Exchanging tax information in the EU: solid foundation, cracks in the implementation

Executive summary

I

It is the objective of the European Union to ensure fair and effective taxation throughout the Single Market, with all taxes collected where they are due. A coordinated EU tax policy increases certainty for taxpayers and enhances investment and competition.

II

Since 2011, the Council has agreed a series of legislative initiatives enabling Member States to share tax information in a variety of ways. Rules on the exchange of information and other forms of administrative cooperation allow Member States to gain access to the information they need to ensure that taxes are collected.

III

The European Commission is responsible for making legislative proposals on the system for exchange of tax information and providing guidance to Member States. If EU legislation is not properly implemented in a Member State, the Commission can bring infringement proceedings against the Member State. Its competences also include monitoring how the system of information exchange is implemented and how effective it is. Member States are responsible for sharing relevant tax information and using it to ensure compliance.

IV

We assessed in this audit, key aspects which determine the effectiveness of the system for exchange of tax information in the EU. We carried out this audit because the system for the exchange of tax information, which is in place since 2011 has not been subject to an audit. The audit covered the period between 2014 and 2019. We assessed the legislative framework that the Commission proposed and put in place, and examined how well it was monitoring the implementation and performance of the system of information exchange. We visited five Member States to assess how they are using the exchanged information and how they are measuring the effectiveness of the system.

V

Our overall conclusion is that the system for exchange of tax information has been well established, but more needs to be done in terms of monitoring, ensuring data quality and using the information received.

VI

Our conclusion on the Commission’s work is that it has established a suitable framework for the exchange of tax information, but it is not proactively providing guidance and is insufficiently measuring the outcomes and impact of the use of the information exchanged. The main conclusion from our visits to Member States is that the information they exchange is of limited quality and is not widely used, and that they do little to monitor the system’s effectiveness.

VII

Although the Commission has provided a clear and transparent legislative framework, some forms of income may still escape taxation in the relevant Member State. The Commission’s monitoring of the system’s implementation in Member States does not go far enough and does not include on the spot visits in the Member States or the effectiveness of sanctions schemes. The Commission provided Member States with useful guidelines in the early stages of the information exchange, but more recently, it has not issued further guidelines aimed at helping Member States to analyse the information received and use it to raise the relevant taxes.

VIII

The information collected by Member States lacks in quality, completeness and accuracy. Most of the Member States we visited do not audit reporting entities to ensure the quality and completeness of data before forwarding it to other Member States. In addition, only a handful of Member States report all the required categories of information, leading to some income escaping taxation. Although Member States identify the relevant taxpayers, information exchanged automatically is under-used.

IX

There is no common EU framework for monitoring the system’s performance and achievements, and the individual assessments carried out by some Member States are short on clarity and transparency.

X

We make several recommendations that are designed to improve the effectiveness of the information exchange system. The recommendations below are addressed to the five Member States we visited (Cyprus, Italy, Netherlands, Poland and Spain), but may also apply to others.

1. The Commission should:
   • take direct and effective actions to address the lack of quality of data sent by Member States;
   • make legislative proposals to ensure that all relevant income information is exchanged;
   • expand its monitoring activities and add to its guidance for Member States.
2. Member States should:
   • ensure that the information they exchange is complete and of sufficient quality;
   • introduce systematic procedures for the risk analysis of incoming information and use them as extensively as possible.
3. The Commission should establish, together with Member States, a reliable common framework for measuring the benefits of the system for exchange of tax information.

Introduction

01

These days, when money can be sent around the world at the click of a button, some taxpayers are tempted to keep investments hidden in offshore accounts. Until fairly recently, many countries and other territories openly operated as tax havens, tolerating secret bank accounts and shell companies and refusing to share information with other governments¹.

02

In the Single Market companies generating profits should pay tax where the money is earned. However, companies use aggressive tax planning to take advantage of differences in the tax rules of different Member States and pay as little tax as possible on their profits.

03

Estimates of the scale of EU tax revenues lost to corporate tax avoidance alone range from €50 to €70 billion per year². The figure rises to almost €190 billion (around 1.7 % of EU GDP at the time of the study) if other factors, such as special tax arrangements and tax collection inefficiencies, are included.

04

To meet demands for social justice and economic growth, the Commission is working with the Member States to make tax systems more transparent, more accountable and more effective across the board. This goes hand in hand with the notion of tax fairness: taxpayers operating across multiple Member States – both multinational companies (MNEs) and individuals – should not enjoy a tax advantage because of limited communication between tax authorities. At the same time, if taxes are paid correctly, and to the correct country, right from the start, later adjustments become unnecessary and tax certainty across the EU improves.

EU action

05

In 1998, the Economic and Financial Affairs Council adopted an EU code of conduct for business taxation³. In acknowledging “the positive effects of fair competition”, the Council called for Member States to cooperate fully in preventing tax avoidance⁴ and evasion⁵, notably through the exchange of tax information⁶.

06

While collecting taxes is a Member State competence, tax fraud and avoidance are tackled also at EU level. The Member States have agreed that it is essential to improve and develop in-depth administrative cooperation among their tax authorities, using the cross-border framework and instruments proposed by the EU.

07

To guarantee the functioning of the internal market and protect the EU’s financial interests, the Commission encourages and facilitates Member State cooperation through its Directorate-General for Taxation and Customs Union (DG TAXUD). In line with its mission statement of developing and monitoring the implementation of tax policy across the EU for the benefit of citizens⁷, businesses and the Member States, and assisting Member States to combat fraud and tax evasion, DG TAXUD:

• drafts legislative proposals for setting up and improving the exchange of tax information between Member States and monitors the implementation of the legislation in the Member States;
• provides mechanisms, systems and electronic interfaces to coordinate the exchange of tax information;
• oversees a central directory for the exchange of information on advance cross-border tax rulings and pricing arrangements;
• manages the Fiscalis 2020 programme for the exchange of best practice and knowledge among Member States and the Commission (see Annex VI).

08

The national governments of the EU Member States are broadly free to design their own tax laws and systems and decision making in this policy sensitive area at an EU level requires the agreement of all Member States. However, all national legislation must respect certain fundamental principles, such as non-discrimination and freedom of movement in the internal market. The tax laws of one country should not allow people to escape taxation in another.

09

Member States in their turn must:

• operate the taxation systems;
• collect and report the required tax related information;
• share best practices; and
• ensure a fair tax competition in the internal market.

Directive on Administrative Cooperation

10

The European Communities first adopted a directive on administrative cooperation in the field of taxation in 1977⁸. It was replaced by the current EU Directive on Administrative Cooperation (DAC)⁹ (see Annex III), which entered into application on 1 January 2013 (see Box 1).

11

The Directive pursues three specific objectives:

• to improve Member States’ ability to prevent cross-border tax fraud, evasion and avoidance; and
• to reduce the scope for incentives and advantages leading to harmful tax competition (see Box 2), including counteracting tax avoidance and aggressive tax planning, by transparency measures related to tax rulings/advance pricing arrangements, and country-by country reports of multinational enterprises; and
• to promote spontaneous tax compliance by making it easier to detect cross-border income and assets.

12

The DAC was a response to the increasing mobility of taxpayers and capital and the internationalisation of financial instruments. These developments had brought an explosion in the number of cross-border transactions, making it difficult for Member States to assess taxes due properly, and encouraging tax fraud. Against that background, there was an urgent need for enhanced tax cooperation.

13

The DAC establishes three main forms of information exchange in the EU. The development of these over time is shown in Figure 1:

• automatic exchange of information (AEOI), which refers to the mandatory exchange of predefined tax and financial accounts data, using predefined formats and at predetermined times;
• exchange of information on request (EOIR), in the form of requests for information regarding one or more taxpayers;
• spontaneous exchange of information (SEOI) – the unsolicited transfer of any information which one country deems to be of interest to another.

The DAC also provides for other forms of administrative cooperation. Of these, this report focuses mainly on simultaneous controls (SMC) of specific taxpayers, in two or more Member States, with a view to exchanging the information thus obtained.

Audit scope and approach

14

In this audit, we assessed the arrangements for the exchange of tax information in the EU, which determine effectiveness and the other tools of administrative cooperation. We focused on the implementation of different amendments of the DAC (DAC1 to DAC5) during the period from 2014 to 2019 (DAC6 did not enter into application until June 2020).

15

We carried out this audit because the system for the exchange of tax information, which is in place since 2011, has not been subject to an audit. In this audit, we address system’s issues and make recommendations for improvements, concerning the design, implementation, and monitoring of the system and usage of the information by Member States, in particular:

• the degree to which the exchange of information contributes to the European Commission’s strategy for fairer taxation;
• the effectiveness of practical arrangements and guidelines provided by the Commission to Member States for the implementation of the DAC Directive;
• the Commission’s monitoring of the system;
• the sufficiency of resources and guidance to properly implement the system;
• the effective use of information received;
• the effectiveness of cooperation arrangements between Member States and between Member States and non-EU countries; and
• the effectiveness of the regulatory and control measures and the allocation of resources.

16

We audited the activity of the Commission and five Member States (Cyprus, Italy, Netherlands, Poland and Spain) which we selected based on quantitative and qualitative risk criteria linked to the exchange of information (see Annex II).

Commission

17

Regarding the Commission, we reviewed the rules on the exchange of tax information under DAC1 to DAC5 and examined how DG TAXUD monitored their implementation. We looked in particular at:

• the relevant legislation, whether the Commission verified the way Member States applied the EU rules through national law, and whether it took the necessary action to address any delays in implementation;
• the existence and quality of guidelines and information relating to the implementation of the DAC, and how the Commission shared that information with Member States;
• the performance of programmes and measures financed by the EU budget to help Member States exchange good practices and ideas for improving the system;
• whether the Commission had put in place a common EU framework for performance monitoring of the system to ensure that it was providing the intended results;

18

The audit approach and methodology we used for performing this audit at the Commission level is presented in Annex I.

Member States

19

The DAC requires Member States to cooperate on the exchange of all relevant tax information. We assessed how Member States:

• ensured that the data they exchanged was accurate, complete and sent within the deadlines;
• used the information they received, and how this could be improved;
• used the other forms of administrative cooperation provided for by the DAC;
• measured the effectiveness of the system and used the results of their measurements to address risks and allocate resources.

20

In the visited Member States, we used risk-based sampling to select automatic and non-automatic information exchanges (AEOI, EOIR and SEOI) and simultaneous controls for our audit. The audit approach we used for auditing Member States is presented in Annex II.

21

Both the information exchanged between Member States and the underlying systems are highly confidential. Thus, this report does not give specific details on Member States unless they are already publicly available.

Observations

While sometimes ahead of international developments, the EU legislative framework presents gaps

22

The legislation in the field of exchange of tax information should ensure that all incomes subject to its provisions are taxed in the right Member State and in full. The Commission should monitor the legislation transposition and implementation and should provide guidelines for its uniform application across the EU. We assessed the legislative process that the Commission put in place for the initial DAC proposal and subsequent amendments. We also looked at how the Commission monitors the DACs’ implementation in Member States, including its sharing of guidelines to that end.

The legislative framework is solid although not comprehensive

The EU is sometimes ahead of international developments

23

In 2013, the Organisation for Economic Co-operation and Development (OECD) adopted an action plan to address the tax avoidance strategy of profit-shifting¹⁰. It reported in 2015 on various aspects of the action plan, such as spontaneous, exchanges of advance cross-border rulings and advanced pricing arrangements¹¹, and country-by-country reporting¹².

24

Many countries have agreed to lift bank secrecy for tax purposes, share tax information with other governments and remove the secrecy surrounding the owners, banking arrangements and financial transactions of legal entities. The starting point was the global EOIR standard developed by the OECD, which first provided for information to be exchanged internationally between two countries at the request of a tax authority, with a new requirement added to obtain and exchange information even if held by a bank.

25

Technological advances have also led to the widespread automatic exchange of information, complementing EOIR. After the first experience provided by the EU Savings Directive, in 2014, the G20 international economic cooperation forum endorsed a proposal for a global AEOI standard, which the OECD published in the same year as the new Common Reporting Standard (CRS)¹³. This specifies what information is to be exchanged and sets out operational rules. As of the end of 2019, more than 100 countries had committed to automatic exchanges of financial account information under the CRS¹⁴.

26

The EU is in the forefront of international developments in this area, at times even ahead of the OECD (see Table 1). It has been quick to implement common practices and has created a harmonised framework which takes account of international benchmarks¹⁵. The DAC is consistent with the OECD provisions¹⁶.

The legislative framework is transparent and logical but incomplete

27

We examined the legislative framework for the system of exchange of tax information and we have identified gaps.

28

The overall legislative framework is based on a clear demarcation of tasks and responsibilities and has well-defined objectives (see Figure 2 and Annex V). The EU rules are enforceable through the Commission and the Court of Justice of the EU.

29

In its 2019 evaluation of DAC¹⁹, the Commission identified no major issues with the legislation. However, the evaluation did not address some legislative gaps described in the following paragraphs.

30

Cryptocurrencies are excluded from the scope of information exchange²⁰. If a taxpayer holds money in electronic cryptocurrencies, the platform or other electronic provider supplying portfolio services for such customers are not obliged to declare any such amounts or gains acquired to the tax authorities. Therefore, money held in such electronic instruments remain largely untaxed.

31

Non-custodial dividend income (e.g. dividends paid through a current, non-custodial bank account, by a company resident in one Member State to its shareholder, resident in another) is not a separate AEOI income category under DAC1. Thus, if the Member State where the company is resident does not exchange such information with the taxpayer’s residence Member State, such amounts might remain untaxed.

32

It is not mandatory for Member States to report all DAC1 income categories, but only the data that is readily available. As a consequence, there are large differences between the number of categories of information reported by each Member State (see paragraph 48), resulting in incomes not being taxed in the Member State where the taxpayer is resident. In a July 2020 legislative proposal²¹, the Commission proposed an amendment to increase harmonisation, where Member States would be obliged to exchange all information that is available, but on at least two categories of income for taxable periods until 2024 and on at least four categories for taxable periods as of 2024.

33

Advance cross-border tax rulings are excluded from the exchange of information if they were issued for natural persons (see paragraph 73). A high net-worth individual (HNWI) obtaining such a ruling from a Member State with favourable tax rates, will avoid paying a fair amount in tax in their Member State of residence.

The Commission monitors the legislative framework

Commission sufficiently monitors transposition of the DAC into national legislation

34

In monitoring the Member States’ transposition of the DAC, the Commission considers all amendments of the Directive, as well as the implementing regulations and other subsequent legislative acts. Member States have to transpose the EU Directives into national law. When delays in the transposition of relevant EU legislation in national law occurs, the Commission opens infringement procedures against individual Member States for such delays. To date there have been 62 such procedures, with two infringement procedures still ongoing as of July 2020.

35

In 2016, DAC5 amendment came into force²² to ensure that tax authorities can access the anti-money laundering information held by other Member States under the fourth Anti-Money Laundering Directive²³, equipping them better to meet their obligations under the DAC with regard to tax evasion and fraud. Most relevant in this regard is access to DAC2 information on financial accounts, especially when the holder of a reportable account is an intermediary structure.

36

While all the five Member States we visited transposed the DAC5 legislation in their national laws, two had not yet implemented the procedures to grant tax authorities legal access to anti-money laundering information, the deadline for which was 1 January 2018. The Commission opened three infringement procedures, which were closed at the date of our audit.

The scope of monitoring the implementation of legislation is not broad enough

37

The Commission is performing a high-level monitoring of the implementation of the legislation. In addition, Member States are required to report to the Commission, in an annual questionnaire, how well they assess the DAC to be working (including statistics on information exchanges and other operational aspects). The Commission is monitoring the exchanges between Member States based on the annual questionnaire and the annual bilateral feedback between Member States. But its monitoring does not include:

• taking direct and effective actions to address the lack of quality of the data sent between Member States, as identified in its 2018 Evaluation²⁴ of the DAC;
• information visits to Member States, both to assess progress and to identify good practices. Such visits could also serve to identify any issues that could be dealt with by improving the legislation or by bringing them to the attention of Fiscalis 2020 (see Annex VI) or administrative cooperation expert groups;
• the assessment of risk-analysis tools developed by Member States for the analysis of the information they receive.

38

According with the DAC provisions²⁵, Member States should implement effective, proportionate and dissuasive penalties for reporting entities. At present, the Commission does not assess the size or the deterrent effect of the penalties that each Member State establishes in law. Nor has it offered any benchmarks for comparison or guidance in this respect.

Opportunities for sharing best practices are not fully exploited and guidance is lacking

39

Article 15 of the DAC requires the Commission and Member States to evaluate and share their experience of administrative cooperation, and expressly provides for the possibility of Member States’ using their experience and best practice to contribute to guidelines. To reach its desired objective of fair taxation in the EU²⁶, the Commission places various fora at the disposal of Member States for the exchange of the good practices identified amongst Member State tax administrations.

40

The Commission has made various tools available for Member States to develop and exchange information and expertise. The main such programme is Fiscalis 2020²⁷, for which the main role is to provide for the IT support enabling the exchanges of information and finances the exchange of best practices and experience through workshops, working visits and project groups. Its role and structure is further detailed in Annex VI.

41

We examined the actions funded by Fiscalis 2020 to improve the exchange of tax information in the EU. These included, between 2015 and 2019, the participation of Member States in nine workshops and five working visits in the field of the DAC. From 2014 to 2018, the programme also funded 660 simultaneous controls initiated in the field of direct and indirect taxation (see paragraph 92).

42

Member States have generally provided positive feedback on the work of the project groups, and on the technical support, the availability and the responsiveness of EU staff. They have also been appreciative of the Commission’s guidelines for the implementation of DAC1 and DAC2.

43

However, due to the complexity and the constant evolution of the DAC, it has become increasingly difficult for Member States to address all discussion points during their meetings, which have not increased in length to match the expansion of the DAC. In the same time, the Commission has not produced further guidelines on the use of information (e.g. DAC3 and DAC4 exchanges and other tools provided by the DAC). In 2019, Member States set up a new Fiscalis project group on the use of advanced analytics to measure data quality within a common framework, but this initiative has yet to produce any results.

44

The tax authorities in the Member States we visited expressed the wish for more exchanges of good practices (through project groups, workshops and working visits) and guidelines, especially in the field of data analysis and use.

Member States only make limited use of the information exchanged automatically

45

To ensure that the system for exchange of information functions correctly, the data that Member States exchange must be accurate, complete and sent when it can be most useful. In the five Member States we visited, therefore, we assessed how the automatic sending of data was organised and how the tax authorities used the information they received.

There are weaknesses related to the timeliness, the accuracy and the completeness of AEOI

DAC1 data is not always complete and information is sometimes shared late

46

For DAC1²⁸, the information is collected by the tax authorities from the national databases of each visited Member State (see Figure 3), but not all income categories of information are collected. The information is sourced from local databases (which contain names, addresses, birth dates, tax identification numbers (TINs), income figures and details of any taxes withheld). Before sending, it is validated automatically in accordance with the technical guidelines for the DAC1 reporting process. In addition to the automatic validation, only one of the five Member States we visited carried out checks of data quality (its completeness, accuracy, and timeliness); these took only the form of manual checks on a limited data sample and they were not implemented as a systematic process.

47

One data completeness issue for DAC1 is that Member States rarely link the information they send to a TIN issued by the taxpayer’s country of residence. In the three-year period 2015-2017, only 2 % of taxpayers concerned by AEOI were associated with a TIN issued by the receiving country²⁹. Member States use other data (name, address, birth date, etc.) to perform the matching of the data towards their own resident taxpayers (see paragraph 59). Nevertheless, the TIN is a unique identification for each taxpayer and providing it should increase Member States’ capability of identifying the relevant taxpayers and correctly asses the related taxes. The issue of making mandatory the TIN reporting for DAC1 data was addressed by the Commission in the proposal for the amendment of the DAC³⁰.

48

Data completeness is affected by the fact that, under the current rules for AEOI, it is not mandatory for Member States to collect information about all DAC1 income categories, but only information “that is available”³¹. The tax authorities generally opposed to report all categories of information. Even the few that agreed to the proposal in theory pointed out that putting it into practice would mean a lot of additional work. As a result, only a handful of Member States collect and report all five categories of DAC1 information (see Figure 4).

49

The omission of a data category is sometimes explained by the difficulty of accessing information e.g. immovable property revenue is commonly recorded in separate local registers where the tax authorities cannot easily track it down. Omission may also be explained by reference to a particular tax regime (e.g. Member States that do not tax life insurance products will not hold the corresponding information readily available in their tax registers). It may also be that a Member State does not keep separate or distinct records, as can occur in the case of income from employment and directors’ fees.

50

Member States are required to share DAC1 information within six months of the end of the tax year during which the information became available. Although the legal deadlines are generally respected by Member States, since DAC1 entered into force, information has been exchanged, on average, twelve months after the end of the relevant tax year³³.

DAC2 information exchange functions generally on time, but still lacks in data quality and completeness

51

For DAC2, the process for gathering the data is slightly different from DAC1 (see Figure 5). Reporting institutions use dedicated web portals to submit standard reports³⁴.

52

In only one of the Member States we visited did the tax authorities perform quality checks on the data submitted by financial institutions. However, we detected issues that could affect DAC2 data completeness in each of those Member States. Three of them had no specific procedures for auditing the financial institutions regarding the quality and completeness of the data sent. Four kept no register of financial institutions concerned by the DAC2 reporting requirement, which would help them to verify that all reports have been submitted.

53

All five Member States operated a system of penalties in respect of the DAC2 reporting obligations, but none imposed any penalty so far. Enforcing penalties means having clear data verification procedures and regularly auditing the reporting institutions. Moreover, there is a risk that the penalties are not large enough to have a deterrent effect.

54

We also found several issues with DAC2 data quality, most notably with regard to the non-reporting of foreign TINs. Financial institutions are legally obliged to record and report account holders’ tax numbers, even the ones issued by other Member States. However, of the data exchanged by Member States, only 70 % of accounts whose holders are natural persons were linked to a TIN and only 73 % of accounts whose holders are legal persons³⁵ were linked to a company registration number.

55

Another data quality issue is the duplication of account records where there are multiple beneficial owners. Moreover, the current reporting format does not allow for the identification of multiple ownership accounts. As a result, tax authorities allocate the full amount to several owners resulting in the duplication of account records.

56

Article 14(2) of the DAC requires Member States to provide bilateral annual feedback on the AEOI information they receive. In practice, however, the only annual feedback given relates to the matching rates obtained between foreign and domestic data (see paragraph 59). There is very little feedback on data quality, completeness and timeliness, and none on other quantitative or qualitative aspects.

Member States receive huge volumes of information, with information generally underused

The effectiveness of matching varies significantly between Member States with unmatched information not used

57

Each year, millions of individual data items are exchanged between Member States. In order to use the data for their taxation purposes, Member States need to match it against their taxpayers’ data and identify the relevant ones. The focus of the matching systems is to identify as many taxpayers as possible, in order to ensure the highest degree of tax compliance.

58

We checked how the five Member States use their IT systems and matching processes to find the relevant taxpayers and collect the required taxes.

59

They used a range of automatic or semi-automatic IT programs to match incoming DAC1 and DAC2 information with tax residents. If data received is not matched, it cannot be used to assess the relevant taxes. The effectiveness of these systems depends on their inherent flexibility and interconnectivity with local tax databases, but also on the quality of the information received. Two Member States also matched data manually.

60

Automatic matching usually makes use of search keys that run algorithms based on combinations of fields (name, address, and date of birth are most common for DAC1, while DAC2 focuses more on TIN matching). The systems in three of the Member States we visited also allowed approximate matches (with approximations made for the received data, such as similar names or addresses). Semi-automatic matching combines manual matching with a simpler range of search keys (e.g. address only) or comparisons with other external databases.

61

Matching rates vary considerably, both from one country to another and by category of information. Failure to obtain a match may be due to ineffective matching processes, incorrect or incomplete incoming data, or to missing taxpayer records in the receiving Member State. When providing feedback (see paragraph 56), the visited Member States do not at present address the reasons for low matching rates or describe any specific issues.

62

In the five Member States we visited, the overall matching rates ranged from 68 % to 99 % for DAC1, and from 70 % to 95 % for DAC2 (one Member State had not yet analysed its 2016 DAC2 data). The highest rates correlated with a significant volume of manual matching, but two Member States had obtained very good results by using only automatic or semi-automatic matching.

63

These matching rates show that large quantities of information are not used, since they are not matched against relevant taxpayers, with none of the Member States we visited currently making further checks of unmatched data. Not using the unmatched data leads to a taxation shortfall.

DAC1 and DAC2 information is not rigorously exploited

64

We assessed how the five Member States we visited used incoming DAC1 and DAC2 information. One Member State highlighted certain data fields in tax returns, but without pre-filling amounts. Another notified taxpayers that information was available and they should complete the corresponding fields. Both Member States told us that their strategies had a deterrent effect on the non-declaration of revenue and increased taxpayer compliance. The remaining three did not provide pre-filled tax declarations.

65

A key prerequisite for efficient data use is to conduct a risk analysis of the received data. In 2015-2016, a Fiscalis project group drew up a guideline for Member States on data matching and analysis. The main purpose of the guideline was to support Member States in establishing risk management strategies or bringing existing strategies into line with the new requirements under mandatory AEOI. The guideline identifies common risks and proposes key risk indicators and concrete approaches to the selection of high-risk taxpayers.

66

Only two of the five Member States carried out a structured risk analysis of incoming data, and a third used only non-standardised procedures at regional/local level. The remaining two Member States had not yet carried out a risk analysis of the DAC1 and DAC2 data they had received for 2015-2018.

67

In one of the Member States we visited, an extensive campaign targeting high-risk taxpayers selected through a central risk analysis had led to high additional tax volumes being assessed. In another Member State, voluntary disclosure campaigns relating to DAC1 and DAC2 foreign income for 2015 and 2017, with no late payment penalties, resulted in a large number of voluntary disclosures and significant additional tax revenue. A separate campaign in the same Member State, involving letters to selected taxpayers about DAC2 information that had been received for 2015 and 2017, was less successful because of the low reply rate. A third Member State had used DAC information sporadically in successive campaigns, but with inconsistent results.

Information is generally under-used

68

In the Member States we visited, we sampled a total of 150 items of information received during 2016 – 75 each under DAC1 and DAC2 – to determine how the tax authorities had used the information. We found that, while the information in 105 audited samples was matched and uploaded to a tax authority’s databases, 60 were disregarded (e. g. without risk analysis/justification) and only 45 resulted in further tax related actions.

69

For DAC1 (see Figure 6):

• 51 items were matched to taxpayers in national databases; 24 were neither matched nor analysed further;
• of the 51 matched items, 40 were uploaded to central databases, but no information was available about how they were then used; 11 items were forwarded to regional/local tax offices for follow-up;
• there was evidence that 13 items were used to review tax returns and calculate additional taxes, or to perform tax audits; local tax offices gave feedback in only seven cases.

70

For DAC2 (Figure 7):

• 55 items were matched to taxpayers in national databases; 20 were neither matched nor analysed further;
• of the 55 matched items, 39 were uploaded to central databases, but no information was available about how they were then used; 11 items were forwarded to regional/local tax offices for follow-up; five items were not used at all;
• there was evidence that 32 items were used to pre-fill tax returns, assess taxpayers’ liabilities and calculate additional taxes, or to perform tax audits; local tax offices gave feedback in only six cases.

71

The paucity of feedback from central/local offices on the use of incoming information was an issue in all of the Member States we visited. Feedback from local offices is essential to measuring performance of the use of the information and its benefits.

Information in the EU’s DAC3 directory is mostly complete but little used by Member States

Exchanges of information have increased, but some information is still not reported

72

DAC3³⁶ expanded the scope of mandatory AEOI to advance cross-border tax rulings and advance pricing arrangements. This information is exchanged through a central EU directory that was set up and is maintained by the Commission. Member States must upload all DAC3 information to the directory, where it can be accessed by all Member States, but the Commission has only limited access, according to the legislation in force.

73

The DAC³⁷ explicitly excludes advance cross-border rulings from AEOI if they exclusively concern and involve the tax affairs of natural persons. The impact of this provision is to exempt rulings issued for HNWI from the mandatory and automatic exchange of information.

74

The European Parliament expressed concern about this taxation loophole in a 2019 resolution³⁸. It raised the issue that HNWI and ultra HNWI shift their earnings through multiple tax jurisdictions, by obtaining non-reportable cross border rulings and avoiding paying their fair share of tax.

75

Prior to 2016, Member States exchanged virtually no information about advance cross-border rulings and pricing arrangements. The number of exchanges peaked in 2017, soon after they became mandatory, with advance rulings and arrangements issued before 2017 disclosed in the first reporting data set (see Figure 8).

DAC3 data is usually very comprehensive

76

We examined Member State uploads of the required information to the directory and found no major issues with the completeness of the database. However, data quality was sometimes deficient:

• Member States often did not name individual taxpayers, which complicates data matching in the target country;
• According to the visited Member States and to the Commission’s evaluation³⁹ of the DAC, the summary of uploaded rulings sometimes lacked sufficient detail for a proper understanding of the underlying information; it was difficult for Member States to know when to request further information and, if they did so, to demonstrate that it was needed for purposes of tax assessment.

Minimal use of DAC3 information

77

None of the Member States we visited systematically carried out a risk analysis of information in the central EU directory. However, one Member State had analysed the DAC3 information that was uploaded in 2017 and selected some for further investigation. Although the Member State was generally able to match the information to taxpayers, it had no procedure for analysing and using it in its taxation follow-up.

78

In each Member State we visited, we took a sample of 10 data items related to years 2013 – 2018, to check how Member States used information from the directory (see Figure 9). Of the 50 items sampled, only 4 % was used:

• 36 were matched to taxpayers, and 14 were either not matched or not used in any way;
• of the 36 matched items, 26 were uploaded to central national databases and 10 were sent to local tax offices;
• only two items were used for taxation follow-up, and a local tax office sent feedback in just one case.

Minimal use of information from DAC4 reports

79

DAC4⁴⁰ introduced country-by-country reporting (Figure 10), a requirement for large multinational groups, as of the 2016 tax year, to publish key information including:

• key details of revenues and assets;
• where they make their profits;
• where (in and outside the EU) they pay their taxes;
• employee numbers, etc.

Member States do not ensure that DAC4 information is complete

80

Member States are required to collect information from DAC4 companies on an annual basis. The Member States we visited had IT systems in place to do this and had issued general reporting guidelines to the companies concerned.

81

All five Member States relied on companies’ goodwill to comply with their reporting obligations. The tax authorities did insufficient checks whether all the entities that should have reported in a given year had actually done so. Moreover, four of the five Member States limited their checks to verifications against technical criteria provided by the OECD and the Commission. Only one ran its own additional risk analysis on the data it was exchanging with other Member States.

DAC4 information is greatly under-used

82

The Member States we visited made very little use of incoming DAC4 information. Only one of them had established a rigorous system of risk analysis, which represents very good practice in the field. Following the risk analysis, the information was passed to regional tax teams, which used it in their taxation procedures. Two of the other Member States carried out data matching and had plans for a risk analysis mechanism, while the last two had not used the information they received.

83

In each Member State we visited, we took a sample of 10 country-by country reports related to fiscal years 2016 – 2017, to check how the information was used (see Figure 11). Of the 50 reports sampled, only 4 % were used:

• 20 had been matched to taxpayers, and 30 had not been used at all;
• the 20 matched reports had been uploaded to central databases; 10 had been subjected to risk analysis, revealing no major risks;
• only two reports had been used for taxation follow-up, and no feedback had been received so far on their use.

Exchanges on request, spontaneous exchanges and simultaneous controls work well

EOIR and SEOI work reasonably well, despite delayed replies and poor feedback

84

All Member States we visited had well-functioning arrangements for the EOIR⁴¹ and SEOI⁴². Across the EU, EOIR and SEOI processes have now been in place for a number of years and are used consistently, with a steady year-on-year increase in volumes for EOIR and an exponential one for SEOI (see Figure 12 and Figure 13).

85

In the visited Member States, most information requests are initiated by local or regional tax offices and sent for review to the national tax authorities. Replies are also first verified centrally, and the sending Member State could then be asked for further details. On receiving an information request from abroad, the central tax authorities reply directly, if information is available centrally. Regional/local tax offices are called in to deal with more specific or complex requests. Spontaneous exchanges are usually triggered when a regional tax office identifies information that is relevant to the tax situation in another Member State.

86

According to the Commission’s statistics for 2014-2018, on average 59 % of replies to information requests were sent within the six-month deadline⁴³ (see Figure 14). Our audit in the five Member States revealed that delays generally occurred when requests were complex or insufficiently clear.

Member States deal systematically with EOIR and SEOI information

87

The five Member States we visited generally lodged and tracked all requests and replies using dedicated IT. With minor variations, the systems we viewed recorded deadlines, feedback, tax office details. Deadlines were monitored and, where possible, tax offices made an effort to reply within the time limits.

88

In these five Member States, we reviewed 50 replies received from other Member States in response to an EOIR request. In most cases, the incoming information was relevant and timely enough to be useful to the requesting tax offices, which used it variously to confirm missing taxpayer information, for tax audits or assessments, or as evidence in legal proceedings (see Figure 15).

89

We also examined 50 items of SEOI information (see Figure 16). In all cases, we conclude that the information was reviewed for its usefulness and followed up by the receiving Member State.

Feedback on information quality is not sufficient given the volume of EOIR and SEOI

90

Feedback is needed in particular to assess the effectiveness of the DAC. It is also a way of incentivising and showing recognition for the tax officials who put in so much work collecting information. In the Member States we visited, there was no system of feedback to demonstrate how and where the information is most useful. Feedback was usually only provided by regional tax offices when requested by the sending Member State and it was not always given.

91

Compared with the total volume of information exchanged (EOIR and SEOI), Member States rarely seek any feedback. Figure 17 and Figure 18 highlight the growing divergence between the number of feedback requests and feedback replies. For the period as a whole, Member States sending information asked for feedback in only 2.2 % of all non-AEOI exchanges.

Simultaneous controls: an effective tool

92

Simultaneous controls coordinated by two or more Member States on taxpayers “of common or complementary interest”⁴⁴ can be more effective than controls by only one Member State.

93

Between 2014 and 2018, Member States initiated 660 simultaneous controls in relation to audits in the areas of direct and indirect taxation or direct taxation combined with VAT.

94

We examined five simultaneous controls initiated by each Member State we visited, and five controls in which it had otherwise participated (in both cases unless fewer were available). Generally, the controls were very effective in terms of:

• supplementary tax assessments;
• the exchange of good practices in the assessment of business structures;
• the early detection of fraud schemes before they could be expanded to multiple Member States.

95

Overall, simultaneous controls proved an effective tool. Even Member States that had participated in very few controls rated them highly as a tool for exchanging information, experience and good practices.

The information needed for measuring the performance of the exchange is largely unavailable

96

Tax authorities need to have clear estimates of the revenue losses due to tax avoidance and evasion and take appropriate corrective measures. We assessed how the EU and Member States gather information about the volume of unpaid tax and the impact of the DAC.

Occasional projects estimated the losses due to tax avoidance and evasion, but a common EU performance monitoring framework is lacking

97

In 2018, the Commission and 15 Member States produced a study on corporate income tax revenue shortfalls⁴⁵. In 2019, the Commission also had a study produced on tax evasion by individuals⁴⁶, which focused on offshore wealth (EU taxpayers are estimated to hold €1.5 trillion offshore) and the related tax revenue losses (averaging an estimated €46 billion in the EU each year). Both these reports refer to the importance of the DAC in fighting tax avoidance and evasion. However, there are currently no plans to estimate the tax gap using the data available from DAC information exchanges (including data on profit-shifting by multinational enterprises).

98

A common performance monitoring framework is central to achieving the three DAC objectives (paragraph 11), as it would give tax authorities a better view of the areas most affected by tax evasion and avoidance and allow them to allocate the necessary resources. Despite this, no single set of performance indicators is used throughout the EU to measure the effectiveness of information exchange in the area of taxation.

99

No Fiscalis project groups or other actions have yet addressed the issue of performance monitoring. The Commission and Member States decided to start such a project in 2019 and had several meetings throughout 2020, but owing to the COVID-19 crisis, no results were available at the time of our audit.

100

In its 2019 evaluation of the DAC⁴⁷, the Commission acknowledged that its overall impression of the costs and benefits of exchanging tax information was incomplete because it relied exclusively on surveys, self-assessments and assumptions, with no common approach across the Member States. For example, Belgium and Finland assessed the tax base for DAC1 supplementary taxes for different reference years, and the estimate for the additional tax income resulting from non-AEOI exchanges during 2014-2017 was limited to just six Member States.

Effective performance frameworks are also lacking in Member States

101

None of the Member States we visited had set up a comprehensive performance framework. We also found that one Member State did not use performance indicators and the remaining four took the following different approaches to measure the impact of information exchange:

• use of very limited statistical data and specific evaluations on additional tax revenues;
• an estimation of the potential for additional tax revenue by sending letters to taxpayers about incoming DAC1 and DAC2 information;
• use of statistical data in a study of the corporate and individual income tax gap, assessing the risks by sector and region;
• estimation of additional tax revenues on the basis of feedback concerning the use of tax information. We were unable to verify whether the data on which the estimates were built was complete.

102

In only one of the five Member States we visited was there evidence that local tax offices systematically collected and reported data on the use of incoming information so that the national authorities could produce rough estimates. All the other Member States lacked automatic feedback between local/regional offices and the central authorities.

103

Two of the five Member States had efficiency indicators (e.g. on data exchanged, matched and pre-filled in tax forms); while one had no indicators. One Member State prepared aggregated statistics on tax assessment amounts, late-payment interest and follow-up measures.

104

Three Member States had not calculated the cost of implementing the DAC; one had prepared a cost estimate just prior to our audit, and another put one together during the audit. One Member State collected data on the additional tax revenue collected as a result of DAC1 and DAC2 exchanges, but did not systematically compile information on benefits.

Conclusions and recommendations

105

Our overall conclusion is that the system for exchange of tax information has been well established, but more needs to be done in terms of monitoring, ensuring data quality and using the information received.

106

Our main conclusion on the Commission’s work is that it has established a suitable framework for the exchange of tax information, but it is not proactively monitoring the implementation of the legislation, providing sufficient guidance nor measuring the outcomes and impact of the system.

107

Based on the five Member States that we visited, we concluded that the information being exchanged was of limited quality and was underused. There was also minimal action being taken to monitor the system’s effectiveness. The recommendations below are addressed to those five Member States (Cyprus, Italy, Netherlands, Poland and Spain), but may also apply to others.

108

The Commission has provided a clear and transparent legislative framework for the system of exchange of tax information (see paragraphs 23 to 28). However, certain forms of income earned by non-resident taxpayers may still avoid taxation in their Member States of residence. This concerns income categories that are not part of mandatory DAC reporting (see paragraphs 29 to 33).

Recommendation 1 – Enhance the coverage of the EU legislative framework

To ensure all income categories are taxed appropriately, the Commission should make legislative proposals to:

1. make it mandatory for Member States to report all DAC1 income categories.
2. expand the scope of mandatory information exchange to include cryptocurrencies, non-custodial dividend income and advance cross-border tax rulings issued for natural persons.

Timeframe: By the end of 2022

109

The Commission monitors the transposition of the DAC legislation in the Member States and takes all the necessary measures to ensure that they meet the implementation deadlines. However, its monitoring role does not include taking direct and effective actions to address the lack of quality of the data sent between Member States, visits to Member States nor cover the effectiveness of sanctions imposed by Member States for breaches of the DAC reporting provisions (see paragraphs 34 to 38).

110

Fiscalis 2020 is working well and provides Member States with forums for the exchange of good practices. The programme also finances a great many simultaneous controls, which help Member States to coordinate their tax assessments (see paragraphs 39 to 41). The Commission provided Member States with guidelines for the implementation of DAC1 and DAC2. However, so far it has issued no guidelines for the subsequent directives, and little guidance on risk analysis and using incoming information (see paragraphs 42 to 44).

Recommendation 2 – Develop monitoring and guidance

The Commission should:

1. take direct and effective actions to address the lack of quality of data sent by Member States, as identified in its 2018 Evaluation of the DAC;
2. expand its monitoring activities to cover issues that go beyond the transposition of the DAC legislation, such as on the spot visits in Member States and penalties;
3. further develop its guidance for Member States on implementing the DAC legislation, performing risk analysis and using tax information received.

Timeframe: By the end of 2023

111

Member States seldom perform quality checks on DAC1 data before sending it to other Member States. In particular, they rarely link this information to a TIN issued by the taxpayer’s country of residence. In addition, there are large differences in the number of categories reported by each Member State (see paragraphs 45 to 50).

112

National tax authorities carry out few checks on the quality of DAC2 data reported by financial institutions. As a result, the reported data is incomplete and may be inaccurate. In addition, there is a risk that penalties that Member States apply for incorrect or incomplete reporting are not deterrent enough to ensure full compliance with the DAC2 reporting requirement (see paragraphs 51 to 55).

113

The annual feedback, which Member States provide on the data they receive, is limited to the matching rates they obtain with their national databases and seldom addresses aspects of data quality (see paragraph 56).

Recommendation 3 – Improve the quality and completeness of DAC1 and DAC2 data

In order to ensure that maximum benefit is obtained from the tax information collected and exchanged, Member States should:

1. report all DAC1 income categories;
2. set up and apply procedures for the audit of DAC2 reporting institutions;
3. establish a system of quality and completeness checks of DAC1 and DAC2 data;
4. include qualitative aspects in their annual bilateral feedback on incoming information.

Timeframe: By the end of 2023

114

The variety of systems employed by Member States for matching information produce differing results (some are more successful than others). In most cases, matching incoming data is hampered by issues of low quality and incompleteness. At present, unmatched data is not used for further analysis (see paragraphs 57 to 63).

115

Although a structured risk analysis of incoming data is central to the effective taxation of the related revenue, standard risk analysis procedures were generally not applied by Member States. Moreover, the vast majority of DAC1 and DAC2 tax information collected and exchanged is used to little effect in raising additional tax assessments (see paragraphs 64 to 71).

116

Member States upload mandatory DAC3 data to the EU’s central directory, but the rules explicitly exclude advance cross-border rulings issued for natural persons from this requirement. Information is sometimes of poor quality (missing taxpayer names, overly superficial summaries of rulings) (see paragraphs 72 to 76). None of the Member States we visited has systematic procedures for the risk analysis of DAC3 data nor the use of the information in the DAC3 directory (see paragraphs 77 and 78).

117

Member States generally do not ensure that all reporting companies comply with DAC4 data reporting, preferring instead to rely on the goodwill of the entities (see paragraphs 79 to 81). Furthermore, very few DAC4 information items are used for taxation follow-up (see paragraphs 82 and 83).

Recommendation 4 – Make better use of received information

Member States should use all incoming information as extensively as possible. To this end, they should in particular:

1. introduce procedures for the systematic risk analysis of information received from other Member States;
2. perform further analysis of unmatched DAC1 and DAC2 data, to match it against relevant taxpayers and use the resulting information to assess the related taxes;
3. include clear data and comprehensive summaries of the cross-border rulings and advance pricing arrangements they upload to the EU central directory;
4. introduce procedures to ensure that DAC4 reported data is complete and of sufficient quality.

Timeframe: By the end of 2022

118

The Member States we visited have made effective arrangements for the exchange of information on request and spontaneous exchanges of information. Despite an exponential increase during the audited period in the number of requests made, Member States are able to honour deadlines, and most delays can be explained by the complexity of requests. However, feedback between Member States is very rare (see paragraphs 84 to 91).

119

Simultaneous controls are a powerful and highly rated tool, which Member States use intensively to assess the taxation of cross-border transactions. They are increasingly common and are funded by Fiscalis 2020 (see paragraphs 92 to 95).

120

Although both the Commission and Member States have worked sporadically on the tax gap resulting from tax evasion and avoidance by individuals, no projects are in place to estimate the gap from DAC information (see paragraphs 96 and 97). There is no common framework for monitoring the impact of the system for information exchange, although some Member States have produced inconclusive assessments of the benefits it brings. Furthermore, arrangements for consistent feedback from regional to central level to allow for a proper analysis of the benefits of receiving information from abroad were largely absent (see paragraphs 98 to 104).

Recommendation 5 – Monitor the impact of information exchanges

The Commission should establish, together with Member States, a reliable common framework for measuring the benefits of the system for exchange of tax information.

Timeframe: By the end of 2022

This report was adopted by Chamber IV, headed by Mr Alex Brenninkmeijer, Member of the Court of Auditors, in Luxembourg on 8 December 2020.

Annexes

Annex I – ECA audit approach at the level of the Commission

We performed the audit at the Commission in two stages, a preparatory stage and the audit fieldwork.

In order to collect information and data that could be useful for the audit fieldwork in the Member States and to benchmark EU legislation against international best practices, during the preparatory stage we carried out a preliminary visit to DG Taxation and Customs Union. We had information gathering visits with the Romanian Tax Authorities (ANAF) and the Luxembourgish Tax Administration and analysed all the available reports produced by SAIs.

We discussed the audit methodology with the Commission (audit questions, criteria and standards). DG TAXUD granted us access to the relevant Administrative Cooperation in Direct Taxation groups in the Commission’s systems. Finally, we examined pertinent performance information, such as the Management plans and annual activity reports of DG TAXUD.

As part of the audit fieldwork, we analysed the responses and information received from a general questionnaire sent to the Commission. The questionnaire, completed by DG TAXUD together with supporting evidence, focused on the extent to which Commission has developed a sound framework for the exchange of tax information in the EU, and covered four key areas: (i) the monitoring of the legislative process and its implementation; (ii) the guidelines provided to Member States; (ii) the performance measurement of the system of exchange of tax information; and (iv) fora and guidelines provided for exchange of best practices.

In addition, we sent them one supplementary questionnaire to clarify pending issues, and we organised videoconferences with DG TAXUD to discuss the preliminary findings and the draft conclusions and recommendations.

Annex II – ECA audit approach in Member States

We carried out the audit in two stages, a preparatory stage and the audit fieldwork.

1. During the preparatory stage, we conducted the following actions:
   
   
   • preliminary work at the Commission (DG TAXUD) and desk work, in order to collect information and data that could be useful for the audit fieldwork in the Member States and to benchmark EU legislation against international best practices;
   • desk analysis on the relevant legislation, reports, other documents and analysis of data and statistics provided by the Commission and Member States;
   • discussions with the Belgian Supreme Audit Institution related to an audit on the automated exchange of tax information performed for the Belgian tax administration;
   • a visit to the Romanian tax authorities in a knowledge gathering visit, in which the practicalities and difficulties of the system were discussed.
2. We selected five Member States: Cyprus, Italy, Netherlands, Poland and Spain. The selection of Member States was based on the following risk criteria:
   
   
   • percentage of replies received for the exchanges of information on request sent to other Member States;
   • number of categories of information exchanged by each Member State for DAC1;
   • number of information messages received by each Member State for DAC2;
   • number of advance cross border tax rulings and APAs uploaded in the central database for DAC3 purposes.
3. Second stage: audit fieldwork in the selected Member States, where we paid attention to the design and implementation of system of exchange of tax information, namely how Member States are exchanging the information and how they use the information they are receiving.

We sent a questionnaire to the selected Member States. This addressed the question of how Member States have implemented the exchange of information and if they make the best use of the information they receive in order to allow a fair taxation. On the spot, we discussed the replies given to the questionnaire with the officials from the tax administrations in the field of exchange of tax information, administrative cooperation, audits and controls.

Prior to our visits, we selected samples from the information received and sent by the visited Member States, which we have verified on the spot:

1. 15 DAC1 messages received;
2. 15 DAC2 messages received;
3. 10 exchange of information on request replies received from other MS and 10 replies sent by the visited Member State;
4. 10 spontaneous exchanges of information received from other MS;
5. 10 advanced cross border rulings received;
6. 10 Country-by-Country reports received;
7. five simultaneous controls on direct taxes initiated by the visited Member State (if less than five were available, we selected all of them) and five simultaneous controls in which the Member States visited has participated.

Annex III – History of the DAC, with legislative amendments and current implementing regulations

DAC legislation timeframe

• DAC1 The Economic and Financial Affairs Council (ECOFIN) adopted Council Directive 2011/16/EU on administrative cooperation in the field of direct taxation, repealing Council Directive 77/799/EEC, on 15 February 2011. The Directive entered into force on 11 March 2011, the date it was published in the Official Journal, but it became binding on Member States on 1 January 2013 (with the exception of the AEOI rules, which became binding on 1 January 2015).
• DAC2 On 9 December 2014, ECOFIN adopted Council Directive 2014/107/EU amending Directive 2011/16/EU as regards mandatory automatic exchange of financial account information. The Directive was published in the Official Journal on 16 December 2014 and entered into force 20 days later, with the first exchanges mandatory by 30 September 2017.
• DAC3 On 8 December 2015, ECOFIN adopted Council Directive (EU) 2015/2376 amending Directive 2011/16/EU as regards mandatory automatic exchange of rulings and advance pricing arrangements in the field of taxation. The Directive entered into force on 18 December 2015, the date it was published in the Official Journal, with the first exchanges mandatory by 30 September 2017.
• DAC4 On 25 May 2016, ECOFIN adopted Council Directive (EU) 2016/881 amending Directive 2011/16/EU as regards mandatory automatic exchange of country-by-country reporting in the field of taxation. The Directive entered into force on 30 June 2016, the date it was published in the Official Journal, with the first exchanges mandatory by 30 June 2018.
• DAC5 On 6 December 2016, ECOFIN adopted Council Directive (EU) 2016/2258 amending Directive 2011/16/EU as regards access of tax authorities to anti-money laundering information. The Directive entered into force on the date it was adopted.
• DAC6 On 25 May 2018, ECOFIN adopted Council Directive 2018/822/EU amending Directive 2011/16/EU as regards mandatory automatic exchange of reportable cross-border arrangements. The Directive was published in the Official Journal on 5 June 2018 and entered into force 20 days later, with the first exchanges mandatory by 31 October 2020.

Implementing regulations for the DAC

• 6 December 2012 the European Commission adopted Implementing Regulation (EU) 1156/2012. This Regulation deals with the forms to be used for the exchange on request, the spontaneous exchanges, the requests for notification and feedback, entered into force on 10 December 2012 and applied from 1 January 2013.
• On 15 December 2014 the Commission adopted Implementing Regulation (EU) 1353/2014 amending Regulation (EU) No 1156/2012. The Regulation deals with the computerised format to be used for the mandatory automatic exchange of information. It entered into force on 22 December 2014 and has applied since 1 January 2015.
• On 15 December 2015 the Commission adopted Implementing Regulation (EU) 2015/2378 laying down detailed rules for implementing certain provisions of Council Directive 2011/16/EU on administrative cooperation in the field of taxation. It replaced Implementing Regulation (EU) No 1156/2012, with new provisions regarding the computerised format to be used for the automatic exchange of financial account information. It entered into force on 21 December 2015 and has applied since 1 January 2016.
• On 9 November 2016 the Commission adopted Implementing Regulation (EU) 2016/1963 amending Implementing Regulation (EU) 2015/2378 as regards standard forms and linguistic arrangements to be used in relation to Council Directives (EU) 2015/2376 and (EU) 2016/881. It entered into force on 1 December 2016 and has applied since 1 January 2017, with some provisions only applicable since 5 June 2017.
• On 22 January 2018 the Commission adopted Implementing Regulation (EU) 2018/99 amending Implementing Regulation (EU) 2015/2378 as regards the form and conditions of communication for the yearly assessment of the effectiveness of the automatic exchange of information and the list of statistical data to be provided by Member States for the purposes of evaluating the Council Directive. The Regulation entered into force on 12 February 2018.
• On 28 March 2019 the Commission adopted Implementing Regulation (EU) 2019/532 amending Implementing Regulation (EU) 2015/2378 as regards the standard forms, including linguistic arrangements, for the mandatory exchange of information on reportable cross-border arrangements. It entered into force on 18 April 2019 and has applied since 1 July 2020.

Annex IV – Comparison of DAC2 and the OECD framework

Element	Global standard	DAC2
Data protection	High-level general provisions	Detailed rules implementing the EU data protection framework
Structure	Only the Competent Authority Agreement (CAA) is signed, the Common Reporting Standard (CRS) is implemented domestically.	Only some elements of the CAA are included; the rest is already part of DAC1. CRS is in Annex I, important elements from the Commentaries are in Annex II (Change in circumstances; Self-certification for New Entity Accounts; Residence of a Financial Institution; Account maintained; Trusts that are Passive NFEs; Address of entity’s principal office). Language is adapted to reflect the multilateral context of the two annexes (e.g. “Member States” instead of “jurisdiction”).
CRS Section I(D)	Includes a scenario in (ii) where the domestic law of the relevant Reportable Jurisdiction does not require the collection of the TIN issued by such Reportable Jurisdiction.	Taking into account their legal systems, none of the Member States considered this to be a situation that was applicable to them.
CRS Section I(E)	One of the conditions for reporting the place of birth is that the place of birth is not required to be reported unless the Reporting Financial Institution is otherwise required to obtain and report it under domestic law and it is available in the electronically searchable data maintained by the Reporting Financial Institution.	The condition for reporting the place of birth has been extended to also include cases where the Reporting Financial Institution is or has been otherwise required to obtain and report it under any Union legal instrument in effect or that was in effect the day before the entry into force of the Directive. This is meant to secure the standard of reporting of place of birth even after any repeal of the EUSD.
CRS Section I(F)	Phase-in for gross proceeds (Foreign Account Tax Compliance Act – FATCA-inspired).	No phase-in for gross proceeds (after the Early Adopters’ timeline)
CRS Section III(A)	Exemption for insurances where the Reporting Financial Institution is effectively prevented by law (relaxed further in the Commentaries) from selling such a contract to residents of a Reportable Jurisdiction.	The exemption is not applicable in the EU since the regulatory framework does not meet the conditions in the CRS and the Commentaries.
CRS Section VII(B), Section VIII(C)(9) Section VIII(E)(4) Section VIII(E)(6)	The Commentaries to those paragraphs contain alternative provisions for: Group Cash Value Insurance Contracts and Group Annuity Contracts new accounts of pre-existing customers Related Entity standardised industry coding system	In order to function properly, the four parts of the Commentary providing for “alternative rules” for financial institutions have been incorporated in Annex I.
CRS Section VIII(B)(1)(c) Section VIII	The lists of Non-Reporting Financial Institutions and Excluded Accounts are defined in domestic law.	The lists of Non-Reporting Financial Institutions and Excluded Accounts are defined in domestic law; Member States inform the Commission of the lists (and changes thereof) and the Commission publishes those in the Official Journal.
CRS Section VIII(D)(5)	The term “Participating Jurisdiction” means a jurisdiction (i) with which an agreement is in place pursuant to which it will provide the information specified in Section I, and (ii) which is identified in a published list.	The term “Participating Jurisdiction” explicitly includes all Member States and any other jurisdiction with which the Union has an agreement in place pursuant to which that jurisdiction will provide the information specified in Section I.

Source: ECA, based on DG TAXUD data.

Annex V – Legislative process: duration and changes (DAC1 to DAC4)

Step/issue	DAC1	DAC2	DAC3	DAC4
Adoption COM	2.2.2009	12.6.2013	18.3.2015	28.1.2016
EP Report	10.2.2010	11.12.2013	27.10.2015	12.5.2016
Political agreement in the Council	7.12.2010	14.10.2014	6.10.2015	8.3.2016
Adoption Council	15.2.2011	9.12.2014	8.12.2015	25.5.2016
Number of compromises in the process	5	2 (at least)48	6	N/A
OJ Publication	15.2.2011	16.12.2014	18.12.2015	3.6.2016
Main changes in the legislative process	Income categories: the initial process did not contain any specific categories. Subsequently, they were later specified, but only one category was made mandatory, whereas the initial proposal would have made all mandatory.	The negotiation of DAC2 was an ad-hoc process: the proposal evolved in parallel with the discussions at the OECD where the CRS was being developed. One Member State was allowed a longer transition period for AEOI	Retroactive element for rulings up to 5 years back instead of 10 years back – Limited access for the COM to the central directory – only for the purpose of monitoring the functioning of the Directive. Application of the Directive as of 1.1.17 instead of 1.1.16 (as per Commission proposal)	Member States may opt to delay the obligation for EU resident subsidiaries of non-EU multinational groups to file CbC reports for the first year (i.e. the first fiscal year that starts on or after 1 January 2016)

Source: ECA, based on European Commission documentation and representations.

Annex VI – Fiscalis 2020

The Fiscalis 2020 action programme was established by Regulation (EU) No 1286/2013.

Between 2015 and 2019, the Commission and Member States organised 11 Fiscalis ‘project groups’ to discuss matters relating specifically to the implementation of the DAC, the associated IT systems and the way the information exchanged is used. The work done in 2015-2016 by the project group on data analysis resulted in a Commission guideline on good practices in the risk analysis of information received under DAC1 and DAC2.

Programme objectives

The programme’s overall objective is “to improve the proper functioning of the taxation systems in the internal market by enhancing cooperation between participating countries, their tax authorities and their officials” (Article 5(1) of Regulation (EU) No 1286/2013).

More specifically, this means supporting the fight against tax fraud, tax evasion and aggressive tax planning and the implementation of EU law in the field of taxation. These aims are achieved by:

• ensuring the exchange of information;
• supporting administrative cooperation;
• where necessary, enhancing the administrative capacity of participating countries.

Achievement of the programme objectives is measured in particular on the basis of:

• the availability of, and full access to European information systems for taxation, through the EU’s common communications network;
• feedback from participating countries on the results of programme actions.

In practice, Fiscalis has the following operational objectives and priorities:

• to implement, improve, operate and support the European information systems for taxation;
• to support administrative cooperation activities;
• to reinforce the skills and competence of tax officials;
• to enhance the understanding and implementation of EU taxation law;
• to support improvements to administrative procedures and the sharing of good administrative practices.

The objectives of all programme actions must be in line with the above objectives and priorities, which are based on the Fiscalis 2020 Regulation and developed through annual work programmes. The Commission draws up and adopts each annual work programme with input from the participating countries.

Actions funded under Fiscalis 2020

Joint actions are organised by and for tax officials from participating countries. They cover:

• seminars and workshops;
• project groups set up for a limited period of time to pursue a predefined objective with a precisely described outcome;
• bilateral or multilateral controls and other activities provided for in EU law on administrative cooperation;
• working visits to enable officials to acquire or increase their expertise or knowledge in tax matters;
• expert teams set up to facilitate longer-term structured operational cooperation among participating countries by pooling expertise;
• public administration capacity-building and supporting actions;
• studies;
• communication projects;
• any other activity in support of the Fiscalis 2020 objectives.

The development, maintenance, operation and quality control of the EU components of new and existing European information systems.

The development of common training activities to support the necessary professional skills and knowledge in the field of taxation.

Implementation of the programme

The European Commission is responsible for implementing the programme. It is assisted by a committee composed of delegates from each EU Member State.

Acronyms and abbreviations

ACDT: Administrative cooperation in direct taxation

AEOI: Automatic exchange of information

CRS: Common Reporting Standard

DAC: Directive for Administrative Cooperation in the field of taxation

DAC1: DAC amendment introducing mandatory AEOI for five categories of incomes and capital

DAC2: DAC amendment extending the scope of mandatory AEOI to financial accounts

DAC3: DAC amendment introducing mandatory AEOI for advance cross-border tax rulings and advance pricing arrangements

DAC4: DAC amendment introducing mandatory country-by-country reporting by multinational groups

DAC5: DAC amendment to ensure that tax authorities can access information on money laundering

DAC6: DAC amendment introducing mandatory AEOI on reportable cross-border tax arrangements

DG TAXUD: Directorate-General for Taxation and Customs Union

ECOFIN: Economic and Financial Affairs Council

EOIR: Exchange of information on request

G20: Group of finance ministers and central bank governors from 19 countries and the EU

GDP: Gross domestic product

HNWI: High net worth individuals

MNE: Multinational enterprise

OECD: Organisation for Economic Co-operation and Development

PAOE: Presence of officials of a Member State in the offices of the tax authorities of another Member State or during administrative enquiries carried out therein

SEOI: Spontaneous exchange of information

SMC: Simultaneous control

TIN: Tax identification number

Glossary

Administrative cooperation: Collaboration between Member States, in the form of the exchange and processing of information, joint action or mutual assistance, for the purposes of implementing EU law.

Advance cross-border ruling: An agreement, communication, or other instrument or action with similar effects, issued by a government authority to a particular person or group of persons, and concerning the legal or administrative interpretation or tax treatment of forthcoming cross-border transactions or activities by those persons in another jurisdiction.

Advance pricing arrangement: An agreement, communication, or other instrument or action with similar effects, issued by a government authority to a particular person or group of persons, that determines, in advance of cross-border transactions between associated enterprises, the criteria for pricing the transfer of goods, intangible assets or services between those enterprises.

Automatic exchange of information: Systematic communication of predefined information from one jurisdiction to another at pre-established regular intervals.

Common communications network: A common platform developed by the EU to allow the transmission of information by electronic means between different Member States’ customs and taxation authorities.

Exchange of information on request: Exchange of information based on a specific request made by one jurisdiction to another.

Exchange of tax information: Information exchanged between Member States for tax related purposes

Fiscalis 2020: EU action programme, part of the Europe 2020 strategy, that funds initiatives by tax authorities to improve the functioning of taxation systems in the EU.

Non-custodial dividends: Dividends that are not paid or cashed in a custodial account.

Simultaneous control: Coordinated checks on the tax situation of taxable persons by two or more Member States with a common or complementary interest.

Spontaneous exchange of information: Non-systematic communication of information from one jurisdiction to another, at any moment and without prior request.

Tax avoidance: Use of a legal arrangement to get around an obligation to pay tax, generally by entering into a set of artificial financial arrangements whose main or sole purpose is to reduce a tax bill.

Tax evasion: Using illegal means to avoid paying taxes, typically by misrepresenting income to the tax authorities. Often overlaps with the notion of tax fraud.

Tax identification number: Unique reference number allocated by the tax authorities to a taxpayer.

Audit team

The ECA’s special reports set out the results of its audits of EU policies and programmes, or of management-related topics from specific budgetary areas. The ECA selects and designs these audit tasks to be of maximum impact by considering the risks to performance or compliance, the level of income or spending involved, forthcoming developments and political and public interest.

This performance audit was carried out by Audit Chamber IV Regulation of markets and competitive economy, headed by ECA Member Alex Brenninkmeijer. The audit was led by ECA Member Ildikó Gáll-Pelcz, supported by Claudia Kinga Bara, Head of Private Office and Zsolt Varga, Private Office Attaché; John Sweeney, Principal Manager; Dan Danielescu, Head of Task; Maria Echanove, Mirko Gottmann and Stefan Razvan Hagianu, Auditors. Thomas Everett provided linguistic support.

Endnotes

¹ List of non-cooperative jurisdictions.

² Dover et al: ‘Bringing transparency, coordination and convergence to corporate tax policies in the European Union’, Part I of a study for the European Parliamentary Research Service, September 2015 (PE 558.773).

³ Conclusions of the ECOFIN meeting on 1 December 1997 concerning taxation policy (OJ C 2/01, 6.1.1998).

⁴ A legal arrangement to get around an obligation to pay tax, generally by entering into a set of artificial financial arrangements whose main or sole purpose is to reduce a tax bill.

⁵ Using illegal means to avoid paying taxes, typically by misrepresenting income to the tax authorities. Often overlaps with the notion of tax fraud.

⁶ The term”exchange of tax information” refers to information exchanged between Member States for tax related purposes.

⁷ DG TAXUD’s mission statement and strategic goals.

⁸ Council Directive (77/799/EEC) of 19 December 1977 concerning mutual assistance by the competent authorities of the Member States in the field of direct taxation.

⁹ Council Directive 2011/16/EU of 15 February 2011 on administrative cooperation in the field of taxation and repealing Directive 77/799/EEC.

¹⁰ Tax base erosion and profit shifting (BEPS) actions.

¹¹ OECD/G20: “Countering Harmful Tax Practices More Effectively, Taking into Account Transparency and Substance”, Action 5 – 2015 Final Report.

¹² OECD/G20: “Transfer Pricing Documentation and Country-by Country Reporting”, Action 13 – 2015 Final Report.

¹³ OECD, “Standard for Automatic Exchange of Financial Account Information in Tax Matters”, Second Edition, 2017.

¹⁴ OECD Common Reporting Standard.

¹⁵ Including Article 26(1) and (3) of the OECD Model Tax Convention (on relevance and time limits), the OECD/Council of Europe Convention on Mutual Administrative Assistance in Tax Matters, and the OECD’s model tax information exchange agreement.

¹⁶ Evaluation of Council Directive 2011/16/EU.

¹⁷ Mutual Assistance Directive (77/799/EEC), most recently amended in 2001 and complemented by the Savings Directive (2003/48/EC).

¹⁸ OECD/Council of Europe Convention on Mutual Administrative Assistance in Tax Matters and amending protocol.

¹⁹ Evaluation of Council Directive 2011/16/EU.

²⁰ According to an EU Parliament study, the total market capitalisation of the 100 largest cryptocurrencies is reported to exceed the equivalent of €330 billion globally by early 2018.

²¹ Commission proposal for a Directive amending the DAC, COM(2020) 314 final.

²² Council Directive (EU) 2016/2258 amending Directive 2011/16/EU as regards access to anti-money-laundering information by tax authorities.

²³ Directive (EU) 2015/849 of the European Parliament and of the Council of 20 May 2015 on the prevention of the use of the financial system for the purposes of money laundering or terrorist financing. The rules on access are in Articles 13, 30, 31 and 40.

²⁴ Evaluation of the DAC, COM(2018) 844 final.

²⁵ Article 25a of the DAC.

²⁶ “Upgrading the Single Market: more opportunities for people and business”, European Commission Directorate-General for Communication, 2015.

²⁷ Regulation (EU) No 1286/2013 of the European Parliament and of the Council.

²⁸ Income from employment, directors’ fees, life insurance products, pensions, ownership of or income from immovable property (see Article 8(1) of Council Directive 2011/16/EU).

²⁹ Report from the Commission to the European Parliament and the Council, COM(2018) 844 final.

³⁰ Proposal for a Directive amending DAC, COM(2020) 314 final, July 2020.

³¹ Article 8(2) of Council Directive 2011/16/EU.

³² Map background ©OpenStreetMap contributors licensed under the Creative Commons Attribution-ShareAlike 2.0 license (CC BY-SA).

³³ Report from the Commission to the European Parliament and the Council, COM(2018) 844 final.

³⁴ These contain the information referred to in Article 8(3a) of Council Directive 2011/16/EU, as amended by Council Directive 2014/107/EU.

³⁵ Report from the Commission to the European Parliament and the Council, COM(2018) 844 final.

³⁶ Council Directive (EU) 2015/2376 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation.

³⁷ Article 8a(4) of Council Directive 2011/16/EU, as amended by Council Directive (EU) 2015/2376.

³⁸ European Parliament resolution of 26 March 2019 on financial crimes, tax evasion and tax avoidance (2018/2121(INI)).

³⁹ Evaluation of the DAC, SWD(2019) 327 final.

⁴⁰ Council Directive (EU) 2016/881 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation.

⁴¹ Article 5 DAC.

⁴² Article 9 DAC.

⁴³ Article 7 DAC.

⁴⁴ Article 12 DAC.

⁴⁵ The Concept of Tax Gaps Report II: Corporate Income Tax Gap Estimation Methodologies, Fiscalis Tax Gap Project Group (FPG/041).

⁴⁶ DG TAXUD: “Estimating International Tax Evasion by Individuals”, Taxation Paper No 76-2019.

⁴⁷ Evaluation of Council Directive 2011/16/EU.

⁴⁸ Commission files were incomplete, however the minimum number of compromises were 2.

Timeline

Event	Date
Adoption of Audit Planning Memorandum (APM) / Start of audit	3.7.2019
Official sending of draft report to Commission (or other auditee)	29.10.2020
Adoption of the final report after the adversarial procedure	8.12.2020
Commission’s (or other auditee’s) official replies received in all languages	15.12.2020

Contact

EUROPEAN COURT OF AUDITORS
12, rue Alcide De Gasperi
1615 Luxembourg
LUXEMBOURG

Tel. +352 4398-1
Enquiries: eca.europa.eu/en/Pages/ContactForm.aspx
Website: eca.europa.eu
Twitter: @EUAuditors

More information on the European Union is available on the internet (http://europa.eu).

Luxembourg: Publications Office of the European Union, 2021

PDF	ISBN 978-92-847-5624-7	ISSN 1977-5679	doi:10.2865/600	QJ-AB-21-003-EN-N
HTML	ISBN 978-92-847-5599-8	ISSN 1977-5679	doi:10.2865/823531	QJ-AB-21-003-EN-Q

COPYRIGHT

© European Union, 2021.

The reuse policy of the European Court of Auditors (ECA) is implemented by Decision of the European Court of Auditors No 6-2019 on the open data policy and the reuse of documents.

Unless otherwise indicated (e.g. in individual copyright notices), the ECA’s content owned by the EU is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence. This means that reuse is allowed, provided appropriate credit is given and changes are indicated. The reuser must not distort the original meaning or message of the documents. The ECA shall not be liable for any consequences of reuse.

You are required to clear additional rights if a specific content depicts identifiable private individuals, e.g. in pictures of the ECA’s staff or includes third-party works. Where permission is obtained, such permission shall cancel and replace the above-mentioned general permission and shall clearly indicate any restrictions on use.

To use or reproduce content that is not owned by the EU, you may need to seek permission directly from the copyright holders.

Pictures 3 and 5: Icons made by Pixel perfect from https://flaticon.com.

Software or documents covered by industrial property rights, such as patents, trade marks, registered designs, logos and names, are excluded from the ECA’s reuse policy and are not licensed to you.

The European Union’s family of institutional Web Sites, within the europa.eu domain, provides links to third-party sites. Since the ECA has no control over them, you are encouraged to review their privacy and copyright policies.

Use of European Court of Auditors’ logo

The European Court of Auditors logo must not be used without the European Court of Auditors’ prior consent.

GETTING IN TOUCH WITH THE EU

In person
All over the European Union there are hundreds of Europe Direct Information Centres. You can find the address of the centre nearest you at: https://europa.eu/european-union/contact_en

On the phone or by e-mail
Europe Direct is a service that your questions about the European Union. You can contact this service

• by freephone: 00 800 6 7 8 9 10 11 (certain operators may charge for these calls),
• at the following standard number: +32 22999696 or
• by electronic mail via: https://europa.eu/european-union/contact_en

FINDING INFORMATION ABOUT THE EU

Online
Information about the European Union in all the official languages of the EU is available on the Europa website at: https://europa.eu/european-union/index_en

EU Publications
You can download or order free and priced EU publications at: https://op.europa.eu/en/web/general-publications/publications. Multiple copies of free publications may be obtained by contacting Europe Direct or your local information centre (see https://europa.eu/european-union/contact_en)

EU law and related documents
For access to legal information from the EU, including all EU law since 1952 in all the official language versions, go to EUR-Lex at: http://eur-lex.europa.eu/homepage.html?locale=en

Open data from the EU
The EU Open Data Portal (https://data.europa.eu/euodp/en/home) provides access to datasets from the EU. Data can be downloaded and reused for free, both for commercial and non-commercial purposes.