Corporate social responsibility (CSR) and tax avoidance have been part of multiple recent research topics. To date, research efforts in these areas have provided competing results ranging from a positive relation (Hoi et al. 2013; Lanis and Richardson 2012) to a negative relation (Davis et al. 2016) between this two terms. Davis et al. 2016 explains that corporate tax payments are socially responsible for some stakeholders while others consider tax payments as irrelevant in terms of CSR. For instance, the British newspaper The Guardian published on February 14, 2009 that “around 60 per cent of the UK directors do understand tax avoidance as an ethical issue”. Conversely to this idea, the same article resported that “respondents to a survey of tax directors were unanimous in saying that the payment of corporation tax is not at present a social issue relevant to CSR”.
Previous scientific researches have encountered different results regarding the relation between CSR and tax avoidance. From these studies diverse theories were developed, as for instance, the risk management theory in which CSR is seen as the avoidance of damages in firm’s reputation (Financial Times, 2004). So, accordingly to this perspective CSR is used to reduce the effect of aggressive tax avoidance practices (Hoi et al. 2013). On the other hand, the corporate culture perspective remark the fact that aggressive tax avoidance practices are viewed by society as irresponsible and unethical (Weisbach 2002), proving that CSR must be inconsistent with aggressive tax avoidance practices. Although these theories try to explain the ambiguity in the results of previous studies, there is still no consistent theory regarding this relationship.
In addition to CSR there are other external factors affecting tax avoidance. Boone et al. (2012) show that culture, more precisely, religiosity is a determinant of tax avoidance by corporate taxpayers. So this study is built to find a consistent relation between tax avoidance and CSR including culture as a country specific factor. Culture can be defined in many ways, but in this study I will focus on the distinction between masculine and feminine environments as the determinant in the relation between CSR and tax avoidance. On the one hand, masculine countries refer to societies in which most relevant values are independence and competitiveness. On the other hand, feminine countries are more related to modesty, tenderness and concerned with the quality of live (Hofstede 2010; Srite and Karahanna 2006). This distinction seems to be appropriate since masculine and feminine dimensions are related to investor protection and the share- or stakeholder orientation of a country (Janssen, 2017).
Therefore, the research question of this thesis is:
What is the effect of corporate social responsibility (CSR) and culture on tax avoidance practices?
In this research I study the relationship between CSR and tax avoidance, which is already studied in previous researches but lacks consistency. Previous literature focuses the attention on this relationship without taking into account other interactions that can eventually modify inconsistencies encountered up to date. Hence, I include culture as a second independent variable together with CSR. By adding culture, we will have new outcomes on the relation between CSR and tax avoidance.
Apart from that, previous studies focus their research in countries as Australia (Lanis and Richardson 2013), US (Hoi et al. 2013) or China (Kenny et al. 2017). The sample of this study will include 147 companies from eight European countries divided into two groups. One of the groups will include the four European countries with higher masculinity scores (Germany, UK, Italy and Austria) and the other one will include the four European countries with lower scores in masculinity (Netherlands, Sweden, Norway and Finland). This distinction is in order to observe the differences more easily.
To examine that, I use two different dependent variables to calculate tax avoidance practices for robustness purpose. On the one hand, GAAP ETR which has been widely used in previous studies but has some limitations. Among others, this measure may not show deferred taxes and it just reflects non-conforming tax avoidance practices since it is based in accounting numbers (Aramide et al. 2013). On the other hand, I use Long Term Cash ETR since it seems to overcome certain limitations presented in the GAAP ETR measure. Moreover, two independent variables are used: one is CSR and the other is culture. CSR values are obtained through Thomson Reuters Asset 4 while culture is stated as a dummy variable where 0 represent feminine countries and 1 masculine countries.
The results of our tests present two new contributions to the research on CSR and taxes. First of all, and in line with previous studies, I found a positive relation between CSR and Long Term Cash ETR. This means that firms involved in high CSR practices are less engaged in tax avoidance practices. Therefore, influential stakeholders in companies with high CSR values do actually care about tax avoidance practices. Secondly, contrary to our predictions I find feminine countries to be more engaged in tax avoidance practices than masculine countries. However, this is just proved to be true when GAAP ETR is used as a dependent variable.
The reminder of this paper is as follows. Section 2 include background related to the study together with the development of the hypothesis. Section 3 of the text describe the sample design. Section 4 describe the research methodology. Section 5 present the empirical results. Lastly, section 6 concludes the study.
1. Literature review and Hypothesis Development
2.1 Tax Avoidance
Tax avoidance has been in the core of multiple studies during the last decade and it is still playing an important role nowadays. Dyering et al. (2008) stated that one-fourth of the US public firms has an effective tax rate around 20% while the corporate tax rate is 35%. Therefore, tax avoidance is an important part of firm’s profit but it can result in huge losses if this practice is highly aggressive. The organization for Economic Co-operation and Development (2017) describes tax avoidance as “taxpayer’s intention to legally reduce tax liabilities although this action is usually in contradiction to the intent of the law it purports to follow”. Likewise, Hanlon and Heitzman (2010) defines tax avoidance as the reduction of explicit taxes. This is determined by specific tax rates, the likelihood of detection and punishment and the intrinsic beliefs of firms in terms of society. For instance, if a firm’s reduction of taxes goes beyond the legality it can be forced by tax authorities to pay penalties, representing a reduction of shareholder wealth. From this definition we should remark that tax avoidance practices are within the boundaries of the legality. Other terms as tax evasion are beyond the spirit of the law and therefore they are illegal actions. In this thesis I focus the attention on tax avoidance, so I will not study other tax practices.
It could be noted from the above paragraph that the legal minimizing of tax payments, in other words tax avoidance, is beneficial for firms. This would be understandable if firms do not take into account social and ethical considerations and instead consider corporate taxes as just being an operating cost of the corporation (Avi-Yonah 2008). However, tax avoidance does have social and ethical implications. This is because taxes are an important part in the provision of publics goods such as public health care and education (Freise et al. 2008). In the same line, Weisbach (2002) expresses that tax avoidance practices affect the welfare of society. Aggressive tax activities are seen by society as irresponsible and unpatriotic. Furthermore, Hoi et al. (2013) state that aggressive tax practices are an opportunistic behavior where firms take advantage in the contract between the firm and society in detrimental of the latter. Therefore, there is a controversial thought: lowering corporate tax payments can be beneficial for shareholders but at the same time it can be detrimental for the society as a whole.
2.2 Corporate Social Responsibility
The term Corporate Social Responsibility was introduced in 1952 by Bowen in the United States. Bowen (1952) realized that companies should pay more attention to the goals and values that society demands. In a brief timeframe, this term was applied in practice by known entrepreneurs of this time such as Andrew Carnegie and Henry Ford, who created a healthcare system and provided houses to their employees.
Nowadays, CSR has become an important concept that has acquired the attention in the global economy. Examples of this are the large number of companies documenting in their webpage how they pursue and achieve CSR goals and the concern of political institutions, as the European Commission, which is tremendously implicate in the development of this term (Goerke 2017). The increase in attention for CSR has been due to globalization and international trades which have derived in the demand for companies to increase transparency and corporate citizenship. Furthermore, the highly competence between firms make more difficult their ability to differentiate themselves from competitors and a way to do this is by engaging in CSR (Jamali and Mirshak 2006). While there is still no consensus about what CSR precisely means, different definitions can be found. For instance, Dahslrud (2008) gathers 37 definitions and all of them are consistent in two facts. Firstly, with the fact that companies chase social goals, in addition to economics benefits. Secondly, they refer to five dimensions which CSR is related to: voluntariness, environment, economy, society and stakeholders. For instance, The Commission of the European Communities (2001) defines CSR as a voluntary social and environmental concern of firms in their business operations and in the interaction with shareholders. Likewise, the Business for Social Responsibility (2001) expresses that CSR is a manner to operate that encounter and surpass legal, ethical and public beliefs that society has of business. Overall, the principal idea of CSR is to contribute positively to increase the value of society and, at the same time, to the value creation of the company in terms of profitability. Therefore, firms have to work in accordance to the needs of a wider array of stakeholders.
2.3 Relation between CSR and tax avoidance
Theory and empirical evidence regarding the relation between CSR and tax avoidance have found mixed results suggesting different views. Margolis and Walsh (2003) suggest that firms sometimes have to engage in social responsible processes that do not provide the maximum profit in order to fulfill the requirements of all stakeholders. Aligned with this, Avi – Yonah (2008) states that firms should not be involved in tax avoidance practices even if the strategy used is legal. This is because taxes are one of the main ways in which firms engage with society (Christensen and Murphy 2004). Taking the preceding arguments together the Stakeholder and Corporate Culture theory is supported. This view suggests that firms have social obligations besides creating value for shareholders. Different researches have approved this theory as for instance, Lanis and Richardson (2015) who investigate the link between CSR and tax avoidance and conclude saying that more socially responsible firms engage in less tax avoidance practices. So, according to this theory there seems to be a positive relation between CSR and tax avoidance.
In contrast to this theory, previous literature has also shown a negative relation between CSR and tax avoidance in which CSR is not used in an ethical and responsible manner. This negative relation has been developed by the risk management view. According to this theory, firms believe that the payment of taxes reduces social welfare because tax payments reduce their monetary resources to invest in innovation and job growth (Davis et al. 2013). Consequently, firms enter in tax avoidance practices. In fact, some empirical and theoretical researches have proved that tax payments reduce their overall investments. For instance, Djankov et al. (2008) state that there is a contradictory effect between corporate taxes and firm’s investments.
In addition to this, several studies also discuss the fact that some firms are more efficient allocating resources than the governments. Therefore, firms believe that social welfare is reduced by corporate tax payments. According to this beliefs, Porter and Kramer (2006) expresses that when a firm works properly, it can use its knowledge and expertise to enhance social wealth in a better manner than a government could do. Supporting this theory Hines et al. (2010) and Henderson and Malani (2009) state similar arguments. So from this perspective, firms would not pay taxes because they believe that social welfare will be reduced.
Finally, existing research suggest that firms engage in CSR activities in order to reduce their consequences for being involved in social negative practices (Fombrun et al. 2000; Minor and Morgan 2011). Consistent with this Godfrey et al. (2009) find that the reduction in shareholder’s value after negative news is lower when firm is engaged with CSR activities.
To sum up, while the Stakeholder and Corporate Culture theory seems to find a positive relation between CSR and tax avoidance, the risk management theory view CSR as a way to reduce negative social sanctions due to tax avoidance practices and CSR is merely used as self-interest mechanism by the management. Based on these two contradictory theories I develop the following hypothesis:
H1A: Higher corporate socially responsible firms are less likely to engage in tax avoidance practices than firms that are less corporate socially responsible.
H1B: Higher corporate socially firms are more likely to engage in tax avoidance practices than firms that are less corporate socially responsible.
2.4 Culture
Besides CSR, there are other external factors affecting tax avoidance. For instance, Boone et al. (2012) show that culture, more precisely, religiosity is a determinant of tax avoidance by corporate tax payers. Culture can be defined as “a collective programming of the mind that distinguishes the members of one group or category of people from others” (Hofstede 2010, p.6). Besides religiosity, culture can be defined and distinguished in many other ways. Gray (1998) characterized cultural environments as: individual/collective, large/small power distance or masculine/feminine. Hofstede (2010) added three new dimensions to the latter mentioned characterization on culture: uncertainty avoidance, indulgence/restraint and long/short term orientation. In this paper I will focus the attention on the distinction between masculine and feminine countries since, as stated in the introduction of this paper, masculine and feminine dimension are related to investor protection and the share- or stakeholder orientation of a country (Janssen 2017). This division involves the different roles between sexes in society. The dominant values in masculine societies include the importance of showing off, achieving visible goals and making money (Hofstede 2001). Conversely, feminine countries are defined by Gray (1998) as being modest and helping others, thus less prone to be involved in unethical practices, such as tax avoidance activities.
In addition, differences between women and men in terms of risk taking behaviors have been studied in previous research. The vast majority suggests that women are in general more risk averse than men and they are also more compliant with the law and regulations (Brinig 1995). There is also a large amount of literature explaining gender differences, both in the general population and among professionals. For example, Brinig (1995) shows that women tend to be less risky than men in order to avoid being caught and convicted. Likewise, Olsen and Cox (2001) find that males are less concerned about downside risks than females. Therefore, culture within a country seems to be related with tax avoidance practices. Based on that I develop the following hypothesis between the relation of culture and tax avoidance:
H2: Firms in feminine countries are less likely to engage in tax avoidance practices than firms in masculine countries.
2.5 Relation between CSR, culture and tax avoidance
From the literature above stated, we can establish that culture and CSR are both related with tax avoidance practices. Consequently, combining culture and CSR may result in the explanation for the controversial outcomes encountered in previous literature. On the one hand, CSR is a determinant on tax avoidance but results are still inconsistent. Some studies find a positive relation between CSR and tax avoidance. This prediction is in accordance with theories explaining that firms are concerned with their social obligations and therefore their actions go beyond maximize shareholder’s wealth (Carroll 1979; Mackey et al. 2007). Conversely, some other researches show a negative relation. According to this theory firms see tax payments as a reduction of social welfare, thus are likely to engage in tax avoidance activities. On the other hand, feminine countries have been proved to be less risk averse and more conservative in financial choices (Jianakoplos and Bernasek 1998). In addition, in feminine countries it is not even likely that tax avoidance activities are practiced in the first place. So, it seems that in feminine countries, CSR will reduce the incentive to engage in tax avoidance activities. Conversely to this, firms within masculine countries in which making money is one of the main values of society, could be encouraged to engage in tax avoidance practices in order to reach higher profits. Based on that, there seems to be a moderation effect of culture in the relation between CSR and tax avoidance. Therefore, firms that are engaged in corporate social responsibility within feminine countries result in fewer tax avoidance practices than corporate social responsible firms within masculine countries. This leads to the hypothesis in which culture acts as a moderating effect between the relation of CSR and tax avoidance.
H3A: Corporate socially responsible firms in countries that are characterized as feminine are less likely to engage in tax avoidance practices compared to corporate socially responsible firms in countries characterized as masculine.
Apart from the moderating effect, a mediating effect between CSR and culture may be identified. Feminine countries are more likely to engage in CSR activities (Strand et al. 2014). From Gjølberg’s index (2009), which measured the representation of companies by countries in three different CSR indexes, Strand et al. (2014) concluded that feminine countries are more prone to support stronger CSR performances than masculine countries are. Moreover, country characterization affects tax avoidance as stated in the hypothesis above mentioned. Countries in which feminine culture is established are less likely to engage in tax avoidance practices since their dominant values are cooperation, caring for others and modesty. On the contrary, masculine countries, where competitiveness and professional ambition is in the core of their values, are obviously more likely to engage in practices that are related to increase wealth, such as tax avoidance activities. So, the culture categorization of a country affects both the use of tax avoidance practices and at the same time the engagement of firms in socially responsible practices. This leads to the next hypothesis, which helps to understand in a better way the relation between CSR and tax avoidance practices.
H3B: Firms in countries that are characterized as feminine are more likely to engage in corporate social responsibility and therefore are less likely to engage in tax avoidance practices compared to firms in countries that are characterized as masculine.
2. Research method
3.1 Conceptual model
(IV) (DV)
CULTURE
TAX AVOIDANCE
CULT
CSR
(2)
(IV) (3)
(4) (1)
CSR
CONTROL VARIABLES: SIZE, MBT, LEV, PPE, INTANG, RESOURCES, GOVERNANCE, R&D, PTROA, INVPRO.
(1) Relation between CSR and Tax Avoidance.
(2) Relation between Culture and Tax Avoidance
(3) Moderation effect of Culture in the relation between CSR and Tax Avoidance
(4) Mediation effect of CSR in the relation between Culture and Tax Avoidance.
3.1.1. Dependent variable: Tax Avoidance
Tax avoidance is the dependent variable of this study. I use two different methods to calculate tax avoidance practices since each measure has its own limitations. Thus, using different methods allow the comparison between the results, and in the case they are consistent, I can be more confident about the robustness of results. The measures used in this study are: GAAP effective tax rate and Long-run cash effective tax rate.
Although there are more measures to calculate tax avoidance practices, as for instance the “book tax gap” used in different previous studies (Manzon and Plesko 2002; Hoi et al. 2013), these measures are more focused on revealing aggressive, shelter and evasion tax practices. However, in this research I widely define tax avoidance as the activities focused on cutting tax liabilities by legal tax planning activities (Dyering et al. 2008). So, the purpose of this research is to capture a broader meaning of tax avoidance rather than to catch tax aggressive or tax evasion practices. Therefore, I use effective tax measures which have been tested to better explain broader tax avoidance activities than other measures used for the same purpose (Davis et al. 2016).
The effective tax rate measure is widely used in prior literature (Minnick and 2010; Fariz and Klamm 2012) since it has been proved to be a helpful proxy in the estimation of firm’s tax planning activities (Phillips, 2003). ETR is calculated dividing a proportion of a tax liability by a profit before taxes or cash flow (Hanlon and Heitzman 2010) and basically measure the average of tax per portion (e.g. a dollar) of profit or cash flow. Different variants of ETR can be found in related literature, which the most used are GAAP ETR and the Cash ETR. So, following Fariz and Klamm (2012) I recognize differences between tax expense and taxes paid. For example, tax credits do affect tax payments by lowering payments but it does not have any effect in the accounting tax expenses. Furthermore, Hanlon and Heitzman (2010) express that accrued liabilities included in the tax expense may be adjusted by earnings management, while cash settlements consider cash outflows.
On the one hand, the GAAP ETR, which needs to be included in the financial statements, is the proportion of tax expenditure to profit before taxes (Dyreng et al. 2008). This measure accounts for the accumulate proportion of the accounting income that is tax payable, so this measure has an effect in accounting earnings. I calculate GAAP ETR with the formula stated below:
GAAP ETR = Tax Expenses / Pre-tax Income
Even though GAAP ETR has been used in multiple studies as an index to calculate tax avoidance, it presents certain boundaries. This measure may not reflect deferred taxes caused by the use of aggregate tax expenses. Moreover, because GAAP ETR calculates tax avoidance practices with respect to accounting numbers, it perhaps just reflects non-conforming tax avoidance activities (Aramide et al, 2013).
On the other hand, I use Long-run cash ETR to measure tax avoidance as an improved alternative to the GAAP ETR. This measure overcomes the failures of the GAAP ETR. Firstly, the use of cash retrieved from tax payments instead of tax expense helps to take tax benefits of employee stock options into account (Minnick and Noga 2010). Furthermore, conversely to the GAAP ETR, changes in valuation allowances and tax cushions do not affect Cash ETR calculations (Dyering et al. 2008). Secondly, the volatility in the measurement of tax avoidance due to temporary differences in the treatment of some accounts, will mostly disappear over time (Dyering et al. 2008). Thus, tax avoidance should be measured using a number of years rather than one year. In this research we will take five year cash ETR following Davis et al. (2016). The five year cash ETR is calculated as the sum of cash taxes paid over a five year period divided by the sum of pre-tax income over the same five year period. So I measure the Long-term cash ETR with the formula below:
Long-term Cash ETR = ∑ⁿ t=1 Cash Taxes Paid / ∑ⁿ t=1 (Pre-tax Income – Special Items)
3.1.2. Independent variables: CSR and Culture
CSR
CSR score is the first independent variable in the research. The database used to collect the data required to measure CSR scores is Thomson Reuters Asset4. However, the most of the previous researches that tried to find the relation between CSR and tax avoidance practices used KLD database as Fariz and Klamm (2012) or Davis et al. (2016), I hereby decide to use Thomson Reuters Asset4 following Preub and Preub (2017) who state that the relation between CSR and tax avoidance should be measured with more than one single rating agency.
Thomson Reuters Asset4 gathers data mainly from CSR reports created by the companies but also from different available public disclosures as company websites, annual reports, newspapers and stock exchange files. To create a CSR score firms are evaluated against more than 750 points and more than 250 Key Performance Indicators. These 250 Key Performance Indicators are divided in 18 categories which further are organized into four main performances (pillars): economic, environmental, corporate governance and social. Therefore, the “Integrated Rating”, which is the overall score of the firm, is calculated combining the four pillars (Heuvel, 2012).
In line with prior researches (Kim et al., 2012; Janssen, 2014; Davis et al., 2016), I exclude corporate governance as one of the main performances since the relation between corporate governance and tax avoidance differs from the relation between CSR and tax avoidance and it has been already researched (Liwosky et al., 2013). Financial institutions are also excluded from the sample. So the “Integrated Rating” in this research is obtained by merging economic performance, environmental performance and social performance. The variable is called CSR.
CULTURE
The second independent variable used in this study is Culture. Following Janssen (2017), the culture of each country in which firms operate are distributed in either feminine or masculine environments. This variable is established as a dummy variable in which a firm operating in a country within a masculine culture get a value equal to 1, whereas a firm which is operating within a feminine country gets a value equal to 0. The selection of the countries is based in Hofstede (2010).
3.1.3. Control variables
In the model there are included control variables that have been proved in prior studies to influence the relation between CSR, culture and tax avoidance. These control variables can be divided in three different groups depending to which relation they have an effect.
Firstly, control variables are applied for both culture and CSR in the relation with tax avoidance practices. In line with Janssen (2014) and Davis et al. (2016) I consider Size (SIZE) and growth (MBT) important factors because they may express the differences in tax avoidance. Furthermore, Pacheco Paredes and Wheatley (2017) control for financial leverage (LEV). Finally, I include property plant and equipment (PPE) calculated as net PPE divided by lagged total assets.
Secondly, control variables can only be introduced in the relation between CSR and tax avoidance activities. Davis et al. (2016), establish intangible assets (INTANG) which is calculated as the intangible assets divided by lagged total assets. Since corporate governance is different from CSR, a control variable for corporate governance (GOVERNANCE) is added. Ultimately, cash flow over total assets (RESOURCES), research and development (R&D) and pretax profitability (PTROA) are included as control variables for this relation.
Thirdly, a control variable is included for the relation between culture and tax avoidance. Following Pacheco Paredes and Wheatley (2017) investor protection (INVPRO) is added to control for culture. I take this measure from the World Bank Doing Business which evaluates three dimensions of investor protection: the extent of director liability, the ease of shareholder suits and the extent of disclosure index. So, the investor protection value is calculated by taken the average of these three indices (Todea and Plesoianu 2010).
Table 1 include the control variables used and how they are calculated in this study:
Table 1: Control Variables
Control Variables Measure