A Study on Action 11 of the OECD/G20 BEPS Action Plan

(1)

A Study on Action 11 of

the OECD/G20 BEPS

Action Plan

MEASURING AND MONITORING BASE EROSION AND

PROFIT SHIFTING; A COMPARISON OF THE FINAL

REPORT ON ACTION 11 OF THE OECD/G20 BEPS

ACTION PLAN AND ECONOMIC SCIENTIFIC

LITERATURE

ANNELIES EILANDER

WINTER 2016

(2)

(3)

Master’s Thesis Economics of Taxation III

A Study on Action 11 of the OECD/G20 BEPS Action Plan

Measuring and Monitoring Base Erosion and Profit Shifting; a Comparison of the Final Report on Action 11 of the OECD/G20 BEPS Action Plan and Economic Scientific Literature

Annelies Eilander BSc.

Master Thesis for Economics of Taxation

Faculty of Economics and Business, University of Groningen Prof. dr. I.J.J. Burgers

(4)

Master’s Thesis Economics of Taxation IV

Abstract

Tax avoidance is a global problem that has been receiving a lot of public attention since the outbreak of the economic crisis in 2008. In July 2013, the OECD and G20 published the BEPS (Base Erosion and Profit Shifting) Action Plan to address tax avoidance. The Action Plan contains fifteen Action Points. This thesis focuses on Action Point 11. The aim of Action 11 is to “establish methodologies to collect and analyse data on BEPS and the actions to address it”. Collection and analysis of BEPS data should result in an estimation of the economic scale and impact of BEPS. The work on Action 11 resulted in the Final Report on Action 11 presented on 5 October 2015. The Final Report contains guidelines for the assessment of currently available data and presents a “dashboard” of indicators. A key finding presented in the Final Report is that currently available data is only fit to make a rough estimation of the economic scale and impact of BEPS. Currently available data lacks detail and is not able to distinguish between economic-driven decisions and tax-motivated income shifting.

This thesis contributes to the work on Action 11 through a literature review of scientific economic literature from the years 2010-2015. The purpose of this thesis is to compare the findings in the Final Report on Action 11 with findings of scientific literature. This thesis provides insight in the alignment between the findings presented in the Final Report on Action 11 and conducted economic research in the past five years.

The first step to achieve this goal is making an overview of used measurements of tax avoidance, found indicators and data sources used in the sample studies. The sample consists of eighteen articles published in seven different journals. The literature review shows that twelve different measures of tax avoidance have been used, nineteen indicators of tax avoidance have been identified and fourteen different data sources have been used in the sample studies. The overview contains measures, indicators and data sources not reviewed by the OECD and therefore provides additional input to the work on Action 11.

Furthermore, this thesis contributes to the work on Action 11 by addressing the flaws on the use of the identified measures. The input for the measurement of tax avoidance is often obtained from data sources that lack detail and the ability to distinguish between economic driven decisions and tax-motivated income shifting. The indicators identified in the sample studies are not similar to those that form the “dashboard of indicators”. However, the indicator “CSR-performance” seems to be a promising candidate to be included in this dashboard as it is easy to measure and the link between CSR-performance and tax avoidance is supported by several scientific studies.

(5)

Master’s Thesis Economics of Taxation 1

5.2 Which data sources have been used in the analysis of BEPS and the actions to address it? 26 5.3 Which limitations are attached to the use of the identified measures, indicators and data sources? ... 29 6 Discussion ... 38 6.1 Study limitations ... 38 6.2 Further research ... 38 7 Conclusion ... 40 7.1 Indicators ... 40 7.2 Data sources ... 41

7.3 Methodologies to measure economic the scale and impact of BEPS ... 41

7.4 Conclusion ... 41

Appendix I ... 45

(6)

1 Introduction

Tax avoidance is an issue that has been widely discussed since the outbreak of the international economic crisis in 2008. The economic crisis has also moved the tax avoidance debate to the public. If multinationals do not pay their fair share of taxes they risk being the subject of a media campaign1

against them. However, the willingness of multinationals to comply not only with the letter but also with the spirit of the law seems to be increased2. This changing attitude towards tax avoidance and tax transparency creates an environment which is probably most suitable to implement further changes to the global tax system with regard to tax avoidance and tax transparency.

Multinational enterprises use several methods to avoid taxes. They erode the taxable base (f.e. by making use of a 'hybrid mismatch') and/or artificially move profits from high tax jurisdictions to low tax jurisdictions. These methods are commonly known as base erosion and profit shifting (BEPS). In 2013, the Organisation for Economic Co-oporation and Development (OECD) and the G20 countries issued the BEPS Action Plan. The BEPS Action Plan contains fifteen Action Points. This thesis will focus on Action 11 of the BEPS Action Plan.

Action 11 is described as follows in the BEPS Action Plan:

"Establish methodologies to collect and analyse data on BEPS and the actions to address it"

Dharmapala (2014) finds evidence for the existence of tax-motivated income shifting. However, the economic scale of the problem is not yet fully known. The OECD estimates that tax-motivated income shifting accounts for a loss of 4-10% of global CIT-revenues in its Final Report on Action 11 (OECD, 2015d). Action 11 of the BEPS Action Plan is an initiative to encourage measuring the scale and finding new methodologies to measure BEPS and the effectiveness of the BEPS Action Plan itself. The Final Report on Action 11 (OECD, 2015d) , issued on 5 October 2015, addresses specific challenges with regard to Action 11. One of these challenges is the lack of useful data and data sources. For example, data and data sources that are currently available often lack the ability to separate real economic activity from BEPS-related activity (OECD, 2014; OECD, 2015d). It is important to know which limitations and caveats are connected with the use of data and data sources because flaws in data affect the reliability of a study. In other words, research is only as good as its underlying data. This also holds for the relationship between data and data sources, data is only as good as its source.

This study answers the following question:

1_{See for example: Google, Amazon, Starbucks: The rise of "tax shaming"} http://www.bbc.com/news/magazine-20560359

(7)

Master’s Thesis Economics of Taxation 3 How do methodologies, indicators and data sources to measure and monitor BEPS that have been used in economic literature relate to the findings and recommendations mentioned in the Final Report on Action 11 of the OECD/G20 BEPS Action Plan?

This study will contribute to the scientific literature on Action 11 in multiple ways. The first part of the study consists of an overview of measures, indicators and data sources that have been used in economic studies related to BEPS in the past five years. The first part generates more information about which methodologies have been used to measure and monitor BEPS in BEPS-related research. The first part consists of the following two sub-research questions:

1. Which methodologies and indicators have been used or identified to measure the economic scale and impact of BEPS in the past five years in a selection of scientific literature?

2. Which data sources have been used in the analysis of BEPS and the actions to address it in the last five years in a selection of scientific literature?

The second and last part of this study answers the following question:

3. Which limitations are attached to the use of the identified measures, indicators and data?

The Final Report on Action 11 serves as a guide to answer the last sub-research question. The Final Report defines potential criteria for evaluating available data on BEPS research (OECD, 2015d). By testing if the data and data sources that are found in the first part of the study satisfy these criteria, this study gives insight in the comparability of findings in scientific literature and the work of the OECD. To the best of the author’s knowledge, such a comparison has not yet been published.

Tax avoidance is a phenomenon that influences society as a whole3_{. This study is therefore of both}

academic and societal relevance. This study will not contain any opinions about tax avoidance by multinational enterprises. The sole purpose of this study is to provide a clear overview and discussion of data and methods that can be used to measure and monitor base erosion and profit shifting objectively. Although ethics can be considered relevant with regard to tax avoidance4_{this academic}

field will not be included in the research.

This thesis continues as follows: In Chapter 2 some relevant concepts and terms are clarified, Chapter 3 discusses the state-of-the-art of scientific literature with regard to Action 11 of the BEPS Action Plan, Chapter 4 sets out the research methods and in Chapter 5 the results are described. The thesis concludes with a discussion and recommendations for further research in Chapter 6 and a conclusion in Chapter 7.

(8)

2 Background information and Glossary

2.1 Glossary relevant terms data analysis

This paragraph provides definitions of some relevant terms. The definitions relate to features of data and data analysis that influence the trustworthiness of scientific research. One of the steps in achieving the goals of Action 11 is the assessment of currently available data, more details are provided in paragraph 3.3.2.1. The definitions of relevant terms are given in the box below:

2.2 Tax avoidance by multinationals

By definition, multinational enterprises conduct business in multiple countries. Each of these countries has their own set of domestic tax rules. Domestic tax rules are often established with the goal to achieve coherence within the domestic tax system. However, interaction between the domestic tax rules of different countries leads to gaps and friction (OECD, 2013b). The lack of global coherence between domestic tax systems of different countries creates potential double taxation. The lack of coherence leaves gaps that create possibilities for double non-taxation as well (OECD, 2013b). Multinational enterprises could use these gaps to lower their tax burden artificially. If a multinational uses these gaps to artificially lower the tax burden it faces it could be considered to be tax-avoidant. However, sometimes a multinational faces a lower tax burden because of non-tax avoidant reasons5_.

This distinction is not always properly made in the news media. News media coverage of tax avoidance has increased in the past years. Media coverage raises awareness of corporate tax avoidance for the general public but can also have adverse effects. A study (Lee, 2015) on the effects of news media coverage of tax avoidance on the quantity6_{and quality}7_{of tax disclosure finds that the increased media}

coverage of tax avoidance does not have a positive effect on the quantity and quality of tax disclosure per se. It is even possible that news media coverage discourages companies to be transparent about tax-related items (Lee, 2015).

5_{F.e. loss relief from previous years}

6_{I.e. volume of tax disclosure in financial audited statements} 7_{I.e. contents of tax information in the corporate annual reports}

Credibility – The degree to which results are believable

(9)

Master’s Thesis Economics of Taxation 5 An example of some tax avoidance techniques will be given in the following paragraph.

2.2.1 Base erosion and profit shifting (BEPS)

Tax avoidance is often done by using a method that erodes the taxable base and/or by a method that artificially shifts profits away from the country in which they are generated. The OECD describes BEPS as follows in the BEPS Action Plan (OECD, 2013b):

“BEPS relates chiefly to instances where the interaction of different tax rules leads to double non-taxation or less than single non-taxation. It also relates to arrangements that achieve no or low non-taxation by shifting profits away from the jurisdictions where the activities creating those profits take place. No or low taxation is not per se a cause of concern, but it becomes so when it is associated with practices that artificially segregate taxable income from the activities that generate it. In other words, what creates tax policy concerns is that, due to gaps in the interaction of different tax systems, and in some cases because of the application of bilateral tax treaties, income from cross-border activities may go untaxed anywhere, or be only unduly lowly taxed.”

Erosion of the taxable base is often accompanied by profit shifting8_{. An example}9_{of a method that}

erodes the taxable base is the so-called “cross border double-dip”. A double-dip structure is illustrated in the figure below (OECD, 2012):

In this double-dip structure the two companies use a hybrid financing instrument. Company A (A Co) finances Company B (B Co) with an instrument that qualifies as debt in country B and as equity in country A. Country B recognises the payment from B Co to A Co as deductible interest. Country A qualifies the payments from B Co to A Co as dividends. If A Co is situated in a country with a dividend exemption A Co doesn't have to include the payments in her taxable income10_{. This structure is also}

known as a hybrid mismatch arrangement (OECD, 2012). Country A allows deduction of the payments while at the same time Country A does not include the income in her taxable base. Double non-taxation

8_{See the example of the double-dip structure}

(10)

Master’s Thesis Economics of Taxation 6 is the consequence. This form of double non-taxation is hard to oppose because both companies comply with the law11_.

Another form of profit shifting in order to avoid taxes uses differences between statutory tax rates. Shifting profits from countries with a high statutory tax rate to countries with a low statutory tax rate can be done in several ways (OECD, 2010; OECD, 2012). First, profit shifting can be accomplished through the organisation of the finance structure of an international concern. It is attractive to finance affiliates in high-tax countries with high loans from affiliates situated in countries with a low statutory tax rate. Shifting profits through transfer pricing is the other major method used by multinational enterprises. The allocation of profits can be altered by adjusting the price of intra-firm trade and services. However, the consideration should be in line with the "at arm's length" principle12_(OECD,

2010). The revised OECD Transfer Pricing Guidelines (OECD, 2010), published in 2010, are intended to provide clarity on how the principle should be applied in practice. However, application of the at arm’s length principle is often problematic in practice, especially if there is no comparable price or market for intra-group trade or services13 _{(OECD, 2010). Application of the arm's length principle has been}

extensively criticised in scientific literature as well (among others, (Keuschnigg & Devereux, 2013; Koomen, 2015; Avi-Yonah & Benshalom, 2010)). Keuschnigg & Devereux contest the view that an independent transfer price is the right price for intra-group transactions. They argue that application of the principle distorts multinational activity by reducing debt capacity and foreign investments (Keuschnigg & Devereux, 2013).

2.3 Tax Avoidance and Corporate Social Responsibility

CSR is a concept developed in the second half of the 21st _century14_{. A company applying CSR typically}15

keeps in mind its influence on society as a whole. Related to tax matters, this could mean that a company pays its fair share. This paragraph discusses briefly the relationship between Corporate Social Responsibility (CSR) and tax avoidance.

The nature of the relationship between CSR and tax avoidance has been the subject of several scientific studies (Lanis & Richardson, 2012; Laguir, Staglianó, & Elbaz, 2015). Lanis and Richardson (2012) find a negative and statistically significant association between the level of CSR disclosure and tax aggressiveness. Thus, the more socially responsible firms among the sample of 408 Australian firms were less likely to engage in aggressive tax planning. A recent French study (Laguir, Staglianó, & Elbaz,

11_{I.e. the letter of the law.}

12_{This principle is set out in art. 9 of the OECD Model Convention} 13_{As is often the case for R&D-goods}

(11)

Master’s Thesis Economics of Taxation 7 2015) confirms the findings of Lanis & Richardson (2012). Additionally, the results of this study indicate that the nature of the CSR dimensions16_{a firm has developed is relevant for the level of tax}

aggressiveness. For example, a high social dimension relates to a low level of tax aggressiveness, whereas the authors find a positive relationship between the economic dimension and the level of tax aggressiveness (Laguir, Staglianó, & Elbaz, 2015). Another study (Lanis & Richardson, 2015) finds that companies with a high score on CSR performance tend to carry out less tax avoiding practices than companies achieving a low score on CSR performance. This study also finds that CSR categories "community relations" and "diversity" represent elements of CSR performance that are particularly important for lowering tax avoidance (Lanis & Richardson, 2015). “Community relations” refers to the extent a company contributes to its direct surroundings, for example charitable giving and support for education, housing and volunteer programs (Lanis & Richardson, 2015). The category “diversity” covers the extent to which women and minorities are represented in the company and the balance between work and life (Lanis & Richardson, 2015).

The above shows that CSR could be of use for tackling BEPS. However, despite encouraging findings in scientific literature consensus on if and how taxation should be a part of the CSR-strategy of multinational enterprises has not been reached yet. The OECD has issued a report (OECD, 2011) for multinational enterprises containing guidelines on how to implement CSR in their day-to-day practice. The OECD formulates the following view on CSR and taxation:

"It is important that enterprises contribute to the public finances of host countries by making timely payment of their tax liabilities. In particular, enterprises should comply with both the letter and spirit of the tax laws and regulations of the countries in which they operate.”

Interpretation of the law in a strict grammatical way is therefore not sufficient according to the OECD. Multinational enterprises are responsible for aligning their tax strategy with the spirit of the law. Scientists have expressed this view as well. Christensen & Murphy (2004) address the problems accompanying tax avoidance among multinational enterprises and suggest that businesses should adopt CSR standards on taxation. Landolf (2006)17_{even calls it a "crime against the nation" if a}

company decides not to pay their fair share of taxes. Happé (2011) argues that it is an ethical duty of companies to pay their fair share of taxes and act in accordance with the “spirit of the law”.

16_{The following CSR dimensions have been used in this study: social, governance, economic and environmental} 17_{article available online via this link:}

(12)

Master’s Thesis Economics of Taxation 8 The following paragraph discusses country-by-country reporting. This concept is expected to be of great importance for the work on Action 11 as it is able to provide additional and more detailed data for the data analysis of BEPS (OECD, 2015d).

2.3.1 Country-by-country reporting (CbCR)

CbCR is a concept based on disclosure of key business information, such as profits and taxes per country (Evers, Meier, & Spengel, 2014). Multinational enterprises that comply with CbCR disclose profit and taxes for each individual country in which they operate. Proponents of CbCR claim that disclosure of key business information on a per country basis encourages multinationals to pay a fair amount of taxes, based on economic activity, in each country they operate in. CbCR could also enhance efficiency of tax administrations and reduce the costs of detecting abusive tax arrangements (OECD, 2015c). On the other hand, opponents claim that CbCR is not useful as an instrument against international profit shifting as financial statements only give insight in future prospects of a company and therefore are not of use in detecting profit shifting in the past (Evers, Meier, & Spengel, 2014). Opponents also express warnings with regard to the level of detail required in a per country report. Disclosure of too many details can lead to high administrative costs for companies and can lead to situations in which a company has to provide details on confidential information18_{(Grau Ruiz, 2014).}

Although a discussion about the possible benefits and disadvantages of CbCR is still going on in scientific literature (f.e. (Lowell & Herrington, 2015; Evers, Meier, & Spengel, 2014; Grau Ruiz, 2014; OECD, 2015a)) the OECD announced in the Final Report on Action 13 of the BEPS Action Plan that new CbCR requirements are to be implemented for fiscal years beginning on or after 1 January 2016 (OECD, 2015c). The requirements apply to multinational enterprises with a consolidated group revenue of at least 750 million euros (OECD, 2015c). The aim of the implementation of CbCR is to simplify identification of transfer pricing agreements and other arrangements that artificially shift profits from the economic activities that create the profits for tax administrations (OECD, 2015c). A template of a report that complies with the new requirements as well as model legislation on CbCR can be found in the Final Report of the OECD on Action 13 (OECD, 2015c).

A form of CbCR has already been implemented in the extractive industries19_{. The Extractive Industries}

Transparency Initiative (EITI) is an international standard that can be adopted by countries on a voluntary basis (Evers, Meier, & Spengel, 2014). The extractive industries are subjected to the Dodds-Frank Act in the United States. This Act enforces listed US companies to file publications of payments

18_{F.e. information on APA’s (Grau Ruiz, 2014)}

(13)

Master’s Thesis Economics of Taxation 9 made to governments in each country it operates in, in a standardised way20_{. Both initiatives only}

require a per country specification of payments to governments, a geographical split of profits and/or taxes does not have to be provided under these regulations. The European Union imposes comparable obligations on the extractive industries through the Accounting and Transparency Directive21_.

Financial institution in the EU are subjected to CRD IV22_{. Regulations with regard to CbCR are part of}

this directive. Therefore financial institutions in the EU have to disclose information on profit and taxes on a per country basis to the European Commission (Evers, Meier, & Spengel, 2014).

20_{See https://eiti.org/} 21_{Directive 2013/34}

(14)

3 State of the art

This chapter provides an overview of publications related to Action 11 of the BEPS Action Plan. First, a short history on the establishment of the BEPS Action is given in paragraph 3.1. In paragraph 3.2 a short outline of the fifteen Actions of the BEPS Action Plan is given. This chapter gives a more detailed description of Action 11 in paragraph 3.3.

3.1 Short history OECD/G20 BEPS Action Plan

The international tax system is partly based on the initial Model Tax Convention published by the League of Nations, almost a century ago in 1928. Model Tax Conventions can be used as an example for bilateral or multilateral tax treaties concluded by states (Burgers & Kosters, 2015). The OECD published its first Model Tax Convention in 1963. This Model Tax Convention was based on the Model Tax Convention published in 1946 by the League of Nations (Burgers & Kosters, 2015). The work on Model Tax Conventions is reflected in hundreds of bilateral tax treaties worldwide. Tax treaties based on Model Tax Conventions of the OECD or the United Nations have two functions. These tax treaties prevent double taxation through allocation of taxing powers and counteract tax avoidance and tax evasion. Tax evasion is illegal whereas tax avoidance is not.

However, increasing globalisation, increased mobility of capital and the rise of the digital economy are factors that have great impact on the sustainability of the original international tax system. The international tax system leaves gaps and mismatches that can be exploited to generate double non-taxation23_{. In order to tackle the adverse effects of tax planning strategies that exploit these gaps and}

mismatches the OECD and G2024_{worked together on the BEPS Action Plan}25_{. The OECD and G20 issued}

the BEPS Action Plan on July 19, 2013 (OECD, 2013b).

The BEPS Action Plan contains fifteen action points which are discussed in more detail in the following paragraph.

23_{http://www.oecd.org/ctp/beps-about.htm}

24_{A dozen developing countries worked together directly on BEPS with the OECD and G20 and non-OECD and} non-G20 countries also provided input (http://www.oecd.org/ctp/beps-about.htm)

(15)

3.2 Actions of the BEPS Action Plan

The BEPS Action Plan contains the following fifteen action points (OECD, 2013b): 1. Address the tax challenges of the digital economy

2. Neutralise the effects of hybrid mismatch arrangements 3. Strengthen CFC-rules

4. Limit base erosion via interest deductions and other financial payments

5. Counter harmful tax practises more effectively, taking into account transparency and substance

6. Prevent treaty abuse

7. Prevent the artificial avoidance of the PE-status

8, 9, 10. Assure that transfer pricing outcomes are in line with value creation  (8) Intangibles

 (9) Risks and capital

 (10) Other high-risk transactions

11. Establish methodologies to collect and analyse data on BEPS and the actions to address it 12. Require taxpayers to disclose their aggressive tax planning arrangements

13. Re-examine transfer pricing documentation

14. Make dispute resolution mechanism more effective 15. Develop a multilateral instrument

The fifteen action points can be divided into three categories of actions and two “stand-alone” action points. The first category of actions (Actions 2-5) pursue international coherence of corporate income taxation (OECD, 2013b). Actions 6-10 are invoked to realign taxation with changing business models and technological developments in order to restore the intended effects and benefit of international standards (OECD, 2013b). The final category of actions contains actions 11-14, the actions in this category have the goal to improve the availability of relevant information to identify risk areas through enhancing transparency26_{while maintaining certainty and predictability for business (OECD, 2013b).}

A digital economy introduces tax challenges for both direct and indirect taxes. For example, an enterprise can have a significant digital presence in a country without being liable to tax under the rules of corporate income taxation. Action 1 aims to identify this challenges and options to address the main difficulties (OECD, 2013b).

(16)

Master’s Thesis Economics of Taxation 12 Action 15 of the BEPS Action Plan is of importance for the implementation of measures developed under the BEPS Action Plan. The purpose of Action 15 is to develop a multilateral instrument to implement the changes proposed by the work on the BEPS Action Plan and to amend bilateral tax treaties accordingly.

On 5 October 2015 the Final Reports on all Actions of the BEPS Action Plan were released. It goes beyond the scope of this thesis with a focus on Action 11 to discuss the findings presented in all of the reports in full detail. The following paragraph focuses on Action 11 and includes findings presented in the Final Report on Action 11.

3.3 Action 11 BEPS Action Plan

Tax avoidance through base erosion and profit shifting is a global problem. The economic scale and impact of the problem is not yet precisely known (OECD, 2015d). The OECD wants to quantify BEPS through the work on Action 11. This paragraph continues with a clarification of the need of Action 11 and a discussion of the findings presented in the Final Report on Action 11.

3.3.1 Motivation Action 11

The OECD states that there is enough anecdotal evidence (OECD, 2015d) to assume tax-motivated income shifting by multinationals does take place. In the BEPS Action Plan (OECD, 2013b) the OECD formulates several goals of Action 11. First, the work on Action 11 should provide tools to measure the economic impact of BEPS and monitor the effectiveness and impact of the actions to address BEPS. The other goal is to improve data analysis by assessing existing data sources and identify new data types as well as new methodologies to measure BEPS (OECD, 2013b).

The need to improve data analysis is also discussed in scientific literature. Dharmapala (2014) provides a review of the empirical literature on tax-motivated income shifting and examines estimations of the economic scale in this literature. He finds that more recent studies estimate a smaller magnitude of the economic impact of income shifting compared to older studies. More recent studies use newer and richer data sources than the earlier studies (Dharmapala, 2014). Dharmapala’s study shows that estimations on the magnitude of BEPS depend on the quality of the data that has been used.

The OECD recognises the need of improved data sources by adopting Action 11 in the BEPS Action Plan. Improvement of data sources and availability of data also has been the subject of several comments27

on the two Discussion Drafts on Action 11 issued by the OECD in 2014 and 2015 (Discussion Draft: OECD, 2015a; Comments: OECD, 2015b; OECD, 2014;). The foremost reason to assess current data

(17)

Master’s Thesis Economics of Taxation 13 analysis is that an improvement of the quality of the BEPS measurements is needed. The so-called signal-to-noise ratio enhances if data analysis is improved. The OECD illustrates this in the following figure (OECD, 2015b):

Figure 1

The illustration above displays the ultimate goal of Action 11. The next step in achieving this goal is going into transition by collecting new data. The Final Report on Action 11 contains recommendations on the new types of data that should be collected and on from which data sources this new data should be retrieved. These findings and recommendations are the subject of the following paragraph.

3.3.2 Final Report Action 11

The OECD published the Final BEPS package, including the Final Report on Action 11, on 5 October 2015. Measuring the economic scale of BEPS proved to be difficult due to complexity and certain data limitations, for example the lack of ability to disentangle real economic effects from tax-motivated income shifting. However, the OECD estimates, based on the findings of the work performed since 2013 that global CIT revenues are 4-10% lower due to BEPS (OECD, 2015d). In dollars, this means a loss of a 100-240 million per year (OECD, 2015d). This paragraph continues with an overview of the findings listed in the Final Report on Action 11.

Assessment of existing data relevant for BEPS analysis

(18)

Master’s Thesis Economics of Taxation 14 be filled with new types of data and new tools and it provides input for the development of “best practices” (OECD, 2015d). To evaluate the usefulness of data for BEPS analysis the OECD used the following criteria (OECD, 2015d)28_:

1. Coverage/Representativeness

2. Usefulness for separating real economic effects from tax effects 3. Ability to focus on specific BEPS activity

4. Level of detail 5. Timeliness 6. Access

The findings based on the testing of currently available data against these criteria show that there are certain limitations on the use of currently available data. A brief summary of these limitations is provided below.

Many studies use aggregated macro data, such as FDI29_{, even though separating economic effects from}

BEPS behaviours proves to be difficult if macro data is used (OECD, 2015d; Dharmapala, 2014). The OECD recommends to use firm level micro data for BEPS analysis. The use of firm level micro data, such as financial statements, gives rise to limitations as well. Coverage and completeness of these statements differ across countries (OECD, 2015d). The information given in these statements only reflects accounting concepts and gives therefore only an indirect indication of BEPS behaviour. The OECD mentions that tax return data in combination with financial account data could provide more direct information about the presence of BEPS (OECD, 2015d).

The lack of detail with regard to information about global activities of multinationals is also a concern. Incomplete information limits researchers to identify and quantify BEPS behaviours. Furthermore, a lack of detail of data concerning global activities of multinationals is especially problematic if the available information is non-random (OECD, 2015d). In that case, the available information is not representative of the real situation. This can distort conclusion drawn by researchers. The OECD also mentions30_{that Bureau van Dijk’s database ORBIS has limitations with regard to level of detail and}

coverage as well. It provides financial account data instead of tax return data and information on companies located in developing countries is limited.

Businesses do provide information to tax authorities that could be useful for BEPS analysis. However, in many cases this information is not made available for analysis purposes (OECD, 2015d). The OECD

28_{Box 1.1}

29_{Foreign direct investment}

(19)

Master’s Thesis Economics of Taxation 15 recommends to make the most of already existing data while keeping in mind confidentiality and administrative costs (OECD, 2015d). With regard to data collection, the OECD emphasises the importance of comparability across countries. Currently, reporting standards for accounting purposes differ across countries31_{. These difference complicate data comparison between countries. The OECD}

recommends to work towards homogeneity in data collection (OECD, 2015d).

Another key element of Action 11 is the development of indicators. The findings and recommendations presented in the Final Report on this issue are summarised in the next sub-paragraph.

3.3.3 Development of BEPS-indicators

Action 11 is implemented in the BEPS Action Plan to get more insight in the quantitative aspect of BEPS and in the effectiveness of the countermeasures. Developing BEPS-indicators is useful to achieve this goal because indicators provide information on the economic impact and scale of BEPS, are able to track changes in BEPS over time and monitor the effectiveness of countermeasures (OECD, 2015d). However, as mentioned before, the precision of indicators relies on the precision of the underpinning data. The previous paragraph shows that currently available data is subject to limitations. Until the quality of data is improved, indicators can only provide a general indication of BEPS (OECD, 2015d). The OECD formulates guidelines for the development of indicators in the Final Report. First of all, multiple indicators, based on both macro data and micro data, are needed due to limitations of currently available data (OECD, 2015d). The way an indicator is presented should also be considered. An indicator can be presented as an index or as a ratio32_{. Furthermore, cash flows, financial and tax}

related, should be linked to measures of economic activity. Indicators linked to economic activity control for changes in factors unrelated to BEPS, such as labour and capital (OECD, 2015d). Indicators should also be able to distinguish between real economic effects and tax-motivated income shifting. To ensure the signal-to-noise ratio is as high as possible, possible limitations on the use of indicators should be highlighted. The use of indicators that do not meet the guidelines should be avoided (OECD, 2015d). The OECD also requires that the indicators are simple, clear and timely to ensure they are easy to interpret by policy makers. It would be ideal if an indicators could be of use from a global perspective but could also be used by policy makers from individual countries without much difficulty (OECD, 2015d). To secure sustainability, the developed indicators should have the ability to provide more precise information when data quality improves.

As mentioned before, the OECD followed the concept of creating a dashboard of indicators, jointly able to identify and monitor consistent trends in the economic scale and impact of BEPS and/or specific

31_{F.e. requiring the reporting of accounting items geographically or per segment (OECD, 2015d)}

(20)

Master’s Thesis Economics of Taxation 16 BEPS behaviours (OECD, 2015d). The following six indicators form the dashboard of indicators presented in the Final Report on Action 11 (OECD, 2015d):

1. Concentration of high levels of FDI relative to GDP 2. Differential profit rates compared to effective tax rates

3. Differential profit rates between low tax locations and worldwide MNE activities

4. Effective tax rates of large MNE affiliates relative to non-MNE entities with similar characteristics

5. Concentration of high levels of royalty receipts relative to R&D spending 6. Interest expense to income ratios of MNE-affiliates in high-tax locations

This dashboard of indicators assists with the measurements and monitoring of BEPS. The OECD states that the indicators should be read like a meter that is capable to follow trends in BEPS- behaviour and as warning lights for the existence of potential BEPS-behaviour (OECD, 2015d). No single indicator is capable of providing full insight in BEPS-behaviour but empirical research33_{shows that all indicators}

are capable to measure and monitor BEPS (OECD, 2015d).

In the future, data analysis should be improved. The OECD makes several recommendations on how to get towards better data and tools to monitor and measure BEPS. The key recommendations are briefly summarised in the following paragraph.

3.3.4 The OECD’s recommendations towards better data and tools

The OECD states that getting a more complete picture of the economic scale and impact of BEPS improved data is crucial (OECD, 2015d). According to the OECD, governments should co-operate to improve data collection and data analysis. This includes improvements of public reporting of business tax revenues statistics and non-tax data relevant for BEPS-analysis (OECD, 2015d). Already existing data should be used more effectively and should be made available for research purposes. Next to that, governments should stimulate more academic research related to BEPS and MNE activities. The OECD itself continues the work on Action 11 as well. The OECD wants to work together with all countries that wish to publish a so-called Corporate Tax Statistics publication containing data on BEPS in a standardised format. Under the provisions of Action 13 CbCR becomes obligatory. CbC-reports should be anonymised and analysed and these analyses should be a part of the Corporate Tax Statistics publication (OECD, 2015d). Another recommendation is to publish reports on the proposed revenue impact of proposed and enacted BEPS countermeasures on a regular basis (OECD, 2015d).

(21)

3.4 Interim-conclusion

The Final Report shows us that Action 11 is still “under construction”. A key component of the Final Report is the expression of the urge to improve data analysis in several ways. New data should be collected and limitations on this data should be highlighted. This thesis identifies measures, indicators and data sources that have been used in economic scientific studies to measure and monitor BEPS in the last 5 years. The limitations on the use of the data are highlighted and the indicators are tested against the guidelines presented in the Final Report on Action 11 (OECD, 2015d).

This study contributes to the work on Action 11 in several ways. First, it provides input with regard to the assessment of currently available data relevant for BEPS-analysis. Second, insights are given with regard to coherence between the ambitions of the OECD and the findings in scientific literature. Third, it provides input for the process of developing indicators and discusses data sources and their limitations in more detail.

(22)

4 Research method

The study starts with an overview of measurements, indicators and data sources that have been used in BEPS-related studies. Due to the relative short time frame of this study the selection of studies which will form the sample is subject to certain limitations.

First of all, only articles published in ten journals selected by the author are used. The ten journals are listed in Table I. These ten journals are selected by the author on the basis of multiple reasons. Journals34_{that reached a high place in a ranking of economic journals constructed in Kalaitzidakis et}

al. (2011) are chosen because these journals are frequently cited35_{and have a high relative impact.}

Journals 1-5 mentioned in Table I are primarily chosen because of a top ranking36_{in the before}

mentioned study. Journals 1-5 are journals with a focus on economic theory. In order to get a more varied selection journals 6-8 are accounting journals. Journals 6-7 are mentioned several times in rankings of top academic accounting journals on the internet37_{. Journals 8-10 are chosen because the}

author knew by practical experience these journals publish articles on BEPS-related issues. Another condition for the selection of journals is institutional access to the content of the journals. If the author was not able to access the content of a journal via the University of Groningen a journal was not selected.

After the journal selection, the author consults the selected journals online and selects the relevant articles personally by using relevant search terms38_{. Given the scale and relatively short time frame of}

this study only articles that have been published in the years 2010-2015 are used for the first part of the study.

The other sub-research question is also answered through a literature review. However, the literature used for this part of research will not be subject to the same limitations in terms of publishing year and selected journals.

34_{F.e. American Economic Review, Quarterly Journal of Economics and Journal of Political Economy} 35_{Kalaitzidakis et al. (2011) adjusts for self-citation and the size of journal}

36_{All 5 are among the top-12 journals}

37_{Example given}_{http://www.byuaccounting.net/mediawiki/index.php?title=Top_ranked_accounting_journals}_;

http://www.byuaccounting.net/tenure/journalsincluded.php;

http://www.mysmu.edu/faculty/rogerloh/links/journal_rankings.htm

(23)

5 Results

This chapter discusses the results of the conducted research. Each sub question will be discussed in a separate paragraph. This chapter therefore starts with a short description and analysis of the scientific research on BEPS published in the selected journals39_{. The chapter continues with a discussion on the}

limitations and caveats on the use of the measures, indicators and data sources that have been used in the selected scientific literature.

5.1 Which methodologies and indicators have been used or identified to measure the economic scale and impact of BEPS in the past five years in a selection of scientific literature?

The goal of this paper is to compare the methods used to measure and monitor tax avoidance in economic scientific literature with the methods to measure and monitor BEPS proposed by the OECD in the Final Report on Action 11. The first step in achieving this goal is performing an analysis of (a part of) the scientific literature published in this scientific area in the past five years.

5.1.1 Sample selection

The final sample comprises of 18 articles from seven journals. Three of the selected journals40_{have not}

published articles that are of relevance for this study in the past five years. A list of the articles that have been selected is provided in Table 1 below. The largest portion of the selected articles are about associations between certain variables and tax avoidance. To the best of the author's knowledge, no articles on the effectiveness of specific actions of the BEPS Action Plan have been published in the past five years in the selected journals.

Table 1, Selected Journals and Articles

Journals Article (first author, publishing year)

1. American Economic Review 2. Journal of Political Economy 3. Quarterly Journal of Economics

4. Journal of Public Economics Dharmapala, 2013; Voget, 2011 5. Journal of International Economics Egger, 2010

6. Journal of Accounting and Economics Armstrong, 2012; Badertscher, 2013; Chyz, 2013; Hanlon, 2010; Hope, 2013; de Simone, 2015

7. Journal of Accounting and Public Policy Kubick, 2015; Lanis, 2011; Lanis, 2012; Richardson, 2013b 8. Journal of Contemporary Accounting

and Economics Richardson, 2013a; Salihu, 2015; Taylor, 2014

9. Journal of Law and Economics Borek, 2014 10. Journal of Business Ethics Lanis, 2015

39_{see Table I}

(24)

5.1.2 Measurements of scale and indirect identification

The literature review shows that scientific literature on tax avoidance can be separated in three fields. The first field focuses on finding indicators that can be used as a means to identify the presence of tax avoidance. The other field measures the scale of tax avoidance and/or develops tools to measure it. Both fields are of importance for measuring and monitoring the global problem of tax avoidance. The first field makes a valid contribution to the development of tools to measure the scale of BEPS by providing possible measures of the scale of the problem. Both fields could provide useful input to the work conducted by the OECD with regard to Action 11.

The third field is of relevance for measuring and monitoring BEPS in a more indirect by keeping close attention to the indicators of tax avoidance that are already identified. These indicators of possible existence of tax avoidance or tax aggressiveness could be kept under continuous observation to look for changes in their values. Changes in the values of the indicators can be a sign of diminishing or enhancing tax avoidance.

It is crucial to mention that a change in the value of the indicators can also be caused by other variables41_{. The indicators are only associated with tax avoidance, they do not cause it. More}

information on the limitations and caveats on the use of the measures and indicators identified in the sample studies is provided in paragraph 5.3. The results of the literature review are displayed in Table 2 (Measures) and Table 3 (Indicators).

5.1.3 Measurement of tax avoidance

There are several ways to measure the scale of tax avoidance. Roughly five ways of measuring tax avoidance42_{can be identified in the sample. The findings of the literature review are summarised in}

Table 2 on page 23. A more thorough description of each of the identified measures is given below.

Effective tax rates

As shown in Table 2 there are at least five different ways to measure tax avoidance through a variation of the effective tax rate (ETR). Each of the ETR-variations is relatively simple to compute and used as a measurement of tax avoidance in many studies that are a part of the sample (f.e. (Armstrong & Blouin, 2012; Badertscher, Katz, & Rego, 2013; Lanis & Richardson, 2011), among others). ETR's are computed by dividing an estimation of tax liability by a measure of before-tax profits (Hanlon & Heitzman, 2010). Tax liability is expressed in two different ways in the sample studies: tax expense and taxes paid. Tax

41_{F.e. CSR-performance can increase without a reduction of tax avoidance}

(25)

Master’s Thesis Economics of Taxation 21 expense refers to an estimation of the company’s real tax burden (Janssen, 2000). Taxes paid is the actual amount of taxes that has been contributed to the tax authorities.

Major deviations of ETR from the statutory tax rate can be a sign of tax avoidance43_{. Computations}

methods and definitions of the different ETR's are written down in Table 2. Which variety of the ETR should be used in measuring tax avoidance depends on the research question (Hanlon & Heitzman, 2010). For example, Cash ETR is the proportion of pre-tax book income of a certain year that has been used to pay taxes. A mismatch arises if taxes are paid for tax liabilities with regard to a previous year due to tax deferrals. The numerator and denominator refer to different time periods in such a situation (Hanlon & Heitzman, 2010)44_{. A deeper analysis of the implications of different types of ETR on}

research will be given in paragraph 5.3.

Ratio operating cash flow

Very similar, in terms of method and interpretation, to measuring tax avoidance through computing an ETR is the method that uses a ratio of tax expenses or taxes paid to operating cash flow to measure tax avoidance. This method uses operating cash flow as denominator instead of pre-tax book income. This change of denominator controls for differences in accounting methods relative to firm size (Zimmerman, 1983) and could therefore provide results with greater comparability. Salihu et al. (2015) use the two ratios mentioned in Table 2 next to two ETR's45_{in their study on the relationship between}

foreign investors' interest and corporate tax avoidance.

Book-tax gap and Permanent discretionary book-tax gap

The book-tax gap is the difference between pre-tax book income and taxable income. A difference between pre-tax book income and taxable income could be caused by BEPS. However, a book-tax gap can also be influenced by other factors, such as provisions in tax legislation. For example, loss relief widens the gap as it lowers taxable income but not pre-tax book income. A restriction on interest deductibility can narrow the gap. The permanent discretionary book-tax gap is used to get more insight in the causes of the book-tax gap.

The use of the permanent discretionary book-tax gap as a measure of tax aggressiveness is widespread (among others, for example (Kubick & Masli, 2015; Armstrong & Blouin, 2012)). Two variations of this measure are developed in Desai and Dharmapala (2006) and in Frank et al. (2009).The permanent discretionary tax gap is defined as the residual that is left after a regression of permanent book-tax differences. The regression controls for items that are known to cause permanent book-book-tax

43_{F.e. base erosion and/or profit shifting}

(26)

Master’s Thesis Economics of Taxation 22 differences but are unlikely associated with tax avoidance (Frank, Lynch, & Olhoft Rego, 2009). Thus, the discretionary permanent book- tax gap can also be defined as the permanent difference between pre-tax income and taxable income that is likely (but not necessarily) caused by tax avoidance.

Tax savings through foreign ownership

Egger et al. (2010) measure the scale of tax savings through foreign ownership econometrically. It goes beyond the scope of this thesis to describe their econometric method in detail. An important finding of this study is that multinationals earn significantly lower profits in high-tax countries than domestic firms and earn significantly higher profits in low tax countries than domestic firms. This implies that profit shifting is a likely source of the tax savings46_.

Positive earnings shock

Dharmapala and Riedel introduced a new approach to estimate the existence and magnitude of tax-motivated income shifting in 2013 (Dharmapala & Riedel, 2013). They investigated the distribution of a positive earnings shock at a parent firm across subsidiaries in foreign countries as a measure of tax avoidance. They find that a positive earnings shock at a parent firm is associated with a significant increase of pre-tax income of subsidiaries in low-tax countries, compared to the effects of the earnings shock on pre-tax income of subsidiaries located in high-tax countries.

(27)

Table 2, Measures of tax avoidance

Tax avoidance measures Computation Definition

1a. GAAP ETR Total tax expense per dollar of

pre-tax book income 1b. Long term GAAP ETR total tax expense past 3 years

pre-tax book income

Total tax expense in the last three years per dollar of pre-tax book

income

1c. Current ETR⁺ Current tax expense per dollar of

pre-tax book income

1d. Cash ETR Current taxes paid per dollar of

pre-tax book income

1e. Long term Cash ETR Taxes paid in the last three years per

dollar of pre-tax book income

1f. Foreign ETR Total foreign tax expense per dollar

of pre-tax foreign income 2a. Ratio total tax expense:

operating cash flow

Tax expense per dollar of operating cash flow

2b. Ratio tax paid: operating cash flow

Taxes paid per dollar of operating cash flow

3a. Book-tax gap Pre-tax book income – Taxable income

Permanent difference between pre-tax book income and pre-taxable income 3b. Discretionary

permanent book-tax gap (DTAX) ⁺⁺

Pre-tax book income - Estimated taxable income or

residual from regression

Residual from a regression of permanent differences on

non-discretionary determinants of permanent differences 4. Tax savings through

foreign ownership ⁺⁺⁺

Savings are econometrically calculated

Tax savings of foreign subsidiaries compared to domestic subsidiaries 5. Earnings shock at a

parent firm ⁺⁺⁺⁺

Earnings parent firm last year-earnings parent firm this year

A change in profits at the level of the parent firm

Notes

⁺ deferred tax expense excluded to capture tax avoidance only

⁺⁺ Developed in Frank (2009)(controls for permanent differences unrelated to tax aggressivess) and by Desai & Dharmapala (2006)

(28)

5.1.4 Indicators associated with tax avoidance

Table 3 displays the indicators of the existence of tax avoidance that have been identified in the scientific literature of the sample. Positive indicators are mentioned in the upper part of Table 3 and the negative indicators in the lower part. Positive and negative refers to the relationship between tax avoidance and the indicator. An increase47_{of the value or an improvement of a positive indicator}

increases the likelihood of tax avoidance whereas an increase in the value or an improvement of a negative indicator decreases the likelihood of tax avoidance.

Salihu et al. (2015) use the three indicators (3, 4 and 5) of transfer pricing aggressiveness found by Richardson et al. (2013a) as control variables in their study on the influence of foreign investors' interest on corporate tax avoidance. Thus, these three variables have a double significance; they can be used as indicators of aggressiveness and as control variables in studies on tax avoidance.

The found associations between indicators and tax avoidance are mostly positive of nature.

The indicators that are negatively associated with tax avoidance are mostly related to the way a company is structured and directed (15, 16, 17, and 18). Indicator 14 is of importance with regard to the Action Point 13 of the BEPS Action Plan as it supports the view that more transparency diminishes the likelihood of tax avoidance.

An important finding of this study is that some indicators can be actively and relatively easy influenced (f.e. 14 and 16) by companies to decrease or increase tax avoidance. For example, the decision to disclose profits per country is one a company could make voluntarily. However, as of 1 January 2016 filing a country-by-country report is obligatory for multinationals that meet the criteria48_{. The}

influenceable indicators could also be of interest for policy makers. Policies could be designed in such a way that tax avoidance becomes less likely. For example, encouraging the improvement of CSR-performance diminishes the likelihood of tax avoidance. The non-influenceable indicators49_{can be}

seen as risk indicators and should be closely monitored. Monitoring can be used for indirect identification of firms that are possibly tax-avoidant.

47_{Or the existence of an indicator, some indicators are on/off indicators} 48_{See paragraph 2.3.1}

(29)

Table 3, Indicators of tax avoidance

Positive indicators of tax avoidance Description

1. Tax haven intensity2 _{Number of tax havens used by a company}

2. Number of years share publicly traded3 Amount of years firm's stock has traded in public

markets

3. Firm size6 _{Natural logarithm of total assets}

4. Profitability6,7 Natural logarithm of pre-tax income (in (Salihu,

2015), return on assets) 5. Leverage6,7,8 _{Ratio of long term debt to total assets}

6. Intangible assets6,7 _{Natural logarithm of R&D expenditure}

7. Multinationality6 _{Relative amount of foreign subsidiaries}

8. Reporting an uncertain tax position9 Reported uncertainties about the tax position in

f.e. notes in financial statements 9. Tax expertise of directors9 _{Presence of board members with tax expertise}

10. Performance based remuneration of key personnel9

(a part of) Remuneration is dependent on (financial) performance

11. Relocation of headquarters11 _{A relocation of headquarters to another country}

12. Tax sheltering12 Predicted value from a tax shelter prediction

model

13. Tax aggressiveness of executives13 _{Personal tax aggressiveness of a firm executive}14

Negative indicators of tax avoidance Description

14. Disclosure of geographic earnings in

financial reports1 Separate reporting of profits per country

15. Board of director composition3 _{Proportion of outside members on the board}

16. CSR performance4 Combined score on the CSR disclosure index

(Lanis, 2015) 17. Board of director oversight

characteristics5

Presence of an effective risk management and internal control mechanisms

18. Board members with a tax related affiliation9

Presence of board members with ties to tax-related bodies (f.e. the IRS)

19. Meaningful change in control10 Change in juridical or economical ownership

after a transaction Notes:

1_{a firm is considered to be disclosing if it reports earnings of at least two foreign segments (Hope, 2013)} 2_{tests income shifting ability only}

3_{Lanis (2011)}

4_{A high score on CSR-performance is associated with a low level of tax aggressiveness (Lanis, 2015)} 5_{F.e. big4-auditor, highly independent internal audit commitee (Richardson, 2013b)}

6_{Richardson (2013a) finds a positive relationship between this determinant and TP aggressiveness} 7_{This indicator is used as a control variable in Salihu (2015)}

8_{long term debt/total assets} 9_{Taylor (2014)}

10_{Borek (2014), useful in distinguishing between economical and tax-driven profit shifting} 11_{Voget (2011)}

12_{Developed in Wilson (2009)} 13_{Chyz (2013)}

(30)

5.2 Which data sources have been used in the analysis of BEPS and the actions to address it?

This paragraph discusses the results of sub-research question 2. Sub-paragraph 5.2.1. provides a summary of the results depicted in Table 4. Sub-paragraph 2 discusses the public availability of the databases.

Knowledge about the source of the identified indicators and input for the identified methods is important for several reasons. First, knowing the origin of the data used in scientific studies is needed to be able to successfully reproduce these studies. Confirmation of results through reproduction strengthens the validity of scientific literature and gives insight in the applicability of effects in other scenarios. For example, Richardson et al. (2013a) find determinants of transfer pricing aggressiveness in Australian firms. Knowing the origin of the used data in their study makes it possible to check if these determinants have the same effects on transfer pricing aggressiveness in other countries. The study can be checked for country- specific50_{biases by reproducing it.}

Second, knowledge about the data sources also provides insights in the coverage of the sample studies. Some databases contain skewed data51_{and this should be kept in mind while interpreting study results.}

An overview of the data sources that are used in the scientific literature is provided in Table 4. This paragraph continues with a discussion of Table 4.

5.2.1 Results

The data sources that are used in the sample studies are depicted in Table 4 on the following page. The column “Article reference” in Table 4 displays the authors that used a specific data source. Fourteen different data sources are used in the sample studies. Most data sources provide financial data about companies. Databases 6, 7, 9 and 14 provide data on specific topics. Execucomp (6), a database of Standard and Poor's (S&P), provides information on compensation for executive personnel. The corporate governance database of UT Sydney (7) contains corporate governance information obtained from annual reports of the 500 largest Australian listed (ASX) firms52_{. The}

KLD-database (9) provides social ratings of over 3.000 public firms53_{. Zephyr (14) contains hourly updated}

information on mergers and acquisitions54_.

Datasets can be accessed via different channels. Bureau van Dijk is a company that manages several databases and provides access to these databases. If a database is managed by Bureau van Dijk it is

50_{F.e. cultural or geographical variables} 51_{F.e. Only data on developed countries}

52_{http://help.sirca.org.au/display/GOV/Corporate+Governance+Database+Review+and+Enhancements} 53_{http://www.library.hbs.edu/go/socrates.html}

(31)

Master’s Thesis Economics of Taxation 27 mentioned between parentheses in Table 4. Badertscher et al. (2013) use SEC-filings as a data source in their study, SEC-filings can be obtained from EDGAR55_{. Wharton Research Database Service (WRDS)}

facilitates the use of many datasets including the datasets of Bureau van Dijk, the KLD database, Execucomp and the Thomson Financial Insider Filing Database56_.

5.2.2 Public availability of data

Almost all sample studies use a publicly available database for data retrieval. Only Armstrong et al. (2012) use a database with confidential information provided by a large HRM consulting firm. The dataset contains information on the components of compensation for executive personnel. Kubick and Masli (2015) use ExecuComp to obtain data on compensation of executives. It is not clear why Armstrong et al. (2012) use confidential information and Kubick and Masli (2015) use public information. It is easier to reproduce a study that uses a public database but private information can be more detailed and therefore give better results.

Table 4, data sources

Data sources Article reference

1. Specific firm data provided by large

HRM consulting firm Armstrong

2. Compustat Armstrong, Badertscher, Hope, Kubick

3. SEC-filings (EDGAR) Badertscher

4. SEC Form 4 (Thomson Financial

Insider Filing Database) Chyz

5. AMADEUS (Bureau van Dijk) Dharmapala, Egger, de Simone

6. Execucomp Kubick

7. UT Sydney corporate governance

database Lanis' 11

8. Aspect Huntley database (Australia) Lanis' 11, Lanis' 12

9. KLD database Lanis' 15

10. Wharton Research Data Service

(WRDS) database Lanis' 15

11. Annual reports public listed firms

(ASX) Richardson' 13a, Richardson'13b, Taylor 12. Annual reports firms in FTSE Bursa

Malaysia Top 100 Index Salihu

13. Orbis (Bureau van Dijk) Voget

14. Zephyr (Bureau van Dijk) Voget

55_{Electronic Data Gathering, Analysis, and Retrieval System}

(32)

Master’s Thesis Economics of Taxation 28 Input that is needed to compute ETR’s is obtained from publicly available annual reports. Academics use annual reports of listed and/or private companies obtained via several channels in the sample studies. Annual reports are hand-collected from websites (Salihu et al., 2015; Richardson et al., 2013a) or collected via the use of databases (AMADEUS, Orbis and Compustat). The used databases are accessible for academics via their institutions. Annual reports of multinational enterprises are sometimes only available in a consolidated format. Computing an ETR from consolidated data can give misleading results due to a lack of detail. Enhanced public availability of unconsolidated annual reports will make it easier to compute ETR’s per country. The implementation of country-by-country reporting under the provisions of Action 13 of the BEPS Action Plan could be of great importance for BEPS research as it facilitates computing ETR’s per country.

Input for monitoring tax avoidance through indicators can also be obtained from annual reports. Information on for example transfer pricing determinants (Richardson, Taylor, & Lanis, 2013a) leverage, profitability and R&D expenditures is publicly available in the annual reports of most firms. However, input for most indicators is not available in annual reports. For example, Lanis and Richardson (2015) find that CSR-performance is negatively associated with tax avoidance57_{. In order to}

measure CSR-performance they use the KLD-database58_{. The database is accessible through WRDS.}

The data sources displayed in Table 4 can be used by other scientists to perform studies as well. Aspect Huntley and the Corporate Governance Database of the University of Technology in Sydney are country-specific for Australia. Alternatives with the same or comparable information (Orbis, KLD-database) for companies located outside Australia are available.

57_{See Table 3 for more indicators}

A Study on Action 11 of the OECD/G20 BEPS Action Plan

A Study on Action 11 of

the OECD/G20 BEPS

Action Plan

MEASURING AND MONITORING BASE EROSION AND

PROFIT SHIFTING; A COMPARISON OF THE FINAL

REPORT ON ACTION 11 OF THE OECD/G20 BEPS

ACTION PLAN AND ECONOMIC SCIENTIFIC

LITERATURE

ANNELIES EILANDER

WINTER 2016

A Study on Action 11 of the OECD/G20 BEPS Action Plan

Abstract

Table of Contents

1 Introduction

2 Background information and Glossary

3 State of the art

5 Results