Shareholder ownership and tax avoidance

(1)

MSc Accountancy & Control

Shareholder ownership and tax

avoidance.

Financial Accounting

Karn Shah

10262156

Draft Version

Date: 22-06-2015

Dr. W.H.P Janssen

Word count: 10.307

(2)

2

Statement of Originality

This document is written by student Karn Shah 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.

(3)

3

Abstract

This paper provides empirical evidence on the relationship between concentration of

ownership and the tax avoidance activities of a firm. More specifically, this paper questions whether there is an association between the concentration of ownership and tax avoidance. This is done to contribute to the growing stream of literature about shareholder concentration and tax avoidance and to answer to the call for more empirical evidence. This paper also stimulates the ongoing discussion about tax avoidance as a societal issue.

To answer the proposed research question, a regression analysis will be conducted to determine the relationship between the concentration of ownership and the tax avoidance of a firm. The sample consists of 492 firm-year observations coming from a total of 125 firms. The results show that there is no association between the concentration of ownership and the amount of tax avoidance of a firm. This contradicts prior literature which states that the concentration of ownership either has a positive or a negative relationship to tax

avoidance. The results can be contradicting because of a few reasons. The first reason is that the sample is too small and did not capture all the effects of the entire population. The second reason is that the measure used for tax avoidance did not capture all the effects of tax

avoidance.

This paper has a few limitations. One of the limitations is that the sample is too small due to range constraints from the Blockholders database. This constraint resulted in another limitation, which was the inability to use the long-run tax avoidance as there were not enough year observations to compute this measure.

(5)

5

1.0 Introduction

Tax studies in general are found in various research fields such as accounting, finance, economics and law, which makes tax studies very varying and comprehensive. This varying aspect makes tax research particularly difficult, because every field has its own particular interests and even their own languages (Hanlon & Heitzman, 2010). One of the topics within tax research is tax avoidance. Tax avoidance is a fairly young and rising topic in the

accounting research (Hanlon & Heitzman, 2010). Dyreng, Hanlon & Maydew define tax avoidance as anything that reduces the firm's cash effective tax rate. This means that tax avoidance does not have to be anything illegal (2008).

Dyreng et al. found that firms are able to avoid taxes over a long period of time. They developed a measure called "long-run cash effective tax rate" which describes the firm's ability to avoid taxes over a long period of time for as long as ten years. They also discovered that annual cash effective tax rates are not good predictors of long-run cash effective tax rates (2008). So this study contributed mainly in a descriptive way by designing a measure for long-run tax avoidance. But the authors did not know why some firms engaged in tax avoidance and others did not and they let this question open for future research (Dyreng et al., 2008). Hanlon & Heitzman state that there are multiple determinants for tax avoidance. One of the reasons for tax avoidance could be the ownership structure of the firm (2010). Desai & Dharmapala state that concentrated ownership leads to a greater incentive to avoid taxes (2008). They base this on the theoretical notions of the agency theory, but there is no empirical evidence to support this notion. The authors also discuss some other

determinants for tax avoidance such as: information systems, corporate control and high powered incentives (Desai & Dharmapala, 2008). They continue by stating that the role of taxes and ownership patterns have yet to be explored and they call for more empirical work (Desai & Dharmapala, 2008). So this study mainly sets up a basis for future research by implying that there is a certain relationship between the ownership structure of a firm and its tax avoiding activities and calls for empirical research.

Chen et al. indirectly reply to that call for more empirical research. They found that family firms with relatively few shareholders are less tax aggressive than their non-family counterparts. These results show that family owners are willing to forgo tax benefits to avoid the non-tax costs such as potential penalties and reputation damage (2009). So this research also studied the link between the shareholder construction and tax avoiding activities and was insightful by helping towards a better understanding of the impact of equity ownership and agency conflicts on firms' tax avoiding activities. But this study was focused on family owned

(6)

6

firms, which differ from non-family owned firms. While this piece of empirical evidence showed the tax aggressiveness of family firms, which are highly concentrated firms, compared to non-family firms, it left the tax aggressiveness of highly concentrated firms compared to lowly concentrated firms open for future research.

The underlying notion for the relationship between the shareholder construction and tax avoiding activities is the agency problem. According to Chen et al., firms with a high ownership structure (i.e. less shareholders but with relatively high ownership per shareholder) have a better managerial and shareholder interest alignment and therefore there are less conflicts between the owners and managers (2009). Because there are less conflicts between owners and managers, there will be more tax avoiding activities within a firm. However, there is also some indication that highly concentrated firms may actually engage less in tax

avoiding activities. According to Hanlon & Heitzman, concentrated firms are more concerned with long-term results and are worried about arising costs, such as reputational costs (2010). Also, concentrated firms could engage in less tax avoiding activities because individuals have intrinsic motivations on tax evasion and will compare these to their risks and highly

concentrated firms act more like individuals than non-concentrated firms.

The aforementioned disagreement in previous literature and the call for empirical research makes it interesting to investigate the relationship between concentration of shareholder ownership and tax avoidance. The following question will be investigated: is there a relationship between the concentration of ownership and the tax avoidance of a firm? Desai and Dharmapala state that tax avoidance and ownership patterns have yet to be explored (2008). This thesis would contribute to the direct call for more empirical evidence on this subject, which is a fairly new and undiscovered topic. To my knowledge there is little empirical research done in the way this research will be done, therefore this research will contribute to the growing stream of literature regarding tax avoidance and ownership structure.

Tax avoidance is also a societal issue. It is seen as unethical to have big corporations like Apple pay a relatively low amount of tax and meanwhile increase the tax burden on civilians to push government bills (Doward, The Guardian, 2014). Conducting this research could give more insight to the thought process of corporations that engage in tax avoiding activities, which could result in more understanding on corporate tax avoiding activities. The rest of the paper proceeds as follows: section two will cover the literature review and hypothesis. This will be followed by section three that describes the sample used in the research and the research design itself. In section four, the main results of the research will be

(7)

7

discussed followed by section five, in which the sensitivity analyses are described. Lastly, section six will conclude this paper.

2.0 Literature review and hypotheses

In this section I will discuss three topics. First, in the literature review, I will discuss tax avoidance, mainly explaining the definition of tax avoidance and the costs and benefits resulting from tax avoidance. The second topic is the agency theory. First, I will give a general explanation of the agency theory. After that I will try to make a link between the agency theory and the problem that is being researched. The third topic discussed in this part is the concentration of ownership. Also here I will begin by defining shareholder structure and continue to discuss different types of investors. With these three topics discussed, a

hypothesis will be constructed. 2.1.1 Tax avoidance definitions

As stated earlier, tax avoidance is hard to define and measure. Tax studies are done across many fields and therefore there is no single definition of tax avoidance (Hanlon & Heitzman, 2010). The most broad definition of tax avoidance comes from Dyreng et al., where they studied a firm's ability to reduce their long-run cash effective tax rate in any way possible. The authors defined tax avoidance as anything that reduced their long-run cash effective tax rate (2008). These are mainly activities done in compliance with the law or those in gray-areas. Hanlon & Heitzman build upon this notion and define tax avoidance as the reduction of explicit taxes (2010). The difference between Hanlon & Hetzman's definition and Dyreng et al.'s definition is that Hanlon & Heitzman do include illegal activities and identify different types of tax avoidance. On the one hand are the types of tax avoidance that are perfectly legal, and on the on the other hand would be the more aggressive and noncompliance types of activities (2010).

These two definitions are accepted in the tax research community and are used in various empirical researches. One of those researches is that from Kim et al. (2011). In this research, the authors studied whether tax avoidance could be used as a predictor for stock price crashes. The authors used the definition for tax avoidance as provided by Hanlon & Heitzman (Kim et al., 2011). A different study that uses this definition is that from Chen et al. (2010). This study was already briefly discussed and the authors of this study studied the tax aggressiveness of family firms relative to the tax aggressiveness of non-family firms. They defined tax aggressiveness as the downward management of taxable income through tax

(8)

8

planning activities, including legal, illegal and grey-area activities (Chen et al., 2009). As we have seen earlier, this definition corresponds with the definition given by Hanlon &

Heitzman (2010).

Something noteworthy is that Hanlon & Heitzman (2010) stated that tax avoidance is hard to define because it exists in multiple research fields and that there is no generally accepted definition. However, this was only brought to attention after Hanlon & Heitzman’s paper (2010). This is because both Kim et al. (2010) and Chen et al. (2009) do not discuss different definitions of tax avoidance. This could imply that there may have been more obscurity on the definition of tax avoidance than previously thought and that Hanlon & Heitzman emphasized on a previously uncharted matter (2010).

2.1.2 Benefits and costs involving tax avoidance

Before deciding whether avoiding taxes will be beneficial, the firm has to assess the tradeoff between the marginal benefits against the marginal costs of avoiding taxes (Chen et al, 2009). Considering tax avoidance, there are more types of costs than benefits. Usually, the benefits are quite clear. The main benefit of avoiding taxes would obviously be saving cash so that it can be used for other purposes such as reinvestments in the firm or for higher dividend payouts. Therefore, in this section, the emphasis will lie on the costs involving tax avoidance. One of the main elements involving tax avoidance is rent seeking behavior, also known as rent extraction (Chen et al., 2009). These two definitions will be used

interchangeably from now on. According to Tullock, the concept of rent seeking is when a firm or individual attempts to gain economic benefits without reciprocating this to the society through wealth creation (1967). The term rent seeking was actually coined in 1974 by

Krueger (1974). Chen et al. refer to rent extraction as the non-value maximizing activities that decision makers follow at the cost of shareholders. Some examples they give are: aggressive financial reporting, perk consumption such as receiving target-related bonuses and related-party transactions (2009).

For decision makers, there are some benefits regarding rent seeking. They can, for instance, manage earnings to meet or beat their earnings benchmarks, which might result in a monetary gain for them (Chen et al., 2009). Firms could also make use of tax shelters to save on tax payments, which also could result in a personal monetary gain (Desai et al., 2006). However, these rent seeking activities also bare costs with them. For instance, if a firm is making use of illegal methods to avoid taxes, the IRS could impose penalties which obviously would hurt the firms earnings (Chen et al., 2009). Another cost of rent seeking behavior is a

(9)

9

potential price cut in the shareholder price. According to Chen et al. this can occur when other shareholders become aware of the tax avoiding activities to extract rents and as a result bid the firm price down (Chen et al., 2009). Basically, firms, or managers, can choose to practice tax avoiding activities such as making use of tax shelters, that result in personal monetary gain, which is a form of rent extraction. Managers can also make use of complex tax avoiding activities to hide their intent of rent extracting activities (Chen et al, 2009; Desai &

Dharmapala, 2006).

Another type of cost that firms have to take into account are political costs. Essentially, political costs are potential costs that can be imposed on the firm by external parties as a result of political actions and or measures (Watts & Zimmerman, 1978).

According to Watts & Zimmerman, the political sector has an incentive to redistribute wealth among various groups, and the corporate sector is particularly vulnerable to these

redistributions (1978). According to the political cost hypothesis, firms will tend to show lower profits to steer away any attention from politicians by using different means, such as: social responsibility campaigns, government lobbying and the selection of certain accounting procedures (Watts & Zimmerman, 1978). By minimizing the attention to the firm's high earnings, firms can minimize their potential costs as a result of politically imposed measures (Watts & Zimmerman, 1978). Tax avoidance can be used as a method to influence potential political costs. This means that when a firm is deciding on their level of tax avoidance, they also should consider the potential political costs involved when the level of tax avoidance is higher. This would mean that their earnings would be higher and would lead to increased scrutiny from politicians, which could result in political costs (Watts & Zimmerman, 1978).

The last type of cost that will be discussed is reputational costs. Existing evidence shows that there is a positive relationship between a firm's long-term performance and having a good reputation (Roberts & Dowling, 2002). Reputation costs are costs that are incurred as a result of a loss in reputation. An example of this is the 2010 BP oil spill, also known as the 'Deepwater Horizon oil spill.' As the events unfolded, starting on the 20th of April, BP's share price dropped lower and lower. It went from $59.88 per share before the event, to $27.02 at its lowest following a few months after the event1. In the study of Chen et al., it was shown that family firms are less tax aggressive than non-family firms. This was a result of two factors: concerns of getting a potential penalty imposed by the IRS and the concerns of losing

1

https://www.google.com/finance?q=NASDAQ%3ABP&ei=I4I_Vdn8PMSr8QOOx4HoDQ, Dates: 16-04-2010, $59.88 & 25-06-2010, $27.02.

(10)

10

reputation or family name (2009). So a firm has to take into account whether the loss of reputation is worth the benefits of tax avoidance.

2.2 Agency Theory

In order to understand the relationship between tax avoidance and the concentration of ownership of a firm, one must have some basic knowledge on the agency theory. This information is also needed to construct a hypothesis. Therefore, I will explain the agency theory in this section, following by the link between tax avoidance and ownership

concentration. After addressing these topics, I will construct a hypothesis.

Just like tax research, the agency theory is researched in various fields of studies such as accounting, economics, finance and political science (Eisenhardt, 1989). The agency theory describes the relationship between the principal and the agent. The agent is hired by the principal to perform services on the principal’s behalf (Jensen & Meckling, 1976). Jensen & Meckling state that there is a good chance that the agent will not act in the best interests of the principal if both the agent and the principal are utility maximizers, meaning they act

according their own interests (1976). This divergence in interest can cause problems that occur between shareholders (principal) and management (agent). These problems occur because they each have their own goals or because the principle is unable to monitor (or afford to monitor) the agent’s work (Eisenhardt, 1989). Another important problem between the principal and the agent is the problem of risk sharing. The agent might have a different attitude towards risk and therefore have different preferences. As a result, this could hurt the principal's firm (Eisenhardt, 1989). In other words, the agency problem describes a conflict of interests between the principal and the agent, where the agent (management) could be interested in maximizing its own wealth instead of the firms wealth.

There are two distinct problems associated with the agency theory. The first one is the moral hazard problem. Moral hazard is essentially the problem that the agent does not put in enough effort in his work (Eisenhardt, 1989). In this case, the principal is taking the risk without having the information whether the agent will work adequately. The agent however, does know how much effort he will put in his work. This difference of information between two parties is called information asymmetry. The second problem associated with the agency theory is adverse selection. Adverse selection in the agency theory is the fact that agents can claim to have certain skills for a job and that the principal is not able to verify these claims. (Eisenhardt, 1989). This is also a form of information asymmetry.

(11)

11

minimize any agency conflicts (Jensen & Meckling, 1976). One way is to incentivize the agent to act towards a behaviour that better fits the principal. This can be done for instance by giving out bonuses when certain targets are met, or giving the agent a share of the equity so that his motives are more in line with the principal’s (Jensen & Meckling, 1976). More specifically, the increase in ownership results in less managerial opportunism according to Jensen & Meckling (1976). The reasoning behind incentivizing is that the outcomes of the agency contract coalign and the rewards are now dependent on the same actions. This results in a reduction of the self-interest conflicts between principal and agent (Eisenhardt, 1989). Another way to minimize agency conflicts is for the principle to verify the agent’s behaviour, which is called monitoring. When the agent knows that the principal is able to verify the agent’s behaviour, the agent is less likely to shirk and will behave in the interests of the principal (Eisenhardt, 1989). An example of monitoring is having a board of directors that monitors and restricts the management’s activity on behalf of the shareholders. Monitoring, however, also includes costs. These monitoring costs make it, in most cases, impossible for the principal to ensure that the agent is only taking decisions that are optimal for the firm (Jensen & Meckling, 1976).

2.3.1 Concentration of ownership definition

The term concentration of ownership refers to the allocation of the shares between the shareholders. If there is a low amount of shareholders with relatively more shares per shareholder, then there is a high concentration of ownership. Therefore, the shares are concentrated, meaning they are close together, and not dispersed (La Porta et al., 1998). The definition of concentration of ownership seems to be unambiguous in previous literature. For instance, Ding et al. do not give a definition of ownership concentration, which implies that they assume that there is but one definition (2007).

The concentration of ownership of a firm has been used to study relationships between various subject matters, but not a lot for the relationship with tax avoidance. One of the pioneering works concerning shareholder concentration was the paper from La Porta et al. This study examined the legal rules surrounding the protection of shareholders and creditors. One of their findings was that the concentration of ownership of the largest public company is negatively related to the investor protection (1998). Following this study, other researches also have studied the concentration of ownership of firms, such as: Burkart & Panunzi (2006) and La Porta et al (1999). After these studies, a few researchers started linking tax avoidance with ownership structures.

(12)

12

2.3.2 Types of investors

There are different types of investors and these different types each have their own

characteristics (Bruton et al., 2010). These characteristics influence the investor’s behaviour and lead to different choices. Therefore, in this section, I will discuss the two main types of investors, institutional and transient investors, their behavior and how this might influence the level of tax avoidance.

The first type of investors that will be discussed are the (non-transient) institutional investors. Institutional investors are institutions that invest funds on behalf of others and manage great amounts of equity. Some examples of institutional investors are: insurance companies, hedge funds, retirement funds and mutual funds. According to Bushee,

institutional investors are more sophisticated than regular, individual investors (1998). This means that they possess more knowledge than other types of investors. This is logical as they are specialized in trading stocks. This sophistication results in more monitoring and more disciplining the managers (Bushee, 1998). This could suggest that, because of more monitoring and disciplining, firms with institutional investors have less tax avoidance.

The second type of investors are transient investors. According to Bushee, transient investors are investors that usually hold a small amount of stocks in various companies and trade for short term profits. They usually make their trading decisions based on short term values such as current earnings (1998). Note that institutional investors can also be transient investors as it is a kind of behaviour of the investor. Transient investors are more likely to pressure managers to meet short term goals, so that the firm's stock price increases and the investor can sell the stocks (Bushee, 1998). This could suggest that, because of more pressure to meet short term goals, firms with transient investors have more tax avoidance.

2.4 Hypothesis

The agency theory as discussed in section 2.2, is the backbone of the relationship between tax avoidance and concentration of ownership. For instance, Chen et al. state that firms with a high ownership structure have a better managerial and shareholder interest alignment and therefore there are less conflicts between the owners and managers (2009). This reduction in conflicts between owners and managers results in more tax avoiding activities (Chen et al., 2009). Desai & Dharmapala state that tax avoiding activities are viewed as an activity that shareholders advocate (Desai & Dharmapala, 2006). Desai and Dharmapala also state that, with regards to the agency theory, a highly concentrated ownership of a firm leads to a greater incentive to avoid taxes (Desai & Dharmapala, 2008). This notion is also mentioned by

(13)

13

Hanlon & Heitzman (2010) who state that firms with concentrated ownership may avoid more taxes because the controlling owners benefit more from the tax savings.

On the other hand, there is also some indication that highly concentrated firms may actually engage less in tax avoiding activities. According to Hanlon & Heitzman,

concentrated firms are more concerned with long-term results and are worried about arising costs, such as reputational costs (2010). As discussed earlier, family firms, which generally are concentrated, show less tax avoiding activities than non-family firms. This is because family firms are willing to forgo tax benefits to avoid concerns from other stakeholders. These concerns are about the family shareholders showing rent seeking behaviour and trying to hide that through tax avoiding activities (Chen et al., 2009). Also, concentrated firms could engage in less tax avoiding activities because individuals have intrinsic motivations for tax evasion and will compare these to their risks. As concentrated shareholders have more to say about a firm’s behaviour, the firm in general acts more like an individual than an non-concentrated firm (Hanlon & Heitzman, 2010).

As we have just seen, there is some disagreement in the literature concerning the relationship between tax avoidance and shareholder concentration. It can either be a positive relationship, where a high concentration of ownership results in more tax avoiding activities, or a negative relationship, where a high concentration of ownership leads to less tax avoiding activities. Therefore, to test this relationship, I have constructed the following null hypothesis: H0: There is no association between the concentration of ownership and the amount of tax

avoiding activities of a firm.

3.0 Research Methodology

In the following section I will describe the sample selection, following with the different measures that can be used as a proxy for tax avoidance. After that, I will discuss the different measures for concentration of ownership. After discussing the different alternatives for tax avoidance and concentration of ownership, I will make a choice between the different measures and explain why those measures were used. Lastly, I will discuss the empirical model used in the research.

3.1 Sample selection

The sample of firms in this research are derived from the S&P SmallCap 600 (S&P 600). The S&P 600 is a stock market index that measures the small-cap segment of the U.S. equity market. This stock market index is ideal for this research as it is more likely that there are

(14)

14

more firms with a lower number of shareholders due to the size of the market capitalization of the firms.

Two databases are used to gather all the data that is needed to conduct this research. Firstly, the database Blockholders is used to gather information about the firms' concentration of ownership. Due to scope limitations in this database, the sample period is 1996-2001. Secondly, the database Compustat is used to gather information about tax avoidance and also the information on control variables are extracted from this database.

The data from the two databases results in a total of 2944 firm year observations. Most of the observations are from Compustat, which had information on 471 firms. Blockholders, however, only had information on 180 firms, making the sample much smaller. The

Blockholders database does not have any information on the remaining 291 firms and there is no explanation why. It is unclear whether these firms did not have any large shareholders during that time period, or that the database was unable to collect relevant information on those firms. Merging the two datasets and removing missing data results in 492 firm year observations coming from a total of 125 firms, forming the actual sample.

With this sample, an archival database research will be conducted. More specifically, a regression analysis will be conducted to determine the relationship between the concentration of ownership and the tax avoidance of a firm.

3.2 Variable alternatives

In the following section I will discuss the different measures that can be used to measure tax avoidance and shareholder concentration. After discussing the different measures, a choice will be made later on in section 3.3.

3.2.1 Tax avoidance measures

There are many ways to measure tax avoidance and each measure has its own advantages and disadvantages. Also, not every measure is suitable for every research question. In this section I will discuss the main, usable measures, their meaning and their advantages and

disadvantages.

One of the more common types of tax avoidance measures used in tax research is the 'effective tax rate' (from now on: ETR) measure. This measure incorporates the average rate of tax per dollar of income or cash flow (Hanlon & Heitzman, 2010). This measure is determined by dividing an estimate of the tax liability of a firm by a measure of before-tax profits or cash flow (Hanlon & Heitzman, 2010). For example, to calculate the GAAP ETR one has to divide the 'worldwide total income tax expense,' which in this case is the estimate

(15)

15

of the tax liability, by the 'worldwide total pre-tax accounting income,' which is the before tax profit (Hanlon & Heitzman, 2010). The GAAP ETR uses the total income tax expense, which means that if a firms defers taxes, this will not alter the GAAP ETR, as it is still a tax expense. Another measure, the Cash ETR, uses the same denominator, but for the numerator it uses 'worldwide cash taxes paid' which means that it is affected by tax deferral strategies. At the same time, Cash ETR is not affected by changes in tax accounting in accruals, but GAAP ETR is (Hanlon & Heitzman, 2010). This shows that even the slightest change in variables can result in an essential change in result.

Additionally, there is another type of ETR, which is more focused on the long term implications of tax avoidance. This is the 'Long-run ETR' and was developed by Dyreng et al. (2008). This measure is computed with the same variables used in the Cash ETR, but here the sum of ten years of those variables is used. The main benefit of using this method is that it takes away the year-to-year volatility in annual effective tax rates (Dyreng et al., 2008). One of the detriments of using all ETR variants is that they do not distinguish between real

activities that are tax-favoured, avoidance activities that are meant to actually avoid taxes and targeted tax benefits from lobbying activities (Hanlon & Heitzman, 2010). This can result in taking into account non-tax avoiding activities.

Another type of measurement for tax avoidance is the unrecognized tax benefit. This measure is used as a proxy for tax avoidance and can be obtained directly from the FIN 48 disclosure of a firm (Hanlon & Heitzman, 2010). FIN 48 is an official interpretation issued by The Financial Accounting Standards Board in 2006 and its purpose is to clarify the

accounting for uncertainty in income taxes. This is done by providing the criteria used to recognize and measure the uncertain tax benefits (Cazier et al., 2009). According to Hanlon & Heitzman, one of the underlying determinants for unrecognized tax benefits are the financial reporting incentives. Managers may use the unrecognized tax benefit to manage earnings upwards and it is therefore also used as a proxy for earnings management (2010). This measure is subjected under the judgement of management which makes it less objective. 3.2.2 Concentration of ownership measurements

According to Sjögren, Overland & Mavruk, choosing your measure for the concentration of ownership of a firm is important. A different choice of measure can lead to different

conclusions (2012). Therefore, I will discuss a few different measures so that I will be able to make a choice between these measures for the concentration of ownership, which will be used when testing the hypothesis.

(16)

16

There are a few different ways to measure the concentration of ownership of a firm. One is to take the average percentage of common shares owned by the three largest

shareholders of the firm. This technique was developed by La Porta et al. (1998), but also used by Ding et al. for their research (2007). An advantage of this type of measure is that one can arrange a sample of firms from a low concentration of ownership to a high concentration of ownership, comparably to Sloan's study, where he ranked firms based on their amount of accruals (1996).

Another type of measurement is called the 'weakest link principle.' With this principle, a firm is concentrated if it has an owner with more than a certain threshold of shares or voting power (Edwards & Weichenrieder, 2009). This threshold can be for example 5%, 10% or 20%. If no one shareholder can be identified meeting this threshold, then the firm is not concentrated but dispersed (Edwards & Weichenrieder, 2009). An advantage for this method is that one can divide a sample of firms into two groups, one group of with firms that meet the threshold and one group of firms that do not meet the threshold. Choosing between these measures will rely on the research methodology and that will be discussed in the following sections.

3.3 Variables measurement 3.3.1 Tax avoidance

As discussed earlier, there are a few different methods to measure tax avoidance. One of those measures is the Long-run ETR as described in Dyreng et al. (2008). The main benefit of using this method is that it takes away the year-to-year volatility in annual effective tax rates (Dyreng et al., 2008). However, this method is only effective when the sample period is a longer term, such as ten years (Dyreng et al., 2008). As stated earlier, due to scope limitations of the Blockholders database, the sample period of this research is only six years, and even then, there are not always six years of data for every firm. For this reason, choosing the Long-run ETR as a tax avoidance measure, will not have any added value in this research.

Therefore, instead of using the Long-run ETR, the yearly GAAP ETR will be used such as in the paper of Chen et al. (2009) and described in the paper of Dyreng et al. (2008). The GAAP ETR is calculated as:

(17)

17

where i stands for firm and t stands for year. This dependent variable will be called

GAAPETR. This measure captures the average rate at wich a firm is taxed as it divides the tax expense by the taxable income (Dyreng et al., 2008). The main benefit of this measure is that if a firm defers taxes, it does not alter the tax expense, capturing any methods of avoiding taxes with the use of accruals (Hanlon & Heitzman, 2010). However, this measure does not take away any year-to-year volatility, which could result in incomplete results (Dyreng et al., 2008). Any values lower than 0 and higher than 1 will be dropped, in line with the research of Chen et al. (2009).

3.3.2 Concentration of Ownership

One of the ways to measure concentration of ownership is to take the average of the largest three shareholders of the firm (La Porta et al., 1998). The database Blockholders, however, does not provide any information on individual shareholders, making this approach unusable.

Another approach to measure concentration of ownership is the weakest link principle (Edwards & Weichenrieder, 2009). With this measurement, a firm is concentrated if a

shareholder has more than a set amount of shares or voting power. The set amount can be for instance 5%, 10%, 25% or anything else that could imply a concentrated ownership. This method will be used in this research. As a threshold, I have chosen 5% because shareholders with at least 5% of shares of a company are called 'major shareholders' (Sjögren et al., 2012). This independent variable will be called CONCENTRATED and is a dummy variable. This means that CONCENTRATED is measured as an indicator variable coded as ‘1’ if the firm has one or more shareholders with 5% or more of total shares and as ‘0’ otherwise. The Blockholder database provides data for the percentage of shares held by a blockholder, corrected for overlapping shareholdings (Dlugosz et al., 2006). The dummy variable is calculated as follows:

where i stands for firm and t stands for year. 3.3.3 Control variables

Besides the independent variable CONCENTRATED, a number of control variables are used. As this research is focused on the relationship between shareholder concentration and tax avoidance, any outside factors influencing the results are unwanted. Therefore, there are a few control variables that have to be taken into account.

(18)

18

The first control variable is the leverage of a firm, which will be called LEV. This variable is calculated as follows:

.

where i stands for firm and t stands for year. Firm leverage is used as a control variable because prior literature shows us that highly leveraged firms have lower effective tax rates than low leveraged firms (Stickney & McGee, 1982). This would mean that firms with a high amount of leverage will show higher tax avoidance in the research model, irrespective of the firm’s concentration of ownership. Therefore, leverage needs to be controlled for. High leveraged firms have lower effective tax rates because there are a lot of tax benefits of debt. This is because firms in the US are taxed on their equity and not on their debt (Graham, 2000).

The next control variable is the size of a firm, which will be called SIZE. This variable is calculated as follows:

where i stands for firm and t stands for year. The natural logarithm of firm size is taken to eliminate the big differences in firm size. Firm size is controlled for because prior literature shows us that larger firms have higher effective tax rates than smaller firms (Zimmerman, 1983). For the research model, this would mean that the larger firms will show lower tax avoidance, whether or not the firm has a high concentration of ownership and therefore, the size of a firm needs to be considered. According to Zimmerman, large firms have higher effective tax rates because they are more subject to political scrutiny, resulting in higher political costs (1983).

The last control variable is the market-to-book ratio of a firm, which proxies the growth of a firm and will be called MARKBOOK (Chen et al., 2009). This variable is calculated as follows:

where i stands for firm and t stands for year. Firm growth is a control variable because

growing firms may make more investments in tax beneficial assets (Chen et al., 2009). In this research model, that would mean that growing firms will show higher tax avoidance, whether

(19)

19

or not the firm has a high concentration of ownership. For this reason, firm growth needs to be controlled for.

3.4 Model

To test whether shareholder concentration is associated with tax avoidance, the following regression model will be used:

where i stands for firm and t stands for year. As stated earlier, the dependent variable is calculated by using the GAAP ETR of a firm. If highly concentrated firms conduct more tax avoiding activities than low concentrated firms, is expected to be negative. This is because the lower the ETR is, the more tax is avoided. If highly concentrated firms conduct less tax avoiding activities, then is expected to be positive. If there is no correlation between tax avoidance and equity of ownership, the is expected to be close to 0. The control variables are used to capture possible relationships between tax avoidance and the concentration of ownership that could otherwise influence the results.

4.0 Descriptive statistics & Empirical results

In this section the descriptive statistics and their meanings will be discussed. Afterwards the empirical results of the research will be discussed.

4.1 Descriptive Statistics

The descriptive statistics for tax avoidance, shareholder ownership and the control variables are illustrated in table 1.

Table 1: Descriptive Statistics

Variable Observations Mean

Standard Deviation Median Min Max GAAPETR 492 .3512 .0866 .3646 0 .7571 CONCENTRATED 492 .8434 .3637 1 0 1 LEV 492 .1827 .1436 .1792 0 .6471 SIZE 492 6.2813 .7150 6.2924 4.3823 9.5648 MARKBOOK 492 2.2252 3.8616 1.9002 -73.5788 17.6199

(20)

20

The GAAPETR is 0.3512 which means that the firms in this sample are, on average, taxed for 35,12% on their pretax income. The standard deviation is relatively low, which means that there is not much of a dispersion regarding the effective tax rate among the different firms in the sample. Just like in the paper of Chen et al. (2009), the range of effective tax rate is between 0 and 1.

The dummy variable CONCENTRATED has a mean of .8434, which means that 84,34% of the firms in the sample have one or more shareholders with 5% or more of the total shares. This shows that the majority of the sample has a concentrated ownership structure, which may result in an under-representation of the non-concentrated firms.

Table 2 gives an overview of the correlation between the variables. The values in parentheses are the p-values of the correlation.

GAAPETR CONCENTRATED LEV SIZE MARKBOOK

GAAPETR 1 - CONCENTRATED -0.0442 1 (0.3279) - LEV 0.1418 -0.1513 1 (0.0016) (0.0008) - SIZE 0.0715 -0.0632 -0.0773 1 (0.1133) (0.1613) (0.0867) - MARKBOOK 0.2039 -0.0046 -0.1784 0.1741 1 (0.0000) (0.9196) (0.0001) (0.0001) -

Table 2: Correlation Matrix

The independent variable and the control variables are not highly correlated with each other, the highest correlation is between the leverage of a firm and its market-to-book ratio, which is -0.1784. This means that a higher market-to-book ratio leads to a lower amount of leverage which corresponds with previous literature from Binks & Ennew. They stated that high growth firms may be adversely affected by credit constraints (1996).

The correlation between the variables LEV and CONCENTRATED is -0.1513 which means that the ownership concentration of a firm is negatively correlated with the amount of leverage. If a firm is more concentrated, it will have less leverage. This could be explained by the fact that larger shareholders are not willing to accept the risk of the debt, because the shareholders bear the majority of the risk (Jensen & Meckling, 1976).

The fact that the independent variable and the control variables are not highly correlated with each other, means that there is no collinearity in the model, thus there is no

(21)

21

need to drop any independent variables.

The correlation between the dependent variable and the independent variable is very low and has a high p-value, which indicates that it is close to zero. This means that there is no correlation between the concentration of ownership and the effective tax rate of a firm. The actual coefficient of CONCENTRATED will be discussed in the next section.

4.2.1 Results

In the following sections I will discuss the results of the regression analysis. I will analyze and discuss the coefficients and their meaning.

Table 3 reports the results of the OLS regression analysis between the effective tax rate of a firm and the concentration of ownership.

coefficient t-statistic p-value

CONCENTRATED -0.0029 -0.28 0.779 LEV 0.1111***2 4.11 0 SIZE 0.0054 1.02 0.31 MARKBOOK 0.0051*** 5.1 0 Constant 0.2876*** 7.91 0 Observations (N) 492 F-value 10.09 Prob > F 0 R-squared 0.0766 Adj R-squared 0.069

Table 3; Results of Regression Analysis

The null hypothesis states there is no association between the concentration of ownership and the amount of tax avoiding activities of a firm. The coefficient

CONCENTRATED is slightly negative, but not significantly less than 0, which means that the null hypothesis is not rejected. This means that, in this sample, there is no effect between the concentration of ownership and tax avoidance activities, which is not in line with the prediction that shareholder concentration is either positively or negatively associated with tax avoidance. This does not align with the current literature, which states that a higher

(22)

22

concentration of ownership either results in more tax avoiding activities (Desai &

Dharmapala, 2008; Chen et al., 2009) or less tax avoiding activities (Hanlon & Heitzman, 2010; Chen et al., 2009).

Another coefficient that contradicts prior literature, is the coefficient for the variable LEV. This is the variable that determines a firm’s leverage. Table 3 shows that the coefficient is significantly positive. This means that firms with a higher leverage have a higher effective tax rate. In other words, firms with higher leverage practice less tax avoiding activities. This is not in line with Stickney & McGee (1982), which stated the opposite.

The coefficient MARKBOOK, which proxies for firm growth, is also significantly positive. This means that if a firms market-to-book ratio increases, they conduct less tax avoiding activities. Growing firms have higher market-to-book ratios because their market value is higher than their book value as a result of investors thinking that these firms are undervalued (Chen & Zhao, 2006). This means that growing firms actually conduct less tax avoiding activities than full-grown firms, which also contradicts previous literature.

4.2.2 Discussion

In the previous section, the results of the regression showed that the concentration of ownership is not associated with tax avoiding activities, which is not in line with the

prediction nor with the previous literature. There might be a few reasons for this disagreement between the results and prior literature.

One of the reason that could explain this difference is that the sample is quite specific and small. The sample is small because the database Blockholders only has data on firms from 1996 to 2001. Also, does it not have data for every firm in the S&P Smallcap 600. This is understandable, as collecting information on the shareholder concentration is difficult and not always available. The small sample might not be able to capture all the effects of the entire population and therefore result in a different outcome. Also, as stated earlier, the independent variable consists for 84,34% of firms with a concentrated ownership structure. This could imply an under-representation of the firms with a non-concentrated ownership structure.

A second reason could be that the measure used for tax avoidance, GAAP ETR, did not take away the year-to-year volatility and it did not reflect deferred taxes as tax avoidance. According to Dyreng et al., there are two disadvantages to using the GAAP ETR. The first one is that this measure does not take away any year-to-year volatility as it is an annual measure (2008). Another disadvantage is that GAAP ETR includes both current and deferred

(23)

23

taxes and that means that the GAAP ETR will not reflect tax avoidance in the form of deferring taxes (Dyreng et al., 2008). It could be that these aspects have caused different results.

Another reason for conflicting results could be that the threshold for the concentration of ownership was chosen too low. For a firm to be concentrated in this research, it has to have one or more shareholders with 5% or more of the total shares, this was done in line with Sjögren et al.’s paper (2012). However, it is possible that the effects of large shareholders are better represented by shareholders with a larger percentage of shares than 5% (Sjögren et al, 2012). For this reason, a few sensitivity tests concerning the threshold will be conducted These will be discussed in the next chapter.

5.0 Sensitivity Analysis

In the following section some additional analyses will be conducted to test whether a different threshold for the concentration of ownership has an effect for the outcome of the regression. There is a lot of subjectivity in determining the threshold for a firm with a concentrated ownership and using a different measure can lead to different results (Sjögren et al, 2012). Therefore, to test whether these outcomes hold true for other thresholds, I conduct these additional sensitivity analyses.

5.1 Sensitivity analysis – 10% threshold

I start by changing the dummy variable CONCENTRATED from being coded as ‘1’ if the firm has one or more shareholders with 5% or more of total shares and as ‘0’ otherwise to being coded as ‘1’ if the firm has one or more shareholders with 10% or more of total shares and as ‘0’ otherwise. As stated earlier, determining the threshold for the concentration of ownership requires some subjectivity and there are different thresholds that can be set (Sjögren et al., 2012). The reason to raise the threshold to 10% specifically is because it is in line with the empirical research of Maury & Pajuste, where they investigate whether having multiple large shareholders has an effect on the valuation of the firm (2005).

Table 4 shows the descriptive statistics of the new regression with the new variable called XCON, where X stands for the Roman numeral 10.

(24)

24 Table 4: Descriptive Statistics Sensitivity Analysis 10% Threshold

The descriptives of the old variables are unchanged. The mean of XCON is noticeably lower at 0.2703 than the mean of CONCENTRATED. This means that, in this sample, 27.03% of the firms have one or more shareholders with at least 10% of total shares. This shows that the new threshold results in a higher representation of the non-concentrated firms. Also, this means that the majority of the large shareholders possess somewhere between 5-10% of total shares.

Table 5 shows the results of the regression between the effective tax rate and the new variable XCON.

Coefficient t-statistic p-value

XCON 0.0013 0.15 0.879 LEV 0.1127***3 4.2 0 SIZE 0.0056 1.03 0.301 MARKBOOK 0.0052*** 5.11 0 CONSTANT 0.2840*** 8.27 0 Observations (N) 492 F-value 10.08 Prob > F 0 R-squared 0.0765 Adj R-squared 0.0689

Table 5: Results of Sensitivity Analysis 10% Threshold

3_{*** indicates that the coefficient is significant at the one per cent level.}

Variable Observations Mean Standard Deviation Median Min Max GAAPETR 492 0.3512 0.0866 .3646 0 0.7571

XCON 492 0.2703 0.4445 0 0 1

LEV 492 0.1827 0.1436 .1792 0 0.6471

SIZE 492 6.2813 0.7150 6.2924 4.3823 9.5648

(25)

25

Consistent with the main finding in section 4.2.1 the coefficient of the dependent variable XCON is statistically close to 0 with the raised threshold, which means that H0 is not rejected. This strengthens the earlier finding where there was no effect between the concentration of ownership and tax avoidance activities. This also holds true when a

concentrated firm is defined as a firm with one or more shareholders with 10% or more of the total shares.

5.2 Sensitivity analysis – 20% threshold

A different threshold that is used to define a concentrated ownership is the 20% threshold. La Porta et al. (1999) define a shareholder with 20% of voting rights a controlling shareholder. Even though voting rights are different from owning shares, voting rights are also used as a measure for concentration of ownership (Sjögren et al., 2012). Berle & Means define a controlling shareholder as a shareholder that owns 20% or more of the total shares (1932). This 20% threshold is a higher measure that is used to define a concentrated ownership (Sjögren et al., 2012) and testing with this threshold will improve the credibility of the main findings.

For this sensitivity analysis, I create a new dummy variable as in the previous sensitivity analysis. I start by changing the dummy variable CONCENTRATED from being coded as ‘1’ if the firm has one or more shareholders with 5% or more of total shares and as ‘0’ otherwise to being coded as ‘1’ if the firm has one or more shareholders with 20% or more of total shares and as ‘0’ otherwise. This new variable is called XXCON, where XX stands for the Roman numeral 20. Table 6 shows the descriptive statistics of the new regression with the variable XXCON.

Variable Observations Mean Std. Dev. Median Min Max

ETR 492 0.3512 0.0866 .3646 0 0.7571

XXCON 492 0.0813 0.2735 0 0 1

LEV 492 0.1827 0.1436 .1792 0 0.6471

SIZE 492 6.2813 0.7150 6.2924 4.3823 9.5648

MARKBOOK 492 2.2252 3.8616 1.9002 -73.5788 17.6199 Table 6: Descriptive Statistics Sensitivity analysis 20% Threshold

Again, the descriptives of the old variables are unchanged. The new variable XXCON has a mean of 0.0813, which means that, in this sample, 8,13% of the firms have one or more

(26)

26

shareholders with 20% or more of the total shares. This is significantly lower than the

previous variables and shows that there are not a lot of firms in the sample that own more than one fifth of a firm.

Table 7 shows the results of the regression between the effective tax rate and the new variable XXCON.

Coefficient t-value p-value

XXCON -0.0083 -0.59 0.557 LEV 0.1088***4 3.98 0 SIZE 0.00526 0.98 0.329 MARKBOOK 0.0051*** 5.1 0 CONSTANT 0.2875*** 8.29 0 Observations (N) 492 F-value 10.17 Prob > F 0 R-squared 0.0771 Adj R-squared 0.0695

Table 7: Results of Sensitivity Analysis 20% Threshold

The coefficient for XXCON is still statistically close to 0. This sensitivity analysis also supports the initial findings. So H0 is not rejected which means that there is no association between the concentration of ownership and the amount of tax avoiding activities of a firm when the threshold is set at 20%.

6.0 Conclusion

There is a great amount of previous literature on tax avoidance to be found as it is a complex and rising topic (Hanlon & Heitzman, 2010). However, the amount of studies on tax

avoidance in relation to the concentration of ownership is limited and there still is a call for empirical evidence on the subject (Desai & Dharmapala, 2008). This research is a reply to that call for empirical evidence.

This research examines whether the concentration of ownership is associated with tax avoidance. More specifically, I test whether a high concentration of ownership is associated

(27)

27

with a change in tax avoiding activities. The measure that is used to determine tax avoiding activities is the GAAP effective tax rate of a firm. The independent variable, firm

concentration, is a dummy variable which is coded as a '1' for firms with one or more shareholders with 5% or more of total shares, and as '0' otherwise.

The findings show that there is no association between shareholder concentration and tax avoiding activities. Thus, the null hypothesis, which stated that there is no association between the concentration of ownership and the amount of tax avoiding activities, is not rejected. This contradicts prior literature, which states that a higher concentration of

ownership either results in more tax avoiding activities (Desai & Dharmapala, 2008; Chen et al., 2009) or less tax avoiding activities (Hanlon & Heitzman, 2010; Chen et al., 2009). The findings hold up against additional sensitivity analyses, where the threshold of concentration of ownership was increased to 10% and 20%.

There could be a few reasons why the findings contradict prior literature. The first reason is that the sample is quite specific and small. The small sample might not be able to capture all the effects of the entire population and therefore result in a different outcome. A second reason could be that the measured used for tax avoidance did not take away year-to-year volatility as it is an annual measure and that it did not capture deferred taxes as tax avoidance (Dyreng et al., 2008).

There are some limitations to this study. As said earlier, the sample size is small because of the range restrictions of the Blockholders database. The small sample size could negatively influence the quality of the findings because a larger sample strengthens the

statistical power of a research. This limitation can be overcome by future researchers by either choosing a different sample or using a different database that is available to them. There is another limitation as a result of the range restriction of the Blockholders database. That is the fact that it is not possible to measure the long-term tax avoidance, as you need data from firms over a period of at least ten years (Dyreng et al., 2008). By not being able to use this measure I was not able to take away the year-to-year volatility in annual effective tax rates, which could have led to qualitatively stronger findings. This limitation can also be overcome by using a different database that does have the appropriate data available.

This study contributes to the growing stream of literature on tax avoidance as it answers to the call of empirical evidence. The findings of this study are not in line with prior literature which means that there still is some ambiguity around the effect of shareholder concentration on tax avoidance and that this topic should be continued to be researched in the future. Another contribution that is paper makes, is that it stimulates the discussion about

(28)

28

corporate tax avoidance as an ethical and societal issue. It is argued that corporate tax

avoidance is unethical and that it costs tens of billions of dollars in lost revenue every year to developing countries (Jones, 2015). This shows that corporate tax avoidance is still a societal issue and this paper contributes to that discussion by providing additional insights and

bringing up this subject.

Bibliography

Berle, A. A., & Means, G. (1932). The Modern Corporation and Private Property. New Jersey: Transaction Publishers.

Binks, M. R., & Ennew, C. T. (1996). Growing firms and the credit constraint. Small Business

Economics, 8(1), 17-25.

Bruton, G. D., Filatotchv, I., Chahine, S., & Wright, M. (2010). Governance, Ownership Structure, and Performance of IPO Firms: The Impact of Different Types of Private Equity Investors and Institutional Environments. Strategic Management Journal, 31(1), 491-509.

Burkart, M., & Panunzi, F. (2006). Agency conflicts, ownership concentration, and legal shareholder protection. Journal of Financial Intermediation, 15(1), 1-31.

Bushee, B. J. (1998). The Influence of Institutional Investors on Myopic R&D Investment Behavior.

The Accounting Review, 73(3), 305-333.

Cazier,R. Rego, S., Tian, X., & Wilson, R. (2009) Early Evidence on the Determinants of Unrec- ognized Tax Benefits. Working paper, University of Iowa and Texas Christian University Chen, L., & Zhao, X. (2006). On the relation between the market-to-book ratio, growth oppurtunity,

and leverage ratio. Finance Research Letters, 3(1), 253-266.

Chen, S., Chen, X., Cheng, Q., & Shevlin, T. (2009). Are family firms more tax aggresive than non-family firms? Journal of Financial Economics, 95(1), 41-61.

Desai, M. A., & Dharmapala, D. (2006). Corporate tax avoidance and high-powered incentives.

Journal of Financial Economics, 79(1), 145-179.

Desai, M. A., & Dharmapala, D. (2008). Tax and corporate governance: an economic approach. Tax

and Corporate governance, 13-30.

Ding, Y., Hope, O.-K., Jeanjean, T., & Stolowy, H. (2007). Differences between domestic accounting standards and IAS: Measurement determinants and implications. Journal of Accounting and

Public Policy, 26(1), 1-38.

Dlugosz, J., Fahlenbach, R., Gompers, P., & Metrick, A. (2006). Large Blocks of Stock: Prevalence, Size and Measurement. Journal of Corporate Finance, 12(1), 594-618.

Doward, J. (2014, October 5). How Apple’s Cork HQ became the centre of a bitter global ware over corporate tax avoidance. The Guardian. Retrieved from http://www.theguardian.com

(29)

29 Dyreng, S. D., Hanlon, M., & Maydew, E. L. (2008). Long-Run Corporate Tax avoidance. The

accounting Review, 83(1), 61-82.

Edwards, J. S., & Weichenrieder, A. J. (2009). Control rights, pyramids, and the measurement of ownership concentration. Journal of Economic Behavior & Organization, 72(1), 489-508. Eisenhardt, K. M. (1989). Agency Theory: An Assessment and Review. The Academy of Management

Review, 14(1), 57-74.

Graham, J. R. (2000). How Big Are the Tax Benefits of Debt. The Journal of Finance, 55(5), 1901-1942.

Hanlon, M., & Heitzman, S. (2010). A review of tax research. Journal of Accounting and Economics,

50(1), 127-178.

Jensen, M. C., & Meckling, W. H. (1976). Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics, 3(4), 305-360.

Jones, S. (2015, June 2). Tax dodging by big firms ‘robs poor countries of billions of dollars a year.’

The Guardian. Retrieved from http://www.theguardian.com

Kim, J.-B., Li, Y., & Zhang, L. (2011). Corporate tax avoidance and stock price crash risk: Firm-level analysis. Journal of Financial Economics, 100(1), 639-662.

Krueger, A. O. (1974). The Political Economy of the Rent-Seeking Society. The American Economic

Association, 64(3), 291-303.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (1998). Law and Finance. Journal of

Political Economy , 106(6), 1113-1155.

La Portal, R., Lopez-De-Silanes, F., & Shleifer, A. (1999). Corporate Ownership Around the World.

The Journal of Finance, 54(2), 471-517.

Maury, B., & Pajuste, A. (2005). Multiple large shareholders and firm value. Journal of Banking &

Finance, 29(1), 1813-1834.

Roberts, P. W., & Dowling, G. R. (2002). Corporate Reputation and Sustained Superior Financial Performance. Strategic Management Journal, 23(2), 1077-1093.

Sjögren, S., Overland, C., & Mavruk, T. (2012). Keeping it real or keeping it simple? Ownership concentration measures compared. The 4th workshop of the Nordic Corporate Governace

Network.

Sloan, R. G. (1996). Do Stock Prices Fully reflect information in Accruals and Cash Flows about Future Earnings? The Accounting Review, 71(3), 289-315.

Stickney, C. P., & McGee, V. E. (1982). Effective corporate tax rates the effect of size, capital intensity, leverage and other factors. Journal of Accounting and Public Policy, 1(2), 125-152. Tullock, G. (1967). The Welfare Costs of Tariffs, Monopolies and Theft. Western Economic Journal,

(30)

30 Watts, R. L., & Zimmerman, J. L. (1978). Towards a Postitive Theory of the Determination of

Accounting Standards. The Accounting Review, 563(1), 112-134.

Shareholder ownership and tax avoidance

MSc Accountancy & Control