profitable have more incentives to avoid taxes. The ROE is calculated using the profit before tax to reduce the noise of corporate tax avoidance within the proxy for current profitability. Firm size influences corporate tax avoidance because large corporations likely enjoy economies of scale in tax planning. All control variables are collected through the use of the database ORBIS. Due to outliers within the variables ROE and ASSETS, winsorizing is applied. Calculating the mean and the standard deviation sets the maximum and minimum of these variables. The mean minus three times the standard deviation is the minimum and the mean plus three times the standard deviation is the maximum of these variables.

4.5 Statistical model

To test the hypotheses, this study uses a linear regression model. The ETR in the statistic model is either the ERT1 or ETR3. The full regression model is as follows:

ETR = β₀+β₁ESOP+β₂NUMBERSHARE+β₃ TOP10SHARE+β₄LEV+β₅ROE+β₆ASSETS+εi

In this statistical model, the βi is the coefficient and the εi is the residual error. In table 1 below, the variables are defined, including how the variables will be measured.

(16)

Table 1: Variables

Variables Description Definition

Dependent variable

ETR1 Effective tax rate 2014 Dividing the TTE through the PTI. The lower the ETR, the higher degree of corporate tax avoidance

ETR3 Average effective tax rate over a three-year time span

Dividing the three-year TTE through the three-year PTI. The lower the ETR, the higher degree of corporate tax avoidance

Independent variables

ESOP Employee Stock Ownership

Plan

Indicator of an ESOP in place within a corporation

NUMBERSHARE The total number of shareholders

The total number of shareholders within a corporation

TOP10SHARE The top 10 shareholders within the corporation

The total percentage of direct ownership from the top ten shareholders within the corporation

Control variables

LEV Leverage Ratio of the total long term

debt to total assets

ROE Return on Equity Ratio of the profit or loss

before tax to shareholders funds

ASSETS Firm size Natural log of total assets

(17)

5. Findings

In this section the results of this research are presented, starting with the descriptive statistics in table 2 below. First, there are differences between ETR1 and ETR3 in terms of the maximum. The mean does not differ that much, but there is a significant difference in the maximum of 38.4 which indicates a higher standard deviation for ETR1. Second, there is a scatter in the ownership structure, in terms of total number of shareholders and direct ownership. The smallest shareholder group consists of five shareholders and the biggest in our sample consists of 152 shareholders. The most concentrated shareholder group has 92.5% direct ownership and the most divided shareholder group has only 2.05% direct ownership. The variable ASSETS is rounded to millions in table 2.

Table 2: Descriptive statistics

To make sure that the independent variables are not strongly correlated, a correlation test was conducted to check for multicollinearity. Table 3 below shows the result of the correlation test. There are some significant relationships, but fortunately, there is no evidence of multicollinearity because the values do not exceed 0.7 or are less than -0.710_{. The results of the correlation test are} two-tailed because it was important to know if there was any relationship, though the direction of the relationship was not important. From now on all other tests will be one-tailed because we suspect a direction within the relationship.

10_{These values are based on the “rule of thumb” used by the Rijksuniversiteit Groningen to check for} multicollinearity in a correlation matrix.

Variables N Average Standard deviation Min Max Dependent ETR1 224 0.2763 0.1370 0 1 ETR3 224 0.2730 0.1092 0.0005 0.6163 Independent ESOP 224 0.44 0.497 0 1 NUMBERSHARE 224 98.86 17.469 5 152 TOP10SHARE 182 0.1971 0.1642 0.0205 0.9250 Control LEV 224 0.2354 0.1389 0 0.6682 ROE 224 0.3253 0.3655 0 2.27188 ASSETS (mln) 224 62810 143584 1647 801086

(18)

Table 3: Correlation coefficient

Tables 4 and 5 show the results of the regression analysis, where table 4 shows the results of the regression with ETR1 and table 5 the results of the regression with ETR3. Both tables work in the same way and have five models incorporated. The first model shows the relationship between the ETR and the control variables. Models 2, 3, and 4 show the relationship between ETR and the control variables plus the three hypotheses, which were tested one by one. Model 5 demonstrates the relationship between ETR, the control variables, and all three hypotheses together. Model 5 is the most comprehensive model in this research. In both tables, the adjusted R-squared is the highest in model 5, which indicates a higher standard deviation. The F-value of both tables is highly significant with a level of 1%, indicating that the independent variables explain the changes in the dependent variable. Unfortunately, even though the adjusted R-squared is the highest in model 5, it is still relatively low, indicating that the independent variables only explain a small proportion of the changes in the ETR. Also, I performed a second test to check for multicollinearity. No variance inflation factor (VIF) value is higher than 10, again indicating no multicollinearity among the independent variables.

Table 4: Regression analysis ETR1 ETR1 ETR3 ESOP NUMBERSHARE TOP10

SHARE

LEV ROE ASSETS

ETR1 1 ETR3 1 ESOP 0.091 0.103 1 NUMBERSHARE 0.036 -0.081 -0.231*** 1 TOP10SHARE -0.086 -0.065 -0.167** -0.086 1 LEV -0.108 -0.159** 0.051 0.025 -0.021 1 ROE 0.157** 0.184*** -0.042 0.019 0.000 0.097 1 ASSETS -0.091 -0.131* 0.155** -0.028 -0.016 -0.168** -0.124* 1

*** Correlation is significant at the 1% level (2-tailed) ** Correlation is significant at the 5% level (2-tailed) * Correlation is significant at the 10% level (2-tailed)

ETR1 1 2 3 4 5 Constant 0.295*** 0.283*** 0.269*** 0.312*** 0.224*** ESOP (H1) 0.034** 0.042** NUMBERSHARE (H2) 0.000264 0.001 TOP10SHARE (H3) -0.081 -0.054 LEV -0.137** -0.147** -0.138** -0.174** -0.193***

ASSETS -8.990E-14* -1.093E-13** -8.916E-14* -9.197E-14* -1.110E-13*

ROE 0.059* 0.061*** 0.059*** 0.062** 0.064**

Adjusted R-squared 0.035 0.046 0.032 0.040 0.051

Highest VIF 1.042 1.070 1.042 1.039 1.113

F-value 3.728*** 3.687*** 2.852** 2.879** 2.621***

N 224 224 224 182 182

(19)

Table 5: Regression analysis ETR3

In both tables 4 and 5, it is clear that the control variables have a significant relationship with ETR. The significance level ranges from the 10% significant level up to and including the 1% significant level. The control variables behave as expected, except for ROE. The relationship between ETR and ROE is exactly opposite of what was expected. At the beginning of the research, the expectation was that more profitable corporations, indicated by a high ROE, would avoid more tax and have a lower ETR. Our research suggests that if the ROE increases, the ETR also increases. Thus, this research indicates that the more profitable corporations in our sample have a higher ETR.

In both tables 4 and 5, a significant relationship, at the 5% level, is visible between ESOP and ETR in both model 2 and 5. The differences between ETR1 and ETR3 impact the relationship. Within ETR1, the impact of an ESOP is bigger than in ETR3. This is probably due to the greater diversification of the different ETR1 compared with the diversification within ETR3. The results of this research suggest that if a corporation has an ESOP in place, the corporation has on average either 3.4% (ETR1) or 3.2% (ETR3) higher ETR, compared with a corporation that has no ESOP in place (model 2). This is exactly the opposite of what was expected. It was expected that there would be a positive relationship between ESOP and corporate tax avoidance (meaning that the presence of an ESOP would indicate a lower ETR, instead of the measured higher ETR) instead of the observed negative relationship. Therefore, hypothesis 1 must be rejected.

The number of shareholders (hypothesis 2) has only a slightly significant relationship at the 10% level with ETR3 (table 5) in model 3. The direction of this relationship is also the opposite of what was expected, based on prior research. The expectation was that if the number of shareholders within a corporation would increase the ETR also would increase, but this research indicate that the ETR actually decreased. In model 5, this relationship is not significant anymore. The number of shareholders does not have a significant relationship with ETR1. Therefore, hypothesis 2 must also be rejected. In this research, I found no relationship between the percentage of direct ownership of the top ten shareholders and the ETR (either the ETR1 or ETR3). Again, hypothesis 3 must be rejected. All the hypotheses in this study must be rejected.

ETR3 1 2 3 4 5 Constant 0.299*** 0.288*** 0.350*** 0.311*** 0.330*** ESOP (H1) 0.032** 0.031** NUMBERSHARE (H2) -0.001* -0.000316 TOP10SHARE (H3) -0.050 -0.038 LEV -0.158*** -0.167*** -.157*** -0.186*** -0.196***

ASSETS -1.075E-13** -1.261E-13*** -1.090E-13** -1.158E-13** -1.308E-13***

ROE 0.055*** 0.057*** 0.056*** 0.055*** 0.056***

Adjusted R-squared 0.072 0.089 0.075 0.079 0.091

Highest VIF 1.042 1.070 1.042 1.039 1.113

F-value 6.772*** 6.473*** 5.517*** 4.866*** 4.037***

N 224 224 224 182 182

(20)

As mentioned in the section “Research design”, the samples are grouped by industrial sector. The different sectors, together with the frequencies in which they appear in the sample, are shown in table 6 below. During this research, the relationship between the four11_{largest sectors (except for} “Other services” because the content of that sector is unknown) and the ETR was examined. The other sectors are being disregarded due to the low number of frequencies (N) in which they occur in the sample. If the other sectors would have been distinguished, the results, due to the low frequencies, of these analysis would not have been robust. Tables 7 and 8 give the outcome, where table 7 addresses the ETR1 and table 8 the ETR3. Both tables have four models, where every model consists of the control variables, the three hypotheses, and one of the four biggest sectors. The sectors “Machinery, equipment, furniture, recycling” and ”Wholesale & retail trade” have significant differences in ETR, compared with the other sectors. On average, the sector “Machinery, equipment, furniture, recycling” has either 6.5% in ETR1 or 4.1% in ETR3 lower ETR compared to the other sectors. This contrasts with the sector “Wholesale & retail trade,” where the ETR is on average either 5.3% in ETR1 or 8.7% in ETR3 higher. For the sector “Gas, water, Electricity,” there was only a slightly significant relationship at the 10% level for the ETR3 and no significant relationship for the ETR1. Although the ETR3 gives a more robust result due to less noise within the ETR variable, the relationship is not very significant, and therefore, I cannot state that there was a relationship. The next section of this report discusses the results and some limitations of the study. It also addresses recommendations for further research.

Table 6: The different sectors

Industry N

Banks 15

Chemical, rubber, plastics, non-metal products* 22

Constructions 4

Education, health 3

Food, beverages, tobacco 14

Gas, water, electricity* 24

Hotels & Restaurants 4

Insurance companies 1

Machinery, equipment, furniture, recycling* 43 Metals & metal products 5 Post & telecommunications 2

Primary sector 7

Publishing, printing 7

Textiles, wearing apparel, leather 3

Transport 4

Wholesale & retail trade* 21

Wood, cork, paper 3

Other services 42

Total: 224

11_{Indicated by an asterisk (*) in table 6.}

(21)

Table 7: Sector differences ETR1

Table 8: Sector differences ETR3

ETR1 1 2 3 4 Constant 0.231*** 0.214*** 0.217*** 0.229*** ESOP (H1) 0.043** 0.040** 0.039** 0.039** NUMBERSHARE (H2) 0.001 0.001 0.001* 0.001 TOP10SHARE (H3) -0.059 -0.046 -0.041 -0.060 Chemical -0.023

Gas, water, electricity 0.038

Machinery -0.065***

Wholesale 0.053*

LEV -0.193*** -0.209*** -0.186*** -0.186***

ASSETS -1.107E-13** -1.083E-13* -1.157E-13** -1.027E-13*

ROE 0.068** 0.068** 0.067** 0.064**

Adjusted R-squared 0.048 0.049 0.080 0.055

Highest VIF 1.121 1.184 1.116 1.129

F-value 2.304** 2.332** 3.239*** 2.503***

N 182 182 182 182

ETR3 1 2 3 4 Constant 0.336*** 0.315*** 0.325*** 0.337*** ESOP (H1) 0.032** 0.028* 0.029** 0.025* NUMBERSHARE (H2) -0.000356 -0.000175 -0.000201 -0.000443 TOP10SHARE (H3) -0.043 -0.026 -0.030 -0.049 Chemical -0.021

Gas, water, electricity 0.057*

Machinery -0.041**

Wholesale 0.087***

LEV -0.196*** -0.220*** -0.192*** -0.185***

ASSETS -1.306E-13*** -1.269E-13*** -1.338E-13*** -1.171E-13**

ROE 0.059*** 0.062*** 0.057*** 0.055***

Adjusted R-squared 0.090 0.099 0.109 0.129

Highest VIF 1.121 1.184 1.116 1.129

F-value 3.550*** 3845*** 4.149*** 4.828***

N 182 182 182 182

(22)

6. Discussion and conclusion

This chapter sums up the conclusions of this research, discusses some limitations of the research, and offers some recommendations for future research.

6.1 Discussion and conclusion

As already noted in the “Findings” section, all the set hypotheses were rejected based on the outcome of this research. Though the significant relationship between having an ESOP and corporate tax avoidance (Hypothesis 1) was interesting. The direction of the relationship was found to be negative instead of positive what was expected based on previous research, suggesting that the presence of an ESOP indicate less corporate tax avoiding measures (based on the measured lower ETR compared to corporation without an ESOP). There are some reasons that could explain why the relationship between an ESOP and corporate tax avoidance was found to be negative instead of positive. One explanation might be that the employees see tax payments as a public good, just like Elizabeth Warren (2011), and that employees adhere to the stakeholder theory (Sikka, 2010). The reason why employees might see corporate tax avoidance as a public good more than shareholders do is that shareholders do not always live in the same region where the corporation’s offices is located, and therefore, do not see the effect of the paid corporate taxes. Employees live in the same region and might see the differences that have been made by the government (improving healthcare, education, or infrastructure), paid by, among other things, corporate tax payments (total corporate tax income respectively 11.9% in the Netherlands and 9.9% in the U.S.). Also most governments present their income and expenditure to the citizens through the use of the media12_{, so} that the citizens (and most importantly the employees) are aware of what happened to the corporate tax payments. Shareholders on the other hand, do not always live in the same country of the corporation in which they are shareholders, and therefore are not aware of what happened to the corporate tax payments. The employees might use their power, gained through their rights provided by the ESOP, to block the corporate tax avoidance intention of the shareholders, ensuring that the corporation pays their “fair share” of tax, just like Vodafone attempts to prosecute (Vodafone, 2015), so that the corporation can assist the society as a whole. In other words, the employee might see corporate tax payments as CSR.

Another explanation might be that the employees fear the consequences of corporate tax avoidance measures. As already mentioned in the section “Hypothesis development”, if the public finds out that a corporation uses corporate tax avoidance measures, it has a negative impact on the corporation, which might affect the job stability of the employees (Hanlon and Slemrod, 2007). Job stability is very important for an employee (Secară, 2014; Cazes and Tonin, 2010), which makes it likely that employees would want to thwart corporate tax avoidance plans of the shareholders. Another possibility is that the employees fear the negative effect of the tax avoidance measures would have a great impact on the share price, which impacts the wealth of the employee through the ESOP. Again, this might cause the employees to thwart the corporate tax avoidance plans. These are perhaps the reasons why the relationship between ESOP and corporate tax avoidance was found to be negative instead of the expected positive.

(23)

Based on the outcome of this research, there is no relationship between the shareholders group in terms of size (Hypothesis 2) or concentration (Hypothesis 3). There might also be multiple reasons for this lack of relationship. First it might be that the shareholders, on average, all agree that corporate taxes are costs and that costs must be minimized to maximize the return on equity or shareholder wealth (Hanlon and Slemrod, 2007; Jones and Felps, 2013). Therefore, the size or the concentration does not have an influence on the discussion of the decision of corporate tax avoidance because all the shareholders have the same opinion (taxes are costs), so the decision need not be discussed. Secondly, the first line of reasoning also applies to the influence of the shareholders on the CEOs. If a CEO disregards the desires of the shareholders by not implementing corporate tax avoidance measures, the shareholders could fire the CEO and hire a new CEO to implement the desire of the shareholders. This decision to fire the CEO is easy to make if all the shareholders holds the same vision. If all shareholders see corporate tax payments as costs and the CEO does not want to minimize those costs, the decision to fire that CEO becomes a lot easier due to the fact that all the shareholders have the same opinion. Thirdly, another possibility is that the shareholders disregard the negative consequences of the corporate tax avoidance measure, which could include a reduction in share value and degradation of the reputation and brand name, because it has a minimal effect in comparison with financial misreporting (Hanlon and Slemrod, 2007). Due to the minimal risk of tax avoidance, the size of a board does not have an influence because the size only has an influence on risky decisions (Upadhyay, 2015; Nakano and Nguyen, 2012). These reasons might explain why there was no significant relationship between either the size or concentration between the shareholder group and corporate tax avoidance.

One unexpected outcome of this research is the significant differences in ETR between the different sectors. There are a number of possible explanations for this outcome. The significant differences between the sectors may have to do with industry related subsidies from the government, which affects the amount of corporate tax paid for that sector. Another possible reason is the difference within the view of corporate taxation. Some sectors may see corporate tax as a cost while other sectors see corporate tax as a public good. Some sectors might do everything within the law to reduce the corporate tax payments while other sectors might want to pay their “fair share” of corporate tax. The explanation could also be found in the differences between shareholders within the different sectors. In some sectors, the government is one of the shareholders of the corporation13_{, or the shareholders within that sector see corporate tax as public good. Therefore,} those sectors also might want to pay their “fair share” of corporate tax. These reasons could explain the significant differences between the sectors and the amount (or proportion) of corporate taxes paid.

13_{For example General Motor during the period 2009 through 2013.}

http://www.usatoday.com/story/money/cars/2013/12/09/government-treasury-gm-general-motors-tarp-bailout-exit-sale/3925515/ accessed on 02-24-2016

Structure of egoism A study on the impact of ownership structure on tax avoidance

Structure of egoism

A study on the impact of ownership structure on tax avoidance

Structure of egoism

A study on the impact of ownership structure on tax avoidance

Abstract

Keywords

Table of contents

1. Introduction

2. Theoretical background

2.1 Differing views of corporate tax and CSR and previous literature

2.2 The influence of ownership structure

2.3 Agency theory and tax avoidance

2.4 Stakeholder theory and tax avoidance

3. Hypothesis development

3.1 Employee stock ownership plan

3.2 The composition of the shareholders group

4. Research design

4.1 Sample

4.2 Dependent variable

4.3 Independent variables

4.4 Control variables

4.5 Statistical model

Dependent variable

Independent variables

Control variables

5. Findings

6. Discussion and conclusion

6.1 Discussion and conclusion