In the past 50–60 years, VAT has become one of the main sources of revenue for many countries, including Estonia. According to the 2011 state budget of the Republic of Estonia, the planned income from VAT is 1.3 billion euros for the year. *1 There is currently no real alternative to VAT that would replace the revenues received from VAT and could be collected as easily. *2 The administration of this tax is relatively easy, since a large burden in tax collection has been placed on persons involved in business who sell goods and services to consumers. For instance, there is no need to conduct complicated and controversial evaluations as in the case of taxation of assets. Some advantages can be highlighted also in comparison to the taxation of income.
However, when turnover tax is applied as a value added tax, several risks emerge, and the literature has pointed out that such a system is susceptible to fiscal fraud. *3 The reason for this lies in the fact that applying the value added system increases the incidence of tax loss for the state and, at the same time, fiscal fraud is hard to detect. With this increase, as a result of unlawful activity, a situation may arise wherein a person reduces his tax liability or creates excess payment by means of unjustified deduction of input VAT. This is a peculiarity of the technical application of VAT. *4 The detection of VAT fraud, however, is complicated by the fact that it is impossible for the state to check all transactions, as well as the fact that cross-border transactions make the identification of circumstances difficult and time-consuming for the tax authority.
This article focuses on the issues of fiscal fraud, closing of transactions, and good faith. The objective of the article is to analyse Estonian legislation and the practice of the Supreme Court and its development in the past 10 years, and to compare it to the practice of the European Court of Justice.
There are no data on the precise scale and percentage of fiscal fraud. On 31 May 2006, the European Commission issued a Communication concerning the need to develop a co‑ordinated strategy to improve the fight against fiscal fraud (COM(2006) 254) *5 , which pointed out, among other things, that ‘on the other hand, there are only very few estimates available on the amount of taxes not collected due to fiscal fraud. In general, the economic literature mentions that fraud accounts for approximately 2 to 2.5% of GDP. Therefore, the interests at stake in the case of fiscal fraud are extremely important’. This is echoed in the 31.5.2006 memorandum 06/221 *6 , but the latter cites German and British data, according to which the uncollected amount of VAT is estimated at 17 billion euros per year in Germany (that number includes carousel fraud as well as the black economy, unjustified deductions of input VAT, and bankruptcies). In the UK, the uncollected portion of the total VAT income due is estimated at 13.5% and the rate of carousel fraud at 1.5–3 billion euros per year. That is 1.5–2.5% of the total VAT intake due. The Communication does not cite economic literature in more detail. One must probably be content with such numbers, given that any data on the black economy, fiscal fraud, and other unlawful activity can only be estimated. *7
There are no reliable sources or analyses in Estonia considering the basis on which to assess the percentage of VAT fraud. In relation to the fuel sector, the amount of VAT from the sale of liquid fuel not collected for the state budget in 2010 has recently been estimated at 732 million kroons (i.e., 46.78 million euros). *8 According to the tax authority, the checks conducted in 2010 determined that persons left a total of 63 million euros’ (980 million kroons’) worth of taxes undeclared and unpaid in 2010. Of that, 70% was accounted for by VAT-related infringements, and the tax authority specifically mentions the use of fictitious invoices as the greatest problem. *9
The simplest type of fraud is the underreporting of supply from sales, which is most commonplace in fields where services can be provided or goods sold for cash. That applies, above all, to small enterprises, and the overall importance of this type of fraud is small. In connection to the European Union, one could mention abuse of the 0% tax rate, in which goods are transferred to another Member State in such a way that they are left untaxed in the country of destination. The deduction of input VAT from non-business-related expenses or expenses related to supply exempt from VAT can also be regarded as fiscal fraud. This article deals primarily with fictitious transactions, as the most widespread fraud in Estonia. With such types of fraud, shadow organisations are used in such a way that the latter organisations issue invoices while the goods or services are actually provided by another person, or invoices do not correspond to any supply of goods or services at all. Transaction chains are sometimes used, consisting of actual organisations (so-called buffers) that make it more difficult to detect the fraud. In this respect, there are similarities to carousel fraud. *10 It can be said right away that carousel fraud has not been dealt with by the Administrative Law Chamber of the Supreme Court of Estonia and hence will also not be addressed further in this article.
The amendment and improvement of tax laws is an ongoing process, because persons are always motivated in the payment of taxes to find discordances or contradictions to use to their advantage. That applies to both legal and illegal tax planning, areas in which regulators often enforce additional rules to fight against fiscal fraud and ensure better receipt of taxes. The main code of relevance here in the field of VAT concerns the reverse charge, but legal regulation of due diligence and its fulfilment has also been attempted in various countries. Furthermore, new options for improving VAT intake are being sought.
As mentioned above, one of the measures often used in fighting fiscal fraud is the reverse charge. With the reverse charge, the liability rests fully with another person and the taxpayer is relieved of all obligations (as well as responsibility) related to the calculation and payment of tax. *11 The taxation of timber was carried out in this manner for a long time in Estonia. Provisions concerning the reverse charge for timber were introduced in the earlier Value Added Tax Act (käibemaksuseadus) with the law that was passed on 17.11.1999 and entered into force on 1.1.2000. In the law effective at that time, regulation was distributed among three provisions: §10 (4), §15 (3)1, and §18 (3)1 of the Value Added Tax Act. These provisions were amended on numerous occasions, with the amendments typically specifying the scope of the object of taxation. *12 On 13.6.2001, the Riigikogu passed the new Value Added Tax Act, which entered into force on 1.1.2002. *13 That law also prescribed the reverse charge for various types of timber (in its §32). In connection with accession to the European Union, yet another Value Added Tax Act entered into force in Estonia, on 1.4.2004 *14 , this time without the reverse charge provisions.
However, reverse charge provisions have been reintroduced to the relevant law by now, and, analysing the reasons behind this move, one sees that it is clearly intended to fight fiscal fraud. *15 Pursuant to §411 of the Amendment Act of the Value Added Tax Act *16 that entered into force on 1.1.2011, immovables *17 and waste metal *18 are subject to reverse charge in accordance with Council of the European Union Directive 2006/112/EC (hereinafter referred to as the ‘VAT Directive’) Article 199 (1).
Although the reverse charge has been considered an effective tool *19 , it is hard to justify its use in certain fields only. Looking at Article 199 (1) of the VAT Directive, one sees that these fields can be considered special in terms of tax risks, but the content of Article 199 (1) is far from perfect. It is time to consider enforcing the reverse charge in as widespread a manner as possible above a certain margin rate for transactions throughout the European Union. *20
In addition to the reverse charge, setting in place legal rules for distinguishing between honest and dishonest VAT-payers may be considered. This concerns the provision and regulation of due diligence in the Value Added Tax Act, on which judicial practice has largely focused so far (more on this below). An example of this is §25 d of the German Value Added Tax Act *21 , pursuant to which the buyer acquires a solidary liability with the seller if he or she is aware, or should have been aware, that the seller was unwilling to pay VAT. A somewhat similar regulation has been in effect in the UK with regard to electronic devices. *22 It is noteworthy that in Germany, a person’s obligation of due diligence (i.e., the person knew or should have known) depends on whether that person concluded the transaction below or at market price. If the person has concluded the transaction below market price and is unable to prove that the price formation is economically justified, circumstances may lead to the conclusion that he has acted in bad faith. *23 Focusing on the price in this case indicates the desire to make it harder to use so-called carousel fraud, wherein goods become cheaper in the sales chain, and that should make the buyer suspicious, according to the regulator. The literature takes a critical stance toward the regulation, since the fact that the person knew or should have known about the intentions of the seller of the goods or the supplier of the service cannot create a tax liability, because unlawful behaviour would have to serve as a ground for imputation (the German Zurechnungsgrund) for the creation of tax liability. In essence, there is an opinion that the fact that a person knew or should have known of another person’s fiscal fraud cannot serve as the basis for liability for non-receipt of taxes. The prevailing opinion in the literature is, therefore, that the provision in this form contradicts the principle of prohibition of arbitrary exercise of state authority (German verfassungsrechtliches Willkürverbot) prescribed in the Constitution. *24
In Estonia, the regulator has not enforced separate regulations with regard to due diligence or solidary liability. One of the reasons might be that judicial practice does not point up the use of carousel schemes or their detection by the tax authority. Secondly, it can be argued that the enforcement of such regulations is problematic, since the evidence and circumstances that occur in practice are extremely varied and typically a set of evidence is evaluated (more on this below). Furthermore, criteria for assessing persons’ involvement in a fraud, the good faith of buyers, and ordinary business care have been developed in the practice of the European Court of Justice as well as that of several Member States and, accordingly, legal regulation would be more or less redundant.
One of the most common measures that is also starting to be used in Estonia is the institution of securities. Institution of securities can take place in different ways (e.g., for start-up entrepreneurs). In connection with the above-mentioned fiscal fraud in the fuel sector, a separate regulation has been enforced in Estonia for fuel-sellers, who need to submit a security to the tax authority. *25 As a result of these amendments, the security required of a fuel-seller upon authorisation of fuel for consumption is one million euros. In the case of fuel already authorised for consumption, the security is 100,000 euros. *26 The purpose of the amendments, which entered into force on 1.4.2011, is to prevent situations in which shadow organisations that do not fulfil their tax liabilities operate in the fuel market. The application of securities to absolutely all entrepreneurs would be unrealistic, because that would start to restrict free enterprise, which is why it can only be applied in select areas.
Given that fiscal frauds are a serious problem, the VAT Green Paper mentions additional four models *27 for improving VAT receipt. Even though such models have not even been discussed in Estonia yet, they deserve to be mentioned here. According to the first model, the payment for goods or services would take place through the bank, and the bank would divide the payment into two parts (the price of the goods or services and the amount of tax), of which the amount of VAT would be paid directly to the tax authority. With the second scheme, a central database is created for invoices taxed with VAT that the tax authority can monitor in real time. Thirdly, separate VAT databases are used that are very easy for the tax authority to access in order to check invoices. Certification is seen as a fourth solution. Aside from the reverse charge, the first solution is considered the most effective; however, it would require a substantial transformation of the existing system. The other three models might improve supervision but do not directly ensure a better receipt of taxes.
Fiscal frauds become evident in judicial practice first because the regulator typically responds to problems with a certain delay. That is why the primary burden in resolving the problems that have occurred in practice lies on judicial practice. Estonia is no exception: numerous cases of fiscal fraud, as well as cases concerning fictitious transactions and good faith of buyers, have been resolved in Supreme Court practice in the last 10 years. Where bad faith or failure to exercise due diligence is concerned, different compositions have explicitly been brought out in Supreme Court practice in a judgment from 5.5.2010, with regard to administrative matter 3-3-1-18-10. In light of that judgment, both the tax authority and the courts need to determine whether the fact that the buyer is involved in fiscal fraud or that the buyer knew that the seller was not the actual seller or the conclusion that the buyer must have known that the seller was not the actual seller serves as a factual basis for tax liability (paragraph 10 of the judgment). Next, I will address three different compositions on the basis of the above-mentioned decision. Although the Supreme Court of Estonia has reached dozens of decisions in such cases, I will refer to only the most important of these in this article.
Starting from the most extreme option, this refers to a situation wherein the buyer is involved or directly participates in fiscal fraud. By fiscal fraud, I do not mean the submission of false data under the Penal Code so much as VAT fraud in the sense accepted in tax law that can also be described as an abuse where the deduction of input VAT is ruled out. *28
As regards participation in fiscal fraud, the following opinion expressed in a 5 May 2003 judgment of the Administrative Law Chamber of the Supreme Court, on administrative matter 3-3-1-39-03, deserves to be quoted at length: ‘Participation in fiscal fraud can mean, above all, that the persons organising the economic activity of the other party to the transaction work to the instructions of, under the control of, in prior agreement with, or with the knowledge of the board members of the company seeking the VAT refund. Participation in fiscal fraud can mean, for example, that the amount of money paid by the buyer to the seller as VAT is returned to the buyer or persons related to the buyer—i.e., the buyer benefits economically from the fiscal fraud. Various circumstances may point to the involvement of the buyer in fiscal fraud, such as close family, work, or business relations between the board members of the buyer and persons who have the right of representation of, or who are responsible for conducting the economic activity of, the seller violating the tax law. Information on who has taken possession of the money paid by the buyer to the seller may also indicate involvement of the buyer in fiscal fraud’ (paragraph 12 of the judgment).
Given that the above-mentioned decision lists diverse options to which it is difficult to add anything, that decision has been cited in Supreme Court practice on numerous occasions. Essentially, the Supreme Court judgment describes the work of the classic ‘invoice mill’ in practice wherein payments made on the basis of fictitious invoices move back to the maker of the payment through a dummy. At the same time, the approach even allows one to presume involvement in fiscal fraud in some cases. If there is a close professional relationship between the persons involved, it may be inferred, from the Supreme Court’s perspective, that it connects these persons to each other. However, the author of this article believes that this must not be overstated, given that, regardless of personal relations, the buyer might be able to prove that he has not participated in fiscal fraud. *29
In addition to the previously mentioned possibilities for a person being involved in fiscal fraud, a situation could arise wherein that person in fact lacks the goods or services. In this case, it may be presumed that the person is submitting false information on purpose and wants to get an unjustified right to deduct input VAT. If there are no goods or services, deduction of input VAT can be ruled out automatically, and in that case it is, in fact, immaterial whether the person participated in fiscal fraud or has exercised sufficient due diligence; in essence, this can be equated with the perpetration of fiscal fraud. Given that the receipt of goods or the supply of service to a taxable person that would serve as the condition for the deduction of input VAT is absent, the person lacks the right to deduct input VAT.
In the deduction of input VAT, objectively and subjectively judged circumstances on which the deduction of input VAT is based can be distinguished. The presence of goods or services and also the existence of an invoice and meeting of the requirements set for invoices are objective circumstances that normally ensure that the person in question may deduct input VAT. However, from judicial practice, subjective elements on which basis the right to deduct input VAT is assessed are additionally derived—involving the taxable person’s awareness of, or obligation to be aware of, circumstances that would indicate the unlawful behaviour of the seller. In that respect, Estonian judicial practice conforms to the practice of the European Court of Justice (in the so-called Kittel case), in which the person’s awareness and obligation of awareness of the unlawful behaviour of the seller are similarly distinguished. *30 The Supreme Court of Estonia (both the Administrative Law Chamber and the Criminal Chamber) has so far issued, in total, 18 decisions that address due diligence in the deduction of input VAT, and 14 decisions can be cited that address the same issue in relation to the imposition of VAT. *31 Next, I will consider the most important cases among these.
In the above-mentioned Kittel judgment, the European Court of Justice has explained: ‘In the same way, a taxable person who knew or should have known that, by his purchase, he was taking part in a transaction connected with fraudulent evasion of VAT must, for the purposes of the Sixth Directive, be regarded as a participant in that fraud, irrespective of whether or not he profited by the resale of the goods’ (paragraph 56 of the judgment). The same opinion has been expressed on numerous occasions by the Supreme Court in its practice. *32 Indeed, if the issuer of the invoice and the transferor of the goods or services do not coincide and the buyer knows this, the person can also not have the right to deduct input VAT. Otherwise the correspondence of VAT would not be guaranteed.
The Administrative Law Chamber of the Supreme Court has further explained, in its 19.5.2009 judgment on administrative matter 3-3-1-32-09, that ‘if an absence of actual economic activity by sellers is detected, the bad faith of the buyer is confirmed already if the buyer knew or should have known that the sellers in fact lacked economic activity. Being aware of that, the buyer should also have known that the seller could not have provided a service to him’. On the basis of this conclusion, a case can be constructed wherein, with a stable transaction partner, another person’s invoice is delivered for the same service or goods at one point, which is how the buyer learns that a so-called shadow organisation is used for issuing invoices while the service is still provided or the goods transferred by the seller known to him from before. Although the buyer is not directly involved in fiscal fraud and does not in any way benefit from it, it follows from judicial practice that the buyer cannot in this case deduct input VAT. The question already addressed earlier (in Section 4.2 of this article)—namely that of the basis on which the buyer develops a greater tax liability by being aware of someone else’s unlawful behaviour—arises here.
It must be said that cases in which it has been concluded that the buyer participated in the fiscal fraud or should have known that the seller is not the real seller are predominant in the practice of the Supreme Court in this field. There are virtually no cases where it has been determined that the person knew that the seller is not the real seller.
It follows from the above-mentioned Kittel case that even if a person does not participate in fiscal fraud and is also unaware of the fiscal fraud committed by the seller, he may develop tax liability (he cannot deduct input VAT). The person would gain such awareness if he exercised the due diligence required of him. A 20.1.2010 judgment of the Administrative Law Chamber of the Supreme Court, concerning administrative matter 3-3-1-74-09, states, among other things, that the ‘should have known’ condition means that the taxable person had information on circumstances indicating that the seller is not the real seller (paragraph 11 of the judgment). That view has been specified in judgment 3-3-1-18-10 of the Supreme Court, dating from 5.5.2010: ‘The conclusion that the taxable person should have known that the seller was not the real seller can be drawn if the taxable person had information on circumstances that indicate that the seller is not the real seller or if the taxable person could have learned of such circumstances by exercising due diligence’ (paragraph 15).
Defining due diligence, or ordinary business care, is extremely problematic. *33 Typically, tax law does not consider circumstances outside tax law to be relevant and instead relies on the actual situation. In the evaluation of economic activity, whether the person has all of the required permits, licences, and registrations is, therefore, irrelevant. Similarly, even the question of whether the person’s activity is lawful or unlawful is irrelevant where taxation is concerned. Taxation is value-neutral, meaning that the activity can even be contrary to good morals while the payment and calculation of taxes nevertheless functions as normal. With due diligence, however, it is precisely the question of whether and to what extent the seller fulfils these requirements derived from other fields of law that is strictly pursued. In Estonian judicial practice, observing the rules for preventing money laundering has been considered important in the case of cash payments *34 , while the importance of having certificates of conformity has been emphasised in the case of fuel. *35 The Supreme Court has indeed pointed out that the person thereby only acquires an additional burden of proof *36 , but in practical terms it still means that the taxable person must carefully follow laws outside the scope of tax law. Given that the limitation period for imposing taxes is six years at most *37 , it may prove virtually impossible to submit additional evidence later.
It should be pointed out that making a distinction between different compositions is significant not only in terms of the imposition of VAT. The imposition of income tax directly depends on the evaluation of circumstances. It has been repeatedly stressed in the practice of the Supreme Court that the burden of proof is much lighter in the accounting of income tax than in the accounting of VAT. The Supreme Court has noted that the fulfilment of due diligence requirements in connection with identifying the identity of the seller is irrelevant in the context of income tax as the identity of the seller is irrelevant in identifying the VAT obligation of the buyer. *38 This distinction results from the fact that the scope of due diligence and the consequences of failure to meet these obligations differ between the cases of income tax and VAT. *39 If the buyer knew that the seller was not the real seller, the tax authority would be entitled to tax the compensation paid to said seller as a non-business-related expense. *40
Analysis shows that the practice of the Supreme Court of Estonia conforms to the practice of the European Court of Justice. Similarly to the practice of the European Court of Justice, it rules out deduction of input VAT for buyers who participate or are otherwise involved in fiscal fraud. Deduction of input VAT is ruled out also if the buyer knew or should have known that the seller is not the real seller.
Analysing the situation as a whole, one finds this not to be the clearest and most reasonable solution for entrepreneurs. Interpreting due diligence is a creative activity that depends on the particular economic field and the scope of the person’s activity, which is why it is extremely difficult for persons to assess their risks. The situation becomes more complicated for those entrepreneurs who wish to pursue economic activity in another Member State. Although they can draw general conclusions on how to act with regard to the VAT system, due diligence may depend greatly on each country’s legal order and practice. As a result, it is very difficult for entrepreneurs to operate in different European Union Member States. Often entrepreneurs can learn of the exact content of their due diligence only in the course of judicial proceedings, but that might entail tax liability. Secondly, the fact that the buyer’s tax liability depends on whether he knew or should have known of the seller’s unlawful behaviour is controversial. It is especially problematic if the idea is expressed that the person should have been aware of these circumstances through the exercise of due diligence. In that case, the fundamental situation is that bad faith is presumed, which is unjustified in the opinion of the author of this article and imposes unreasonable obligations on taxable persons, forcing them to perform the work of the tax authority in order to check their transaction partners.
The situation could be resolved by means of the reverse charge, which should be more extensive and not be based on a select few areas of activity. Rather, it should occur above a certain taxable value. The fight against VAT fraud has so far been characterised by measures applied to only specific areas, and, as a result, their effect is restricted. A more extensive reverse charge would rule out the enforcement of a legal regulation that is used to regulate the presumption of good faith and the bases for buyer liability. The application of the reverse charge is currently restricted by the VAT Directive, whose Article 199 restricts the reverse charge to specific areas. The reverse charge would be much more effective if due diligence were legally regulated and the bases for restricting the deduction of input VAT were to be enforced. The reverse charge would also be more effective or easier to apply than other models proposed by the European Commission to date.