Finance and Accounting
Prague University of Economics and Business Faculty of Finance and Accounting
MASTER THESIS
Analysis of VAT gap in the EU
Author: Kübra Okumuş
Supervisor: Ing. Hana Zídková, Ph.D.
Academic year: 2021
ii
Declaration of Authorship
The author hereby declares that he compiled this thesis independently, using only the listed resources and literature. The thesis has not been used to obtain a different or the same degree.
The author grants to Prague University of Economics and Business permission to reproduce and to distribute copies of this thesis document in whole or in part.
Prague, date
Signature
iii
Acknowledgements
I express my deep gratitude to my supervisor, Ing. Hana Zídková, PhD, for her patient guidance, support, and readiness to help me with any question during the thesis preparation. I am thankful for her time dedicated to me and my study.
Finally, I would like to thank my partner, close friends, and family for their support and encouragement throughout my work.
iv
Abstract
Value-added tax is the most significant indirect tax revenue for the state's income and one of the most critical tax policy tools. This tax spread rapidly to many countries and allowed trade market participants to find ways to avoid the Value Added Tax. This issue makes the governments doubt the reliability of the current Value Added Tax system in countries which is applying VAT as the gap between the VAT expect and actual revenue, nowadays, are going to be higher.
This thesis aims at the possible determinants such as real GDP growth, GDP per capita, GINI, final consumption expenditure of households and nonprofit institutions serving households, final consumption expenditure of households, population, VAT on GDP, number of the VAT rates, standard VAT rate, share of the shadow economy, research and development expenditure, and corruption perception index to examine the VAT gap problem to see the importance of determinants for improving the issue. The research has been done by applying regression analysis on the EU countries in 2010 and 2018.
JEL Classification Keywords
H26
Tax evasion, tax avoidance, determinants of VAT gap
v
TABLE OF CONTENTS
1. Introduction ... 1
2. Literature Review... 6
2.1. VAT Revenues in the EU ... 6
2.2. VAT Structures in the EU Countries... 8
2.3. Definition of VAT Gap ... 11
2.4. Determinants of VAT Gap ... 14
3. Methodology ... 18
3.1. Research Mentions ... 18
3.2. Research Design ... 24
3.3. Selected Variables ... 25
4. Analysis ... 26
5. Results ... 29
6. Potential VAT Gap Increase Due to Coronavirus ... 31
7. Limitations and Opportunities ... 32
8. Conclusion ... 33
Annex ... 37
References ... 55
vi
List of Tables
Table – 1: VAT Gap in the EU-28 Member States between 2018 and 2017 Table – 2: VAT Rates in the EU in 2018
Table – 3: VAT Gap by Member State in 2018 Table – 4: Audit methods in the EU
Table – 5: Candidate Explanatory Variables
Table – 6: Descriptive statistics of independent variables in 2010 Table – 7: Descriptive statistics of independent variables in 2018
Table – 8: Cross-Sectional Analysis Result with the Relative Variables for 2010 Table – 9: Cross-Sectional Analysis Result with the Relative Variables for 2018
List of Figures
Figure – 1: Revenues in the EU between 2000 and 2018 (million EUR) Figure – 2: VAT Gap in the EU-28 Member States between 2018 and 2017 Figure – 3: Tax and VAT Revenue in the EU in 2018 (as % of GDP)
vii
List of Abbreviations
CASE Center for Social and Economic Research CPI Corruption Perception Index
EC European Commission
EU European Union
FAIA Fichier Audit Informatisé AED FEC Fichier des Écritures Comptables
GDP Gross Domestic Product
GNI Gross National Income
MIMIC Multiple Indicator Multiple Cause
MS Member States
MTIC Missing Trader Intra Community
NPISH Non-Profit Institutions Serving Households
OECD Organisation for Economic Cooperation and Development
OLS Ordinary Least Squares
R&D Research and Development SAF-T Standard Audit File-Tax
TAXUD Taxation and Customs Union Directorate
TOR Traditional Own Resources
VAT Value Added Tax
VTTL VAT Total Tax Liability
1
1. Introduction
“Value added tax is the most important event in the evolution of tax structure in the last half of the 20th century.”
Sijbren Cnossen
The basic idea of the Value Added Tax (VAT) appears to own originated with a German businessman, Wilhelm von Siemens, within the early 20th century. There have been other developments in those years made by Thomas S. Adams. However, VAT is fathered within the early 1950s by Maurice Lauré, remained confined to some states until the late 1960s. Today, most countries have a VAT, which raises, on average, about 25 per cent of their tax income (Ebrill, 2001).
VAT is a general, broadly based consumption tax assessed on the value added to goods and services within the European Union. It more or less applies to all, or any goods and services purchased and sold for use or consumption within the economy.
Therefore, VAT usually is not levied on goods sold for export or services sold to customers in other countries. On the other hand, imports are taxed to keep the system fair for European Union producers so that they can compete on equal terms on the European market with suppliers located outside the Union (EC, 2020).
VAT is the most considerable indirect tax revenue for the state’s income. Many countries in the world have a VAT in their tax system: as of April 2001, about 123 had one. They raise revenues, on average, equal to nearly 27 per cent of total tax revenue and over 5 per cent of GDP. Annex – 1 provides information on these VATs, and Annex – 2 presents some comparative data on countries with and without a VAT. Not surprisingly, those countries that have implemented a VAT are relatively more developed as gauged by per capita GDP (Ebrill, 2001).
2
VAT has a significant role in the EU budget. It is an essential source for the EU budget, not only for the public budgets of individual member states. The revenues of the budget of the EU are shown in Figure - 1. They consist of the surplus from the previous year, gross national income or GNI own resource, own traditional resources or TOR, and VAT owns the resource. Additionally, the distribution of funds has shown that VAT has a considerable proportion in Figure – 1. The total revenue generated by the VAT's resource for the EU in 2018 was EUR 17.625 million (11,1 per cent of total revenue).
Figure – 1: Revenues in the EU between 2000 and 2018 (million EUR)
Source: European Commission (financial_report_web.pdf (europa.eu))
The VAT gap is an issue faced by both developed and developing countries (European Commission, 2019). Therefore, the EU Member States lose extremely high VAT revenues annually due to fraud and inadequate tax collection systems. The gap, often referred to as the VAT gap, where this loss can be defined as the difference between expected VAT revenues and VAT actually collected, includes estimates of revenue losses from tax fraud and avoidance and corporate bankruptcy, financial and corporate bankruptcies, and miscalculations. (Lamensch, M., & Ceci, E. 2018).Many other factors impact the VAT gap. European Commission defined the causes of the VAT gap being tax evasion, maladministration, and tax optimisation.
3
The VAT gap is also defined by Zídková as the difference between the theoretical VAT liability and accrued VAT receipts in a respective state and year as follows from Zídková, 2014.
The smallest gaps were found in Sweden (0,7 %), Croatia (3,5 %), and Finland (3,6
%), while the largest were found in Romania (33,8 %), Greece (30,1 %), and Lithuania (25,9 %). Overall, half of the EU-28 Member States had a spot gap greater than 9,2 per cent (see Figure - 2 and Table - 1). The largest nominal gaps were recorded in Italy (EUR 35,4 billion), the UK (EUR 23,5 billion), and Germany (EUR 22,1 billion).
Figure – 2: VAT Gap in the EU-28 Member States between 2018 and 2017 (as a per cent of the VTTL)
Source: Poniatowski & Bonch-Osmolovskiy & Śmietanka, 2020
Therefore, measuring the VAT gap is significant for the country due to the VAT Gap measures. The effectiveness of VAT enforcement and compliance can be assessed in each Member State because it provides an estimate of revenue loss because of fraud and evasion, tax avoidance, bankruptcy, financial insolvencies as well as miscalculations. Quantifying the size of the VAT gap can aid in the development of well-targeted policies and the monitoring of their effectiveness (EC, 2020).
4
Table – 1: VAT Gap in the EU-28 Member States between 2018 and 2017 (as a per cent of the VTTL)
2017 2018 VAT
Gap
MS Revenues VTTL VAT
Gap
VAT Gap (%)
Revenues VTTL VAT Gap
VAT Gap (%)
Change (pp)
BE 29763 33619 3856 11,5 31053 34670 3617 10,4 -1
BG 4664 5313 649 12,2 5097 5711 614 10,8 -1,5
CZ 14703 16694 1991 11,9 16075 18261 2187 12 0
DK 27966 30475 2509 8,2 29121 31369 2248 7,2 -1,1
DE 226582 248382 21800 8,8 235130 257207 22077 8,6 -0,2
EE 2149 2286 137 6 2331 2458 127 5,2 -0,8
IE 13060 14652 1592 10,9 14175 15857 1682 10,6 -0,3
EL 14642 21898 7256 33,1 15288 21858 6570 30,1 -3,1
ES 73970 79003 5033 6,4 77561 82470 4909 6 -0,4
FR 162011 173840 11829 6,8 167618 180406 12788 7,1 0,3
HR 6465 6843 378 5,5 6946 7198 252 3,5 -2
IT 107576 142939 35363 24,7 109333 144772 35439 24,5 -0,3
CY 1765 1859 93 5 1951 2028 77 3,8 -1,2
LV 2164 2512 348 13,9 2449 2705 256 9,5 -4,4
LT 3310 4422 1111 25,1 3522 4754 1232 25,9 0,8
LU 3433 3525 92 2,6 3729 3928 199 5,1 2,5
HU 11729 13564 1835 13,5 12950 14140 1190 8,4 -5,1
MT 810 984 174 17,7 920 1084 164 15,1 -2,5
NL 49833 52329 2496 4,8 52619 54897 2278 4,2 -0,6
AT 28304 30949 2645 8,5 29323 32231 2908 9 0,5
PL 36330 42374 6044 14,3 40411 44862 4451 9,9 -4,3
PT 16810 18872 2062 10,9 17865 19754 1889 9,6 -1,4
RO 11650 17727 6077 34,3 12890 19485 6595 33,8 -0,4
SI 3482 3640 159 4,4 3765 3913 148 3,8 -0,6
SK 5919 7362 1443 19,6 6319 7899 1579 20 0,4
FI 20404 21510 1106 5,1 21364 22171 807 3,6 -1,5
SE 44115 44987 872 1,9 43433 43739 306 0,7 -1,2
UK 162724 184706 21982 11,9 168674 192126 23452 12,2 0,3
Total EU - 28
1086332 1227226 140935 11,5 1131912 1271953 140042 11 -0,5
Median 10.9% 9.2%
Source: Poniatowski & Bonch-Osmolovskiy & Śmietanka, 2020
5
This thesis purposes of examining the effects of the determinants of the VAT gap in 2010 and 2018. This research will improve the problem's solution and enrich the literature on the topic by discussing the importance of the determinants. The chosen determinants are real GDP growth, GDP per capita, GINI, final consumption expenditure of households, final consumption expenditure of households and nonprofit institutions serving households, population, VAT on GDP, number of VAT rate, standard VAT rate, share of shadow economy, research and development expenditure, and corruption perception index, whose impact is usually discussed.
Based on the reasoning above, the purpose of this study is to answer the following research question:
How did the real GDP growth, GDP per capita, GINI, final consumption expenditure of households, final consumption expenditure of households and nonprofit institutions serving households, population, VAT on GDP, number of VAT rate, standard VAT rate, share of shadow economy, research and development expenditure, and corruption perception index influence on VAT gap in 2010 and 2018?
In the first part of the thesis, it will be introduced the subject, and the chapters of the thesis will be explained. Furthermore, the importance, purpose, and ways of examining the matter will be mentioned. In the second part of the thesis, I will review the relevant literature about the subject of my analysis. Then, I will deal with these topics in detail:
the VAT revenues and the VAT structures in the EU countries, the definition of the VAT gap, and the last but not the minor determinants of the VAT gap. This chapter will play a significant role in determining the determinants and research method point of view in the following chapters of the thesis. In the third part, I will be mentioning the literature from the methodology point of view. Therefore, this part will include the methods and other details from the previous studies.
Additionally, I will explain my methodology in the following chapter. Then there will be the selected independent variables and the reason for choosing them in the last methodology chapter. In the fourth part, I will be applying the regression analysis for
6
analysing the subject econometrically. The variables selected in the study aim to see the effect of these variables on the VAT gap. Specific variables are VAT gap as a dependent variable, real GDP growth, GDP per capita, GINI, final consumption expenditure of households, final consumption expenditure of households and nonprofit institutions serving households, population, VAT on GDP, number of the VAT rate, standard VAT rate, the share of the shadow economy, research and development expenditure, and corruption perception index as independent variables. It is crucial to summarise that before solving the VAT gap issue. In the fifth part, I will be presenting and discussing the results. It is also significant that to see the current situation of the problem and subject. Because of this reason, there will be a chapter that will be explaining the potential VAT gap increase due to coronavirus. In the seventh part, I will touch on the limitations, challenges, and opportunities for future studies about this research. Ultimately, in the eighth part, the thesis will be concluded and summarised.
Moreover, used sources and annexes will be listed.
2. Literature Review
This section presents the theoretical frameworks relevant to my study and a review of previous sources on the topic of the VAT gap. To see the previous studies is significant for the understanding importance of the topic point of view, which is behind the research. The section supports the context for the purpose of my study and the reasoning behind my research question. Furthermore, I present the background and justification to my hypotheses.
2.1. VAT Revenues in the EU
The state revenue that consists primarily of tax revenue is the primary mechanism for ensuring economic development. Regarding that, fiscal policy controls the global factors affecting the national market and also the economy. At the same time, it becomes the interior guarantee obtaining the social development of each particular state (Bikas & Andruskaite, 2013).
7
A value added tax is a critical component of the tax system. The value-added tax is one of the essential taxes in terms of budget revenue, comprising the most significant share of the state's tax revenue. In the case of the economy's cyclical fluctuations, the system of value added tax is primarily used to stabilise state revenue and ensure the performance of public functions. The recent economic and financial downturn confirms this. The burden of value added tax imposed by the pricing system is transferred to all consumers, regardless of their legal status, solvency options, or other factors. It is a significant factor influencing the competitiveness of the state. The fact that the VAT is widely spread across different countries supports value added tax efficiency, which involves collecting the possible revenue for public budgets without significantly affecting the economy or one of the industries or groups of consumers.
Because this tax has an impact on the development of the internal market without borders, the procedure for collecting is strictly regulated in the European Union.
Furthermore, value added tax is given special attention to avoid price differences for consumers and opportunities to provide exceptional advantages to a single national market (Bikas & Andruskaite, 2013).
The VAT is a pillar of the EU's tax and economic system. VAT is a significant source of a revenue generator for all governments in the EU because it is one of the most broadly based taxes. With high government deficits and the current economic climate, several governments are looking to increase the share of VAT collected as a proportion of total taxes collected (Lejeune, 2011).
VAT revenues have an essential percentage on GDP and create a significant part of tax revenues in all EU Member States, as shown in Figure – 3 below.
8
Figure – 3: Tax and VAT Revenue in the EU in 2018 (as % of GDP)
Source: own presentation of data based on OECD
2.2. VAT Structures in the EU Countries
VAT is an indirect tax levied on public and private consumption. Despite the fact that the tax burden is shifted to consumers, corporations pay this tax (Bendikienė, Šaparnis, 2006), as manufacturers and service providers include it in their prices (Dilius, Kareivaitė, 2010).
Goods and services are commonly charged with VAT aiming to collect as many funds into the state budget as possible. The VAT administration is not complex and relatively cheap. VAT is a critical element of the tax system, a significant source of revenue in many states. According to the VAT scheme, the tax is levied on added value created in each phase of the manufacturing process so that each stage brings revenue to the state budget (Bendikienė, Šaparnis, 2006). VAT is introduced at the beginning of the manufacturing process and is counted in each product or service production and marketing phase until it reaches the consumer, who pays this tax (Štreimikienė, Mikalauskienė, 2006).
Value added tax does not affect production and distribution and is collected after final consumer prices are applied (Dilius, Kareivaitė, 2010). On the other hand, in terms of
30.51 31.14 30.92 27.61 28.46 27.88 30.00 31.90 26.84 25.46 35.23 30.20 28.14 30.62 32.80 30.52 29.14 28.34 30.18 32.00 29.85 31.48 28.21
11.70 11.07 11.29 14.61 13.75 14.33 12.21 10.32 15.37 16.76 6.98 12.01 14.08 11.59 9.42 11.70 13.07 13.87 12.03 10.21 12.36 10.73 14.01
Tax revenue VAT revenue
9
consumption, people with lower incomes spend a large portion of their income than those with higher incomes. Therefore, VAT is defined as a regressive tax (Jenkins et al., 2006).
In the EU, VAT is based on an invoice or, in other words, tax credit method. In all transactions, first, the seller issues an invoice and charges the output tax to the buyer.
Then this amount of output tax minus the amount of VAT paid by the seller at the prices for goods and services purchased by him, which is an input tax, must be transferred to the tax administration. The basis of this practice is based on voluntary tax compliance, which supports the self-enforcement mechanism (Pomeranz, 2015).
Accordingly, the vendor has incentives to charge the output tax on the sales he had paid regarding his purchases to collect the money from the input taxes. An exception to this rule is the sales of goods and services to final consumers. Since they will not deduct input tax, they will have to deal with specific incentives to avoid paying taxes.
Nevertheless, this type of tax evasion requires that retailers agree not to collect an output tax (Fedeli and Forte, 1999). Hence, they both play a role in the decision to evade taxes.
The VAT system can be clarified by parameters that determine its scope, especially the level of the standard rate and reduced rates, the amount, and types of exemptions.
Also, other parameters are significant for this clarification, such as several administrative provisions regarding how economic agents must behave, such as thresholds for registration as taxpayers, frequency of declarations and payments, rules on cross-border trade. The European Union has endeavoured over the years, following the objectives of the Single Market, to harmonise these parameters with several Directives. The VAT Directive, which was adopted on January 1, 2007, replaces the Sixth Directive, contains all provisions regarding the standard system of VAT. The Directive does not provide a uniform percentage rate for the entire Union but sets limits for the Member States. In the EU VAT Directive (2006/112/EC), for example, it limits the minimum standard rate to 15% and provides for two reduced rates of at least 5%
for the listed goods and services (see Annex – 3). Specific derogations and exceptions are in place for the Member States, implying the existence of exemptions, zero rates, and super-reduced rates (CASE, 2013).
10 Table – 2: VAT Rates in the EU in 2018
Member States Standard Rate Reduced Rates
Austria 20 10/13
Belgium 21 0/6/12
Bulgaria 20 0/5
Croatia 25 0/5/13
Cyprus 19 0/5/9
Czech Republic 21 10/15
Denmark 25 0
Estonia 20 0/9
Finland 24 10/14
France 20 2.1/5.5/10
Germany 19 7
Greece 24 6/13
Hungary 27 5/18
Ireland 23 0/4.8/9/13.5
Italy 22 4/5/10
Latvia 21 5/12
Lithuania 21 5/9
Luxembourg 17 3/8/14
Malta 18 0/5/7
Netherlands 21 6
Poland 23 5/8
Portugal 23 6/13
Romania 19 0/5/9
Slovakia 20 10
Slovenia 22 9.5
Spain 21 4/10
Sweden 25 0/6/12
United Kingdom 20 0/5
Source: own presentation of data based on EC.
11
2.3. Definition of VAT Gap
According to the European Commission, the VAT gap is the overall difference between the expected VAT revenue and the collected amount. The VAT gap is defined as the difference between the VAT amount actually collected and the VAT Total Tax Liability (VTTL), in absolute or percentage terms.
In other words, The VAT gap is the difference between the theoretical VAT liability and accrued VAT receipts in a respective state and year as follows from Zídková, 2014:
(1)
European countries are losing billions of euros in VAT revenues due to tax fraud and inadequate tax collection systems. The VAT Gap calculation (1) estimates revenue loss due to tax fraud, tax evasion, and tax avoidance, but also due to bankruptcy, financial insolvencies, or miscalculations.
European Commission (2019) defined the causes of the VAT gap as below:
- Fraud and tax evasion - Corporate insolvency - Corporate bankruptcy - Maladministration - Tax optimisation
12 Table – 3: VAT Gap by Member State in 2018
Member State VAT Gap % VAT Gap (mil. of €)
Belgium 10,40% 3.617
Bulgaria 10,80% 614
Czechia 12,00% 2.187
Denmark 7,20% 2.248
Germany 8,60% 22.077
Estonia 5,20% 127
Ireland 10,60% 1.682
Greece 30,10% 6.570
Spain 6,00% 4.909
France 7,10% 12.788
Croatia 3,50% 252
Italy 24,50% 35.439
Cyprus 3,80% 77
Latvia 9,50% 256
Lithuania 25,90% 1.232
Luxembourg 5,10% 199
Hungary 8,40% 1.190
Malta 15,10% 164
The Netherlands 4,20% 2.278
Austria 9,00% 2.908
Poland 9,90% 4.451
Portugal 9,60% 1.889
Romania 33,80% 6.595
Slovenia 3,80% 148
Slovakia 20,00% 1.579
Finland 3,60% 807
Sweden 0,70% 306
United Kingdom 12,20% 23.452
Source: EC, “Study and Reports on the VAT Gap in the EU-28 Member States: 2020 Final Report.”
13
The member states act individually to reduce the VAT gap in a number of ways. One of the ways is to strengthen their audit by relying more on electronic auditing and advising or requiring taxpayers to have or submit standard tax audit files in accordance with OECD recommendations. For instance, Portugal and Luxembourg have been doing so since January 2011. In addition, Lithuania and Poland use the standard audit file developed by the OECD and adjusted for tax purposes, SAF-T, for regular VAT lists. Austria, Luxembourg, and France also request SAF-T (E&Y, 2021). In Table – 4, more EU countries’ audit file methods are shown. Another way is to present voluntary compliance programs and tax control frameworks, such as in the Netherlands. Finally, penalties for non-compliance are raised in the United Kingdom.
The creation of the EuroFisc helps EU-wide action and cooperation between member states to fight fraud. As part of its VAT strategy, the European Commission proposes more measures to limit both the VAT gap and the costs of collection and enforcement (Lejeune, 2011).
Table – 4: Audit methods in the EU Member
States Name of the measure Since
Austria SAF-T reporting 2009
Belgium Annual List of Customers 1972 Bulgaria Sales and Purchase Journal 2003 Croatia Form U-RA of received invoices 2019
Czechia VAT Control Statement 2016
Estonia Annex of VAT return 2014
France FEC (SAF-T reporting) 2014
Greece E-invoicing 2020
Hungary
Real-time invoice reporting 2018 (Summary Report for VAT Returns) (2013)
Italy
Real-time e-invoices 2019 (Quarterly Declaration of VAT Sales and
Purchase Invoices) (2011)
14
Latvia Control Statement as an annex of VAT
return 2001
Lithuania i.MAS (SAF-T) reporting 2016
Luxembourg FAIA (SAF-T reporting) 2011
Poland SAF-T –VAT evidence 2016
Portugal SAF-T list of sales and purchase invoices 2013
Romania Form 394 2014
Slovakia National Recapitulative Statement 2014
Spain Form SII 2017
Source: VAT legislation of the EU member states, documents of local tax administrations
2.4. Determinants of VAT Gap
According to Poniatowski & Bonch-Osmolovskiy & Śmietanka, 2020, there is a fundamental problem of ineffective tax collection, especially from a VAT point of view, in the EU. It is a challenge to assess the scale of tax evasion, which probably explains the lack of data published so far. As a result, the study of tax non-compliance determinants is no longer novel in the economic literature. The majority of the literature on such factors focuses solely on personal income taxes, voluntary tax compliance, and deterrence effects. This focus is related to data availability. A limited number of studies looking at such cross-country variations focus on interpreting dynamics in tax revenue (e.g., Aizenman and Jinjarak, 2018) or have a qualitative nature (e.g., Keen and Smith, 2007). Then, as more data sequences become available over a while long enough to cover economic ups and downs, they become more accessible. As a result, the finding provides an opportunity to conduct econometric analyses looking at the determinants of tax non-compliance from a new perspective.
Some researchers such as Barbone et al. (2013), Zídková (2017), Lešnik et al. (2018), Poniatowski et al. (2018 and 2019), Szczypińska (2019), and Carfora et al. (2020) have already used panel data for their VAT gap studies. European Commission's VAT Gap Studies, which were published in 2013, 2014, 2015, 2016, 2017, 2018, and 2019 have categorised variables such as:
15
- One of the categories of tax policy characteristics demonstrates how various tax administration efforts are linked to the VAT gap in each country. The amount spent on tax administration concerning GDP is insufficient to cover how effectively the funds are used. Furthermore, the IT expenditure variable is expected to determine the impact of development processes implemented in administrative procedures. Similarly, the administrative effectiveness variable refers to the tax administration's independence from political pressures and the quality of policy formulation and implementation.
- Another category is macroeconomic variables, which aim to explain cyclical conditions as they affect taxpayer behaviour. For example, consider unemployment, GDP per capita.
- Other variables describe the economy's sectoral and corporate structure. In particular, it is emphasised the retail sector, which could be the key for the shadow economy and tax evasion, and the other labour-intensive industries such as real estate, construction, industry, telecommunications, and art. This model also considers the structure of companies based on employment size and the relative size of the shadow economy.
- Since the variability of tax fraud, a significant component of the VAT Gap, may be linked to particular factors that are not included in the list of covariates.
Three methods are used to estimate the scope of the fraud. As one of the possible indicators of fraud, international trade is immediate changes in intra- Community purchase figures, which would indicate an increasing scale of Missing Trader Intra-Community (MTIC) fraud.
According to the estimation results, GDP growth, general government surplus, IT expenditure, trade at risk, and the shares of the agriculture, communication services, and financial sectors are statistically significant at the 5% level of significance.
Furthermore, GDP, the general government balance, the share of IT expenditure in total tax administration expenditure, or the percentage of risky goods imports in GDP all have a positive relationship with the VAT Gap (Poniatowski & Bonch- Osmolovskiy & Śmietanka, 2020).
16
In another study (Zídková & Pavel, 2016), GDP, standard VAT rate, and the difference between reduced rate, the share of household consumption in GDP are analysed with a regression analysis. According to the results, the research revealed that the rise within the ratio of VAT revenue to GDP causes a discount in the VAT gap. Further findings were that if the standard VAT rate and the difference between the standard and reduced VAT rate increase, the VAT gap grows. These findings were an exciting result with a different result compared to the previously mentioned literature. According to Szczypińska, the number of VAT rates or their spread is no longer a vital determinant of the VAT gap, implying that the Council's proposal to limit the use of reduced VAT rates is difficult to justify. The reform of the tax system indicating a reduction in VAT rates, or their spread may or may not contribute to a reduction in the VAT gap. Agha and Haughton (1996) explained the same issue, stating that the standard VAT rate increases the VAT gap, as discovered and predicted by tax theory. According to Agha and Haughton's approach, the higher the VAT rate, the lower compliance with tax obligations. Similarly, the number of tax rates has a negative impact on VAT payments.
In contrast, VAT revenue increases with the length of the country's VAT operation, and smaller countries have lower levels of tax evasion. As a result of this theory, the reason is that a higher tax burden would most likely discourage people from complying with VAT. At a specific tax rate, the amount of tax saved would be large enough to outweigh the risk of punishment in the event of detection by a tax audit. Finally, the control variable – the share of household consumption in GDP – is causing the VAT gap to widen.
In another Zídková study was examining the determinants of the VAT gap. Although the VAT gap is no longer only precipitated through tax evasion, it is an indicator of this, following the application of a regression analysis of potential variables explaining the VAT gap in the 24-EU Member States in two selected years between 2002 and 2006. Data on the VAT gap was once available; two factors common for each examined year affecting the VAT gap in the surveyed states were discovered. Namely, the final consumption of households and nonprofit organisations in each state has a positive effect on the VAT gap and the share of VAT in GDP, lowering the VAT gap.
17
Furthermore, it was determined that the various identified variables that would provide an explanation for the measurement of the VAT gap were the share of the shadow economy and the standard VAT rate, which had a positive impact, and GDP per capita, the percentage of intra-community trade, final consumption of restaurant and hotel service, and the variety of VAT rates, which had a negative impact (Zídková, 2014).
Kasnauskiene & Krimisieraite (2015) examined the VAT gap in Lithuania, one of the biggest in the EU, to figure out what determinants restrict the country’s ability to mobilise revenue with the aid of the useof the MIMIC model method. According to their MIMIC model method results, two factors familiar with government consumption expenditure and inflation statistically affect the VAT gap in the long run.
This factor on the VAT gap is influenced by changes in general government consumption expenditure and, as a result, changes in inflation. Accordingly, the expansion of changes in overall government consumption expenditure determines the growth of the VAT gap. Because most of these expenditures were allocated to social protection in absolute terms, the authors believe that increasing sponsorship to the current sector determines that people with low income are less interested in working because the profit from work is insufficient to become active labour market participants. Such people spend less money on consumption, which is reducing the opportunity for business growth. In the long run, the VAT gap growth can be explained by changes in inflation growth. Individuals are forced to seek additional sources of income through other means, such as engaging in illegal activities, hiding income, and thus avoiding tax compliance to maintain a minimum equal level of consumption. In the short run, however, the two determinants, inflation and household deposits, have a statistically significant influence on the gap. The short-run results show that only two of the five causal variables statistically impact the VAT gap: inflation and household deposits. When there is an increase in inflation, the VAT gap widens. As a result, it is significant to ensure price stability in the country to maintain at least the equal stage of the VAT gap. It reduces the VAT gap as well as consumer purchasing power fluctuations.
18
According to CASE, 2012, the expected influence of the Corruption Perception Index (CPI) was negative on the VAT gap, which means that the increasing value of the corruption index (positive perception of corruption) reduces tax evasion. In terms of the results of this index within the European Union, the Nordic countries have the highest values for an average of twelve years. Finland has a value of approximately 9,51, followed by Denmark with an average value of 9,46 and Sweden with a mean value of 9,24. Conversely, the lowest measured value was reached by Romania with an average score of 3,23, Bulgaria with a value of around 3,82, and Latvia with an average score of 4,17 (Majerová, 2016).
3. Methodology
This section outlines the empirical methodology used to examine the research question and test my hypotheses. First, I describe the determinants used in previous works, their research models, and their results. Then, I explain the methodology used to collect, format, and process the underlying dataset for the study. Afterwards, I present chosen independent and dependent variables which are used in regression analysis. In order to examine my research question in an appropriate manner, yielding the most insightful and accurate results, this association study has a deductive, quantitative approach.
3.1. Research Mentions
In the previous studies, most of the researchers chose to use the panel data estimation method on this topic, as discussed in the literature part, or other regression analysis techniques. The panel data derived from the VAT Gap Study have already been used by a number of researchers – such as Barbone et al. (2013), Zídková (2017), Lešnik et al. (2018), Poniatowski et al. (2018 and 2019), Szczypińska (2019), and Carfora et al.
(2020), (EC,2020). With the panel data estimation method, users are able to observe the variables pairs of each jurisdiction over different periods, which means that only the variations over time between country pairs are observed. However, differences across these pairs are ignored (Woodridge, 2012). This technique also enables users to control for all unobserved (approximately) time-invariant factors influencing the
19
dependent variable, such as geography, language, climate, and cultural proximity in this study. Moreover, the research design of this thesis is adapted from the ‘Study and Reports on the VAT Gap in the EU-28 Member States - 2020 Final Report’, ‘What Causes the VAT Gap?, Hana Zídková – Jan Pavel’ and ‘What drives the VAT gap in the European Union?, Agnieszka Szczypińska’, in which the authors used the regression model or explained the determinants of the VAT gap (available to see in the table – 5).
Table – 5: Candidate Explanatory Variables
Variable
Underlying Factor Captured by
Variable
Hypothesis of Relationship with VAT Gap
Reason for Including Variable to the
Model/
Authors that used this Variable
Source of Data
Economic variables
GDP per capita
Wealth/Level of development
Decreases Aizenmann and Jinjarak (2005)
Eurostat (national accounts)
GDP Size of
economy Decreases Reckon (2009) Eurostat
Growth of GDP Business cycle Decreases
Sancak, Velloso and Xing (2010), CASE (2013)
Eurostat
Unemployment
Business cycle and income
inequality
Increases CASE (2013) Eurostat
Final
Consumption of Households and
NPISH on GDP
Size of potential VAT base
Increases
CASE (2013), D’Agosto, Marigliani
and Pisani (2014)
Eurostat (national accounts)
Final
Consumption per capita
Size of potential VAT base and
also level of development
Increases with size of tax base/decreases
with level of development
Variable expresses the spending capacity
of the citizens and is not influenced by the size of population
Eurostat (national accounts)
20 Government
Consumption Expenses on GDP
Size of public
sector Decreases
D’Agosto, Marigliani and Pisani (2014),
Reckon (2009)
Eurostat (national accounts)
Household Final
Consumption of Restaurants and
Hotel Services on
Total
Consumption
Proxy for effect of tourism
Decreases
Reckon (2009), Christie and Holzner (2006)
Eurostat (national accounts)
Share of Intra- community Trade-in Total Imports
Exposure to carousel fraud/Opennes
s of economy
Increases if carousel fraud
takes place/
decreases if impact of openness of economy prevails
Aizenmann and Jinjarak (2005), Bodin et al.
(2001)
Eurostat (national accounts)
Value Added in Construction on GDP
Relative size of construction
industry
Decreases Reckon (2009)
Eurostat (national accounts)
Value Added in Agriculture on GDP
Share of
agriculture Increases
Bird, Martinez- Vazquez and Torgler
(2004), Aizenmann and Jinjarak (2005)
Eurostat (national accounts)
Tax variables
Standard VAT
Rate VAT burden Increases
Agha and Haughton (1996), CASE (2013),
Bodin et al. (2001), Reckon (2009)
European Commission
(VAT rates in EU)
VAT Revenues on
GDP
Tax quota (VAT burden)
Increases Agha and Haughton (1996)
European Commission
(Taxation Trends in
EU
21 VAT to Total
Tax Revenues
Significance of VAT in tax
structure
Increases
Variable included as the authors believe that
if VAT is a significant source of the state
budget, the tax authorities might collect it better.
European Commission
(Taxation Trends in
EU)
Tax Quota (total tax revenue incl.
social security) on GDP
Total tax
burden Increases
Aizenmann and Jinjarak
(2005)
European Commission
(Taxation Trends in
EU)
E-filing in VAT Compliance (percentage of VAT
returns filed electronically)
Access of tax administrators
to on-line data
and simplification
for VAT payers
Decreases
Extended E-filing could simplify the work of tax administration and tax compliance for the taxpayers, so it can reduce the tax evasion
or mistakes in tax compliance
OECD (2011)
Number of VAT Rates
Complexity of VAT system/
fiscal policy
Increases due to complexity/
Decrease if impact of more effective taxation
of goods with lower demand elasticity prevails
Christie and Holzner (2006)
European Commission
(VAT rates in EU)
Difference between Standard and Reduced Rate (if multiple reduced rates, then average thereof)
Tax policy and complexity of
VAT system
Increases Bodin et al. (2001)
European Commission
(VAT rates)
22
Social and other
factors
Share of Tertiary Education
Level of
education Decreases?
More educated society would in the opinion
of
the authors are less inclined with tax
evasion and more able to comply
with complicated VAT
rules; also Bodin et al. (2001)
Eurostat
GINI Coefficient
Income
inequality Increases
Bird, Martinez- Vazquez, and Torgler (2004), Christie and Holzner
(2006)
Eurostat (indicators
of life conditions) Share of
Shadow Economy
Significance of shadow
economy
Increases
Bird, Martinez- Vazquez, and Torgler (2004)
Schneider (2012)
Perception of Corruption Index
Level of corruption, confidence of people in public sector
Decreases (higher CPI indicates less corruption in
country)
Bird, Martinez- Vazquez, and Torgler (2004), Christie and Holzner (2006), Reckon (2009)
Transparenc y International
Report
Membership in EU
Application of harmonised
VAT rules
Decreases? Reckon (2009), CASE (2013)
European Commission,
EU Countries
Source: Zídková & Pavel, 2016
In another study, based on their literature review, they analysed potential explanatory variables, classified into the following categories (Szczypińska, 2019):
23 Macroeconomic and demographic factors:
output gap, % GDP (OECD)
GDP (Eurostat)
population size (Eurostat)
share of the shadow economy, % GDP (World Bank)
Gini index (World Bank)
share of small firms /1-9 employees/ (OECD)
trade exchange within/outside the EU, % GDP (Eurostat)
exports discrepancy, % exports (Comext)
share of cashless transactions (ECB)
Institutional factors:
quality of institutions /Corruption Perception Index/ (Transparency International)
Factors related to the design of the tax system and its effectiveness:
weighted average VAT rate (CASE)
standard VAT rate (EC)
the lowest (non-zero) VAT rate (EC)
the number of VAT rates (EC)
spread of VAT rates /difference between the highest and the lowest VAT rate/ (EC)
collection cost /administrative costs for tax administration in relation to net revenue /(OECD)
the complexity of tax forms /time to prepare and pay taxes in hours/ (World Bank)
IT expenditure /share of total IT expenditure in total revenue body revenue/ (OECD)
HR expenditure /share of total human resource management support functions in total expenditure/ (OECD)
In the following part, the analysis method in this thesis will be described.
24
3.2. Research Design
This chapter will be devoted to the methods used in the thesis to conduct the research.
The chosen subject requires the use of the appropriate form: qualitative or quantitative.
In this case, it was decided to apply the quantitative approach to reach the results statistically. The study aims to identify the most relevant determinant in the VAT gap and produce a realistic result with the best approach.
To examine the relationship or vice versa between two or more variable factors, which are determinants of the VAT gap, I would use regression analysis and cross-sectional method. Two types of variables are needed in such research, which is (1) dependent variable, of which value is changed due to changes in the causation variable; (2) an explanatory or independent variable that is known as the cause of the changes in the dependent variable. The variables used in this thesis will be discussed in detail in the following part. Due to the use of multiple explanatory variables in this thesis, it is suitable to apply the cross-sectional analysis method, known as regression analysis, one of the methods. The relationship of determinants can be compared to that between other independent variables in 2010 and 2018. The cross-sectional work has the advantage that it is not so tricky as the panel regression from the econometrical point of view. For the panel regression, there are numerous tests necessary due to the time series of data. If the data do not fulfil those tests, the panel regression could be spurious and could give biased results. That was the reason why I decided on simple cross- sectional regression.
The main tools of the thesis are defining the relations between variables to find reasons for the VAT gap in 2010 and 2018. For the research, the necessary sample data will be collected from CASE, the OECD reports, EU Commission, and other essential institutions' reports, which will be used for this thesis.
The selected variables from the factors mentioned in the research review part will be determined in the next part.
25
3.3. Selected Variables
Dependent Variable: The only dependent variable that is studied in this thesis is the VAT gap. The VAT gap data covered is in the European Commission, "Study and Reports on the VAT Gap in the EU-28 Member States - 2020 Final Report." for the EU countries in 2018. For 2010, the VAT gap data, which is available in the Eurostat data browser, was used. The reason for choosing this independent variable is the fact that it is an approximation of tax evasion and the efficiency of VAT collection. The primary purpose of this thesis is to examine the factors influencing tax evasion, so this variable is satisfactory.
VAT gap, % (Study and Reports on the VAT Gap in the EU-28 Member States:
2020 Final Report and Eurostat)
Explanatory Variables: Selected explanatory variables and their data sources are as listed below
Real GDP growth, % (EC: final report 2020 and Eurostat)
GDP per capita, mil. of EUR, (Eurostat, national accounts)
GINI/ (1-100), (Eurostat, indicators of life conditions)
Final consumption expenditure of households, mil. of EUR, (Eurostat, national accounts)
Final consumption expenditure of households and nonprofit institutions serving households, mil. of EUR, (Eurostat, national accounts)
Population, number of people (Eurostat)
VAT on GDP, % (EC, taxation trends in EU)
Number of VAT Rate, number, (EC, VAT rates)
Standard VAT Rate, % (EC, taxation trends in EU)
Share of the shadow economy, % GDP (Schneider, 2019)
Research and development expenditure, % of GDP (World Bank)
Quality of institutions, CPI/ (1-10), (Transparency International)
26
It is crucial that all of these variables are from different categories in their selection.
Real GDP growth, GDP per capita, final consumption expenditure of households, and final consumption expenditure of households and nonprofit institutions serving households which are macroeconomic variables, aim to explain the cyclical conditions that affect taxpayer behaviour. Population, share shadow economy, GINI index, and CPI aim to describe the economic structure and institutional variables. VAT on GDP and the standard VAT rate is significant to see the effect on the VAT gap from the VAT burden point of view. The number of VAT rates and R&D expenditure has an aim similar to others. The goal of selecting them is to check the relationship between the tax administration variable and the complexity of the VAT system.
4. Analysis
The analysis will pursue the model based on cross-country OLS regression in the years 2010 and 2018. For the research to analyse the influence of independent variables on the dependent variable, the EViews software will be used. The econometric methods used are based on Zídková, 2014. In the model, the dependent variable in the analysis will be the VAT gap. The independent variables will also be entered in the same order in which they were listed in the last part of table – 5. They were chosen for the analysis based on the literature summarised above and their own considerations. Also, it contains a brief description of the reason for their inclusion and sources from which they were obtained. The chosen explanatory variables are mainly inspired by Zídková, 2014, and Poniatowski & Bonch-Osmolovskiy & Śmietanka, 2020.
In Table – 6 and Table - 7, descriptive statistics are shown below.
27
Table – 6: Descriptive statistics of independent variables in 2010
Mean Median Maximum Minimum Std. Dev.
Real GDP
growth 1,80 1,85 6,00 -5,50 2,83
GDP per
capita 25.309,62 23.220,00 79.160,00 5.050,00 15.873,58
GINI 29,61 29,55 37,00 23,80 3,80
Final consumption exp. of
households 273.158,60 102.953,30 1.348.202,00 4.512,60 399.965,50 Final
consumption exp. of households
and NISH 279.337,90 107.398,30 1.413.207,00 3.994,30 413.747,10 Population 19.155.717,00 9.677.503,00 81.802.257,00 414.027,00 23.614.227,00
VAT on GDP 7,33 7,30 9,50 5,20 1,06
Number of
VAT rate 2,81 3,00 4,00 2,00 0,80
Standard
VAT rate 20,16 20,00 25,00 15,00 2,39
Share of shadow
economy 19,31 18,30 32,60 8,20 7,37
R&D exp. 1,58 1,52 3,73 0,46 0,90
CPI 6,30 6,25 9,30 3,50 1,92
Source: own calculation
Table – 7: Descriptive statistics of independent variables in 2018
Mean Median Maximum Minimum Std. Dev.
Real GDP
growth 3,21 2,95 9,00 0,90 1,80
GDP per
capita 28.093,08 23.300,00 83.470,00 6.550,00 17.527,86
GINI 29,85 28,80 39,60 20,90 4,40
Final con.
exp. of
households 330.312,40 131.081,70 1.670.779,00 6.860,40 484.150,90 Final con.
exp. of households
and NISH 336.784,50 127.458,70 1.755.393,00 5.835,70 501.762,40 Population 19.519.708,00 9.949.307,00 82.792.351,00 475.701,00 24.258.143,00
VAT on GDP 7,65 7,60 9,70 4,30 1,28
