25.1.2008   

EN

Official Journal of the European Union

C 20/1

I-VII

1-11

1-11

12-17

18-93

18-49

19-23

24-42

43-49

50-56

57-70

60-61

62-65

66-68

69-70

71-77

71-72

73-77

78-83

84-93

85-87

88-91

92-93

94-112

94-103

104-112

I.

Large scale value added tax evasion and fraud distort the functioning of the single market and affect the financial interests of the Member States and the financing of the Community budget.

II.

In 2004, new Community legislation came into force to speed up and strengthen cross-border cooperation between Member State authorities, mainly through clearer procedures, more comprehensive exchange of information and increased direct contacts between local tax offices.

III.

The audit of the Court aimed to assess whether information exchanges between Member States take place in a timely and effective manner and are supported by sound procedures and adequate administrative structures.

IV.

The audit showed that information exchanges between Member States can help Member States to assess taxation correctly and to prevent and detect fraud.

V.

However, the Court found that:

VI.

There is a need for more intensive and rapid cooperation, more direct contacts between local tax offices and better monitoring to ensure that Member States provide efficient assistance to each other.

VII.

The weaknesses of VIES should be urgently addressed, for example by radically shortening the timescale for collecting and capturing data and by granting broader direct access to data to enable multilateral consultations.

1.

Value added tax (VAT) evasion and fraud distort fair competition in the single market and reduce Member States' tax revenue. As the Council has emphasised (4), such fraud must be tackled effectively and decisively in the interest of honest entrepreneurs and Member State budgets. In order to complement national efforts in this regard, the Council agreed on 28 November 2006 on the urgent need to establish at Community level an anti-fraud strategy to combat tax fraud.

2.

VAT evasion and fraud also impact on the financing of the budget of the European Union, as they result in an increased need to call from Member States own resources based on gross national income (GNI). The GNI resource is intended to cover the balance of total expenditure not covered of other resources. It follows that losses caused by VAT fraud affect the overall balance of the own resources system as laid down by the Community legislator (5). As the criteria for calculating VAT and GNI resources differ, those losses might also affect the incidence of the financial burden on the Member States (6), thus undermining the principle of equity.

3.

Though investigations into the magnitude of VAT evasion and fraud have not been carried out in all Member States, several estimates have been published. The International VAT Association (7) quotes estimates of VAT losses ranging from 60 billion to 100 billion euro per annum across the European Union. In the United Kingdom alone, HM Revenue and Customs (HMRC) estimate that in the tax year 2005-2006 VAT revenue losses amounted to 18,2 billion euro (8). In Germany, the Ministry of Finance published (9) the results of a study which estimated, for 2005, VAT losses of 17 billion euro.

4.

Most of the VAT evasion is linked to undeclared economic activities (shadow economy). There is, however, also a significant part of VAT evasion that occurs as a side-effect of the VAT arrangements put in place when the single market was introduced in 1993. For intra-Community trade, goods supplied are exempted from VAT. The VAT due is then payable in the Member State where the goods arrive. These arrangements can give rise to VAT evasion either in the supplier country or in the country of destination. Three main types of evasion take place:

5.

Whilst goods move freely across internal borders, the action of national tax authorities still remains largely constrained by those same borders. If a trader makes intra-Community supplies, he has the right to obtain a refund of the input VAT. The VAT due is then payable by the trader in the Member State of destination. Cooperation between Member States' authorities is therefore necessary in order to exchange the information needed to assess VAT correctly. The common system for the exchange of information between the Member States is governed by Council Regulation (EC) 1798/2003 of 7 October 2003 on administrative cooperation in the field of value added tax (1) (hereafter called the regulation), which entered into force on 1 January 2004, repealing the former Regulation (EEC) No 218/92 (12). The regulation lays down the conditions under which tax authorities in the Member States are to cooperate with one another and with the Commission. The regulation does not preclude more comprehensive mutual assistance on the basis of other legal acts, including bilateral or multilateral agreements. Where Member States conclude additional bilateral arrangements on administrative cooperation, they are required to inform the Commission and the other Member States thereof without delay.

6.

When adopting the regulation in October 2003, the Council stated (13) that there had been until then an under-use of the arrangements for administrative cooperation. The Council called for exchange of information to be made more intensive to combat fraud more effectively.

7.

Three types of exchange of information are foreseen under the regulation (see Figure 1):

8.

Exchanges of information under the regulation are mainly channelled through Central Liaison Offices (CLO). Each Member State has to designate a single CLO with principal responsibility for contacts with other Member States. In addition, competence to exchange information directly with other Member States may be delegated to other departments or to individual officials.

9.

Furthermore, the regulation lays down rules for controls conducted simultaneously in two or more Member States and allows for tax officials to be present in other Member States, for example in order to have access to documentation held there or to attend in ongoing enquiries.

10.

The regulation also provides a role for the Commission in promoting and evaluating administrative cooperation between Member States. In this it is assisted by a committee made up of representatives of Member States, the Standing Committee on Administrative Cooperation (SCAC).

11.

Furthermore, the regulation is, together with Decision No 2235/2002/EC on the Fiscalis programme, the legal basis for Community expenditure (14) to develop and keep operational electronic systems for information exchange between national administrations, multilateral controls, training seminars for tax officials, and the exchange of officials between administrations. The financial framework for Fiscalis for the period from 2003 to 2007 was set at 67,3 million euro.

12.

The objective of the Court's audit was to assess whether the information exchanges between Member States are carried out in a timely and effective manner and whether adequate administrative structures and procedures are in place to support administrative cooperation. The audit covered:

13.

In 2006, CLOs in seven Member States (15) were visited. A sample of 420 exchanges of information on request, sent or received in 2005 by these Member States, was examined and the timeliness and quality of the responses was assessed. However, the accuracy of individual tax assessments was not itself subjected to audit.

14.

Audit missions were carried out pursuant to Article 248(3) of the Treaty which provides that the Court's audit shall be, if necessary, performed on the premises of any body which manages revenue or expenditure on behalf of the Community and in the Member States, including on the premises of any natural or legal person in receipt of payments from the budget. The German Ministry of Finance, however, did not comply with an audit request by the Court stating that, in its opinion, there was no legal basis for such an audit. Observations concerning Germany in this report are therefore based on findings from audit missions to other Member States, information obtained during audit missions at the Commission, and publicly available reports.

15.

In 23 cases (16) selected from the sample, the Court invited Member States to provide details on how the information obtained through exchanges was followed up by the local tax offices which had initiated the requests.

16.

At the Commission, the operation of the relevant departments was examined including a review of the management of the Fiscalis programme. The operation of the Standing Committee on Administrative Cooperation (SCAC) was also reviewed.

17.

In addition, use has been made of reports and work of National Audit Offices in Member States that have performed similar audits. Issues raised in previous Court reports (17) were also followed up.

18.

Article 5 of the regulation provides for Member States to exchange information on request. As a rule, any information that may help to effect a correct assessment of VAT may be subject to a request, including any information relating to specific cases. In order to obtain the information sought by the tax administration in another Member State, the requested authority is required to proceed as though acting on its own account.

19.

Since the regulation came into force on 1 January 2004, the number of requests for information has increased by 62 %. It went up from 18 162 requests in 2003 to 29 381 in 2006.

20.

However, an analysis of the figures provided by Member States to the Commission (18) shows that this increase appears to be mainly attributable to enlargement (19) with the number of requests sent amongst EU-15 in 2005 and 2006 being below the level reached in 2004 (see Table 1).

Table 1

Number of requests for exchange of information (according to requesting Member States)

21.

In 2006, more than 3,3 million traders across the EU declared that they had made acquisitions from other Member States. This means that even with the conservative assumption that each of the 29 381 exchanges of information on request in 2006 concerned different traders, the probability of being subject to such an exchange of information was below 0,9 %. In reality, the probability was even lower, as a number of companies were subject to several requests.

22.

An important reason why cooperation has not become more intensive is the limited exploitation of the possibilities of decentralisation offered by the regulation. When adopting the regulation, the Council stated (20) that, up to then, there had not been enough direct contact between local tax offices, with communication between central liaison offices being the rule and that this centralisation had led to under-use of the arrangements for administrative cooperation. The Council concluded that provisions should be made to bring about more direct contacts between services.

23.

Limited use has been made so far by Member States of this option. Only a few Member States (21) have devolved competence to territorial services. If such decentralisation is implemented only in a minority of Member States, it fails to produce its full effects as there are no counterparts for decentralised cooperation in most of the other Member States.

24.

Rapid replies to requests for information are critical for effective cooperation. They allow fraudulent companies to be quickly identified and make it more difficult for fraudsters to hide their gains from VAT fraud. Therefore the regulation obliges Member States to exchange information as quickly as possible and no later than three months after the receipt of a request. Where the requested authority is already in possession of that information, the time limit is reduced to one month.

25.

Statistics for 2006 provided by Member States show that almost 50 % of requests were not answered within the deadline. This figure was even higher than in 2005 when 42 % of the replies were late.

26.

Late replies occur in all Member States but their frequency varies significantly between Member States. In 2006, two Member States (22) replied within the deadline in almost 90 % of cases, whereas eight Member States (23) replied late in more than 50 % of cases (see Table 2).

Table 2

Late replies by Member State

27.

As far as the replies given within one month are concerned, the statistics showed a decrease in the number of times for which this deadline has been met from 2 154 cases in 2004 down to 1 885 in 2006. Compared to the total number of requests, this means that the share of early replies fell from 9,7 % in 2004 to 6,4 % in 2006.

28.

In addition, in certain Member States visited, a number of requests were found to have remained outstanding for one year or more. In Italy for example, there were, at the time of the Court's audit visit in May 2006, 182 unanswered requests received in the last four months of 2004 alone. In the United Kingdom, in early 2006, 121 requests were still outstanding from 2003 and 2004.

29.

The regulation stipulates explicitly that where the requested authority is unable to respond by the deadline it shall inform the requesting authority in writing forthwith of the reasons for this and when it would be able to respond. Nevertheless, prior notice of probable delays is only rarely given. Late replies' notifications were sent in 3 out of 57 cases not answered on time which were examined by the Court.

30.

Statistics from Member States for 2006 showed that in only 5 % of all cases (737 out of 14 565 late replies) had delays been notified to the requesting authorities. Absence of such notification prevents the requesting Member State from taking alternative action in the knowledge that the request will not be promptly answered.

31.

Furthermore, during the audit only very few cases were found where partial information was given that was already in the possession of the requested authority and therefore rapidly available, followed by a more comprehensive answer once tax control visits had taken place.

32.

The audit revealed a number of factors which impact negatively on the timeliness of replies. Most of the delays in responding to requests occur at Local Tax Office (LTO) level. Mechanisms are not always in place to ensure that LTOs give as much priority to administrative cooperation as to their other activities (24). A similar observation was made in a report provided by the National Audit Office of Denmark (NAOD) (25). Danish customs and tax authorities had set target figures for compliance with case processing time limits, but, at the time of the audit, requests from the tax authorities of other countries were not included in these target figures. According to a follow-up report by the NAOD, the minister responsible has taken initiatives to deal with the issue.

33.

A number of other causes were found which have a negative impact on the treatment of requests for information. These included: inadequate use of the information request forms (e.g. multiple requests in a single standard form, over use of ‘free-text’ boxes of these forms and use of non-electronic communications) by the sending Member States, difficulties in understanding what was actually requested (e.g. due to unclear information, language problems or mistakes in using the standard forms), and doubts about the usefulness of certain requests in the Member States receiving them.

34.

Delays can also be caused by the fact that not all Member States have set up adequate systems for monitoring and managing timely response to requests.

35.

Positive examples of such management of requests received were found in Luxembourg, Poland, Slovenia and the UK, where CLOs set internal deadlines and monitor them effectively.

36.

In Italy, where the software put in place for this purpose by the CLO was not effectively used, cases of input errors and missing data were found which led to mismanagement of requests. In the CLO reminders to liaison departments were programmed for only 10 days before the expiry of deadlines, too late for necessary action to be undertaken.

37.

Though both the Dutch and French CLOs set intermediate deadlines for the LTOs to respond to requests received, in neither CLO was there evidence of monitoring of these timetables. Only where reminders were received from the requesting Member State was any action taken towards the LTO which was late in providing answers.

38.

In some Member States visited with a high number of late replies, the Court found that complicated organisational set-ups also contribute to delays and create problems in monitoring.

39.

In Italy, in addition to the CLO, there are three liaison departments at central Member State level, all of them equally competent in the field of VAT cooperation. Other Member States are free to contact each time any one of the three liaison departments and send the requests directly to the department selected. Working procedures of the three liaison departments are not harmonised and there is insufficient coordination amongst them and weak supervision by the CLO. Furthermore, such an organisational set-up is not in line with the regulation which provides (26) that liaison departments have to be given specific territorial competence or specialised operational responsibility.

40.

The Dutch CLO comprises two independent operating units, one located in Amsterdam and the other in Almelo, each with its own management structure. The Court found cases where it took nearly three months before replies to requests provided by LTOs were passed on to the requesting Member States. In these cases, LTOs had delivered replies to the CLO in Almelo which were transmitted to the CLO Amsterdam from where they were sent to the requesting Member States.

41.

In Germany, according to the information made available to other Member States (27), the tasks of the CLO have been split between three units of the Central Federal Tax Office, one based in Bonn, the two others in Saarlouis. No information was made available to other Member States about who was heading the CLO.

42.

The coordination problems caused by the complicated organisational structures and the weaknesses of monitoring systems contribute to the significant discrepancies in the statistics produced by different Member States. In several cases there are significant differences between the number of requests which a Member State claims to have received and the number of requests which other Member States claim to have sent it (see Table 3).

Table 3

Discrepancies by Member State (2005)

43.

Examination of the sampled requests and answers showed that though the information provided was sometimes inadequately presented, in most cases it was comprehensive and detailed, answering the questions posed. In response to the Court's questions on the follow-up of a selection of exchanges (see paragraph 15), Member States affirmed that the answers received had enabled correct assessment of VAT in the cases concerned.

44.

Currently there is no legal requirement to assess the results of administrative cooperation but a few Member States have conducted such exercises. These evaluations demonstrate the benefits of administrative cooperation.

45.

In France, though the financial impact of the cooperation has not been evaluated, an analysis by the competent authorities showed that 34 % of replies received in 2005 confirmed or revealed the existence of fraud in France or another Member State.

46.

An evaluation by the CLO in Slovenia of 89 requests for administrative cooperation received by their authorities in 2005 showed that in the case of 21 requests, administrative enquiries resulted in the collection of additional VAT of approximately 1,5 million euro.

47.

An analysis made by the Polish CLO in 2006 showed that 713 exchanges of information on request led to the detection of 64 fraud cases.

48.

However, it was noted that there are some concerns with regard to the quality of information exchanged on suspected Missing Trader Fraud. Germany reported to the SCAC in 2005 problems in obtaining sufficiently detailed information from other Member States to use in legal action before German courts. The matter has been repeatedly discussed by the SCAC since and a supplementary questionnaire has been developed to tackle cases where more detailed information is needed for legal action in Member States. So far, however, Member States have not reached agreement about the circumstances in which this questionnaire should be used or the requirements for its use.

49.

Furthermore, there is no mechanism whereby the user of replies to requests for information give feedback on the usefulness of the information supplied. Any possible shortcomings in the reply are not reported back and cannot be taken into account when dealing with further requests.

50.

Article 17 of the regulation provides for Member States to exchange information without prior request. It specifies that each Member State should forward information to any other Member State concerned in three situations. Firstly, where the effectiveness of the control system in the other Member State necessarily depends on the information provided by the Member State of origin; secondly, where there is a suspicion of a breach of VAT legislation in the other Member State; and thirdly, in cases which present a risk of tax loss (due to fraud or avoidance) in the other Member State. This represents a very broad obligation.

51.

Article 18, however, limits the obligation provided for in Article 17, stating that the Member States may determine by choice whether they will exchange certain categories of information without prior request.

52.

In total, five categories with 13 subcategories of cases have been defined (28), where information can be forwarded by automatic exchange (29) or structured automatic exchange (30). These categories include, for example, information about VAT identification numbers allocated to taxable persons established in another Member State or about taxable persons who are (potential) missing traders but whose VAT identification number has not been cancelled.

53.

The exchanges of information show great differences in the use of these categories by Member States. Germany, for example, communicated to the Commission that it would participate in all categories, whilst France announced its participation only in a single subcategory. The statistics reported by Member States on these automatic exchanges of information without request are of poor quality. Discrepancies mainly due to differing understandings of what constitutes a reportable event make it impossible to draw any meaningful conclusions from the data provided.

54.

In addition to the exchanges of information in predefined categories, Article 19 of the regulation stipulates that Member States may forward to each other spontaneously any information that may help them to effect a correct assessment of VAT. An example for such cooperation is the European Carousel Network (Eurocanet) which was launched at Belgian initiative and receives Commission support through the European Anti-Fraud Office OLAF. The network aims to speed up information exchanges on a limited number of targeted companies which are suspected of being involved in missing trader fraud. For this type of information exchange, in which not all Member States participate, the Commission has no access to collect statistical data.

55.

During the audit it was found that information forwarded without prior request was not always adequately used. The Luxembourg CLO, for example, reported that they had repeatedly received requests for data that they had already provided spontaneously to other Member States.

56.

Extra information is also often given as part of a reply to a request for information, when information is provided which was not explicitly requested but is considered to be useful, for example information about other traders involved in business with a specific company. An audit by the Lithuanian National Audit Office (31) found that such additional information provided by other Member States in replies to requests for information was not forwarded to the appropriate authorities.

57.

VIES is a common computer network for a largely automated exchange of data between tax administrations. It also enables traders to obtain confirmation that their trading partners in other Member States are registered.

58.

The VIES network provides information on the VAT registration numbers issued by the Member States, including date of issue, trader's name, trader's address and, where applicable, date of cessation of validity of a VAT number. It is possible to check the history of a given VAT number, i.e. all modifications made to its attributes.

59.

Furthermore, Member States communicate through VIES details of all intra-community supplies by their traders to other Member States. This is done on the basis of recapitulative statements which are to be drawn up for each calendar quarter by each intra-Community supplier and record the total value of goods supplied to each purchaser in other Member States. The tax administrations can cross-check this information on intra-Community supplies with declarations made by their national traders to identify breaches of VAT legislation.

60.

Economic operators have to provide recapitulative statements of all their intra-Community supplies of goods to customers holding a VAT identification number. The statements cover calendar quarters. According to the regulation, this information is to be entered into VIES as soon as possible and, at the latest, within three months of the end of the calendar quarter to which they relate. This means that even when traders submit their statements on time, information on intra-community supplies made in January may only be available in VIES in June, i.e. almost six months later.

61.

Several Member States already collect information on intra-Community supplies on a monthly basis, and the reduction of timescales for the transmission of information has been discussed repeatedly by the SCAC. However, an agreement amongst Member States on a general application of shorter transmission periods has not been reached.

62.

Each Member State has only access in VIES to transactions involving its own traders. Because of this, whenever Member States, in order to identify possible fraud networks, need information about trade operations between other Member States, they have to send a request for information to the Member States concerned. This takes time to be answered and commits human resources in both Member States.

63.

Article 22 of the regulation requires Member States to ensure that their databases are kept up to date, and are complete and accurate. As discussions in the SCAC show, there is currently no common approach amongst Member States with regard to the question of how to ensure quality and reliability of data in VIES. Although required by Article 22 of the regulation, no criteria have as yet been defined to determine which changes are not pertinent, essential or useful and therefore need not to be made.

64.

Work carried out by National Audit Offices also highlights the problem of reliability of data. As an example, discrepancies in reporting were identified during an audit conducted in parallel by the Supreme Audit Offices of the Czech Republic and of the Slovak Republic (32). Differences between the values of supplies of goods reported by taxable persons from other Member States and the values of acquisitions of goods reported by taxable persons from the Czech Republic or the Slovak Republic were caused, amongst other things, by differing methods of reporting of data on intra-Community supplies and acquisitions in the individual Member States. This involved, for example, the declaration of intra-Community supplies in the supplier's recapitulative statement for a different quarter than for intra-Community acquisitions in the acquirer's VAT return.

65.

A report (33) by the Polish Supreme Chamber of Control drew attention to the fact that the availability of recapitulative statements on intra-Community acquisitions significantly improves the possibility for checks and facilitates the detection of attempted fraud. Although some Member States already collect such data from their traders, only the input in VIES of recapitulative statements on supplies is currently mandatory.

66.

Article 27(4) of the regulation provides that suppliers can obtain confirmation of the validity of the VAT identification number of any specified person. All Member States have their own systems to validate VAT numbers. Depending on the Member States, the requests for confirmation can be made by telephone, by fax or electronically on a website.

67.

Since 2002, the Commission has made a tool available on its website known as ‘VIES VAT number validation’, which enables traders to perform an on-line check of the validity of a VAT number. Eleven Member States allow the display of a trader's name and address when checking the number, while a further three Member States display only the name of the trader. The other thirteen confirm only the validity of a VAT number without providing any information on the trader.

68.

The information provided is not fully reliable due to different interpretations of basic notions such as the start or end of an activity, and due to the fact that some Member States update their databases with retroactive effect. Where such updates with retroactive effect have taken place, it is difficult for a tax administration to verify whether a given VAT number was shown as being valid at the time when transactions under examination took place.

69.

VIES has been in place since 1992. In 2004, it was decided that the system needed an upgrade to be ready for new needs and to incorporate technological improvements. This would result in VIES II, with the new features being gradually added to the existing VIES system.

70.

The feasibility study for VIES II mentioned six areas of development:

71.

Article 11 of the regulation provides a legal basis for designated officials to be present in other Member States, for example in order to have access to documentation held there or to attend enquiries that take place with traders.

72.

Limited use has been made of these possibilities (see Table 4). For 2006, only six Member States reported cases where their officials had been present in administrative offices in other Member States (15 cases in total). Only three Member States reported cases where their officials had participated in administrative enquiries in other Member States (3 cases in total).

Table 4

Use of other instruments

73.

Article 12 of the regulation allows for two or more Member States to conduct simultaneous controls, in their own territory, on the tax situation of one or more operators who are of common or complementary interest. The regulation provides that a Member State shall identify the taxable person whom it intends to propose for a simultaneous control. The other Member States concerned shall then decide whether they wish to take part in the simultaneous controls.

74.

In 2006, a total of 25 (36) such simultaneous controls were reported.

75.

Simultaneous controls can take place in the form of multilateral controls as provided by the decision on the Fiscalis Programme. Under this decision, funding for travel and subsistence expenses is available. The procedure to launch and conduct such controls which typically involve a high need for coordination, has been significantly simplified and clarified (37).

76.

In 2005, a total of 12 such multilateral controls were organised by 5 different Member States. Although this was a higher figure than in the years 2001 to 2004 (38), it falls short of the peak reached in 2000 when 16 multilateral controls took place.

77.

According to the Dutch authorities, such controls can lead to the imposition of significant amounts of additional VAT. In one of the three multilateral controls organised by the Netherlands and finalised in 2005, a further VAT amount of 35 million euro was established. The tax administration in Luxembourg reported (39) a case where a multilateral control conducted over more than three years together with the UK authorities resulted in the successful dismantling of a network behind an 80 million euro fraud with mobile phones.

78.

The regulation attributes to the Commission the role of supporting and promoting administrative cooperation between Member States and to evaluate it together with them. Furthermore, the Commission is required to pool the Member States' experience with the aim of improving cooperation.

79.

In the SCAC, the Commission has made proposals for improving administrative cooperation. Examples are the initiative of the Commission to reduce the timeframes for the production of recapitulative statements of intra-Community supplies or the suggestion to provide more information about traders in VIES. Most of these proposals were not followed.

80.

At the request of the SCAC, the Commission has designed a specific form (Trend form) to indicate developments in the operations of Missing Trader Fraud. However, so far no Member State has used the form to provide information. Likewise, the Commission has not been informed of all bilateral agreements between Member States on matters covered by the regulation.

81.

In its 2006 communication concerning the need to develop a co-ordinated strategy to improve the fight against fiscal fraud (40), the Commission stated that it considered that the moment had come to set up a monitoring system, on the basis of quantifiable indicators, in order to ensure that each Member State is able to and actually does provide efficient assistance to other Member States. However, so far the Commission has not made any concrete legislative proposal for the introduction of such a monitoring system.

82.

Furthermore, three years after the entry into force of the regulation, the organisational set-up for administrative cooperation in the Member States and its compliance with the regulation had not yet been assessed by the Commission.

83.

Pursuant to Article 41(2) of the regulation, access of Commission staff to the information exchanged is limited to what is necessary for maintenance and development of the communication network. Without access to the contents of the operational information exchanged, the Commission cannot gain on its own all the insights necessary to detect the reasons for problems and put forward solutions.

84.

During the course of the audit, a number of other problems were identified which prevent the potential benefits of cooperation from being fully exploited or constrain possible improvements.

85.

The prompt withdrawal of a trader's VAT identification number is an important tool to stop fraudulent activities and to protect honest entrepreneurs.

86.

From an analysis of the relevant Community legislation (41), it can be concluded that traders who do not carry out any real economic activity, but only engage in transactions simulating such activity in the context of VAT fraud are not entitled to have a VAT identification number.

87.

However, Community legislation does not lay down procedures or conditions under which VAT identification numbers must be cancelled, for example in cases where traders are involved both in legal economic activity and in fraudulent operations. Such cases are subject to diverse national legislation and practice.

88.

During its audit missions, the Court was informed that in some of the Member States (42), it is very difficult to prosecute persons involved in fraudulent schemes, for example participating in chains set up for carousel fraud, where only other Member States are defrauded.

89.

The UK National Audit Office (NAO) reported (43) that Danish tax authorities had informed them that they suspected some Danish companies were involved in supply chains with missing traders in the UK. However, there was no tax loss in Denmark and therefore only limited criminal investigations could be undertaken.

90.

The German Federal Court of Auditors reported (44) that no mutual agreements with other Member States had been reached in order to prosecute intra-Community VAT fraud. The existence of such agreements were, according to German legislation, the precondition for prosecuting traders involved in this type of fraud.

91.

The Convention on the protection of the European Communities' financial interests (45) stipulates that each Member State is to establish effective, proportionate and dissuasive criminal penalties in their national law to combat fraud and to cooperate effectively with other Member States. Article 1 of the Convention provides that it also applies in cases where fraud results in the illegal diminution of the resources of the EU budget. However, according to an explanatory report (46) on the Convention adopted by the Council, VAT was excluded from its scope (47).

92.

Action to prevent intra-Community VAT fraud cannot be effectively targeted by Member States, or the Commission, until they have reliable estimates of fraud occurring, including information about the economic sectors most concerned. The Commission has not yet developed a successful common approach on how to achieve this.

93.

In its resolution of 12 December 2006 on promoting activities in the field of value added tax, the Contact Committee (48) mandated a working group to produce recommendations for a single model for estimating VAT losses. Such a common approach would also allow Member States to determine whether the fight against fraud was genuinely successful, or whether tackling VAT fraud just resulted in it being displaced to other economic sectors or other Member States.

94.

Despite new arrangements introduced in 2004, administrative cooperation between Member States in the field of VAT is still not intensive enough to cope with intra-Community VAT evasion and fraud.

95.

So far, most Member States have not fully exploited the possibilities provided by the Community legislation to devolve competence and responsibility for direct contacts with tax authorities in other Member States to their local or regional tax offices (see paragraphs 19 to 23).

96.

The exchange of information upon request, a major instrument of administrative cooperation between Member States, is affected by a high proportion of late replies and by the absence of interim replies. The frequency of late replies varies, however, considerably between Member States (see paragraphs 24 to 31).

97.

Most of the delays in responding to requests occur due to a lack of priority given to requests by local tax offices. Weaknesses in the monitoring systems and/or in the organisational set-up of Central Liaison Offices in some Member States are other reasons which also contribute to delays (see paragraphs 32 to 42).

98.

Evaluations show that cooperation leads to results. Where requests are answered in accordance with the rules, they often allow to reveal or confirm the existence of fraud and may result in the collection of additional VAT. However, there are some concerns that the quality of information exchanged on suspected missing traders is not always sufficient to serve as a basis for successful legal action against the fraudsters (see paragraphs 43 to 49).

99.

The framework for exchanges of information without prior request is not well defined. Broad obligations to inform other Member States have been set, but in practice exchange is limited to certain categories of information which are at the discretion of Member States. Furthermore, information provided spontaneously is not always systematically exploited (see paragraphs 50 to 56).

100.

VIES, the system for information exchange, has serious weaknesses. Delays in collecting and capturing accurate data and problems in correcting wrong data undermine its usefulness. The system contains data on intra-Community supplies only, but not on intra-Community acquisitions. This constrains the possibilities for cross-checks. Though a decision to upgrade the system was taken in 2004, progress in implementing new features is slow (see paragraphs 57 to 70).

101.

Tools available for a multilateral control approach are rarely used, although Community funding is provided through the Fiscalis programme (see paragraphs 71 to 77).

102.

Most of the proposals made by the Commission through the Standing Committee on Administrative Cooperation to improve exchange of information between Member States were not followed (see paragraphs 78 to 83).

103.

A number of other factors prevent the potential benefits of cooperation from being fully exploited. These include absence of common rules for withdrawing VAT numbers, difficulties in cross-border prosecution and insufficient tools for quantifying and analysing VAT fraud (see paragraphs 84 to 93).

104.

For combating intra-Community VAT fraud successfully, higher priority needs to be given by Member States to administrative cooperation, in respect of both the operational information exchanges and their administrative management.

105.

Member States should encourage more direct communication between local inspection staff as an effective way to speed up the exchange of information. It would at the same time help to increase the intensity of cooperation and the quality of the information exchanged.

106.

More efficient monitoring of exchanges of information between Member States is necessary to ensure that problems are swiftly identified and tackled and that each Member State provides effective assistance to other Member States. Improvements in Member States own monitoring of exchanges of information would also be needed.

107.

The procedures for exchanges of information without prior request need to be clarified. Information which is provided spontaneously should be systematically exploited by Member States.

108.

To improve VIES, action should be taken to

109.

The introduction of harmonised rules for withdrawing VAT numbers from traders involved in fraudulent activities should be considered.

110.

Having comparable data on intra-Community VAT evasion would contribute to a better targeted cooperation between Member States. A common approach needs to be developed by the Commission together with the Member States to quantifying and analysing VAT evasion.

111.

It should be considered how to improve cross-border prosecution of intra-Community VAT fraud in the Member States, for example by changing the Council's current interpretation of the Convention on the protection of the European Communities' financial interests which does not include VAT revenue.

112.

With regard to the Commission proposal for an additional horizontal regulation on mutual administrative assistance for the protection of the financial interests of the Community (49), the Court recalls its recommendation (50) that the Commission should make an effort to propose a simplification and consolidation of Community anti-fraud legislation with a view to avoiding duplications and overlapping or contradictory provisions. Existing weaknesses in cooperation between the Commission and the Member States could be addressed in the framework of such an overhaul.