Download to read offline
![The views and opinions expressed in this report are not necessarily shared by the
European Commission or CASE Network, nor does the report anticipate decisions taken
by the European Commission.
Keywords:
consumption taxation, VAT, tax fraud, tax evasion, tax avoidance, tax gap,
tax non-compliance, policy gap
JEL Codes:
H24, H26
© CASE – Center for Social and Economic Research, Warsaw, 2020
Graphic Design:
Katarzyna Godyń-Skoczylas | grafo-mania
ISBN: 978-83-7178-703-4
Publisher:
CASE – Center for Social and Economic Research
Al. Jana Pawla II 61, office 212, 01-031 Warsaw, Poland
tel.: (48 22) 206 29 00, fax: (48 22) 206 29 01
e-mail: case@case-research.eu
www.case-research.eu
This report was commissioned by the Directorate General for Taxation and Customs Union
(TAXUD) of the European Commission under project No. TAXUD/2019/AO-14, and written
by a team of experts from CASE – Center for Social and Economic Research (Warsaw)
directed by Grzegorz Poniatowski, and composed of Mikhail Bonch-Osmolovskiy and Adam
Śmietanka. The Project was coordinated by Roberto Zavatta (Economisti Associati, Bologna).
It remains the property of TAXUD.
We acknowledge valuable comments from reviewers, Hana Zídková and Michael Udell. We
also acknowledge discussions with several officials of tax and statistical offices of the Member
States, who offered valuable information, comments, and suggestions. All responsibility for
the estimates and the interpretation in this Report remains with the authors.
Adam Śmietanka
No. 503 (2020)
[editorial page]
Acknowledgments
This report was commissioned by the Directorate General for TaxaRon and Customs Union (TAXUD) of
the European Commission under project No. TAXUD/2019/AO-14, and wri[en by a team of experts
from CASE (Center for Social and Economic Research, Warsaw) directed by Grzegorz Poniatowski, and
composed of Mikhail Bonch-Osmolovskiy and Adam Śmietanka. The Project was coordinated by
Roberto Zava[a (EconomisR AssociaR, Bologna). It remains the property of TAXUD.
We acknowledge valuable comments from reviewers, Hana Zídková and Michael Udell. We also
acknowledge discussions with several officials of tax and staRsRcal offices of the Member States, who
offered valuable informaRon, comments, and suggesRons. All responsibility for the esRmates and the
interpretaRon in this Report remains with the authors.
The views and opinions expressed in this report are not necessarily shared by the European
Commission or CASE Network, nor does the report anRcipate decisions taken by the European
Commission.
Keywords: consumpRon taxaRon, VAT, tax fraud, tax evasion, tax avoidance, tax gap, tax non-
compliance, policy gap
JEL codes: H24, H26
© CASE – Center for Social and Economic Research, Warsaw, 2020
Graphic Design: ….](https://image.slidesharecdn.com/reportno503-201207153322/75/Study-and-Reports-on-the-VAT-Gap-in-the-EU-28-Member-States-2020-Final-Report-2-2048.jpg)
![CASE Working Paper | No 1 (2015)
5
List of Tables
Table 1.1. Real and Nominal Growth in the EU-28 in 2018
(in national currencies [NAC]). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Table 1.2. VAT Rate Structure as of 31 December 2017
and Changes during 2018 (%). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Table 1.3. Change in VAT Revenue Components (2018 over 2017). . . . . . . . . . 21
Table 2.1. VAT Gap as a percent of the VTTL
in EU-28 Member States, 2018 and 2017. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Table 3.1. Belgium: VAT Revenue, VTTL, Composition of VTTL,
and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Table 3.2. Bulgaria: VAT Revenue, VTTL, Composition of VTTL,
and VAT Gap, 2014–2018 (BGN million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Table 3.3. Czechia: VAT Revenue, VTTL, Composition of VTTL,
and VAT Gap, 2014–2018 (CZK million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Table 3.4. Denmark: VAT Revenue, VTTL, Composition of VTTL,
and VAT Gap, 2014–2018 (DKK million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Table 3.5. Germany: VAT Revenue, VTTL, Composition of VTTL,
and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Table 3.6. Estonia: VAT Revenue, VTTL, Composition of VTTL,
and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Table 3.7. Ireland: VAT Revenue, VTTL, Composition of VTTL,
and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Table 3.8. Greece: VAT Revenue, VTTL, Composition of VTTL,
and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37](https://image.slidesharecdn.com/reportno503-201207153322/75/Study-and-Reports-on-the-VAT-Gap-in-the-EU-28-Member-States-2020-Final-Report-5-2048.jpg)
![17
CASE Reports | No. 503 (2020)
Table 1.1. Real and Nominal Growth in the EU-28 in 2018 (in national currencies [NAC])
Source: Eurostat.
Member State
Real GDP
Growth (%)
General
Government
Balance (%)
Change in
Unemployment
Rate (pp)
Nominal Growth (%)
GDP
Final
Consumption
GFCF
Belgium 1.5 −0.8 −1.1 3.0 3.3 6.2
Bulgaria 3.1 2.0 −1.0 7.2 7.7 9.7
Czechia 2.8 0.9 −0.7 5.5 6.6 9.1
Denmark 2.4 0.7 −0.7 3.3 3.0 7.3
Germany 1.5 1.9 −0.4 3.1 2.9 6.3
Estonia 4.8 −0.6 −0.4 9.5 8.1 5.3
Ireland 8.2 0.1 −0.9 9.1 6.0 −18.7
Greece 1.9 1.0 −2.2 2.5 0.9 −12.0
Spain 2.4 −2.5 −1.9 3.5 3.4 7.7
France 1.8 −2.3 −0.4 2.8 2.2 4.6
Croatia 2.7 0.2 −2.7 4.5 4.5 4.7
Italy 0.8 −2.2 −0.6 1.7 2.0 3.8
Cyprus 4.1 −3.7 −2.7 5.5 5.0 −4.5
Latvia 4.3 −0.8 −1.3 8.4 7.3 18.0
Lithuania 3.6 0.6 −0.9 7.1 6.8 10.1
Luxembourg 3.1 3.1 0.1 5.7 6.1 −5.3
Hungary 5.1 −2.1 −0.5 9.9 7.6 24.4
Malta 7.3 1.9 −0.3 9.5 10.2 0.8
Netherlands 2.4 1.4 −1.1 4.9 4.6 6.3
Austria 2.4 0.2 −0.6 4.2 3.3 6.0
Poland 5.3 −0.2 −1.0 6.6 6.4 10.8
Portugal 2.6 −0.4 −1.9 4.3 3.9 9.0
Romania 4.4 −2.9 −0.7 11.0 13.2 3.9
Slovenia 4.1 0.7 −1.5 6.4 5.4 11.4
Slovakia 3.9 −1.0 −1.6 6.0 6.0 4.9
Finland 1.5 −0.9 −1.2 3.4 3.1 6.6
Sweden 2.0 0.8 −0.3 4.4 4.4 4.6
United Kingdom 1.3 −2.2 −0.3 3.5 3.8 1.6
EU-28 (EUR) 2.0 −0.7 −0.9 3.3 3.1 4.2](https://image.slidesharecdn.com/reportno503-201207153322/75/Study-and-Reports-on-the-VAT-Gap-in-the-EU-28-Member-States-2020-Final-Report-17-2048.jpg)
![CASE Reports | No. 503 (2020)
20
c. Sources of Change in VAT Revenue Components
The value of the actual VAT revenue can be decomposed into components, which is helpful
in understanding the underlying sources of its evolution. Since revenue is a product of the
VTTL and the compliance ratio4
, VAT collection could be expressed as:
Actual Revenue = VTTL × Compliance Ratio,
where Compliance Ratio is: 1 – VAT Gap (%).
As the VTTL is a product of the base and the effective rate, the actual revenue could be
further decomposed and expressed as:
Actual Revenue = Net Base × Effective Rate × Compliance Ratio,
where Effective Rate is the ratio of the theoretical VTTL to the Net Base. The Net
Base (which is the sum of the final consumption and investment by households, non-profit
institutions serving households [NPISH], and government), in turn, is calculated as the
difference between the Gross Base, which includes VAT, and the VAT revenues actually
collected.
Table 1.3 and Figure 1.1 present the decomposition of the total changes in nominal
VAT revenues into these three components: change in net taxable base, change in the
effective rate applied to the base, and change in the compliance ratio.5
4 In other words, VAT collection efficiency.
5 In other words, VAT collection efficiency.
5](https://image.slidesharecdn.com/reportno503-201207153322/75/Study-and-Reports-on-the-VAT-Gap-in-the-EU-28-Member-States-2020-Final-Report-20-2048.jpg)
![CASE Working Paper | No 1 (2015)
63
a. Introduction
The examination of tax non-compliance determinants is not new to the economic
literature. Most of the literature dealing with such factors focuses on personal income
taxes, voluntary tax compliance, and deterrence effects. This focus is clearly related to
data availability. The empirical studies are based mostly on micro-data gathered in surveys
and audit statistics. Thus, they concentrate on the impact of individuals’ characteristics
(see e.g. Feinstein [1991]). Similarly, studies scrutinising the determinants of compliance
in corporate and consumption taxation usually look at micro-level revenue figures from
fiscal registers or audit data (see e.g. Casey and Castro [2015]). The studies based on fiscal
registers and audit and survey data face an important limitation, i.e. the inability to observe
the variability of determinants across tax systems and economies. A rather limited num-
ber of studies looking at such cross-country variations focus on the variation of dynamics
in tax revenue (see e.g. Aizenman and Jinjarak, [2018]) or have a qualitative nature (see e.g.
Keen and Smith [2007]).
The European Commission’s VAT Gap Study made available a large set of standardised
data on tax compliance from a group of countries with varying economic and institution-
al characteristics. The series are available across a time period long enough to cover eco-
nomic upturns and downturns. As a result, the Study provides an opportunity to conduct
econometric analyses looking at the determinants of tax non-compliance from a new
perspective. 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).
The econometric analysis outlined in this Study extends the above-mentioned studies
several-fold. Concerning the data preparation procedure, we eliminate potential bias in
the data by correcting the VAT Gap series for each country for revisions in subsequent
vintages of the Study. Moreover, we account for measurement errors, i.e. changes in
the VAT Gap not related to change in compliance but rather to specific one-off factors.
To deal with the scarcity of observations of exogenous variables, we perform a dummy
5. Econometric Analysis
of VAT Gap Determinants](https://image.slidesharecdn.com/reportno503-201207153322/75/Study-and-Reports-on-the-VAT-Gap-in-the-EU-28-Member-States-2020-Final-Report-63-2048.jpg)
![CASE Reports | No. 503 (2020)
95
The Policy Gap could be further decomposed to account for the loss of revenue. Such
components are the Rate Gap and the Exemption Gap, which capture the loss in VAT
liability due to the application of reduced rates and the loss in liability due to the im-
plementation of exemptions.
The Rate Gap is defined as the difference between the VTTL and what would be obtained
in a counterfactual situation, in which the standard rate, instead of the reduced, parking,
and zero rates, is applied to final consumption. Thus, the Rate Gap captures the loss in
revenue that a particular country incurs by adopting multiple VAT rates instead of a single
standard rate (Barbone et al., 2015).
The Exemption Gap is defined as the difference between the VTTL and what would
be obtained in a counterfactual situation, in which the standard rate is applied to exempt
products and services, and no restriction of the right to deduct applies24
. Thus, the
Exemption Gap captures the amount of revenue that might be lost because of exempted
goods and services. Note that the Exemption Gap is composed of the loss in the VAT
on the value added of exempt sectors, minus the VAT on their inputs, minus the VAT
on GFCF inputs for these sectors. Thus, in principle, the Exemption Gap might be positive
or negative (if the particular sector had negative value added, or if it had large GFCF
expenditures relative to final consumption) (Barbone et al., 2015).
In algebraic terms, we have the following:
Definitions:
VAT Gap in the EU-28 Member States
products and services, and no restriction of the right to deduct applies24
. Thus, the
Exemption Gap captures the amount of revenue that might be lost because of exempted
goods and services. Note that the Exemption Gap is composed of the loss in the VAT on
the value added of exempt sectors, minus the VAT on their inputs, minus the VAT on GFCF
inputs for these sectors. Thus, in principle, the Exemption Gap might be positive or negative
(if the particular sector had negative value added, or if it had large GFCF expenditures
relative to final consumption) (Barbone et al., 2015).
In algebraic terms, we have the following:
Definitions:
𝑇𝑇𝑖𝑖
∗,𝐸𝐸
=
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the case where the standard rate
instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and
the GFCF of non-business activities).
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
– liability from final consumption and GFCF of the non-business activities of group
i of products, in the case where the standard rate instead of the zero rate, parking rate, or
reduced rate is applied. Actual liability from intermediate consumption and the GFCF of
business activities is assumed.
𝑇𝑇𝑖𝑖
∗,𝑅𝑅
=
𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the event where exempt products
within the group are taxed at the standard rate.
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝑅𝑅
– liability from the final consumption of group i when exempt products within the
group are taxed at the standard rate. Actual liability from final consumption GFCF of non-
business activities is assumed.
𝜏𝜏𝑠𝑠 – statutory rate.
𝑖𝑖 ∈ (1; 65) – sectors of the economy.
– effective rate for group i of products in the case where the standard
rate instead of the zero rate, parking rate, or reduced rate is applied
(for final consumption and the GFCF of non-business activities).
VAT Gap in the EU-28 Member States
products and services, and no restriction of the right to deduct applies24
. Thus, the
Exemption Gap captures the amount of revenue that might be lost because of exempted
goods and services. Note that the Exemption Gap is composed of the loss in the VAT on
the value added of exempt sectors, minus the VAT on their inputs, minus the VAT on GFCF
inputs for these sectors. Thus, in principle, the Exemption Gap might be positive or negative
(if the particular sector had negative value added, or if it had large GFCF expenditures
relative to final consumption) (Barbone et al., 2015).
In algebraic terms, we have the following:
Definitions:
𝑇𝑇𝑖𝑖
∗,𝐸𝐸
=
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the case where the standard rate
instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and
the GFCF of non-business activities).
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
– liability from final consumption and GFCF of the non-business activities of group
i of products, in the case where the standard rate instead of the zero rate, parking rate, or
reduced rate is applied. Actual liability from intermediate consumption and the GFCF of
business activities is assumed.
𝑇𝑇𝑖𝑖
∗,𝑅𝑅
=
𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the event where exempt products
within the group are taxed at the standard rate.
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝑅𝑅
– liability from the final consumption of group i when exempt products within the
group are taxed at the standard rate. Actual liability from final consumption GFCF of non-
business activities is assumed.
𝜏𝜏𝑠𝑠 – statutory rate.
𝑖𝑖 ∈ (1; 65) – sectors of the economy.
– liability from final consumption and GFCF of the non-business activities
of group i of products, in the case where the standard rate instead of
the zero rate, parking rate, or reduced rate is applied. Actual liability from
intermediate consumption and the GFCF of business activities is assumed.
24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this Report differs from that
in Keen (2013). Keen (2013) defines the Rate Gap as the loss from applying reduced and zero rates to the final consumption
liability, measured as a percentage of the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumu-
lated in the production process as a percentage, on the contrary, of final consumption liability. Due to these definitions, the
Policy Gap can be split multiplicatively into gaps attributable to reduced rates and exemptions. Since the numerator of the “
[1 – Rate Gap]” and denominator of the “[1 – Exemption Gap]” are equal, multiplication of these two components yields
– VAT revenue as a percentage of Notional Ideal Revenue, which equals “[1 – Policy Gap]” (Barbone et al., 2015).](https://image.slidesharecdn.com/reportno503-201207153322/75/Study-and-Reports-on-the-VAT-Gap-in-the-EU-28-Member-States-2020-Final-Report-95-2048.jpg)
![CASE Reports | No. 503 (2020)
96
page 83 of 99
In algebraic terms, we have the following:
Definitions:
𝑇𝑇𝑖𝑖
∗,𝐸𝐸
=
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the case where the standard rate
instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and
the GFCF of non-business activities).
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
– liability from final consumption and GFCF of the non-business activities of group
i of products, in the case where the standard rate instead of the zero rate, parking rate, or
reduced rate is applied. Actual liability from intermediate consumption and the GFCF of
business activities is assumed.
𝑇𝑇𝑖𝑖
∗,𝑅𝑅
=
𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the event where exempt products
within the group are taxed at the standard rate.
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝑅𝑅
– liability from the final consumption of group i when exempt products within the
group are taxed at the standard rate. Actual liability from final consumption GFCF of non-
business activities is assumed.
𝜏𝜏𝑠𝑠 – statutory rate.
𝑖𝑖 ∈ (1; 65) – sectors of the economy.
24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this
Report differs from that in Keen (2013). Keen (2013) defines the Rate Gap as the loss from
applying reduced and zero rates to the final consumption liability, measured as a percentage of
the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumulated in
the production process as a percentage, on the contrary, of final consumption liability. Due to
these definitions, the Policy Gap can be split multiplicatively into gaps attributable to reduced
rates and exemptions. Since the numerator of the “[1 - Rate Gap]” and denominator of the “[1 -
Exemption Gap]” are equal, multiplication of these two components yields – VAT revenue as a
percentage of Notional Ideal Revenue, which equals “[1 - Policy Gap]” (Barbone et al., 2015).
– effective rate for group i of products in the event where exempt products
within the group are taxed at the standard rate.
page 83 of 99
Definitions:
𝑇𝑇𝑖𝑖
∗,𝐸𝐸
=
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the case where the standard rate
instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and
the GFCF of non-business activities).
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
– liability from final consumption and GFCF of the non-business activities of group
i of products, in the case where the standard rate instead of the zero rate, parking rate, or
reduced rate is applied. Actual liability from intermediate consumption and the GFCF of
business activities is assumed.
𝑇𝑇𝑖𝑖
∗,𝑅𝑅
=
𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the event where exempt products
within the group are taxed at the standard rate.
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝑅𝑅
– liability from the final consumption of group i when exempt products within the
group are taxed at the standard rate. Actual liability from final consumption GFCF of non-
business activities is assumed.
𝜏𝜏𝑠𝑠 – statutory rate.
𝑖𝑖 ∈ (1; 65) – sectors of the economy.
24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this
Report differs from that in Keen (2013). Keen (2013) defines the Rate Gap as the loss from
applying reduced and zero rates to the final consumption liability, measured as a percentage of
the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumulated in
the production process as a percentage, on the contrary, of final consumption liability. Due to
these definitions, the Policy Gap can be split multiplicatively into gaps attributable to reduced
rates and exemptions. Since the numerator of the “[1 - Rate Gap]” and denominator of the “[1 -
Exemption Gap]” are equal, multiplication of these two components yields – VAT revenue as a
percentage of Notional Ideal Revenue, which equals “[1 - Policy Gap]” (Barbone et al., 2015).
– liability from the final consumption of group i when exempt products
within the group are taxed at the standard rate. Actual liability from final
consumption GFCF of non-business activities is assumed.
page 83 of 99
Definitions:
𝑇𝑇𝑖𝑖
∗,𝐸𝐸
=
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the case where the standard rate
instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and
the GFCF of non-business activities).
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
– liability from final consumption and GFCF of the non-business activities of group
i of products, in the case where the standard rate instead of the zero rate, parking rate, or
reduced rate is applied. Actual liability from intermediate consumption and the GFCF of
business activities is assumed.
𝑇𝑇𝑖𝑖
∗,𝑅𝑅
=
𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the event where exempt products
within the group are taxed at the standard rate.
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝑅𝑅
– liability from the final consumption of group i when exempt products within the
group are taxed at the standard rate. Actual liability from final consumption GFCF of non-
business activities is assumed.
𝜏𝜏𝑠𝑠 – statutory rate.
𝑖𝑖 ∈ (1; 65) – sectors of the economy.
24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this
Report differs from that in Keen (2013). Keen (2013) defines the Rate Gap as the loss from
applying reduced and zero rates to the final consumption liability, measured as a percentage of
the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumulated in
the production process as a percentage, on the contrary, of final consumption liability. Due to
these definitions, the Policy Gap can be split multiplicatively into gaps attributable to reduced
rates and exemptions. Since the numerator of the “[1 - Rate Gap]” and denominator of the “[1 -
Exemption Gap]” are equal, multiplication of these two components yields – VAT revenue as a
percentage of Notional Ideal Revenue, which equals “[1 - Policy Gap]” (Barbone et al., 2015).
– statutory rate.
page 83 of 99
Definitions:
𝑇𝑇𝑖𝑖
∗,𝐸𝐸
=
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the case where the standard rate
instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and
the GFCF of non-business activities).
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝐸𝐸
– liability from final consumption and GFCF of the non-business activities of group
i of products, in the case where the standard rate instead of the zero rate, parking rate, or
reduced rate is applied. Actual liability from intermediate consumption and the GFCF of
business activities is assumed.
𝑇𝑇𝑖𝑖
∗,𝑅𝑅
=
𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
– effective rate for group i of products in the event where exempt products
within the group are taxed at the standard rate.
𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖
∗,𝑅𝑅
– liability from the final consumption of group i when exempt products within the
group are taxed at the standard rate. Actual liability from final consumption GFCF of non-
business activities is assumed.
𝜏𝜏𝑠𝑠 – statutory rate.
𝑖𝑖 ∈ (1; 65) – sectors of the economy.
24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this
Report differs from that in Keen (2013). Keen (2013) defines the Rate Gap as the loss from
applying reduced and zero rates to the final consumption liability, measured as a percentage of
the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumulated in
the production process as a percentage, on the contrary, of final consumption liability. Due to
these definitions, the Policy Gap can be split multiplicatively into gaps attributable to reduced
rates and exemptions. Since the numerator of the “[1 - Rate Gap]” and denominator of the “[1 -
Exemption Gap]” are equal, multiplication of these two components yields – VAT revenue as a
percentage of Notional Ideal Revenue, which equals “[1 - Policy Gap]” (Barbone et al., 2015).
– sectors of the economy.
Policy Gap:
Exemption Gap:
Rate Gap:
By definition we have:
Thus:
VAT Gap in the EU-28 Member States
Policy Gap:
1 − 𝑃𝑃 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Exemption Gap:
1 − 𝑃𝑃𝐸𝐸 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Rate Gap:
1 − 𝑃𝑃𝑅𝑅 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
By definition we have:
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Thus:
𝑃𝑃 = 1 − (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸
VAT Gap in the EU-28 Member States
Policy Gap:
1 − 𝑃𝑃 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Exemption Gap:
1 − 𝑃𝑃𝐸𝐸 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Rate Gap:
1 − 𝑃𝑃𝑅𝑅 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
By definition we have:
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Thus:
𝑃𝑃 = 1 − (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸
Using the above convention, one can decompose the Rate Gap and the Exemption Gap
into components indicating the loss of the Notional Ideal Revenue due to the implementation
VAT Gap in the EU-28 Member States
Policy Gap:
1 − 𝑃𝑃 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Exemption Gap:
1 − 𝑃𝑃𝐸𝐸 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Rate Gap:
1 − 𝑃𝑃𝑅𝑅 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
By definition we have:
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Thus:
𝑃𝑃 = 1 − (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸
Using the above convention, one can decompose the Rate Gap and the Exemption Gap
into components indicating the loss of the Notional Ideal Revenue due to the implementation
of reduced rates and exemptions on specific goods and services. Such additive
VAT Gap in the EU-28 Member States
Policy Gap:
1 − 𝑃𝑃 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Exemption Gap:
1 − 𝑃𝑃𝐸𝐸 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Rate Gap:
1 − 𝑃𝑃𝑅𝑅 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
By definition we have:
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Thus:
𝑃𝑃 = 1 − (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸
Using the above convention, one can decompose the Rate Gap and the Exemption Gap
into components indicating the loss of the Notional Ideal Revenue due to the implementation
of reduced rates and exemptions on specific goods and services. Such additive
decomposition is carried out for the computation of, as defined by Barbone et al. (2015),
VAT Gap in the EU-28 Member States
Policy Gap:
1 − 𝑃𝑃 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Exemption Gap:
1 − 𝑃𝑃𝐸𝐸 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Rate Gap:
1 − 𝑃𝑃𝑅𝑅 = (
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
By definition we have:
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
+ (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
− ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
Thus:
𝑃𝑃 = 1 − (
∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
) = (
2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝐸𝐸
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖
∗,𝑅𝑅
𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖
𝑁𝑁
𝑖𝑖=1
)
= 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸
Using the above convention, one can decompose the Rate Gap and the Exemption Gap
into components indicating the loss of the Notional Ideal Revenue due to the implementation
of reduced rates and exemptions on specific goods and services. Such additive
decomposition is carried out for the computation of, as defined by Barbone et al. (2015),
the Actionable Exemption Gap, which excludes the services and notional values that are
unlikely to be taxed even in an ideal world.](https://image.slidesharecdn.com/reportno503-201207153322/75/Study-and-Reports-on-the-VAT-Gap-in-the-EU-28-Member-States-2020-Final-Report-96-2048.jpg)
Uploaded byCASE Center for Social and Economic Research
Study and Reports on the VAT Gap in the EU-28 Member States: 2020 Final Report
The report analyzes the VAT gap in the EU-28 member states in 2018. It finds that the total VAT gap in the EU was an estimated €137 billion, representing 12.2% of the total VAT liability. This is an increase compared to 2017, when the gap was €117 billion or 11.2% of the total liability. The report examines VAT revenue, total VAT liability, and VAT gap estimates for each member state from 2014 to 2018. It also conducts econometric analysis to identify factors influencing VAT gap levels across countries.
More Related Content
Similar to Study and Reports on the VAT Gap in the EU-28 Member States: 2020 Final Report
More from CASE Center for Social and Economic Research
Study and Reports on the VAT Gap in the EU-28 Member States: 2020 Final Report
- 1. Study and Reports onthe VAT Gap in the EU-28 Member States: 2020 Final Report Grzegorz Poniatowski Mikhail Bonch-Osmolovskiy Adam Śmietanka No. 503 (2020) CASE Reports
- 2. The views andopinions expressed in this report are not necessarily shared by the European Commission or CASE Network, nor does the report anticipate decisions taken by the European Commission. Keywords: consumption taxation, VAT, tax fraud, tax evasion, tax avoidance, tax gap, tax non-compliance, policy gap JEL Codes: H24, H26 © CASE – Center for Social and Economic Research, Warsaw, 2020 Graphic Design: Katarzyna Godyń-Skoczylas | grafo-mania ISBN: 978-83-7178-703-4 Publisher: CASE – Center for Social and Economic Research Al. Jana Pawla II 61, office 212, 01-031 Warsaw, Poland tel.: (48 22) 206 29 00, fax: (48 22) 206 29 01 e-mail: [email protected] www.case-research.eu This report was commissioned by the Directorate General for Taxation and Customs Union (TAXUD) of the European Commission under project No. TAXUD/2019/AO-14, and written by a team of experts from CASE – Center for Social and Economic Research (Warsaw) directed by Grzegorz Poniatowski, and composed of Mikhail Bonch-Osmolovskiy and Adam Śmietanka. The Project was coordinated by Roberto Zavatta (Economisti Associati, Bologna). It remains the property of TAXUD. We acknowledge valuable comments from reviewers, Hana Zídková and Michael Udell. We also acknowledge discussions with several officials of tax and statistical offices of the Member States, who offered valuable information, comments, and suggestions. All responsibility for the estimates and the interpretation in this Report remains with the authors. Adam Śmietanka No. 503 (2020) [editorial page] Acknowledgments This report was commissioned by the Directorate General for TaxaRon and Customs Union (TAXUD) of the European Commission under project No. TAXUD/2019/AO-14, and wri[en by a team of experts from CASE (Center for Social and Economic Research, Warsaw) directed by Grzegorz Poniatowski, and composed of Mikhail Bonch-Osmolovskiy and Adam Śmietanka. The Project was coordinated by Roberto Zava[a (EconomisR AssociaR, Bologna). It remains the property of TAXUD. We acknowledge valuable comments from reviewers, Hana Zídková and Michael Udell. We also acknowledge discussions with several officials of tax and staRsRcal offices of the Member States, who offered valuable informaRon, comments, and suggesRons. All responsibility for the esRmates and the interpretaRon in this Report remains with the authors. The views and opinions expressed in this report are not necessarily shared by the European Commission or CASE Network, nor does the report anRcipate decisions taken by the European Commission. Keywords: consumpRon taxaRon, VAT, tax fraud, tax evasion, tax avoidance, tax gap, tax non- compliance, policy gap JEL codes: H24, H26 © CASE – Center for Social and Economic Research, Warsaw, 2020 Graphic Design: ….
- 3. CASE Working Paper| No 1 (2015) 3 Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1. Background: Economic and Policy Context in 2018. . . . . . . . . . . . . . . . . . . . . . 16 a. Economic Conditions in the EU during 2018. . . . . . . . . . . . . . . . . . . . . . . 16 b. VAT Regime Changes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 c. Sources of Change in VAT Revenue Components. . . . . . . . . . . . . . . . . . . 20 2. The VAT Gap in 2018. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3. Individual Country Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 4. Policy Gap Measures for 2018 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 5. Econometric Analysis of VAT Gap Determinants. . . . . . . . . . . . . . . . . . . . . . . . . . 63 a. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 b. Data and Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 c. Methods and Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 d. Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 6. The Potential Impact of the Coronavirus Recession on the Evolution of the VAT Gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Table of Contents
- 4. CASE Working Paper| No 1 (2015) 4 Annex A. Methodological Considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 a. Source of Revisions of VAT Gap Estimates. . . . . . . . . . . . . . . . . . . . . . . . . 88 b. Decomposition of VAT Revenue. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 c. Data Sources and Estimation Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 d. Fast VAT Gap Estimates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 e. Derivation of the Policy Gap. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 f. Tests of the Econometric Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Annex B. Statistical Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Annex C. Additional Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Table of Contents
- 5. CASE Working Paper| No 1 (2015) 5 List of Tables Table 1.1. Real and Nominal Growth in the EU-28 in 2018 (in national currencies [NAC]). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Table 1.2. VAT Rate Structure as of 31 December 2017 and Changes during 2018 (%). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Table 1.3. Change in VAT Revenue Components (2018 over 2017). . . . . . . . . . 21 Table 2.1. VAT Gap as a percent of the VTTL in EU-28 Member States, 2018 and 2017. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Table 3.1. Belgium: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Table 3.2. Bulgaria: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (BGN million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Table 3.3. Czechia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (CZK million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Table 3.4. Denmark: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (DKK million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Table 3.5. Germany: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Table 3.6. Estonia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Table 3.7. Ireland: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Table 3.8. Greece: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
- 6. CASE Working Paper| No 1 (2015) 6 Table 3.9a. Spain: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Table 3.9b. Spain: Alternative Estimates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Table 3.10. France: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Table 3.11. Croatia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (HRK million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Table 3.12a. Italy: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Table 3.12b. Italy: Alternative Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Table 3.13. Cyprus: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2015–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Table 3.14. Latvia: VAT Revenue VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Table 3.15. Lithuania: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Table 3.16. Luxembourg: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Table 3.17. Hungary: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (HUF million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Table 3.18. Malta: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 List of Tables
- 7. CASE Working Paper| No 1 (2015) 7 Table 3.19. Netherlands: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Table 3.20. Austria: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Table 3.21. Poland: VAT Revenue VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (PLN million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Table 3.22. Portugal: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Table 3.23. Romania: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (RON million) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Table 3.24. Slovenia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Table 3.25. Slovakia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Table 3.26. Finland: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Table 3.27. Sweden: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (SEK million) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Table 3.28. United Kingdom: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (GBP million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Table 4.1. Policy Gap, Rate Gap, Exemption Gap, and Actionable Gaps. . . . . . 62 Table 5.1. Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Table 5.2. Descriptive Statistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 List of Tables
- 8. CASE Working Paper| No 1 (2015) 8 Table 5.3. Econometric Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Table 5.4. Robustness Check . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Table A1. Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Table B1. VTTL (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Table B2. Household VAT Liability (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Table B3. Intermediate Consumption and Government VAT Liability (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Table B4. GFCF VAT Liability (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Table B5. VAT Revenues (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Table B6. VAT Gap (EUR million). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Table B7. VAT Gap (percent of VTTL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 List of Tables
- 9. CASE Working Paper| No 1 (2015) 9 List of Graphs Figure 1.1. Change in VAT Revenue Components (2018 over 2017, %). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Figure 2.1. Evolution of the VAT Gap in the EU, 2014–2018 and Fast Estimate for 2019. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Figure 2.2. VAT Gap as a percent of the VTTL in EU-28 Member States, 2018 and 2017. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Figure 2.3. Percentage Point Change in VAT Gap, 2018 over 2017. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 2.4. VAT Gap in EU Member States, 2014–2018. . . . . . . . . . . . . . . . . . . . . . 27 Figure 5.1. Comparison of Results (VAT Gap as % of the VTTL in EU-28). . . . 70 Figure 5.2. Backcasting of EU-wide Estimates Presented in Figure 5.1 (VAT Gap as % of the VTTL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Figure 5.3. Backcasting of Individual Estimates (VAT Gap as % of the VTTL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Figure 5.4. Individual Estimates in Consecutive Studies (VAT Gap as % of the VTTL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Figure 5.5. Linear Predictions Broken Out by Member State. . . . . . . . . . . . . . . . 82 Figure 5.6. Contributions to VAT Gap Change. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Figure 6.1. 2020 Spring Forecasts of the European Commission (%). . . . . . . . . 86 Figure 6.2. Change in the VAT Gap and Prediction Intervals (increments, percentage points). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
- 10. CASE Working Paper| No 1 (2015) 10 Figure 6.3. VAT Gap and Prediction Intervals (% of the VTTL). . . . . . . . . . . . . . . 87 Figure A1. Components of Ideal Revenue, VTTL, and VAT Collection . . . . . . . 98 Figure C1. VAT Gap Forecasts for 2020 (increments, pp). . . . . . . . . . . . . . . . . . 107 List of Graphs
- 11. List of Acronymsand Abbreviations CASE Center for Social and Economic Research (Warsaw) COICOP Classification of Individual Consumption according to Purpose CPA Statistical Classification of Products by Activity in accordance with Regulation (EC) No 451/2008 of the European Parliament and of the Council of 23 April 2008 establishing a new statistical classification of products by activity EC European Commission ESA European System of Accounts EU European Union EU-28 Member States of the European Union, UK inclusive FE Fixed Effects GDP Gross Domestic Product GFCF Gross Fixed Capital Formation IC Intermediate Consumption MFI Monetary Financial Institution MOSS Mini One Stop Shop MTIC Missing Trader Intra-Community NAC National Currency NPISH Non-Profit Institutions Serving Households OECD Organisation for Economic Cooperation and Development ORS Own Resource Submissions o/w of which pp. percentage points SUT Supply and Use Tables TAXUD Taxation and Customs Union Directorate-General of the European Commission VAT Value Added Tax VTTL VAT Total Tax Liability
- 12. CASE Working Paper| No 1 (2015) 12 This Report has been written for the European Commission, DG TAXUD, for the project TAXUD/2019/AO-14, “Study and Reports on the VAT Gap in the EU-28 Member States”, and is a follow-up to the seven reports published between 2013 and 2019. This Study contains Value Added Tax (VAT) Gap estimates for 2018, fast estimates using a simplified methodology for 2019, the year immediately preceding the analysis, and includes revised estimates for 2014–2017. It also includes the updated and extended results of the econometric analysis of VAT Gap determinants initiated and initially reported in the 2018 Report (Poniatowski et al., 2018). As a novelty, the econometric analysis to forecast potential impacts of the coronavirus crisis and resulting recession on the evolution of the VAT Gap in 2020 is reported. In 2018, most European Union (EU) Member States (MS) saw a slight decrease in the pace of gross domestic product (GDP) growth, but the economic conditions for increasing tax compliance remained favourable. We estimate that the VAT total tax liability (VTTL) in 2018 increased by 3.6 percent whereas VAT revenue increased by 4.2 percent, leading to a decline in the VAT Gap in both relative and nominal terms. In relative terms, the EU-wide Gap dropped to 11 percent and EUR 140 billion. Fast estimates show that the VAT Gap will likely continue to decline in 2019. Of the EU-28, the smallest Gaps were observed in Sweden (0.7 percent), Croatia (3.5 percent), and Finland (3.6 percent), the largest – in Romania (33.8 percent), Greece (30.1 percent), and Lithuania (25.9 percent). Overall, half of the EU-28 MS recorded a Gap above 9.2 percent. In nominal terms, the largest Gaps were recorded in Italy (EUR 35.4 billion), the United Kingdom (EUR 23.5 billion), and Germany (EUR 22 billion). The Policy Gap and its components remained stable. For the EU overall, the average Policy Gap level was 44.24 percent. Of this, in 2018, 10.07 percentage points were due to the application of various reduced and super-reduced rates (the Rate Gap) and 34.17 were due to the application of exemptions without the right to deduct. The results of the econometric analysis show that the VAT Gap is influenced by a group of factors relating to the current economic conditions, institutional environment, and economic structure as well as to the measures and actions of tax administrations. Executive Summary
- 13. CASE Reports |No. 503 (2020) 13 Out of a broad set of tested variables, GDP growth and general government balance ap- peared to explain a substantial set of VAT Gap variation across time and countries. Within the control of tax administrations, share of IT expenditure proved to have the highest statistical significance in explaining the size of the VAT Gap. In addition, the VAT Gap appeared to be inter-related with the values of risky imports of goods, indicating the role of fraud in driving the overall share of the VAT Gap. Since the COVID-19 recession will likely have a dire impact on the EU economies, the VAT Gap in 2020 is forecasted to increase. If the EU economy contracts by 7.4 percent in 2020 and the general government deficit jumps as forecasted in the Spring Forecast of the Euro- pean Commission, the Gap could increase by 4.1 percentage points year-over year up to 13.7 percent and EUR 164 billion in 2020. The hike in 2020 could be more pronounced than the gradual decrease of the Gap observed over the three preceding years. Moreover, a return to the VAT Gap levels observed in 2018 and 2019 will take time and require significant action from tax administrations.
- 14. CASE Working Paper| No 1 (2015) 14 This Report presents the findings of the 2020 “Study to quantify the VAT Gap in the EU Member States”, which is the seventh publication following the original Study conducted by Barbone et al. in 20131 . We present Value Added Tax (VAT) Gap estimates for 2018, fast estimates using a simplified methodology for 2019, the year immediately preceding the analysis, and include revised estimates for 2014–20172 . We also include updated and extended results of the econometric analysis of VAT Gap determinants initiated and initially reported in the 2018 Report (Poniatowski et al., 2018). As a novelty, we operationalise the econometric analysis to forecast potential impacts of the coronavirus crisis and resulting recession on the evolution of the VAT Gap in 2020 and 2021. The VAT Gap, which is addressed in detail by this Report shall be understood as the Compliance Gap. It is the difference between the expected and actual VAT revenues and represents more than just fraud and evasion and their associated policy measures. The VAT Gap also covers VAT lost due to, for example, insolvencies, bankruptcies, administrative errors, and legal tax optimisation. It is defined as the difference between the amount of VAT collected and the VAT Total Tax Liability (VTTL) – namely, the tax liability accord- ing to tax law. The VAT Gap can be expressed in absolute or relative terms, commonly as a ratio of the VTTL or gross domestic product (GDP). In this Report, we refer to the VAT Gap as the ratio of the VTTL. In addition to the analysis of the Compliance Gap, this Report also updates the Policy Gap estimates from 2018 as well as the contribution that reduced rates and exemptions made to these theoretical VAT revenue losses. The structure of this Report builds on the previous publications. Chapter 1 presents the main economic and policy factors that affected European Union (EU) Member States (MS) during the course of 2018. It also includes a decomposition of the change in VAT 1 The first study of the VAT Gap in the EU was conducted by Reckon (2009); however, due to differences in methodology, it cannot be directly compared to these latter studies. 2 The estimates for 2019 are referred to as “fast” since they use different method described in Section d in Annex A and could be associated with larger estimation error. Introduction
- 15. CASE Reports |No. 503 (2020) 15 revenues. The overall results are presented and briefly described in Chapter 2. Chapter 3 provides detailed results and outlines trends for individual countries coupled with analytical insights. In Chapter 4, we examine the Policy Gap and the contribution that VAT reduced rates and exemptions have made to this Gap. Chapter 5 is devoted to the econometric analysis. It provides an overview of the literature, highlights the most important novelties introduced with this update, and discusses and visualises the results which are complemented by a robustness check. The final chapter presents the impact of the coronavirus recession on the evolution of the VAT Gap. Annex A contains the methodological considerations underlying all components of the analysis. Annex B provides statistical data and a set of comparative tables, whereas Annex C provides additional graphs.
- 16. CASE Working Paper| No 1 (2015) 16 a. Economic Conditions in the EU during 2018 In 2018, most EU MS saw a moderate decrease in the pace of GDP growth. Overall, growth of the EU economy fell from 2.5 percent in 2017 down to 2.0 percent in 2018 in real terms. Positive economic tailwinds provided particularly good conditions for an increase in VAT collections in Ireland (GDP growth of 8.2 percent), Poland (5.3 percent), and Hungary (5.1 percent). The lowest GDP growth rates were observed in Italy (0.8 percent) and the United Kingdom (1.5 percent). In nominal terms, GDP increased by 3.3 percent and consumer prices by 1.9 percent. Final consumption, which is the core of the VAT base (68 percent of the VTTL in 2018), ncreased by 3.1 percent in total. Investment in gross fixed capital formation (GFCF, which made up 14 percent of the VTTL in 2018) increased by 4.2 percentage points for the entire EU. The change in GFCF was volatile across countries and varied from −18.7 percent in Ireland to 24.4 percent in Hungary. Due to the volatility and frequent revisions of GFCF figures by Statistical Offices, GFCF is the main source of VAT Gap revisions. Whenever new information on the actual investment figures of exempt sectors becomes available, the estimates of VAT Gap are revised backwards. General government budgets and the labour markets remained relatively sound. The average general government balance amounted to −0.7 percent with half of EU MS observing a nominal surplus. The unemployment rate fell in nearly all EU MS and by −0.9 percent on average. 1. Background: Economic and Policy Context in 2018
- 17. 17 CASE Reports |No. 503 (2020) Table 1.1. Real and Nominal Growth in the EU-28 in 2018 (in national currencies [NAC]) Source: Eurostat. Member State Real GDP Growth (%) General Government Balance (%) Change in Unemployment Rate (pp) Nominal Growth (%) GDP Final Consumption GFCF Belgium 1.5 −0.8 −1.1 3.0 3.3 6.2 Bulgaria 3.1 2.0 −1.0 7.2 7.7 9.7 Czechia 2.8 0.9 −0.7 5.5 6.6 9.1 Denmark 2.4 0.7 −0.7 3.3 3.0 7.3 Germany 1.5 1.9 −0.4 3.1 2.9 6.3 Estonia 4.8 −0.6 −0.4 9.5 8.1 5.3 Ireland 8.2 0.1 −0.9 9.1 6.0 −18.7 Greece 1.9 1.0 −2.2 2.5 0.9 −12.0 Spain 2.4 −2.5 −1.9 3.5 3.4 7.7 France 1.8 −2.3 −0.4 2.8 2.2 4.6 Croatia 2.7 0.2 −2.7 4.5 4.5 4.7 Italy 0.8 −2.2 −0.6 1.7 2.0 3.8 Cyprus 4.1 −3.7 −2.7 5.5 5.0 −4.5 Latvia 4.3 −0.8 −1.3 8.4 7.3 18.0 Lithuania 3.6 0.6 −0.9 7.1 6.8 10.1 Luxembourg 3.1 3.1 0.1 5.7 6.1 −5.3 Hungary 5.1 −2.1 −0.5 9.9 7.6 24.4 Malta 7.3 1.9 −0.3 9.5 10.2 0.8 Netherlands 2.4 1.4 −1.1 4.9 4.6 6.3 Austria 2.4 0.2 −0.6 4.2 3.3 6.0 Poland 5.3 −0.2 −1.0 6.6 6.4 10.8 Portugal 2.6 −0.4 −1.9 4.3 3.9 9.0 Romania 4.4 −2.9 −0.7 11.0 13.2 3.9 Slovenia 4.1 0.7 −1.5 6.4 5.4 11.4 Slovakia 3.9 −1.0 −1.6 6.0 6.0 4.9 Finland 1.5 −0.9 −1.2 3.4 3.1 6.6 Sweden 2.0 0.8 −0.3 4.4 4.4 4.6 United Kingdom 1.3 −2.2 −0.3 3.5 3.8 1.6 EU-28 (EUR) 2.0 −0.7 −0.9 3.3 3.1 4.2
- 18. CASE Reports |No. 503 (2020) 18 b. VAT Regime Changes 2018 was another stable year in terms of both EU-wide and country-specific changes affecting the VTTL. The temporary measure of the Mini One Stop Shop (MOSS) retention fee, which is the revenue retained in the country of origin of service providers obliged to pay VAT in the country of residence of their customers, was maintained in 2018 at the level of 15 percent. For this reason, the rule for estimating the VTTL of electronic services remained unchanged. As for country-specific changes, only one MS implemented significant changes to the structure of its VAT rates in 2018. As of January 2018, Latvia introduced a super-reduced rate of 5 percent applicable to a range of common vegetables and fruits. There were also a few examples of the reclassification of rates applicable to certain products. Among those, Lithuania applied a reduced rate of 9 percent on accommodation services (down from 21 percent). Similarly, starting from November, Romania applied a reduced rate of 5 percent to accommodation, restaurants, and catering services. In Hungary, the rate applicable to Internet access services was reduced from 18 percent to 5 percent. Overall, the average effective rate remained unchanged compared to 2017 and accounted for 12 percent3 . 3 Changes in the effective rat compared to the 2017 Report also result from the revision of the VTTL estimates and the statistical data underlying the estimates.
- 19. 19 CASE Reports |No. 503 (2020) Table 1.2. VAT Rate Structure as of 31 December 2017 and Changes during 2018 (%) Source: TAXUD, VAT Rates Applied in the Member States of the European Union: Situation of 1st January 2018. Member State Standard Rate (SR) Reduced Rate(s) (RR) Super− Reduced Rate Parking Rate Changes during 2018 Effective Rate 4 Belgium 21 6 / 12 − 12 10.1 Bulgaria 20 9 − − 14.0 Czechia 21 10 / 15 − 12.6 Denmark 25 − − − 14.9 Germany 19 7 − − 10.6 Estonia 20 9 − − 12.9 Ireland 23 9 / 13.5 4.8 13.5 12.3 Greece 24 6 / 13 − − 13.1 Spain 21 10 4 − 8.8 France 19.6 5.5 / 10 2.1 − 9.6 Croatia 25 5 / 13 − − 16.4 Italy 22 10 4 / 5 − 10.2 Cyprus 19 5 / 9 − − 10.5 Latvia 21 12 5 − Super−Reduced Rate introduced (5%) 11.8 Lithuania 21 5 / 9 − − 13.6 Luxembourg 17 8 3 14 12.2 Hungary 27 5 / 18 − − 14.8 Malta 18 5 / 7 − − 12.1 Netherlands 21 6 − − 10.0 Austria 20 10 / 13 − 12 11.3 Poland 23 5 / 8 − − 12.1 Portugal 23 6 / 13 − 13 11.5 Romania 20 5 / 9 − − 12.1 Slovenia 22 9.5 − − 11.8 Slovakia 20 10 − − 11.6 Finland 24 10 / 14 − − 12.2 Sweden 25 6 / 12 − − 13.4 United Kingdom 20 5 − − 9.6 4 The effective rate is the ratio of the VTTL and the tax base. See methodological considerations in Section c in Annex A. Source: TAXUD, VAT Rates Applied in the Member States of the European Union: Situation of 1st January 2018. 4 The effective rate is the ratio of the VTTL and the tax base. See methodological considerations in Section c in Annex A.
- 20. CASE Reports |No. 503 (2020) 20 c. Sources of Change in VAT Revenue Components The value of the actual VAT revenue can be decomposed into components, which is helpful in understanding the underlying sources of its evolution. Since revenue is a product of the VTTL and the compliance ratio4 , VAT collection could be expressed as: Actual Revenue = VTTL × Compliance Ratio, where Compliance Ratio is: 1 – VAT Gap (%). As the VTTL is a product of the base and the effective rate, the actual revenue could be further decomposed and expressed as: Actual Revenue = Net Base × Effective Rate × Compliance Ratio, where Effective Rate is the ratio of the theoretical VTTL to the Net Base. The Net Base (which is the sum of the final consumption and investment by households, non-profit institutions serving households [NPISH], and government), in turn, is calculated as the difference between the Gross Base, which includes VAT, and the VAT revenues actually collected. Table 1.3 and Figure 1.1 present the decomposition of the total changes in nominal VAT revenues into these three components: change in net taxable base, change in the effective rate applied to the base, and change in the compliance ratio.5 4 In other words, VAT collection efficiency. 5 In other words, VAT collection efficiency. 5
- 21. CASE Reports |No. 503 (2020) 21 Table 1.3. Change in VAT Revenue Components (2018 over 2017) Source: own calculations. Member State Change in Revenue Change in the VTTL Change in ComplianceChange in Base Change in Effective Rate Belgium 4.3% 3.1% 3.6% −0.5% 1.2% Bulgaria 9.3% 7.5% 8.0% −0.4% 1.7% Czechia 6.5% 6.6% 7.8% −1.1% −0.1% Denmark 4.3% 3.1% 3.2% 0.0% 1.2% Germany 3.8% 3.6% 3.3% 0.2% 0.2% Estonia 8.5% 7.5% 8.8% −1.2% 0.9% Ireland 8.5% 8.2% 7.4% 0.8% 0.3% Greece 4.4% −0.2% −0.6% 0.5% 4.6% Spain 4.9% 4.4% 3.8% 0.5% 0.4% France 3.5% 3.8% 2.2% 1.6% −0.3% Croatia 6.8% 4.5% 4.3% 0.2% 2.1% Italy 1.6% 1.3% 2.0% −0.7% 0.3% Cyprus 10.5% 9.1% 8.0% 1.0% 1.3% Latvia 13.2% 7.7% 8.4% −0.7% 5.1% Lithuania 6.4% 7.5% 7.5% 0.0% −1.0% Luxembourg 8.6% 11.4% 5.9% 5.2% −2.5% Hungary 13.9% 7.5% 9.4% −1.8% 5.9% Malta 13.5% 10.1% 9.8% 0.3% 3.1% Netherlands 5.6% 4.9% 5.2% −0.3% 0.7% Austria 3.6% 4.1% 3.2% 0.9% −0.5% Poland 11.4% 6.0% 6.4% −0.4% 5.1% Portugal 6.3% 4.7% 4.0% 0.6% 1.5% Romania 12.7% 12.0% 14.3% −2.0% 0.7% Slovenia 8.1% 7.5% 6.1% 1.3% 0.6% Slovakia 6.8% 7.3% 7.0% 0.3% −0.5% Finland 4.7% 3.1% 3.8% −0.7% 1.6% Sweden 4.8% 3.5% 4.2% −0.6% 1.3% United Kingdom 4.6% 5.0% 4.0% 1.0% −0.3% EU-28 (total) 4.2% 3.6% 3.3% 0.4% 0.5%
- 22. CASE Reports |No. 503 (2020) 22 Figure 1.1. Change in VAT Revenue Components (2018 over 2017, %) Source: own calculations. As depicted by Table 1.3 and Figure 1.1 and highlighted in the preceding section, the growth of the base was the main driver of VAT revenue growth in 2018. An increase in the base contributed to approximately 78 percent of the total VAT revenue growth in the EU. The effect of increased compliance contributed to approximately 10 percent of the growth, which translated to 0.4 percent of the overall VAT revenue. For the vast majority of EU MS, both the tax base and compliance effect were positive. In five countries, namely Hungary, Romania, Latvia, Malta, and Poland, the overall effect of the increase in the tax base and compliance exceeded 10 percent of VAT revenue. VAT Gap in the EU-28 Member States page 15 of 99 Figure 1.1. Change in VAT Revenue Components (2018 over 2017, %) Source: own calculations. As depicted by Table 1.3 and Figure 1.1 and highlighted in the preceding section, the growth of the base was the main driver of VAT revenue growth in 2018. An increase in the base contributed to approximately 78 percent of the total VAT revenue growth in the EU. The effect of increased compliance contributed to approximately 10 percent of the growth, which translated to 0.4 percent of the overall VAT revenue. For the vast majority of EU MS, both the tax base and compliance effect were positive. In five countries, namely Hungary, Romania, Latvia, Malta, and Poland, the overall effect of the increase in the tax base and compliance exceeded 10 percent of VAT revenue. 2. The VAT Gap in 2018 The estimates of the VAT Gap presented in this section were derived using the same methodology as in the previously cited VAT Gap Studies. The VAT Gap is defined as the difference between the VTTL and the amount of VAT actually collected over the same period. We compute the VTTL using a top-down “consumption-side” approach by deriving the expected VAT liability from the observed national accounts data, such as supply and use tables (SUT). For this reason, the methodology used in this Study relies on the availability and quality of SUT data, which vary country to country. The VAT liability is estimated for final household, government, and NPISH expenditures; non- deductible VAT from the intermediate consumption of exempt industries; and VAT from the GFCF of exempt sectors. We also account for country-specific tax regulations, such as exemptions for small businesses under the VAT thresholds (if applicable); non-deductible business expenditures on food, drinks, and accommodation; and restrictions to deduct VAT on leased cars, among others. The precise formula is given in Section c in Annex A. -4 -2 0 2 4 6 8 10 12 14 16 18 BE BG CZ DK DE EE IE EL ES FR HR IT CY LV LT LU HU MT NL AT PL PT RO SI SK FI SE UK Effective rate Compliance Base Revenue
- 23. CASE Working Paper| No 1 (2015) 23 The estimates of the VAT Gap presented in this section were derived using the same methodology as in the previously cited VAT Gap Studies. The VAT Gap is defined as the difference between the VTTL and the amount of VAT actually collected over the same period. We compute the VTTL using a top-down “consumption-side” approach by deriving the expected VAT liability from the observed national accounts data, such as supply and use tables (SUT). For this reason, the methodology used in this Study relies on the availability and quality of SUT data, which vary country to country. The VAT liability is estimated for final household, government, and NPISH expenditures; non-deductible VAT from the intermediate consumption of exempt industries; and VAT from the GFCF of exempt sectors. We also account for country-specific tax regulations, such as exemptions for small businesses under the VAT thresholds (if applicable); non-deductible business expenditures on food, drinks, and accommodation; and restrictions to deduct VAT on leased cars, among others. The precise formula is given in Section c in Annex A. The results presented in this report are not fully comparable with the results presented in the earlier Reports, as each year some figures are revised backwards. The main source of the revisions are the updates of national accounts and revenue figures compiled by Member States. Moreover, in the course of our computations, some expenditure and investment figures that are not available for the most recent years are estimated. Thus, whenever actual national accounts data is published or new information on taxable investment becomes available, VAT Gap estimates need to be revised. A detailed discussion on the sources of the revisions is presented in Section a in Annex A. In nominal terms, in 2018, the VTTL and VAT revenue amounted to EUR 1,272 billion and EUR 1,132 billion, respectively. Compared to 2017, VAT revenue increased by 4.2 percent whereas the VTTL increased by 3.6 percent, leading to decline in the VAT Gap in both relative and nominal terms. In relative terms, the EU-wide Gap dropped to 11 percent. Fast estimates show that the VAT Gap will likely continue to decline in 2019 and could fall below EUR 130 billion and 10 percent of the VTTL6 . 6 As discussed in Section d in Annex A fast estimates use a simplified methodology and their accuracy is lower. 2. The VAT Gap in 2018
- 24. CASE Reports |No. 503 (2020) 24 Figure 2.1. Evolution of the VAT Gap in the EU, 2014–2018 and Fast Estimate for 2019 Source: own calculations. The smallest Gaps were observed in Sweden (0.7 percent), Croatia (3.5 percent), and Finland (3.6 percent), the largest – in Romania (33.8 percent), Greece (30.1 percent), and Lithuania (25.9 percent). Overall, half of the EU-28 MS recorded a Gap above 9.2 percent (see Figure 2.2 and Table 2.1). In nominal terms, the largest Gaps were recorded in Italy (EUR 35.4 billion), the United Kingdom (EUR 23.5 billion), and Germany (EUR 22.1 billion). page 16 of 99 revisions are the updates of national accounts and revenue figures compiled by Member States. Moreover, in the course of our computations, some expenditure and investment figures that are not available for the most recent years are estimated. Thus, whenever actual national accounts data is published or new information on taxable investment becomes available, VAT Gap estimates need to be revised. A detailed discussion on the sources of the revisions is presented in Section a in Annex A. In nominal terms, in 2018, the VTTL and VAT revenue amounted to EUR 1,272 billion and EUR 1,132 billion, respectively. Compared to 2017, VAT revenue increased by 4.2 percent whereas the VTTL increased by 3.6 percent, leading to decline in the VAT Gap in both relative and nominal terms. In relative terms, the EU-wide Gap dropped to 11 percent. Fast estimates show that the VAT Gap will likely continue to decline in 2019 and could fall below EUR 130 billion and 10 percent of the VTTL6 . Figure 2.1. Evolution of the VAT Gap in the EU, 2014-2018 and Fast Estimate for 2019 Source: own calculations. The smallest Gaps were observed in Sweden (0.7 percent), Croatia (3.5 percent), and Finland (3.6 percent), the largest – in Romania (33.8 percent), Greece (30.1 percent), and Lithuania (25.9 percent). Overall, half of the EU-28 MS recorded a Gap above 9.2 percent (see Figure 2.2 and Table 2.1). In nominal terms, the largest Gaps were recorded in Italy (EUR 35.4 billion), the United Kingdom (EUR 23.5 billion), and Germany (EUR 22.1 billion). 6 As discussed in Section d in Annex A fast estimates use a simplified methodology and their accuracy is lower. 14.3% 13.0% 12.1% 11.5% 11.0% 9.6% 162 154 143 141 140 125 0 20 40 60 80 100 120 140 160 180 0% 2% 4% 6% 8% 10% 12% 14% 16% 2014 2015 2016 2017 2018 2019* % of VTTL (left axis) EUR billion (right axis)
- 25. CASE Reports |No. 503 (2020) 25 Figure 2.2. VAT Gap as a percent of the VTTL in EU-28 Member States, 2018 and 2017 Source: own calculations. The rank of MS with respect to the relative size of the Gap remained relatively stable, with the largest changes in position observed for Hungary and Latvia (improvement by eight and six positions, respectively). The VAT Gap share decreased in 21 countries. The most significant decreases in the VAT Gap occurred in Hungary (–5.1 percentage points), Latvia (–4.4 percentage points), and Poland (–4.3 percentage points), whereas the biggest increases were observed for Luxembourg (+2.5 percentage points), Lithuania (+0.8 percentage points), and Austria (+0.5 percentage points) (see Figure 2.3). VAT Gap in the EU-28 Member States page 17 of 99 Figure 2.2. VAT Gap as a percent of the VTTL in EU-28 Member States, 2018 and 2017 Source: own calculations. The rank of MS with respect to the relative size of the Gap remained relatively stable, with the largest changes in position observed for Hungary and Latvia (improvement by eight and six positions, respectively). The VAT Gap share decreased in 21 countries. The most significant decreases in the VAT Gap occurred in Hungary (-5.1 percentage points), Latvia (-4.4 percentage points), and Poland (-4.3 percentage points), whereas the biggest increases were observed for Luxembourg (+2.5 percentage points), Lithuania (+0.8 percentage points), and Austria (+0.5 percentage points) (see Figure 2.3). Figure 2.3. Percentage Point Change in VAT Gap, 2018 over 2017 Source: own calculations. 0 5 10 15 20 25 30 35 40 SE HR FI SI CY NL LU EE ES FR DK HU DE AT LV PT PL BE IE BG CZ UK MT SK IT LT EL RO 2017 2018 median -6 -5 -4 -3 -2 -1 0 1 2 3 HU LV PL EL MT HR FI BG PT SE CY DK BE EE NL SI RO ES IE IT DE CZ FR UK SK AT LT LU
- 26. CASE Reports |No. 503 (2020) 26 Figure 2.3. Percentage Point Change in VAT Gap, 2018 over 2017 Source: own calculations. page 17 of 99 Source: own calculations. The rank of MS with respect to the relative size of the Gap remained relatively stable, with the largest changes in position observed for Hungary and Latvia (improvement by eight and six positions, respectively). The VAT Gap share decreased in 21 countries. The most significant decreases in the VAT Gap occurred in Hungary (-5.1 percentage points), Latvia (-4.4 percentage points), and Poland (-4.3 percentage points), whereas the biggest increases were observed for Luxembourg (+2.5 percentage points), Lithuania (+0.8 percentage points), and Austria (+0.5 percentage points) (see Figure 2.3). Figure 2.3. Percentage Point Change in VAT Gap, 2018 over 2017 Source: own calculations. SE HR FI SI CY NL LU EE ES FR DK HU DE AT LV PT PL BE IE BG CZ UK MT SK IT LT EL RO 2017 2018 median -6 -5 -4 -3 -2 -1 0 1 2 3 HU LV PL EL MT HR FI BG PT SE CY DK BE EE NL SI RO ES IE IT DE CZ FR UK SK AT LT LU
- 27. CASE Reports |No. 503 (2020) 27 Figure 2.4. VAT Gap in EU Member States, 2014–2018 Source: own calculations. Figure 2.4. VAT Gap in EU Member States, 2014-2018 Source: own calculations.
- 28. CASE Reports |No. 503 (2020) 28 Table 2.1. VAT Gap as a percent of the VTTL in EU-28 Member States, 2018 and 2017 Source: own calculations. 2017 2018 VAT Gap Change (pp)MS Revenues VTTL VAT Gap VAT Gap (%) Revenues VTTL VAT Gap VAT Gap (%) BE 29,763 33,619 3,856 11.5% 31,053 34,670 3,617 10.4% −1.0 BG 4,664 5,313 649 12.2% 5,097 5,711 614 10.8% −1.5 CZ 14,703 16,694 1,991 11.9% 16,075 18,261 2,187 12.0% 0.0 DK 27,966 30,475 2,509 8.2% 29,121 31,369 2,248 7.2% −1.1 DE 226,582 248,382 21,800 8.8% 235,130 257,207 22,077 8.6% −0.2 EE 2,149 2,286 137 6.0% 2,331 2,458 127 5.2% −0.8 IE 13,060 14,652 1,592 10.9% 14,175 15,857 1,682 10.6% −0.3 EL 14,642 21,898 7,256 33.1% 15,288 21,858 6,570 30.1% −3.1 ES 73,970 79,003 5,033 6.4% 77,561 82,470 4,909 6.0% −0.4 FR 162,011 173,840 11,829 6.8% 167,618 180,406 12,788 7.1% 0.3 HR 6,465 6,843 378 5.5% 6,946 7,198 252 3.5% −2.0 IT 107,576 142,939 35,363 24.7% 109,333 144,772 35,439 24.5% −0.3 CY 1,765 1,859 93 5.0% 1,951 2,028 77 3.8% −1.2 LV 2,164 2,512 348 13.9% 2,449 2,705 256 9.5% −4.4 LT 3,310 4,422 1,111 25.1% 3,522 4,754 1,232 25.9% 0.8 LU 3,433 3,525 92 2.6% 3,729 3,928 199 5.1% 2.5 HU 11,729 13,564 1,835 13.5% 12,950 14,140 1,190 8.4% −5.1 MT 810 984 174 17.7% 920 1,084 164 15.1% −2.5 NL 49,833 52,329 2,496 4.8% 52,619 54,897 2,278 4.2% −0.6 AT 28,304 30,949 2,645 8.5% 29,323 32,231 2,908 9.0% 0.5 PL 36,330 42,374 6,044 14.3% 40,411 44,862 4,451 9.9% −4.3 PT 16,810 18,872 2,062 10.9% 17,865 19,754 1,889 9.6% −1.4 RO 11,650 17,727 6,077 34.3% 12,890 19,485 6,595 33.8% −0.4 SI 3,482 3,640 159 4.4% 3,765 3,913 148 3.8% −0.6 SK 5,919 7,362 1,443 19.6% 6,319 7,899 1,579 20.0% 0.4 FI 20,404 21,510 1,106 5.1% 21,364 22,171 807 3.6% −1.5 SE 44,115 44,987 872 1.9% 43,433 43,739 306 0.7% −1.2 UK 162,724 184,706 21,982 11.9% 168,674 192,126 23,452 12.2% 0.3 Total EU−28 1,086,332 1,227,266 140,935 11.5% 1,131,912 1,271,953 140,042 11.0% −0.5 Median 10.9% 9.2%
- 29. CASE Working Paper| No 1 (2015) 29 3. Individual Country Results Country Page Belgium 30 Bulgaria 31 Czechia 32 Denmark 33 Germany 34 Estonia 35 Ireland 36 Greece 37 Spain 38 France 40 Croatia 41 Italy 42 Cyprus 44 Latvia 45 Lithuania 46 Luxembourg 47 Hungary 48 Malta 49 Netherlands 50 Austria 51 Poland 52 Portugal 53 Romania 54 Slovenia 55 Slovakia 56 Finland 57 Sweden 58 United Kingdom 59
- 30. CASE Reports |No. 503 (2020) 30 Table 3.1. Belgium: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Belgium 2014 2015 2016 2017 2018 2019* VTTL 30,272 31,416 32,263 33,619 34,670 35,534 o/w liability on household final consumption 17,326 17,714 18,522 19,230 19,688 o/w liability on government and NPISH final consumption 1,424 1,435 1,272 1,317 1,358 o/w liability on intermediate consumption 6,103 6,675 7,017 7,289 7,520 o/w liability on GFCF 4,739 4,957 4,808 5,106 5,440 o/w net adjustments 680 634 644 676 663 Highlights · In 2018, the VAT Gap accounted for 10.4 percent of the VTTL (a decline of 1.1 percentage points compared to 2017). · The VAT revenue reported by Eurostat contains VAT assessed but unlikely to be collected. This component was removed from the reference figures to ensure comparability with other EU MS. VAT Revenue 27,518 27,594 28,750 29,763 31,053 31,679 VAT GAP 2,755 3,822 3,513 3,856 3,617 VAT GAP as a percent of VTTL 9.1% 12.2% 10.9% 11.5% 10.4% 9.4% VAT GAP change since 2014 +1.3 pp VAT Gap in the EU-28 Member States page 21 of 99 Table 3.1. Belgium: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 30,272 31,416 32,263 33,619 34,670 35,534 o/w liability on household final consumption 17,326 17,714 18,522 19,230 19,688 o/w liability on government and NPISH final consumption 1,424 1,435 1,272 1,317 1,358 o/w liability on intermediate consumption 6,103 6,675 7,017 7,289 7,520 Highlights In 2018, the VAT Gap accounted for 10.4 percent of the VTTL (a decline of 1.1 percentage points compared to 2017). The VAT revenue reported by Eurostat contains VAT assessed but unlikely to be collected. This component was removed from the reference figures to ensure comparability with other EU MS. o/w liability on GFCF 4,739 4,957 4,808 5,106 5,440 o/w net adjustments 680 634 644 676 663 VAT Revenue 27,518 27,594 28,750 29,763 31,053 31,679 VAT GAP 2,755 3,822 3,513 3,856 3,617 VAT GAP as a percent of VTTL 9.1% 12.2% 10.9% 11.5% 10.4% 9.4% VAT GAP change since 2014 +1.3 pp 9.1% 12.2% 10.9% 11.5% 10.4% 9.4% 0.0% 5.0% 10.0% 15.0% 20.0% 0 10000 20000 30000 40000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 31. CASE Reports |No. 503 (2020) 31 Table 3.2. Bulgaria: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (BGN million) Bulgaria 2014 2015 2016 2017 2018 2019* VTTL 9,576 9,867 9,852 10,391 11,169 12,363 o/w liability on household final consumption 6,910 7,071 7,257 7,779 8,279 o/w liability on government and NPISH final consumption 302 275 284 298 341 o/w liability on intermediate consumption 1,111 1,110 1,151 1,256 1,413 o/w liability on GFCF 1,174 1,328 1,143 1,044 1,110 Highlights · The VAT Gap in Bulgaria in 2018 amounted to 10.8 percent, which is about the EU total. · After a considerable improvement in 2016, the VAT Gap in Bulgaria has remained stable and is expected to remain so in 2019 based on fast estimates. o/w net adjustments 79 82 16 14 25 VAT Revenue 7,451 7,940 8,639 9,121 9,968 10,988 VAT GAP 2,124 1,927 1,213 1,270 1,201 VAT GAP as a percent of VTTL 22.2% 19.5% 12.3% 12.2% 10.8% 11.1% VAT GAP change since 2014 −11.4 pp VAT Gap in the EU-28 Member States page 22 of 99 Table 3.2. Bulgaria: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (BGN million) 2014 2015 2016 2017 2018 2019* VTTL 9,576 9,867 9,852 10,391 11,169 12,363 o/w liability on household final consumption 6,910 7,071 7,257 7,779 8,279 o/w liability on government and NPISH final consumption 302 275 284 298 341 o/w liability on intermediate consumption 1,111 1,110 1,151 1,256 1,413 Highlights The VAT Gap in Bulgaria in 2018 amounted to 10.8 percent, which is about the EU total. After a considerable improvement in 2016, the VAT Gap in Bulgaria has remained stable and is expected to remain so in 2019 based on fast estimates. o/w liability on GFCF 1,174 1,328 1,143 1,044 1,110 o/w net adjustments 79 82 16 14 25 VAT Revenue 7,451 7,940 8,639 9,121 9,968 10,988 VAT GAP 2,124 1,927 1,213 1,270 1,201 VAT GAP as a percent of VTTL 22.2% 19.5% 12.3% 12.2% 10.8% 11.1% VAT GAP change since 2014 -11.4 pp 22.2% 19.5% 12.3% 12.2% 10.8% 11.1% -1.0% 4.0% 9.0% 14.0% 19.0% 24.0% 0 2000 4000 6000 8000 10000 12000 14000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 32. CASE Reports |No. 503 (2020) 32 Table 3.3. Czechia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (CZK million) Cech Republic 2014 2015 2016 2017 2018 2019* VTTL 384,062 409,703 417,820 439,493 468,350 488,365 o/w liability on household final consumption 245,538 253,991 264,293 277,353 291,006 o/w liability on government and NPISH final consumption 19,387 21,179 21,705 21,091 23,755 o/w liability on intermediate consumption 71,811 75,118 78,614 83,448 88,367 o/w liability on GFCF 48,021 59,799 53,287 57,802 64,161 Highlights · The VAT Gap in Czechia as a percent of the VTTL remained nearly unchanged in 2018 as compared to 2017. · The revenue was amended to more accurately reflect tax accrued to taxation period on the basis of information received from the Tax Authorities. For 2018, VAT revenue reported by Eurostat was revised upwards by CZK 3.8 billion. o/w net adjustments −695 −384 −78 −201 1,061 VAT Revenue 319,485 337,774 354,181 387,074 412,271 439,441 VAT GAP 64,577 71,929 63,639 52,419 56,079 VAT GAP as a percent of VTTL 16.8% 17.6% 15.2% 11.9% 12.0% 10.8% VAT GAP change since 2014 −4.8 pp VAT Gap in the EU-28 Member States page 23 of 99 Table 3.3. Czechia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (CZK million) Cech Republic 2014 2015 2016 2017 2018 2019* VTTL 384,062 409,703 417,820 439,493 468,350 488,365 o/w liability on household final consumption 245,538 253,991 264,293 277,353 291,006 o/w liability on government and NPISH final consumption 19,387 21,179 21,705 21,091 23,755 o/w liability on intermediate consumption 71,811 75,118 78,614 83,448 88,367 Highlights The VAT Gap in Czechia as a percent of the VTTL remained nearly unchanged in 2018 as compared to 2017. The revenue was amended to more accurately reflect tax accrued to taxation period on the basis of information received from the Tax Authorities. For 2018, VAT revenue reported by Eurostat was revised upwards by CZK 3.8 billion. o/w liability on GFCF 48,021 59,799 53,287 57,802 64,161 o/w net adjustments -695 -384 -78 -201 1,061 VAT Revenue 319,485 337,774 354,181 387,074 412,271 439,441 VAT GAP 64,577 71,929 63,639 52,419 56,079 VAT GAP as a percent of VTTL 16.8% 17.6% 15.2% 11.9% 12.0% 10.8% VAT GAP change since 2014 -4.8 pp 16.8% 17.6% 15.2% 11.9% 12.0% 10.8% 0.0% 5.0% 10.0% 15.0% 20.0% 0 100000 200000 300000 400000 500000 600000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 33. CASE Reports |No. 503 (2020) 33 Table 3.4. Denmark: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (DKK million) Denmark 2014 2015 2016 2017 2018 2019* VTTL 208,401 213,396 218,207 226,691 233,799 240,382 o/w liability on household final consumption 120,503 123,843 128,717 132,514 137,422 o/w liability on government and NPISH final consumption 5,283 5,395 5,114 5,198 5,308 o/w liability on intermediate consumption 52,826 53,321 51,615 54,632 561,47 o/w liability on GFCF 24,421 25,372 27,095 28,457 28,991 Highlights · The VAT Gap in Denmark fell down to 7.2 percent of the VTTL in 2018. · Since 2014, the VAT Gap has followed a slight downward trend of about 1 percentage point per year. o/w net adjustments 5,368 5,465 5,668 5,890 5,931 VAT Revenue 185,994 191,479 199,306 208,025 217,046 221,523 VAT GAP 22,407 21,917 18,901 18,666 16,753 VAT GAP as a percent of VTTL 10.8% 10.3% 8.7% 8.2% 7.2% 7.8% VAT GAP change since 2014 -3.6 pp VAT Gap in the EU-28 Member States page 24 of 99 Table 3.4. Denmark: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (DKK million) 2014 2015 2016 2017 2018 2019* VTTL 208,401 213,396 218,207 226,691 233,799 240,382 o/w liability on household final consumption 120,503 123,843 128,717 132,514 137,422 o/w liability on government and NPISH final consumption 5,283 5,395 5,114 5,198 5,308 o/w liability on intermediate consumption 52,826 53,321 51,615 54,632 561,47 Highlights The VAT Gap in Denmark fell down to 7.2 percent of the VTTL in 2018. Since 2014, the VAT Gap has followed a slight downward trend of about 1 percentage point per year. o/w liability on GFCF 24,421 25,372 27,095 28,457 28,991 o/w net adjustments 5,368 5,465 5,668 5,890 5,931 VAT Revenue 185,994 191,479 199,306 208,025 217,046 221,523 VAT GAP 22,407 21,917 18,901 18,666 16,753 VAT GAP as a percent of VTTL 10.8% 10.3% 8.7% 8.2% 7.2% 7.8% VAT GAP change since 2014 -3.6 pp 10.8% 10.3% 8.7% 8.2% 7.2% 7.8% 0.0% 5.0% 10.0% 15.0% 20.0% 0 50000 100000 150000 200000 250000 300000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 34. CASE Reports |No. 503 (2020) 34 Table 3.5. Germany: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Germany 2014 2015 2016 2017 2018 2019* VTTL 229,881 232,507 239,911 248,382 257,207 264,502 o/w liability on household final consumption 142,430 141,011 144,979 149,029 152,971 o/w liability on government and NPISH final consumption 6,207 6,553 6,823 7,039 7,382 o/w liability on intermediate consumption 42,450 44,876 46,857 48,567 50,544 o/w liability on GFCF 37,176 37,843 39,483 41,458 44,070 Highlights · Over the period 2015–2018, the VAT Gap in Germany has remained nearly constant, amounting to ca. 9 percent of the VTTL. · The estimates for Germany were revised backwards due to an improved methodology for imputing missing and confidential values in Eurostat’s SUT. o/w net adjustments 1,618 2,223 1,769 2,290 2,239 VAT Revenue 203,081 211,616 218,779 226,582 235,130 244,111 VAT GAP 26,800 20,891 21,132 21,800 22,077 VAT GAP as a percent of VTTL 11.7% 9.0% 8.8% 8.8% 8.6% 7.7% VAT GAP change since 2014 −3.1 pp VAT Gap in the EU-28 Member States page 25 of 99 Table 3.5. Germany: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 229,881 232,507 239,911 248,382 257,207 264,502 o/w liability on household final consumption 142,430 141,011 144,979 149,029 152,971 o/w liability on government and NPISH final consumption 6,207 6,553 6,823 7,039 7,382 o/w liability on intermediate consumption 42,450 44,876 46,857 48,567 50,544 Highlights Over the period 2015-2018, the VAT Gap in Germany has remained nearly constant, amounting to ca. 9 percent of the VTTL. The estimates for Germany were revised backwards due to an improved methodology for imputing missing and confidential values in Eurostat’s SUT. o/w liability on GFCF 37,176 37,843 39,483 41,458 44,070 o/w net adjustments 1,618 2,223 1,769 2,290 2,239 VAT Revenue 203,081 211,616 218,779 226,582 235,130 244,111 VAT GAP 26,800 20,891 21,132 21,800 22,077 VAT GAP as a percent of VTTL 11.7% 9.0% 8.8% 8.8% 8.6% 7.7% VAT GAP change since 2014 -3.1 pp 11.7% 9.0% 8.8% 8.8% 8.6% 7.7% 0.0% 5.0% 10.0% 15.0% 20.0% 0 50000 100000 150000 200000 250000 300000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 35. CASE Reports |No. 503 (2020) 35 Table 3.6 Estonia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Estonia 2014 2015 2016 2017 2018 2019* VTTL 1,911 1,986 2,090 2,286 2,458 2,609 o/w liability on household final consumption 1,338 1,374 1,436 1,530 1,652 o/w liability on government and NPISH final consumption 34 35 64 69 77 o/w liability on intermediate consumption 232 244 262 282 305 o/w liability on GFCF 298 323 318 392 418 Highlights · Over the period 2015–2018, the VAT Gap in Estonia has remained stable in the range between 5 and 6 percent of the VTTL. · No substantial change in the size of the VAT Gap is expected based on fast estimates. o/w net adjustments 9 9 10 12 5 VAT Revenue 1,711 1,873 1,975 2,149 2,331 2,483 VAT GAP 200 113 115 137 127 VAT GAP as a percent of VTTL 10.4% 5.7% 5.5% 6.0% 5.2% 4.8% VAT GAP change since 2014 −5.3 pp VAT Gap in the EU-28 Member States page 26 of 99 Table 3.6. Estonia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 1,911 1,986 2,090 2,286 2,458 2,609 o/w liability on household final consumption 1,338 1,374 1,436 1,530 1,652 o/w liability on government and NPISH final consumption 34 35 64 69 77 o/w liability on intermediate consumption 232 244 262 282 305 Highlights Over the period 2015-2018, the VAT Gap in Estonia has remained stable in the range between 5 and 6 percent of the VTTL. No substantial change in the size of the VAT Gap is expected based on fast estimates. o/w liability on GFCF 298 323 318 392 418 o/w net adjustments 9 9 10 12 5 VAT Revenue 1,711 1,873 1,975 2,149 2,331 2,483 VAT GAP 200 113 115 137 127 VAT GAP as a percent of VTTL 10.4% 5.7% 5.5% 6.0% 5.2% 4.8% VAT GAP change since 2014 -5.3 pp 10.4% 5.7% 5.5% 6.0% 5.2% 4.8% 0.0% 5.0% 10.0% 15.0% 20.0% 0 500 1000 1500 2000 2500 3000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 36. CASE Reports |No. 503 (2020) 36 Table 3.7. Ireland: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Ireland 2014 2015 2016 2017 2018 2019* VTTL 12,406 13,543 14,027 14,652 15,857 15,978 o/w liability on household final consumption 7,418 7,732 7,815 8,101 8,522 o/w liability on government and NPISH final consumption 173 183 202 207 187 o/w liability on intermediate consumption 3,200 3,808 3,820 3,957 4,446 o/w liability on GFCF 1,443 1,649 1,995 2,173 2,498 Highlights · The estimates for Ireland were revised backwards due to an improved methodology for imputing missing and confidential values in Eurostat’s SUT. · The VAT Gap in Ireland is expected to fall substantially in 2019 due to increased revenues. This might be an overestimation as previous years’ fast estimates were eventually revised upwards by 2 percentage points because of more precise revenue numbers. o/w net adjustments 173 172 195 214 205 VAT Revenue 11,528 11,831 12,603 13,060 14,175 15,037 VAT GAP 878 1,712 1,425 1,592 1,682 VAT GAP as a percent of VTTL 7.1% 12.6% 10.2% 10.9% 10.6% 5.9% VAT GAP change since 2014 +3.5 pp VAT Gap in the EU-28 Member States page 27 of 99 Table 3.7. Ireland: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 12,406 13,543 14,027 14,652 15,857 15,978 o/w liability on household final consumption 7,418 7,732 7,815 8,101 8,522 o/w liability on government and NPISH final consumption 173 183 202 207 187 o/w liability on intermediate consumption 3,200 3,808 3,820 3,957 4,446 Highlights The estimates for Ireland were revised backwards due to an improved methodology for imputing missing and confidential values in Eurostat’s SUT. The VAT Gap in Ireland is expected to fall substantially in 2019 due to increased revenues. This might be an overestimation as previous years’ fast estimates were eventually revised upwards by 2 percentage points because of more precise revenue numbers. o/w liability on GFCF 1,443 1,649 1,995 2,173 2,498 o/w net adjustments 173 172 195 214 205 VAT Revenue 11,528 11,831 12,603 13,060 14,175 15,037 VAT GAP 878 1,712 1,425 1,592 1,682 VAT GAP as a percent of VTTL 7.1% 12.6% 10.2% 10.9% 10.6% 5.9% VAT GAP change since 2014 +3.5 pp 7.1% 12.6% 10.2% 10.9% 10.6% 5.9% 0.0% 5.0% 10.0% 15.0% 20.0% 0 5000 10000 15000 20000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 37. CASE Reports |No. 503 (2020) 37 Table 3.8. Greece: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Grece 2014 2015 2016 2017 2018 2019* VTTL 17,287 18,545 20,591 21,898 21,858 22,441 o/w liability on household final consumption 12,750 13,695 15,673 16,386 16,653 o/w liability on government and NPISH final consumption 424 603 673 691 689 o/w liability on intermediate consumption 1,759 1,858 2,008 2,115 2,196 o/w liability on GFCF 2,114 2,143 1,948 2,404 2,012 Highlights · VAT compliance in Greece showed a significant improvement in 2018 (a decrease of the VAT Gap by 3.1 percentage points down to 30.1 percent). · Fast estimate suggests that next year the VAT Gap will increase above 31%. o/w net adjustments 239 246 290 302 308 VAT Revenue 12,676 12,885 14,333 14,642 15,288 15,390 VAT GAP 4,611 5,660 6,258 7,256 6,570 VAT GAP as a percent of VTTL 26.7% 30.5% 30.4% 33.1% 30.1% 31.4% VAT GAP change since 2014 +3.4 pp VAT Gap in the EU-28 Member States page 28 of 99 Table 3.8. Greece: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 17,287 18,545 20,591 21,898 21,858 22,441 o/w liability on household final consumption 12,750 13,695 15,673 16,386 16,653 o/w liability on government and NPISH final consumption 424 603 673 691 689 o/w liability on intermediate consumption 1,759 1,858 2,008 2,115 2,196 Highlights VAT compliance in Greece showed a significant improvement in 2018 (a decrease of the VAT Gap by 3.1 percentage points down to 30.1 percent). Fast estimate suggests that next year the VAT Gap will increase above 31%. o/w liability on GFCF 2,114 2,143 1,948 2,404 2,012 o/w net adjustments 239 246 290 302 308 VAT Revenue 12,676 12,885 14,333 14,642 15,288 15,390 VAT GAP 4,611 5,660 6,258 7,256 6,570 VAT GAP as a percent of VTTL 26.7% 30.5% 30.4% 33.1% 30.1% 31.4% VAT GAP change since 2014 +3.4 pp 26.7% 30.5% 30.4% 33.1% 30.1% 31.4% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 0 5000 10000 15000 20000 25000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 38. CASE Reports |No. 503 (2020) 38 Table 3.9a. Spain: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Spain a 2014 2015 2016 2017 2018 2019* VTTL 69,824 72,283 74,791 79,003 82,470 83,515 o/w liability on household final consumption 50,920 52,864 55,178 57,795 59,613 o/w liability on government and NPISH final consumption 2,413 2,433 2,494 2,567 2,667 o/w liability on intermediate consumption 8,525 8,451 8,552 9,229 9,881 o/w liability on GFCF 7,311 7,777 7,891 8,708 9,576 Highlights · Between 2015 and 2018, the VAT Gap has remained relatively stable at a level of 6 percent of the VTTL. · The results were revised due to the update of Eurostat’s revenue figures. o/w net adjustments 655 759 675 704 733 VAT Revenue 62,825 67,913 70,214 73,970 77,561 79,224 VAT GAP 6,999 4,370 4,577 5,033 4,909 VAT GAP as a percent of VTTL 10.0% 6.0% 6.1% 6.4% 6.0% 3.1% VAT GAP change since 2014 −4.1 pp VAT Gap in the EU-28 Member States page 29 of 99 Table 3.9a. Spain: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 69,824 72,283 74,791 79,003 82,470 83,515 o/w liability on household final consumption 50,920 52,864 55,178 57,795 59,613 o/w liability on government and NPISH final consumption 2,413 2,433 2,494 2,567 2,667 o/w liability on intermediate consumption 8,525 8,451 8,552 9,229 9,881 Highlights Between 2015 and 2018, the VAT Gap has remained relatively stable at a level of 6 percent of the VTTL. The results were revised due to the update of Eurostat’s revenue figures. o/w liability on GFCF 7,311 7,777 7,891 8,708 9,576 o/w net adjustments 655 759 675 704 733 VAT Revenue 62,825 67,913 70,214 73,970 77,561 79,224 VAT GAP 6,999 4,370 4,577 5,033 4,909 VAT GAP as a percent of VTTL 10.0% 6.0% 6.1% 6.4% 6.0% 3.1% VAT GAP change since 2014 -4.1 pp 10.0% 6.0% 6.1% 6.4% 6.0% 3.1% 0.0% 5.0% 10.0% 15.0% 20.0% 0 20000 40000 60000 80000 100000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 39. CASE Reports |No. 503 (2020) 39 Table 3.9b. Spain: Alternative Estimates Note: Adjusting revenues for the continuing reduction in the stock of claims and adjusting the VTTL for the difference between national accounting and tax conventions in the construction sector based on the data received from Spanish Tax Authorities led to a downward revision of the VAT Gap for the entire period 2014–2018. Spain 2014 2015 2016 2017 2018 VAT Gap based on alternative data 2,946 2,177 2,680 2,925 1,737 VAT Gap based on alternative data, as a percent of VTTL 4.3% 3.1% 3.7% 3.8% 2.2%
- 40. CASE Reports |No. 503 (2020) 40 Table 3.10. France: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) France 2014 2015 2016 2017 2018 2019* VTTL 165,520 167,521 168,611 173,840 180,406 181,524 o/w liability on household final consumption 98,441 98,826 100,505 102,189 105,477 o/w liability on government and NPISH final consumption 1,606 1,631 1,695 1,734 1,750 o/w liability on intermediate consumption 27,176 30,159 30,503 31,365 32,205 o/w liability on GFCF 32,852 31,667 30,719 33,308 35,550 Highlights · The VAT Gap in 2018 remained stable compared to 2017 and amounted to 7.1 percent of the VTTL and EUR 12.8 billion. · In 2019, the VAT Gap is likely to decline. o/w net adjustments 5,445 5,238 5,189 5,244 5,424 VAT Revenue 148,454 151,680 154,490 162,011 167,618 174,356 VAT GAP 17,066 15,841 14,121 11,829 12,788 VAT GAP as a percent of VTTL 10.3% 9.5% 8.4% 6.8% 7.1% 3.9% VAT GAP change since 2014 −3.2 pp VAT Gap in the EU-28 Member States page 31 of 99 Table 3.10. France: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 165,520 167,521 168,611 173,840 180,406 181,524 o/w liability on household final consumption 98,441 98,826 100,505 102,189 105,477 o/w liability on government and NPISH final consumption 1,606 1,631 1,695 1,734 1,750 o/w liability on intermediate consumption 27,176 30,159 30,503 31,365 32,205 Highlights The VAT Gap in 2018 remained stable compared to 2017 and amounted to 7.1 percent of the VTTL and EUR 12.8 billion. In 2019, the VAT Gap is likely to decline. o/w liability on GFCF 32,852 31,667 30,719 33,308 35,550 o/w net adjustments 5,445 5,238 5,189 5,244 5,424 VAT Revenue 148,454 151,680 154,490 162,011 167,618 174,356 VAT GAP 17,066 15,841 14,121 11,829 12,788 VAT GAP as a percent of VTTL 10.3% 9.5% 8.4% 6.8% 7.1% 3.9% VAT GAP change since 2014 -3.2 pp 10.3% 9.5% 8.4% 6.8% 7.1% 3.9% 0.0% 5.0% 10.0% 15.0% 20.0% 0 50000 100000 150000 200000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 41. CASE Reports |No. 503 (2020) 41 Table 3.11. Croatia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (HRK million) Croatia 2014 2015 2016 2017 2018 2019* VTTL 45,718 48,187 48,511 51,073 53,394 55,366 o/w liability on household final consumption 33,715 34,679 35,333 37,098 38,876 o/w liability on government and NPISH final consumption 1,596 1,615 1,644 1,874 1,953 o/w liability on intermediate consumption 5,667 6,722 7,025 7,158 7,356 o/w liability on GFCF 4,485 4,508 4,274 4,737 4,958 Highlights · The VAT Gap in Croatia fell in 2018 by 2 percentage points down to 3.5 percent of the VTTL. · Since 2015, the Gap has followed a downward trend and is expected to do so in 2019 as well. o/w net adjustments 255 663 234 205 251 VAT Revenue 41,647 43,387 45,143 48,251 51,526 55,040 VAT GAP 4,071 4,800 3,368 2,822 1,868 VAT GAP as a percent of VTTL 8.9% 10.0% 6.9% 5.5% 3.5% 0.6% VAT GAP change since 2014 −5.4 pp VAT Gap in the EU-28 Member States page 32 of 99 Table 3.11. Croatia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (HRK million) 2014 2015 2016 2017 2018 2019* VTTL 45,718 48,187 48,511 51,073 53,394 55,366 o/w liability on household final consumption 33,715 34,679 35,333 37,098 38,876 o/w liability on government and NPISH final consumption 1,596 1,615 1,644 1,874 1,953 o/w liability on intermediate consumption 5,667 6,722 7,025 7,158 7,356 Highlights The VAT Gap in Croatia fell in 2018 by 2 percentage points down to 3.5 percent of the VTTL. Since 2015, the Gap has followed a downward trend and is expected to do so in 2019 as well. o/w liability on GFCF 4,485 4,508 4,274 4,737 4,958 o/w net adjustments 255 663 234 205 251 VAT Revenue 41,647 43,387 45,143 48,251 51,526 55,040 VAT GAP 4,071 4,800 3,368 2,822 1,868 VAT GAP as a percent of VTTL 8.9% 10.0% 6.9% 5.5% 3.5% 0.6% VAT GAP change since 2014 -5.4 pp 8.9% 10.0% 6.9% 5.5% 3.5% 0.6% 0.0% 5.0% 10.0% 15.0% 20.0% 0 10000 20000 30000 40000 50000 60000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 42. CASE Reports |No. 503 (2020) 42 Table 3.12a. Italy: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Italy 2014 2015 2016 2017 2018 2019* VTTL 137,817 139,703 140,400 142,939 144,772 146,855 o/w liability on household final consumption 97,232 99,621 99,890 100,918 102,246 o/w liability on government and NPISH final consumption 2,054 2,207 2,269 2,281 2,308 o/w liability on intermediate consumption 21,543 21,350 21,086 22,350 22,440 o/w liability on GFCF 13,305 13,318 13,883 14,005 14,366 Highlights · Over the analysed period, the VAT Gap in Italy has followed a downward sloping trend, reaching 24.5 percent of the VTTL in 2018. · Thanks to information provided by the Tax Authorities, the time break in the intermediate consumption of public administration in Eurostat’s SUT was corrected. o/w net adjustments 3,682 3,208 3,272 3,385 3,412 VAT Revenue 96,567 100,345 102,086 107,576 109,333 111,793 VAT GAP 41,250 39,358 38,314 35,363 35,439 VAT GAP as a percent of VTTL 29.9% 28.2% 27.3% 24.7% 24.5% 23.9% VAT GAP change since 2014 −5.5 pp VAT Gap in the EU-28 Member States page 33 of 99 Table 3.12a. Italy: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 137,817 139,703 140,400 142,939 144,772 146,855 o/w liability on household final consumption 97,232 99,621 99,890 100,918 102,246 o/w liability on government and NPISH final consumption 2,054 2,207 2,269 2,281 2,308 o/w liability on intermediate consumption 21,543 21,350 21,086 22,350 22,440 Highlights Over the analysed period, the VAT Gap in Italy has followed a downward sloping trend, reaching 24.5 percent of the VTTL in 2018. Thanks to information provided by the Tax Authorities, the time break in the intermediate consumption of public administration in Eurostat’s SUT was corrected. o/w liability on GFCF 13,305 13,318 13,883 14,005 14,366 o/w net adjustments 3,682 3,208 3,272 3,385 3,412 VAT Revenue 96,567 100,345 102,086 107,576 109,333 111,793 VAT GAP 41,250 39,358 38,314 35,363 35,439 VAT GAP as a percent of VTTL 29.9% 28.2% 27.3% 24.7% 24.5% 23.9% VAT GAP change since 2014 -5.5 pp 29.9% 28.2% 27.3% 24.7% 24.5% 23.9% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 0 50000 100000 150000 200000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 43. CASE Reports |No. 503 (2020) 43 Table 3.12b Italy: Alternative Estimates Note: The estimates above are based on adjusted revenues for the changes in outstanding stocks of net reimbursement claims (to better approximate accrued revenues) and Italy’s own estimates of illegal activities, namely illegal drugs and prostitution activities. 38,194 2014 2015 2016 2017 2018 VAT Gap based on alternative data 38,256 38,880 38,294 38,194 34,743 VAT Gap based on alternative data, as a percent of VTTL 28.1% 28.1% 27.0% 27.0% 24.0%
- 44. CASE Reports |No. 503 (2020) 44 Table 3.13. Cyprus: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2015–2018 (EUR million) Cyprus 2014 2015 2016 2017 2018 2019* VTTL N/A 1,681 1,761 1,859 2,028 o/w liability on household final consumption N/A 1,079 1,130 1,188 1,245 o/w liability on government and NPISH final consumption N/A 28 27 30 29 o/w liability on intermediate consumption N/A 437 452 447 485 o/w liability on GFCF N/A 108 134 172 243 Highlights · Thanks to information from the Tax Authorities, revenue figures were corrected to account for the expected backward revisions of Eurostat’s figures. · Due to expected revision of national accounts and an important component of the country-specific adjustments and a potentially large estimation error, fast estimates for Cyprus are not published. o/w net adjustments N/A 29 17 22 25 VAT Revenue N/A 1,517 1,664 1,765 1,951 VAT GAP N/A 165 97 93 77 VAT GAP as a percent of VTTL N/A 9.8% 5.5% 5.0% 3.8% VAT GAP change since 2014 −6.0 pp VAT Gap in the EU-28 Member States page 35 of 99 Table 3.13. Cyprus: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2015-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL N/A 1,681 1,761 1,859 2,028 o/w liability on household final consumption N/A 1,079 1,130 1,188 1,245 o/w liability on government and NPISH final consumption N/A 28 27 30 29 o/w liability on intermediate consumption N/A 437 452 447 485 Highlights Thanks to information from the Tax Authorities, revenue figures were corrected to account for the expected backward revisions of Eurostat’s figures. Due to expected revision of national accounts and an important component of the country-specific adjustments and a potentially large estimation error, fast estimates for Cyprus are not published. o/w liability on GFCF N/A 108 134 172 243 o/w net adjustments N/A 29 17 22 25 VAT Revenue N/A 1,517 1,664 1,765 1,951 VAT GAP N/A 165 97 93 77 VAT GAP as a percent of VTTL N/A 9.8% 5.5% 5.0% 3.8% VAT GAP change since 2015 -6.0 pp 9.8% 5.5% 5.0% 3.8% 0.0% 5.0% 10.0% 15.0% 20.0% 0 500 1000 1500 2000 2500 2015 2016 2017 2018 VAT GAP as a percent of VTTL VAT Revenue VTTL
- 45. CASE Reports |No. 503 (2020) 45 Table 3.14. Latvia: VAT Revenue VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Latvia 2014 2015 2016 2017 2018 2019* VTTL 2,248 2,348 2,329 2,512 2,705 2,819 o/w liability on household final consumption 1,748 1,801 1,847 1,965 2,074 o/w liability on government and NPISH final consumption 43 49 53 58 63 o/w liability on intermediate consumption 293 317 316 325 342 o/w liability on GFCF 211 238 175 227 290 Highlights · In 2018, Latvia recorded the second fastest decline of the VAT Gap in the EU by 4.4 percentage points down to 9.5 percent. · It is expected to fall further in 2019 by around 2 percentage points. o/w net adjustments −47 −57 −61 −63 −64 VAT Revenue 1,787 1,876 2,032 2,164 2,449 2,632 VAT GAP 460 472 297 348 256 VAT GAP as a percent of VTTL 20.5% 20.1% 12.8% 13.9% 9.5% 6.6% VAT GAP change since 2014 −11.0 pp VAT Gap in the EU-28 Member States page 36 of 99 Table 3.14. Latvia: VAT Revenue VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 2,248 2,348 2,329 2,512 2,705 2,819 o/w liability on household final consumption 1,748 1,801 1,847 1,965 2,074 o/w liability on government and NPISH final consumption 43 49 53 58 63 o/w liability on intermediate consumption 293 317 316 325 342 Highlights In 2018, Latvia recorded the second fastest decline of the VAT Gap in the EU by 4.4 percentage points down to 9.5 percent. It is expected to fall further in 2019 by around 2 percentage points. o/w liability on GFCF 211 238 175 227 290 o/w net adjustments -47 -57 -61 -63 -64 VAT Revenue 1,787 1,876 2,032 2,164 2,449 2,632 VAT GAP 460 472 297 348 256 VAT GAP as a percent of VTTL 20.5% 20.1% 12.8% 13.9% 9.5% 6.6% VAT GAP change since 2014 -11.0 pp 20.5% 20.1% 12.8% 13.9% 9.5% 6.6% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 0 500 1000 1500 2000 2500 3000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 46. CASE Reports |No. 503 (2020) 46 Table 3.15. Lithuania: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Lithuania 2014 2015 2016 2017 2018 2019* VTTL 3,879 3,876 4,015 4,422 4,754 4,910 o/w liability on household final consumption 3,168 3,164 3,315 3,590 3,839 o/w liability on government and NPISH final consumption 41 43 44 48 50 o/w liability on intermediate consumption 373 403 404 434 463 o/w liability on GFCF 442 461 470 505 552 Highlights · Over the period 2015–2018, the VAT Gap in Lithuania remained stable, amounting to 25 percent of the VTTL, on average. · Based on fast estimates, it is expected that the VAT Gap will fall significantly in 2019 – by about 4 percentage points. o/w net adjustments −145 −195 −218 −155 −150 VAT Revenue 2,764 2,889 3,028 3,310 3,522 3,850 VAT GAP 1,115 987 988 1,111 1,232 VAT GAP as a percent of VTTL 28.7% 25.5% 24.6% 25.1% 25.9% 21.6% VAT GAP change since 2014 −2.8 pp VAT Gap in the EU-28 Member States page 37 of 99 Table 3.15. Lithuania: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 3,879 3,876 4,015 4,422 4,754 4,910 o/w liability on household final consumption 3,168 3,164 3,315 3,590 3,839 o/w liability on government and NPISH final consumption 41 43 44 48 50 o/w liability on intermediate consumption 373 403 404 434 463 Highlights Over the period 2015-2018, the VAT Gap in Lithuania remained stable, amounting to 25 percent of the VTTL, on average. Based on fast estimates, it is expected that the VAT Gap will fall significantly in 2019 – by about 4 percentage points. o/w liability on GFCF 442 461 470 505 552 o/w net adjustments -145 -195 -218 -155 -150 VAT Revenue 2,764 2,889 3,028 3,310 3,522 3,850 VAT GAP 1,115 987 988 1,111 1,232 VAT GAP as a percent of VTTL 28.7% 25.5% 24.6% 25.1% 25.9% 21.6% VAT GAP change since 2014 -2.8 pp 28.7% 25.5% 24.6% 25.1% 25.9% 21.6% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 0 1000 2000 3000 4000 5000 6000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 47. CASE Reports |No. 503 (2020) 47 Table 3.16. Luxembourg: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) Luxemburg 2014 2015 2016 2017 2018 2019* VTTL 3,888 3,510 3,736 3,525 3,928 o/w liability on household final consumption 1,237 1,289 1,331 1,361 1,469 o/w liability on government and NPISH final consumption 30 32 33 44 89 o/w liability on intermediate consumption 875 1,070 1,138 1,160 1,215 o/w liability on GFCF 348 411 626 541 726 Highlights · In 2018, the VAT Gap was 5.1 percent of the VTTL, which was a 2.5 percentage point incline year-over-year. · Due to an important component of the country-specific adjustments related to e-commerce and financial intermediation services and a potentially large estimation error, fast estimates for Luxemburg are not published. o/w net adjustments 1,398 709 608 419 429 VAT Revenue 3,749 3,420 3,422 3,433 3,729 VAT GAP 139 90 314 92 199 VAT GAP as a percent of VTTL 3.6% 2.6% 8.4% 2.6% 5.1% VAT GAP change since 2014 +1.5 pp VAT Gap in the EU-28 Member States page 38 of 99 Table 3.16. Luxembourg: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 3,888 3,510 3,736 3,525 3,928 o/w liability on household final consumption 1,237 1,289 1,331 1,361 1,469 o/w liability on government and NPISH final consumption 30 32 33 44 89 o/w liability on intermediate consumption 875 1,070 1,138 1,160 1,215 Highlights In 2018, the VAT Gap was 5.1 percent of the VTTL, which was a 2.5 percentage point incline year-over-year. Due to an important component of the country-specific adjustments related to e-commerce and financial intermediation services and a potentially large estimation error, fast estimates for Luxemburg are not published. o/w liability on GFCF 348 411 626 541 726 o/w net adjustments 1,398 709 608 419 429 VAT Revenue 3,749 3,420 3,422 3,433 3,729 VAT GAP 139 90 314 92 199 VAT GAP as a percent of VTTL 3.6% 2.6% 8.4% 2.6% 5.1% VAT GAP change since 2014 +1.5 pp 3.6% 2.6% 8.4% 2.6% 5.1% 0.0% 5.0% 10.0% 15.0% 20.0% 0 1000 2000 3000 4000 5000 2014 2015 2016 2017 2018 VAT GAP as a percent of VTTL VAT Revenue VTTL
- 48. CASE Reports |No. 503 (2020) 48 Hungary 2014 2015 2016 2017 2018 2019* VTTL 3,695,038 3,934,985 3,842,561 4,193,962 4,509,050 4,847,886 o/w liability on household final consumption 2,561,233 2,667,644 2,813,513 2,928,236 3,037,227 o/w liability on government and NPISH final consumption 114,447 121,681 112,677 123,619 131,027 o/w liability on intermediate consumption 495,980 529,845 527,033 562,286 608,761 o/w liability on GFCF 464,953 560,845 340,200 520,047 690,748 Highlights · In 2018, Hungary recorded the fastest decline of the VAT Gap in the EU – 5.1 percentage points down to 8.4 percent. · It is expected to decline further in 2019, but only by 1 percentage point. o/w net adjustments 58,426 54,969 49,138 59,774 41,287 VAT Revenue 3,011,162 3,309,540 3,299,838 3,626,566 4,129,537 4,526,757 VAT GAP 683,876 625,445 542,723 567,396 379,513 VAT GAP as a percent of VTTL 18.5% 15.9% 14.1% 13.5% 8.4% 6.6% VAT GAP change since 2014 −10.1 pp Table 3.17. Hungary: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (HUF million) VAT Gap in the EU-28 Member States page 39 of 99 Table 3.17. Hungary: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (HUF million) 2014 2015 2016 2017 2018 2019* VTTL 3,695,038 3,934,985 3,842,561 4,193,962 4,509,050 4,847,886 o/w liability on household final consumption 2,561,233 2,667,644 2,813,513 2,928,236 3,037,227 o/w liability on government and NPISH final consumption 114,447 121,681 112,677 123,619 131,027 o/w liability on intermediate consumption 495,980 529,845 527,033 562,286 608,761 Highlights In 2018, Hungary recorded the fastest decline of the VAT Gap in the EU – 5.1 percentage points down to 8.4 percent. It is expected to decline further in 2019, but only by 1 percentage point. o/w liability on GFCF 464,953 560,845 340,200 520,047 690,748 o/w net adjustments 58,426 54,969 49,138 59,774 41,287 VAT Revenue 3,011,162 3,309,540 3,299,838 3,626,566 4,129,537 4,526,757 VAT GAP 683,876 625,445 542,723 567,396 379,513 VAT GAP as a percent of VTTL 18.5% 15.9% 14.1% 13.5% 8.4% 6.6% VAT GAP change since 2014 -10.1 pp 18.5% 15.9% 14.1% 13.5% 8.4% 6.6% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 0 1000000 2000000 3000000 4000000 5000000 6000000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 49. CASE Reports |No. 503 (2020) 49 Malta 2014 2015 2016 2017 2018 2019* VTTL 935 861 925 984 1,084 1,110 o/w liability on household final consumption 460 488 517 538 582 o/w liability on government and NPISH final consumption 16 18 49 55 60 o/w liability on intermediate consumption 393 253 277 301 337 o/w liability on GFCF 63 82 58 72 88 Highlights · The VAT Gap in Malta fell by approximately 2.5 percent- age points in 2018 down to 15.1 percent of the VTTL. · As a net exporter of electronic services, VTTL and revenue in Malta was af- fected by the withdrawal of the MOSS retention fee as of 2019. · The VTTL in Malta was revised significantly upwards thanks to the availability of data from fiscal registers allowing for more accurate estimations of the effective rates and propexes for financial and gambling services. o/w net adjustments 2 20 24 18 18 VAT Revenue 642 673 712 810 920 934 VAT GAP 293 188 213 174 164 VAT GAP as a percent of VTTL 31.3% 21.8% 23.0% 17.7% 15.1% 16.8% VAT GAP change since 2014 −16.2 pp Table 3.18 Malta: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) VAT Gap in the EU-28 Member States page 40 of 99 Table 3.18. Malta: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 935 861 925 984 1,084 1,110 o/w liability on household final consumption 460 488 517 538 582 o/w liability on government and NPISH final consumption 16 18 49 55 60 o/w liability on intermediate consumption 393 253 277 301 337 Highlights The VAT Gap in Malta fell by approximately 2.5 percentage points in 2018 down to 15.1 percent of the VTTL. As a net exporter of electronic services, VTTL and revenue in Malta was affected by the withdrawal of the MOSS retention fee as of 2019. The VTTL in Malta was revised significantly upwards thanks to the availability of data from fiscal registers allowing for more accurate estimations of the effective rates and propexes for financial and gambling services. o/w liability on GFCF 63 82 58 72 88 o/w net adjustments 2 20 24 18 18 VAT Revenue 642 673 712 810 920 934 VAT GAP 293 188 213 174 164 VAT GAP as a percent of VTTL 31.3% 21.8% 23.0% 17.7% 15.1% 16.8% VAT GAP change since 2014 -16.2 pp 31.3% 21.8% 23.0% 17.7% 15.1% 16.8% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 0 200 400 600 800 1000 1200 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 50. CASE Reports |No. 503 (2020) 50 Netherlands 2014 2015 2016 2017 2018 2019* VTTL 47,199 49,756 50,500 52,329 54,897 o/w liability on household final consumption 25,363 25,953 26,218 27,101 28,290 o/w liability on government and NPISH final consumption 556 595 571 590 621 o/w liability on intermediate consumption 12,853 13,718 13,687 14,052 14,696 o/w liability on GFCF 7867 8962 9481 10,038 10,744 Highlights · In 2018, the VAT Gap fell by 0.6 percentage points down to nearly 4 percent of the VTTL. · Due to a substantial change in the VAT rates in 2019 and a potentially large estimation error, fast estimates for the Netherlands are not published. o/w net adjustments 560 528 543 547 546 VAT Revenue 42,951 44,746 47,849 49,833 52,619 VAT GAP 4,248 5,010 2,651 2,496 2,278 VAT GAP as a percent of VTTL 9.0% 10.1% 5.3% 4.8% 4.2% VAT GAP change since 2014 −4.8 pp Table 3.19. Netherlands: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) VAT Gap in the EU-28 Member States page 40 of 99 Table 3.18. Malta: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 935 861 925 984 1,084 1,110 o/w liability on household final consumption 460 488 517 538 582 o/w liability on government and NPISH final consumption 16 18 49 55 60 o/w liability on intermediate consumption 393 253 277 301 337 Highlights The VAT Gap in Malta fell by approximately 2.5 percentage points in 2018 down to 15.1 percent of the VTTL. As a net exporter of electronic services, VTTL and revenue in Malta was affected by the withdrawal of the MOSS retention fee as of 2019. The VTTL in Malta was revised significantly upwards thanks to the availability of data from fiscal registers allowing for more accurate estimations of the effective rates and propexes for financial and gambling services. o/w liability on GFCF 63 82 58 72 88 o/w net adjustments 2 20 24 18 18 VAT Revenue 642 673 712 810 920 934 VAT GAP 293 188 213 174 164 VAT GAP as a percent of VTTL 31.3% 21.8% 23.0% 17.7% 15.1% 16.8% VAT GAP change since 2014 -16.2 pp 31.3% 21.8% 23.0% 17.7% 15.1% 16.8% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 0 200 400 600 800 1000 1200 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 51. CASE Reports |No. 503 (2020) 51 Austria 2014 2015 2016 2017 2018 2019* VTTL 27,955 28,736 29,768 30,949 32,231 32,910 o/w liability on household final consumption 18,992 19,259 19,885 20,623 21,321 o/w liability on government and NPISH final consumption 957 943 947 954 1,493 o/w liability on intermediate consumption 4,093 4,188 4,183 4,322 4,176 o/w liability on GFCF 2,585 2,890 3,284 3,467 3,676 Highlights · Over the period 2014–2018, the VAT Gap in Austria remained nearly constant, amounting to ca. 8-9 percent of the VTTL, on average. · In 2019, the VAT Gap is expected to decrease by about 1.5 percentage points. o/w net adjustments 1,328 1,456 1,469 1,583 1,566 VAT Revenue 25,386 26,247 27,301 28,304 29,323 30,446 VAT GAP 2,569 2,489 2,466 2,645 2,908 VAT GAP as a percent of VTTL 9.2% 8.7% 8.3% 8.5% 9.0% 7.5% VAT GAP change since 2014 +0.2 pp Table 3.20. Austria: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) VAT Gap in the EU-28 Member States page 42 of 99 Table 3.20. Austria: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 27,955 28,736 29,768 30,949 32,231 32,910 o/w liability on household final consumption 18,992 19,259 19,885 20,623 21,321 o/w liability on government and NPISH final consumption 957 943 947 954 1,493 o/w liability on intermediate consumption 4,093 4,188 4,183 4,322 4,176 Highlights Over the period 2014-2018, the VAT Gap in Austria remained nearly constant, amounting to ca. 8-9 percent of the VTTL, on average. In 2019, the VAT Gap is expected to decrease by about 1.5 percentage points. o/w liability on GFCF 2,585 2,890 3,284 3,467 3,676 o/w net adjustments 1,328 1,456 1,469 1,583 1,566 VAT Revenue 25,386 26,247 27,301 28,304 29,323 30,446 VAT GAP 2,569 2,489 2,466 2,645 2,908 VAT GAP as a percent of VTTL 9.2% 8.7% 8.3% 8.5% 9.0% 7.5% VAT GAP change since 2014 +0.2 pp 9.2% 8.7% 8.3% 8.5% 9.0% 7.5% 0.0% 5.0% 10.0% 15.0% 20.0% 0 5000 10000 15000 20000 25000 30000 35000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 52. CASE Reports |No. 503 (2020) 52 Poland 2014 2015 2016 2017 2018 2019* VTTL 162,348 167,037 168,993 180,386 191,180 201,610 o/w liability on household final consumption 112,465 115,495 119,692 127,010 132,706 o/w liability on government and NPISH final consumption 7,103 7,356 7,605 8,007 8,626 o/w liability on intermediate consumption 22,939 24,786 25,508 27,079 27,866 o/w liability on GFCF 16,875 17,038 13,695 15,757 19,397 Highlights · In 2018, Poland recorded the third most significant decline of the VAT Gap in the EU of 4.3 percentage points down to 9.9 percent. · The trend of significant decreases in the VAT Gap started in 2015 is expected to end in 2018 as the rate in 2019 will remain nearly identical. o/w net adjustments 2,967 2,361 2,493 2,534 2,585 VAT Revenue 122,671 125,836 134,554 154,656 172,210 182,147 VAT GAP 39,678 41,201 34,439 25,730 18,970 VAT GAP as a percent of VTTL 24.4% 24.7% 20.4% 14.3% 9.9% 9.7% VAT GAP change since 2014 −14.5 pp Table 3.21. Poland: VAT Revenue VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (PLN million) VAT Gap in the EU-28 Member States page 43 of 99 Table 3.21. Poland: VAT Revenue VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (PLN million) 2014 2015 2016 2017 2018 2019* VTTL 162,348 167,037 168,993 180,386 191,180 201,610 o/w liability on household final consumption 112,465 115,495 119,692 127,010 132,706 o/w liability on government and NPISH final consumption 7,103 7,356 7,605 8,007 8,626 o/w liability on intermediate consumption 22,939 24,786 25,508 27,079 27,866 Highlights In 2018, Poland recorded the third most significant decline of the VAT Gap in the EU of 4.3 percentage points down to 9.9 percent. The trend of significant decreases in the VAT Gap started in 2015 is expected to end in 2018 as the rate in 2019 will remain nearly identical. o/w liability on GFCF 16,875 17,038 13,695 15,757 19,397 o/w net adjustments 2,967 2,361 2,493 2,534 2,585 VAT Revenue 122,671 125,836 134,554 154,656 172,210 182,147 VAT GAP 39,678 41,201 34,439 25,730 18,970 VAT GAP as a percent of VTTL 24.4% 24.7% 20.4% 14.3% 9.9% 9.7% VAT GAP change since 2014 -14.5 pp 24.4% 24.7% 20.4% 14.3% 9.9% 9.7% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 0 50000 100000 150000 200000 250000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 53. CASE Reports |No. 503 (2020) 53 Portugal 2014 2015 2016 2017 2018 2019* VTTL 17,020 17,598 17,890 18,872 19,754 20,253 o/w liability on household final consumption 12,823 13,190 13,345 13,843 14,397 o/w liability on government and NPISH final consumption 229 444 487 535 554 o/w liability on intermediate consumption 2,625 2,433 2,732 2,928 3,088 o/w liability on GFCF 1,017 1,170 941 1,194 1,295 Highlights · The VAT Gap in Portugal was just below the EU total (9.6 percent of the VTTL). It followed a downward trend over the analysed period. Between 2014 and 2018, the Gap fell by approximately one percentage point yearly, on average. o/w net adjustments 326 361 385 372 420 VAT Revenue 14,682 15,368 15,767 16,810 17,865 18,828 VAT GAP 2,338 2,230 2,123 2,062 1,889 VAT GAP as a percent of VTTL 13.7% 12.7% 11.9% 10.9% 9.6% 7.0% VAT GAP change since 2014 −4.2 pp Table 3.22. Portugal: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) VAT Gap in the EU-28 Member States page 44 of 99 Table 3.22. Portugal: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 17,020 17,598 17,890 18,872 19,754 20,253 o/w liability on household final consumption 12,823 13,190 13,345 13,843 14,397 o/w liability on government and NPISH final consumption 229 444 487 535 554 o/w liability on intermediate consumption 2,625 2,433 2,732 2,928 3,088 Highlights The VAT Gap in Portugal was just below the EU total (9.6 percent of the VTTL). It followed a downward trend over the analysed period. Between 2014 and 2018, the Gap fell by approximately one percentage point yearly, on average. o/w liability on GFCF 1,017 1,170 941 1,194 1,295 o/w net adjustments 326 361 385 372 420 VAT Revenue 14,682 15,368 15,767 16,810 17,865 18,828 VAT GAP 2,338 2,230 2,123 2,062 1,889 VAT GAP as a percent of VTTL 13.7% 12.7% 11.9% 10.9% 9.6% 7.0% VAT GAP change since 2014 -4.2 pp 13.7% 12.7% 11.9% 10.9% 9.6% 7.0% 0.0% 5.0% 10.0% 15.0% 20.0% 0 5000 10000 15000 20000 25000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 54. CASE Reports |No. 503 (2020) 54 Romania 2014 2015 2016 2017 2018 2019* VTTL 85,971 88,269 78,520 80,993 90,682 98,353 o/w liability on household final consumption 51,889 53,728 48,986 51,803 59,786 o/w liability on government and NPISH final consumption 4,177 3,745 3,560 3,541 4,027 o/w liability on intermediate consumption 9,760 9,646 7,765 8,478 9,230 o/w liability on GFCF 16,978 18,640 16,338 15,890 16,479 Highlights · In 2018, the VAT Gap remained nearly unchanged. · Overall, between 2014 and 2018, the Gap fell by roughly 7 percentage points. · The effective rates for certain categories (such as agricultural products, restaurants, and hotels) were modified based on legislation in order to improve consistency with other countries. o/w net adjustments 3,167 2,510 1,871 1,281 1,160 VAT Revenue 51,086 57,520 49,253 53,229 59,990 65,461 VAT GAP 34,885 30,750 29,267 27,764 30,693 VAT GAP as a percent of VTTL 40.6% 34.8% 37.3% 34.3% 33.8% 33.4% VAT GAP change since 2014 −6.7 pp Table 3.23. Romania: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (RON million) VAT Gap in the EU-28 Member States page 45 of 99 Table 3.23. Romania: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (RON million) 2014 2015 2016 2017 2018 2019* VTTL 85,971 88,269 78,520 80,993 90,682 98,353 o/w liability on household final consumption 51,889 53,728 48,986 51,803 59,786 o/w liability on government and NPISH final consumption 4,177 3,745 3,560 3,541 4,027 o/w liability on intermediate consumption 9,760 9,646 7,765 8,478 9,230 Highlights In 2018, the VAT Gap remained nearly unchanged. Overall, between 2014 and 2018, the Gap fell by roughly 7 percentage points. The effective rates for certain categories (such as agricultural products, restaurants, and hotels) were modified based on legislation in order to improve consistency with other countries. o/w liability on GFCF 16,978 18,640 16,338 15,890 16,479 o/w net adjustments 3,167 2,510 1,871 1,281 1,160 VAT Revenue 51,086 57,520 49,253 53,229 59,990 65,461 VAT GAP 34,885 30,750 29,267 27,764 30,693 VAT GAP as a percent of VTTL 40.6% 34.8% 37.3% 34.3% 33.8% 33.4% VAT GAP change since 2014 -6.7 pp 40.6% 34.8% 37.3% 34.3% 33.8% 33.4% -5.0% 5.0% 15.0% 25.0% 35.0% 45.0% 0 20000 40000 60000 80000 100000 120000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 55. CASE Reports |No. 503 (2020) 55 Slovenia 2014 2015 2016 2017 2018 2019* VTTL 3,490 3,491 3,504 3,640 3,913 3,982 o/w liability on household final consumption 2,442 2,448 2,573 2,682 2,820 o/w liability on government and NPISH final consumption 69 76 85 83 89 o/w liability on intermediate consumption 491 468 469 461 523 o/w liability on GFCF 401 419 303 346 406 Highlights · The VAT Gap in Slovenia followed a downward trend over the analysed period. Between 2014 and 2018, the Gap fell by six percentage points, in total. · This trend is expected to continue into 2019 with a decrease of another 2 percentage points. o/w net adjustments 87 79 74 68 76 VAT Revenue 3,155 3,220 3,319 3,482 3,765 3,889 VAT GAP 335 271 186 159 148 VAT GAP as a percent of VTTL 9.6% 7.8% 5.3% 4.4% 3.8% 2.3% VAT GAP change since 2014 · -5.8 pp Table 3.24. Slovenia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) VAT Gap in the EU-28 Member States page 46 of 99 Table 3.24. Slovenia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 3,490 3,491 3,504 3,640 3,913 3,982 o/w liability on household final consumption 2,442 2,448 2,573 2,682 2,820 o/w liability on government and NPISH final consumption 69 76 85 83 89 o/w liability on intermediate consumption 491 468 469 461 523 Highlights The VAT Gap in Slovenia followed a downward trend over the analysed period. Between 2014 and 2018, the Gap fell by six percentage points, in total. This trend is expected to continue into 2019 with a decrease of another 2 percentage points. o/w liability on GFCF 401 419 303 346 406 o/w net adjustments 87 79 74 68 76 VAT Revenue 3,155 3,220 3,319 3,482 3,765 3,889 VAT GAP 335 271 186 159 148 VAT GAP as a percent of VTTL 9.6% 7.8% 5.3% 4.4% 3.8% 2.3% VAT GAP change since 2014 -5.8 pp 9.6% 7.8% 5.3% 4.4% 3.8% 2.3% 0.0% 5.0% 10.0% 15.0% 20.0% 0 1000 2000 3000 4000 5000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 56. CASE Reports |No. 503 (2020) 56 Slovakia 2014 2015 2016 2017 2018 2019* VTTL 7,133 7,398 6,866 7,362 7,899 8,187 o/w liability on household final consumption 5,303 5,136 5,111 5,421 5,744 o/w liability on government and NPISH final consumption 93 96 98 101 107 o/w liability on intermediate consumption 883 971 904 930 1,051 o/w liability on GFCF 869 1,206 763 916 992 Highlights · The VAT Gap in Slovakia remained stable in 2018 at just below 20 percent of the VTTL. · Over the 2014-2018 period, the Gap fell by approximately 10 percentage points. o/w net adjustments -14 -12 -10 -6 4 VAT Revenue 5,021 5,423 5,424 5,919 6,319 6,826 VAT GAP 2,112 1,975 1,443 1,443 1,579 VAT GAP as a percent of VTTL 29.6% 26.7% 21.0% 19.6% 20.0% 16.6% VAT GAP change since 2014 −9.6 pp Table 3.25 Slovakia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million) VAT Gap in the EU-28 Member States page 47 of 99 Table 3.25. Slovakia: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 7,133 7,398 6,866 7,362 7,899 8,187 o/w liability on household final consumption 5,303 5,136 5,111 5,421 5,744 o/w liability on government and NPISH final consumption 93 96 98 101 107 o/w liability on intermediate consumption 883 971 904 930 1,051 Highlights The VAT Gap in Slovakia remained stable in 2018 at just below 20 percent of the VTTL. Over the 2014-2018 period, the Gap fell by approximately 10 percentage points. o/w liability on GFCF 869 1,206 763 916 992 o/w net adjustments -14 -12 -10 -6 4 VAT Revenue 5,021 5,423 5,424 5,919 6,319 6,826 VAT GAP 2,112 1,975 1,443 1,443 1,579 VAT GAP as a percent of VTTL 29.6% 26.7% 21.0% 19.6% 20.0% 16.6% VAT GAP change since 2014 -9.6 pp 29.6% 26.7% 21.0% 19.6% 20.0% 16.6% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 0 2000 4000 6000 8000 10000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL
- 57. CASE Reports |No. 503 (2020) 57 Finland 2014 2015 2016 2017 2018 2019* VTTL 20,181 20,069 20,679 21,510 22,171 22,599 o/w liability on household final consumption 11,074 11,386 11,575 11,830 12,198 o/w liability on government and NPISH final consumption 465 478 504 490 506 o/w liability on intermediate consumption 4,545 4,276 4,396 4,589 4,654 o/w liability on GFCF 3,498 3,316 3,513 3,839 4,096 Highlights · The VAT Gap in Finland has fallen gradually throughout the entire analysed period. In 2018, it fell below 4 percent of the VTTL and EUR 1 billion. o/w net adjustments 598 613 691 761 717 VAT Revenue 18,948 18,974 19,694 20,404 21,364 21,876 VAT GAP 1,233 1,095 985 1,106 807 VAT GAP as a percent of VTTL 6.1% 5.5% 4.8% 5.1% 3.6% 3.2% VAT GAP change since 2014 −2.5 pp VAT Gap in the EU-28 Member States page 48 of 99 Table 3.26. Finland: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (EUR million) 2014 2015 2016 2017 2018 2019* VTTL 20,181 20,069 20,679 21,510 22,171 22,599 o/w liability on household final consumption 11,074 11,386 11,575 11,830 12,198 o/w liability on government and NPISH final consumption 465 478 504 490 506 o/w liability on intermediate consumption 4,545 4,276 4,396 4,589 4,654 Highlights The VAT Gap in Finland has fallen gradually throughout the entire analysed period. In 2018, it fell below 4 percent of the VTTL and EUR 1 billion.o/w liability on GFCF 3,498 3,316 3,513 3,839 4,096 o/w net adjustments 598 613 691 761 717 VAT Revenue 18,948 18,974 19,694 20,404 21,364 21,876 VAT GAP 1,233 1,095 985 1,106 807 VAT GAP as a percent of VTTL 6.1% 5.5% 4.8% 5.1% 3.6% 3.2% VAT GAP change since 2014 -2.5 pp 6.1% 5.5% 4.8% 5.1% 3.6% 3.2% 0.0% 5.0% 10.0% 15.0% 20.0% 17000 18000 19000 20000 21000 22000 23000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL Table 3.26. Finland: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (EUR million)
- 58. CASE Reports |No. 503 (2020) 58 Sweden 2014 2015 2016 2017 2018 2019* VTTL 365,287 390,123 411,285 433,453 448,689 o/w liability on household final consumption 188,086 197,435 203,952 213,174 222,949 o/w liability on government and NPISH final consumption 19,872 20,547 22,014 22,671 23,703 o/w liability on intermediate consumption 89,135 95,434 98,416 102,223 103,940 o/w liability on GFCF 62,428 70,346 80,354 88,311 90,937 Highlights · Sweden recorded the lowest VAT Gap in the EU in 2018 of about 0.7 percent of the VTTL. · Fast estimates are not reported for Sweden as they suggest a slightly negative VAT Gap. o/w net adjustments 5,766 6,360 6,548 7,075 7,160 VAT Revenue 353,439 378,830 404,987 425,053 445,550 VAT GAP 11,848 11,293 6,298 8,400 3,139 VAT GAP as a percent of VTTL 3.2% 2.9% 1.5% 1.9% 0.7% VAT GAP change since 2014 −2.5 pp Table 3.27. Sweden: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (SEK million) VAT Gap in the EU-28 Member States page 49 of 99 Table 3.27. Sweden: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (SEK million) 2014 2015 2016 2017 2018 2019* VTTL 365,287 390,123 411,285 433,453 448,689 o/w liability on household final consumption 188,086 197,435 203,952 213,174 222,949 o/w liability on government and NPISH final consumption 19,872 20,547 22,014 22,671 23,703 o/w liability on intermediate consumption 89,135 95,434 98,416 102,223 103,940 Highlights Sweden recorded the lowest VAT Gap in the EU in 2018 of about 0.7 percent of the VTTL. Fast estimates are not reported for Sweden as they suggest a slightly negative VAT Gap. o/w liability on GFCF 62,428 70,346 80,354 88,311 90,937 o/w net adjustments 5,766 6,360 6,548 7,075 7,160 VAT Revenue 353,439 378,830 404,987 425,053 445,550 VAT GAP 11,848 11,293 6,298 8,400 3,139 VAT GAP as a percent of VTTL 3.2% 2.9% 1.5% 1.9% 0.7% VAT GAP change since 2014 -2.5 pp 3.2% 2.9% 1.5% 1.9% 0.7% 0.0% 5.0% 10.0% 15.0% 20.0% 0 100000 200000 300000 400000 500000 2014 2015 2016 2017 2018 VAT GAP as a percent of VTTL VAT Revenue VTTL
- 59. CASE Reports |No. 503 (2020) 59 United Kingdom 2014 2015 2016 2017 2018 2019* VTTL 143,308 147,570 153,759 161,926 169,976 172,377 o/w liability on household final consumption 95,192 97,237 102,317 108,064 112,940 o/w liability on government and NPISH final consumption 2,560 3,420 3,045 3,085 3,159 o/w liability on intermediate consumption 31,681 32,604 33,037 33,957 35,972 o/w liability on GFCF 12,255 13,468 14,255 14,923 15,654 Highlights · The VAT Gap in the United Kingdom remained relatively stable over the 2014–2018 period. · Effective rates were revised based on the new treatment of illegal goods smuggling and the rate of exemption for education services. o/w net adjustments 1,621 840 1,105 1,898 2,252 VAT Revenue 127,647 132,948 137,531 142,655 149,228 155,104 VAT GAP 15,661 14,622 16,228 19,271 20,748 VAT GAP as a percent of VTTL 10.9% 9.9% 10.6% 11.9% 12.2% 10.0% VAT GAP change since 2014 +1.3 pp VAT Gap in the EU-28 Member States page 50 of 99 Table 3.28. United Kingdom: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014-2018 (GBP million) 2014 2015 2016 2017 2018 2019* VTTL 143,308 147,570 153,759 161,926 169,976 172,377 o/w liability on household final consumption 95,192 97,237 102,317 108,064 112,940 o/w liability on government and NPISH final consumption 2,560 3,420 3,045 3,085 3,159 o/w liability on intermediate consumption 31,681 32,604 33,037 33,957 35,972 Highlights The VAT Gap in the United Kingdom remained relatively stable over the 2014-2018 period. Effective rates were revised based on the new treatment of illegal goods smuggling and the rate of exemption for education services. o/w liability on GFCF 12,255 13,468 14,255 14,923 15,654 o/w net adjustments 1,621 840 1,105 1,898 2,252 VAT Revenue 127,647 132,948 137,531 142,655 149,228 155,104 VAT GAP 15,661 14,622 16,228 19,271 20,748 VAT GAP as a percent of VTTL 10.9% 9.9% 10.6% 11.9% 12.2% 10.0% VAT GAP change since 2014 +1.3 pp 10.9% 9.9% 10.6% 11.9% 12.2% 10.0% 0.0% 5.0% 10.0% 15.0% 20.0% 0 50000 100000 150000 200000 2014 2015 2016 2017 2018 2019* VAT GAP as a percent of VTTL VAT Revenue VTTL Table 3.28. United Kingdom: VAT Revenue, VTTL, Composition of VTTL, and VAT Gap, 2014–2018 (GBP million)
- 60. CASE Working Paper| No 1 (2015) 60 In this chapter, we present an update of the series of estimates of the Policy Gap and its components for the EU-28. As discussed in the previous Reports, the Policy Gap captures the effects of applying multiple rates and exemptions on the theoretical revenue that could be levied in a given VAT system. In other words, the Policy Gap is an indicator of the additional VAT revenue that could theoretically (i.e. under the assumption of perfect tax compliance) be generated if a uniform VAT rate is applied to the final domestic use of all goods and services. Due to the idealistic assumption of perfect tax compliance and a very broad base that captures entire final consumption and households’ GFCF, the practical interpretation of the Policy Gap draws criticism. Nonetheless, the assumption of perfect VAT collectability is indispen- sable, as interdependencies between tax compliance and rate structure are not straight- forward. In order to learn how different components contribute to revenue losses, we compose the Policy Gap into different components of revenue loss, as we show in Annex A.e. Such lements are, for instance, the Rate Gap and the Exemption Gap, which capture the loss in VAT liability due to the application of reduced rates and the loss in liability due to the implementation of exemptions, respectively. Moreover, following Barbone et al. (2013), the Policy Gap and its components could be further adjusted to address the issue of the extent to which the loss of theoretical revenue depends on the decisions of policymakers. Measures that exclude liability from the final consumption of “imputed rents” (the notional value of home occupancy by home- owners), the provision of public goods and services, and financial services. For these specific groups of services, charging VAT is impractical or currently goes beyond the control of national authorities. The estimates of the Policy Gap, Rate Gap, Exemption Gap, Actionable Policy Gap, and Actionable Exemption Gap for the EU-28 MS for 2018 are presented in Table 4.1. For the EU overall, the average Policy Gap level was 44.24 percent. This means that the VAT that could currently be levied in the case of full compliance generates 44.24 percent of what could have been generated if all the exemptions and reduced rates were 4. Policy Gap Measures for 2018
- 61. CASE Reports |No. 503 (2020) 61 abolished and all final use according to national accounts’ definition was taxed. Of this 44.24 percent, in 2018, 10.07 percentage points were due to the application of various reduced and super-reduced rates (the Rate Gap) and 34.17 were due to the application of exemptions without the right to deduct. According to the Rate Gap estimates, reduced rates are least applied in Denmark (0.77 percent), Latvia (2.37 percent), and Estonia (2.68 percent). On the other side of spectrum are Cyprus (25.97 percent) and Italy (15.86 percent). The MS with the highest values of the Exemption Gap are Spain (43.59 percent), due to the application of other than VAT indirect taxes in the Canary Islands, Ceuta, and Melilla, and the United King- dom (43.18 percent). The lowest value of the Exemption Gap was observed in Malta (15.79 percent). The largest part of the Exemption Gap is composed of exemptions on services that cannot be taxed in principle, i.e. imputed rents and the provision of public goods (26.06 percent). The remaining level of the Exemption Gap is financial services (2.33 percent) and the “Actionable” Exemption Gap, which is 5.77 percent, on average. The Actionable Policy Gap – a combination of the Rate Gap and the Actionable Exemption Gap – is 15.85 percent on average. This figure shows the combined reduction of Ideal Revenue due to reduced rates (10.07 percent) and exemptions (5.77 percent) which could possibly be removed. In three cases, i.e. the financial services Gaps in Cyprus, Ireland and Malta and the Actionable Exemption Gap in Malta, negative gaps were observed. Although theoretically possible, this likely results from a measurement error7 . 7 The Exemption Gap could become negative in periods when input VAT exceeds potential output VAT, like periods of increased investment or when losses are incurred. The measurement error may result from difficulties in decomposing the components of the base, such as sectoral GFCF and net adjustments, and inaccuracies in the underlying data and parameters.
- 62. CASE Reports |No. 503 (2020) 62 Table 4.1. Policy Gap, Rate Gap, Exemption Gap, and Actionable Gaps Source: own calculations. A B C D E F G H Policy Gap (%) Rate Gap (%) Exemption Gap (%) o/w Imputed Rents (%) o/w Public Services (%) o/w Financial Services (%) Actionable Exemption Gap (C - D - E - F) (%) Actionable Policy Gap (G + B) (%) BE 52.32 11.91 40.42 7.39 25.49 3.69 3.84 15.75 BG 29.74 3.18 26.56 10.13 14.61 1.75 0.06 3.24 CZ 39.21 5.57 33.64 8.22 17.02 2.10 6.31 11.87 DK 40.90 0.77 40.13 7.54 24.27 4.98 3.35 4.12 DE 44.15 6.76 37.39 6.72 21.30 2.78 6.58 13.35 EE 35.27 2.68 32.59 6.86 15.69 1.94 8.10 10.78 IE 48.63 12.23 36.40 10.44 23.58 -1.20 3.57 15.80 EL 45.84 8.44 37.39 9.22 16.65 1.28 10.24 18.68 ES 58.17 14.57 43.59 9.67 18.74 2.78 12.40 26.97 FR 52.92 12.93 39.99 9.37 22.01 3.14 5.47 18.39 HR 34.30 8.82 25.48 7.61 11.90 2.29 3.68 12.49 IT 53.79 15.86 37.93 10.82 18.45 1.34 7.31 23.17 CY 44.55 25.97 18.58 6.93 13.84 -5.49 3.29 29.26 LV 42.12 2.37 39.75 10.00 15.61 2.14 12.00 14.37 LT 32.97 3.83 29.14 4.49 14.52 1.73 8.40 12.23 LU 35.84 11.86 23.98 8.65 3.72 2.71 8.90 20.76 HU 45.31 8.01 37.30 7.06 17.91 3.32 9.01 17.02 MT 32.39 16.60 15.79 4.24 16.98 2.36 -7.80 8.80 NL 52.46 11.16 41.30 7.30 25.44 5.99 2.56 13.72 AT 45.07 14.76 30.32 7.66 18.76 2.74 1.15 15.91 PL 48.06 14.91 33.15 3.84 14.49 3.64 11.18 26.09 PT 50.75 14.11 36.64 8.22 19.33 3.25 5.84 19.95 RO 36.49 14.23 22.27 8.79 11.21 0.10 2.17 16.40 SI 46.94 11.71 35.23 7.66 17.27 2.70 7.60 19.31 SK 41.60 2.34 39.26 10.06 17.01 2.82 9.37 11.71 FI 50.29 9.73 40.57 10.10 21.27 3.20 6.00 15.72 SE 46.67 7.90 38.77 5.47 26.69 3.19 3.42 11.32 UK 51.97 8.78 43.18 11.70 19.79 4.00 7.68 16.47 EU-28 44.24 10.07 34.17 8.08 17.98 2.33 5.77 15.85
- 63. CASE Working Paper| No 1 (2015) 63 a. Introduction The examination of tax non-compliance determinants is not new to the economic literature. Most of the literature dealing with such factors focuses on personal income taxes, voluntary tax compliance, and deterrence effects. This focus is clearly related to data availability. The empirical studies are based mostly on micro-data gathered in surveys and audit statistics. Thus, they concentrate on the impact of individuals’ characteristics (see e.g. Feinstein [1991]). Similarly, studies scrutinising the determinants of compliance in corporate and consumption taxation usually look at micro-level revenue figures from fiscal registers or audit data (see e.g. Casey and Castro [2015]). The studies based on fiscal registers and audit and survey data face an important limitation, i.e. the inability to observe the variability of determinants across tax systems and economies. A rather limited num- ber of studies looking at such cross-country variations focus on the variation of dynamics in tax revenue (see e.g. Aizenman and Jinjarak, [2018]) or have a qualitative nature (see e.g. Keen and Smith [2007]). The European Commission’s VAT Gap Study made available a large set of standardised data on tax compliance from a group of countries with varying economic and institution- al characteristics. The series are available across a time period long enough to cover eco- nomic upturns and downturns. As a result, the Study provides an opportunity to conduct econometric analyses looking at the determinants of tax non-compliance from a new perspective. 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). The econometric analysis outlined in this Study extends the above-mentioned studies several-fold. Concerning the data preparation procedure, we eliminate potential bias in the data by correcting the VAT Gap series for each country for revisions in subsequent vintages of the Study. Moreover, we account for measurement errors, i.e. changes in the VAT Gap not related to change in compliance but rather to specific one-off factors. To deal with the scarcity of observations of exogenous variables, we perform a dummy 5. Econometric Analysis of VAT Gap Determinants
- 64. CASE Reports |No. 503 (2020) 64 variable adjustment. Although this operation rises the number of explanatory variables, overall it increases the degrees of freedom due to higher number of observations includ- ed in the estimation. In regard to the specification of the models, we extend the list of covariates relating to tax policy characteristics, macroeconomic variables, variables describing the structure of the economy, and proxies of tax fraud. b. Data and Variables Our endogenous variable is the VAT Gap of country i in year t taken from each of the European Commission’s VAT Gap Studies (i.e. the 2013, 2014, 2015, 2016, 2017, 2018, and 2019 Studies). To ensure the comparability of vintages across time, the data was transformed using the methodology described in the following section. The wide set of covariates included in the analysis originates from the 2019 Study but includes around 16 new variables8 . The covariates could be grouped as those describing tax policies, indicators of the macroeconomic situation, variables describing the exogenous factors to the tax administration economic characteristics of a country, and proxies of VAT fraud. The inclusion of tax policy characteristics is expected to show how the various efforts of tax administrations relate to the VAT Gap in each country. It could be expected that the greater the efforts of the administration are, the higher the level of tax compliance, both voluntary and involuntary. Expenditure on tax administration in relation to GDP alone might not be enough to capture how effectively the funds are used – the “IT expenditure” variable is expected to pick up the effect of innovative processes intro- duced into administrative processes. Similarly, the “Administrative effectiveness” variable, meaning the independence of the tax administration from political pressures as well as the quality of policy formulation and implementation, should account for general proficiency in collecting taxes and the credibility of government. The set of macroeconomic variables aims to explain the cyclical conditions that affect taxpayer behaviour. For example, the “Unemployment” variable should be able to capture situations when taxpayers face stronger incentives to evade tax liabilities due to the increased number of bankruptcies and liquidity constraints. Similarly, “GDP per capita” is expected to capture periods of economic stress as well as decreasing with wealth incentives not to comply. We also expect that the level of government debt could comple- ment the list of core determinants by accounting for the economic constraints and prudence of public finance. 8 See Table 5.1, EC (2019).
- 65. CASE Reports |No. 503 (2020) 65 We suspect that certain economic characteristics which show large variation across countries and rather low variation in time are also related to VAT compliance. Thus, we include variables describing the sectoral and company structure of the economy. In particular, we distinguish the retail sector, which could be the key sector, along with other labour-intensive sectors, as well as real estate, construction, industry, telecommunications, and art. The moel also takes into consideration the structure of com- panies by size of employment and the relative size of the shadow economy. One of the newly introduced variables is the value of credit transfer payments involving non-MFIs – this variable should help to explain how advanced the financial system is in terms of cashless transactions, which are more secure and easier to control by the tax administration. Since the variability of tax fraud, a significant component of the VAT Gap, may be related to very specific factors not included in the covariates list, we proxy the scale of fraud using three alternative approaches9. As one of the possible indicators of fraud, we look at international trade, as sudden changes – mostly in intra-Community purchase figures – may indicate an increasing scale of Missing Trader Intra-Community (MTIC) fraud. We also create a more refined indicator of trade at risk. This indicator was constructed by applying an algorithm which examined the differences over time in the reported values of traded goods known for being targeted by fraud (we used a list of goods that were placed under a reverse charge procedure). The relative differences between the values of trade reported by both sides were first smoothed using a moving average to limit the influence of short-term fluctuations. In the next step, this time series were treated with the k-means algorithm in order to identify possible “odd” values. In the last step, a set of filters was applied to these values in order to make sure that the discrepancies were significant and not an isolated event. The goal of this process was to identify periods where these differences were non-systematic, which in turn may indicate the emergence of fraud. In the final step, the values of the discrepancies were aggregated for each country and related to the total value of trade for goods under scrutiny. In addition, we look at the frequency of use of specific customs procedures (CPCs 42 and 63) which could be regarded as risky10 . The full list of variables is included in Table 5.1 below. 9 For a detailed analysis of fraud indicators, see EC (2018). 10 Customs Procedure Codes 42 and 63 are the regimes an importer uses in order to obtain a VAT exemption when the imported goods will be transported to another MS.
- 66. CASE Reports |No. 503 (2020) 66 Table 5.1. Variables Variable Source No. of Obs. Remarks Expected Relationship Endogenous variable VAT Gap VAT Gap reports, EC Yearly data of 26−28 MS observed between 2000 and 2017 The data will be gathered from published VAT Gap reports utilising the most recent vintage available - Tax administration variables Standardised fiscal rules index EC Full coverage Negative Number of staff OECD Available from 2003 but with missing data Data available with two-year lag (https://www.oecd-il- ibrary.org/taxation/ tax-administra- tion_23077727) Negative Number of audits completed OECD Unclear Other verification actions OECD Unclear Total administrative costs OECD Negative VAT electronic filing rate % OECD Negative IT expenditure share OECD Negative Dispersion of stat- utory tax rates EC Full coverage Taxation trends (https:// ec.europa.eu/taxa- tion_customs/business/ economic-analysis-taxa- tion/data-taxation_en) Positive Policy Gap EC 2012−2017 Positive Rate Gap EC 2012−2017 Positive Exemption Gap EC 2012−2017 Positive Macroeconomic variables Real GDP Growth EUROSTAT Full coverage Negative Debt-to-GDP Ratio EUROSTAT Full coverage Unclear General gov. sur- plus (deficit) EUROSTAT Full coverage Negative GDP at market prices EUROSTAT Full coverage Negative GDP per capita EUROSTAT Full coverage Negative Final consumption expenditure EUROSTAT Full coverage Negative Final consumption ex- penditure of households EUROSTAT Full coverage Negative Unemployment rate EUROSTAT Full coverage Positive Output gap OECD Full coverage Positive
- 67. CASE Reports |No. 503 (2020) 67 Variable Source No. of Obs. Remarks Expected Relationship Economic structure and institutional variables Economic Risk Rating ICRG Full coverage https://epub.prsgroup. com/products/icrg/ countrydata, the higher the risk the lower the value of the indexes Negative Financial Risk Rating ICRG Full coverage Negative Political Risk Rating ICRG Full coverage Negative Population EUROSTAT Full coverage Unclear Age structure EUROSTAT Full coverage Unclear Immigration EUROSTAT Full coverage Unclear Political Regime Character- istics: Political Competition INSCR Full coverage https://www.system- icpeace.org/inscrdata.html Negative Political Regime Char- acteristics: Constraint on Executive Power INSCR Full coverage Negative The Worldwide Govern- ance Indicators: Voice and Accountability World Bank Full coverage The Worldwide Govern- ance Indicators (https:// info.worldbank.org/ governance/wgi/ Home/Reports) Negative The Worldwide Gov- ernance Indicators: Political Stability World Bank Negative Government effectiveness World Bank Negative The Worldwide Gov- ernance Indicators: Regulatory Quality World Bank Negative The Worldwide Governance Indica- tors: Rule of Law World Bank Negative The Worldwide Gov- ernance Indicators: Control of Corruption World Bank Negative Population at risk of poverty EUROSTAT Full coverage Positive Share of companies with no employees EUROSTAT 2006−2017 Overall negative relation to firm size Share of companies with 1-4 employees EUROSTAT 2006−2017 Share of companies with 5-9 employees EUROSTAT 2006−2017 Share of companies with over 10 employees EUROSTAT 2006−2017
- 68. CASE Reports |No. 503 (2020) 68 Variable Source No. of Obs. Remarks Expected Relationship Share of Gross Value Added – companies with 0-9 employees EUROSTAT Full coverage Overall negative relation to firm size Share of Gross Value Added – companies with 10-19 employees EUROSTAT Full coverage Share of Gross Value Added – companies with 20-49 employees EUROSTAT Full coverage Share of Gross Value Added - companies with over 50 employees EUROSTAT Full coverage Agriculture, forestry, and fishing - sector share EUROSTAT Full coverage Unclear Industry - sector share EUROSTAT Full coverage Unclear Manufacturing - sector share EUROSTAT Full coverage Unclear Construction - sector share EUROSTAT Full coverage Unclear Wholesale and retail trade, transport, accommoda- tion, and food service activities - sector share EUROSTAT Full coverage Unclear Information and commu- nication - sector share EUROSTAT Full coverage Unclear Financial and insurance activities - sector share EUROSTAT Full coverage Unclear Real estate activi- ties - sector share EUROSTAT Full coverage Unclear Professional, scientific, and technical activities; administrative and support service activities - sector share EUROSTAT Full coverage Unclear Public administration, defence, education, human health, and social work activities - sector share EUROSTAT Full coverage Unclear Arts, entertainment and recreation…- sector share EUROSTAT Full coverage Unclear
- 69. CASE Reports |No. 503 (2020) 69 Variable Source No. of Obs. Remarks Expected Relationship Size of the shadow economy IMF 2000−2016 https://www.imf.org/ en/Publications/WP/ Issues/2019/12/13/ Explaining-the-Shad- ow-Economy-in-Eu- rope-Size-Caus- es-and-Policy-Op- tions-48821 Positive Gini Index World Bank Full coverage Unclear Electronic payments ECB Available from 2014 https://sdw.ecb. europa.eu/reports. do?node=1000001961 Negative Corruption Per- ception Index Transparency International Full coverage Higher values are related to lower per- ceived corruption https://www.transpar- ency.org/cpi2018 Negative Fraud proxies Imports with Customs Procedure Codes 42 and 63 EC 2007−2017 EC’s Surveillance Database Positive Intra-EU import at risk (share in GDP) EUROSTAT Full coverage Positive Intra-EU export at risk (share in GDP) EUROSTAT Full coverage Positive Total import EUROSTAT Full coverage Positive Import (only alcohol and tobacco) EUROSTAT Full coverage Positive Trade-at-risk Own calculation 2000-2017 Broken to importation, intra-Community acquisition, export and intra-Community supply. Positive Source: own elaboration; expected relationships based on analysis of descriptive statistics, intuition, and literature review including summary by Carfora et al. (2020).
- 70. CASE Reports |No. 503 (2020) 70 c. Methods and Approach The VAT Gap estimates presented in each release of the Study have been updated recursively whenever new information became available. Specifically, there are three different sources of VAT Gap revisions11. However, the revisions have one important property. As shown in Figure 5.1, they have a minor impact on the dynamics of the Gap for periods when full information is available. Figure 5.1. Comparison of Results (VAT Gap as % of the VTTL in EU-28) Source: own elaboration based on EC (2013, 2014, 2015, 2016, 2017, 2018, and 2019). 11 See Annex A.a. for more details. VAT Gap in the EU-28 Member States c. Methods and Approach The VAT Gap estimates presented in each release of the Study have been updated recursively whenever new information became available. Specifically, there are three different sources of VAT Gap revisions11 . However, the revisions have one important property. As shown in Figure 5.1, they have a minor impact on the dynamics of the Gap for periods when full information is available. Figure 5.1. Comparison of Results (VAT Gap as % of the VTTL in EU-28) Source: own elaboration based on EC (2013, 2014, 2015, 2016, 2017, 2018, and 2019). As the updates do not impact year-over-year changes in the VAT Gap, but only in magnitudes, we derived past estimates of the VAT Gap for each and every MS using a backcasting procedure. The backcasting procedure relies on the magnitude of values for a period of 5 years covered by the most recent estimates. At the same time, the dynamics, i.e. year-over-year changes in percentage points, for the years not covered by the full estimates are based on previous Studies (the most recent Study available including specific years). For instance, the estimates for 2000-2013 included in 2020 Study rely on the seven 11 See Annex A.a. for more details. 0 2 4 6 8 10 12 14 16 18 20 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2019 Study 2018 Study 2017 Study 2016 Study 2015 Study 2014 Study 2013 Study
- 71. CASE Reports |No. 503 (2020) 71 As the updates do not impact year-over-year changes in the VAT Gap, but only in magnitudes, we derived past estimates of the VAT Gap for each and every MS using a backcasting procedure. The backcasting procedure relies on the magnitude of values for a period of 5 years covered by the most recent estimates. At the same time, the dynamics, i.e. year-over-year changes in percentage points, for the years not covered by the full estimates are based on previous Studies (the most recent Study available including specific years). For instance, the estimates for 2000–2013 included in 2020 Study rely on the seven studies published between 2013 and 2019 but were adjusted to the magnitude of full estimates for 2014–2019. Such a procedure has not been used in any of the previous studies. In our view, despite using fixed effects specifications, such a procedure eliminates potential problems stemming from the revisions, which might be correlated both in time and across entities. For aggregate EU-wide figures, this backcasting is depicted by Figure 5.2, whereas the time series for each country are depicted by Figure 5.3. Figure 5.4 shows estimates for each country published in consecutive vintages of the Study. Figure 5.2. Backcasting of EU-wide Estimates Presented in Figure 5.1 (VAT Gap as % of the VTTL) Source: own elaboration based on EC (2013, 2014, 2015, 2016, 2017, 2018, and 2019). VAT Gap in the EU-28 Member States studies published between 2013 and 2019 but were adjusted to the magnitude of full estimates for 2014-2019. Such a procedure has not been used in any of the previous studies. In our view, despite using fixed effects specifications, such a procedure eliminates potential problems stemming from the revisions, which might be correlated both in time and across entities. For aggregate EU-wide figures, this backcasting is depicted by Figure 5.2, whereas the time series for each country are depicted by Figure 5.3. Figure 5.4 shows estimates for each country published in consecutive vintages of the Study. Figure 5.2. Backcasting of EU-wide Estimates Presented in Figure 5.1 (VAT Gap as % of the VTTL) Source: own elaboration based on EC (2013, 2014, 2015, 2016, 2017, 2018, and 2019). 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 VAT Gap estimates
- 72. CASE Reports |No. 503 (2020) 72 Figure 5.3 Backcasting of Individual Estimates (VAT Gap as % of the VTTL) Source: own elaboration based on EC (2013, 2014, 2015, 2016, 2017, 2018, and 2019) page 61 of 9 Figure 5.3. Backcasting of Individual Estimates (VAT Gap as % of the VTTL) Source: own elaboration based on EC (2013, 2014, 2015, 2016, 2017, 2018, and 2019)
- 73. CASE Reports |No. 503 (2020) 73 Figure 5.4 Individual Estimates in Consecutive Studies (VAT Gap as % of the VTTL) Source: own elaboration based on EC (2013, 2014, 2015, 2016, 2017, 2018, and 2019) VAT Gap in the EU-28 Member States Figure 5.4. Individual Estimates in Consecutive Studies (VAT Gap as % of the VTTL) Source: own elaboration based on EC (2013, 2014, 2015, 2016, 2017, 2018, and 2019)
- 74. CASE Reports |No. 503 (2020) 74 As shown in Table 5.1, the explanatory variables are often available for only a subset of observations. The nature of missing data varies across variables. Some data sources cover only specific MS (e.g. OECD), other sources are available for the most recent years only (Surveillance database) or were discontinued (e.g. Verification actions). However, there is one important similarity – data is not missing at random in most instances. The problem of the unavailability of observations markedly decreases the number of degrees of freedom in the models with numerous exogenous side variables introduced. This creates a trade-off between two econometric problems – omitted variables and insufficient degrees of freedom. To reduce the scale of the problem, we impute the values of the missing variables. We use a simple and intuitive method that partially controls the bias created by the non-random character of the missing data (Allison, 2001). The procedure for missing predictors in regression analysis that we use is called dummy variable adjustment or the missing indicator method. In this approach, if X is an incompletely observed predictor in a regression model, then a binary response indicator for X is created (RX = 1, if the value in X is miss- ing; RX = 0, if the corresponding value in X is present). Then, it is included in the regression model together with missing values in X, which are filled in with any constant value c. The method that we use increases the number of observations substantially but also creates a bias (Kleinke et al., 2011). Allison (2001) concluded that the method generally yields biased coefficient estimates and should only be applied in certain situations, for example when the unobserved value simply could not exist. The imputation could not use more refined techniques like the procedure proposed by Little and Rubin (1987) since the multivariate data are neither missing completely at random nor the conditionality of missing data could be controlled. In accordance with the Data and variables section, the basic regression takes the form12 : VGit =a1 TAVit +a2 MVit +a3 ESVit +a4 FPit +at +ai +uit The endogenous variable is the VAT Gap for country i in year t, VGit , which might be explained by the variables related directly to the actions taken by tax administrations (TAVit ), control variables describing the current macroeconomic situation (MVit ), control variables describing the characteristics of specific MS (economic structure variables - ESVit ), and fraud proxies (FPit ). These variables are characterised by a small variation over time and a relatively large variation across countries. Apart from these variables, we include 12 We also tested the alternative structure of the equation, i.e. the logarithmic form. However, the measures of the model’s fit pointed to selecting the non-log form of the model.
- 75. CASE Reports |No. 503 (2020) 75 fixed effects by country (ai ), such that the expression above is a fixed effects model, and year time effects (at ) (within estimator). Finally, is the error term with the classical statistical properties. A fixed effects model seems particularly appropriate, as one could argue some explanatory factors like the efforts of the tax administration or institutional variables might be correlated with many other factors that are not included in the regressions. The drawback is that the estimates of the fixed effects are uninterpretable, meaning that part of the variation cannot be attributed to specific factors. We are also unable to estimate the impact of the variables that show little within-country variation, as for example, level of VAT tax rates or firm size. As some of the listed variables are significantly correlated with others, we bear in mind the potential collinearity and endogeneity problem, which is tackled by the careful selection of variables for each specification. d. Results Due to the multiplicity of covariates and the enormous number of potential combinations of model specifications, we have proceeded parsimoniously. The approach consisted of three stages. In the first stage, we have run Bayesian Model Averaging to learn which variables are not significant in the majority of specifications’ variations. In the second stage, we created a correlation matrix of the remaining variables to learn which are collinear and cannot be presented in common specifications. Finally, we eliminated specifications on the basis of tests presented in Annex A. The narrow dataset obtained after the first stage consisted of 27 explanatory variables. A summary of the statistics of these variables is shown in Table 5.2.
- 76. CASE Reports |No. 503 (2020) 76 Table 5.2. Descriptive Statistics Source: own elaboration. n Mean Minimum Maximum Standard Deviation VAT Gap (endogenous) 471 0.16 0.01 0.46 0.10 Real GDP Growth 485 0.02 –0.15 0.12 0.04 Unemployment rate 485 0.09 0.02 0.28 0.04 Debt–to–GDP Ratio 483 0.57 0.04 1.79 0.33 General gov. surplus (deficit) 485 –0.03 –0.32 0.07 0.04 IT expenditure share 246 0.09 0 0.28 0.07 Policy Gap 135 0.44 0.12 0.60 0.09 Effective rate 471 0.13 0.08 0.21 0.03 Size of the shadow economy 440 0.23 0.09 0.40 0.08 Share of companies with no employees 233 0.54 0.09 0.82 0.16 Share of companies with 1–4 employees 233 0.33 0.10 0.72 0.13 Share of companies with 5–9 employees 233 0.13 0.06 0.27 0.05 Share of Gross Value Added – companies with 0–9 employees 181 0.22 0.12 0.37 0.04 Share of Gross Value Added – companies with 10–19 employees 170 0.08 0.04 0.12 0.01 Share of Gross Value Added – companies with 20–49 employees 172 0.11 0.05 0.16 0.02 Share of Gross Value Added – companies with over 50 employees 170 0.59 0.39 0.73 0.06 Agriculture, forestry and fishing – sector share 485 0.03 0.00 0.14 0.02 Construction – sector share 485 0.06 0.01 0.13 0.02 Industry – sector share 485 0.21 0.06 0.39 0.06 Wholesale and retail trade, transport, accommodation, and food service activities – sector share 485 0.21 0.10 0.32 0.04 Wholesale and retail trade, transport, accommodation, and food service activities – sector share 485 0.05 0.03 0.11 0.01 Financial and insurance activities – sector share 485 0.06 0.02 0.30 0.04 Real estate activities – sector share 485 0.09 0.05 0.19 0.02 Professional, scientific, and technical activities; administrative and support service activities – sector share 485 0.08 0.02 0.15 0.02 Public administration, defence, education, human health, and social work activities – sector share 485 0.17 0.10 0.24 0.03 Arts, entertainment, and recreation... – sector share 485 0.03 0.01 0.15 0.01 Imports with Customs Procedure Code 42 and 63 (log) 150 0.16 –2.58 4.85 1.60 Intra–EU import at risk (share in GDP) 485 0.01 0.00 0.07 0.01
- 77. CASE Reports |No. 503 (2020) 77 The results of our regressions are shown in Table 5.3. The simplest model, the baseline specification, which is later used for predictions and robustness checks, is described in column (1). As can be seen in the Table, GDP growth, general government surplus, IT expenditure, trade at risk, and the shares of the agriculture, communication services, and financial sectors are all statistically significant at the 5 percent level of significance. According to the estimation results of the baseline specification, in order to decrease the VAT Gap by one percentage point, GDP needs to increase by 3.6 percentage points more, the general government balance needs to improve by 3.4 percentage points, the share of IT expenditure in the overall expenditure of tax administrations needs to ncrease by roughly 5.4 percentage points, or the share of risky imports of goods in GDP needs to increase by one percentage point13 . The alternative specifications (columns (2) to (9)) show that a number of variables that were suspected to be related to changes in the VAT Gap appeared to be statistically insignificant at the p=0.05 level. This concerns some of the tax administration variables, i.e. the frequency of verification actions, the Fiscal Rules Index, and the frequency of electronic payments. The alternative fraud proxies, namely discrepancies in Intrastat registers and the frequency of using CPCs 42 and 64 appeared to be more weakly inter-related with the Gap as compared to the cross-border trade in risky goods. The alter native specifications also show that the share of small and medium-sized companies if measured by their share in overall employment could have a positive impact on the VAT Gap. However, due to the inter-relation between the sectoral structure of the economy and firm size, we decided to remove the firm size variable from the baseline equation. The equation with sectoral share variables appeared to translate larger proportion of variation than the equation with firm-size variables (column (5) and (6)). 13 The impact of changes in the value of exogenous variables is derived under ceteris paribus assumption, by dividing one over the respective coefficient value.
- 78. CASE Reports |No. 503 (2020) 78 Table 5.3. Econometric Specifications14 14 For illustrative purposes, Table 5.3 does not report the coefficients of fixed effects as well as two dummies that were introduced to account for the shifts of the VTTL in Malta and Ireland unrelated to a change in actual tax compliance (i.e. to filter VAT Gap measurement errors). 15 Fixed Effects (FE) specification. (1) (2) (3) (4) (5) (6) (7) (8) (9) FE15 (Baseline) FE (Shadow economy) FE (Sectors) FE (Tax admin- istration) FE (Firm size, employees) FE (Firm size, GVA) FE (CPC) FE (Trade discrepancies) FE (Fiscal prudence) Macroeconomic variables GDP growth –0.279*** –0.264*** –0.216** –0.275*** –0.322*** –0.285*** –0.294*** –0.308*** –0.277*** General gov. surplus (deficit) –0.291*** –0.279*** –0.309*** –0.302*** –0.226*** –0.206** –0.254*** –0.241*** –0.295*** Tax administration variables IT expenditure –0.184*** –0.173*** –0.182*** –0.190*** –0.148*** –0.147*** –0.172*** –0.17532*** –0.18532*** Verification actions –0.034 Electronic payments –0.838 Fiscal Rules Index 0.001 Economic structure and institutional variables Agriculture share 0.817*** 0.796*** 0.896** 0.850*** 0.836*** 0.819*** 0.840*** Manufacturing share –0.696* Construction share –0.458* Retailers share –0.103 Communication share –1.174*** –1.117*** –1.534*** –1.202*** –1.142*** –1.159*** –1.184*** Financial share –0.889*** –0.898*** –0.746* –0.852*** –0.797*** –0.826*** –0.887*** Real estate share 0.649 R&D share 0.903* Public administration share –0.641 Shadow economy size 0.163*
- 79. CASE Reports |No. 503 (2020) 79 (1) (2) (3) (4) (5) (6) (7) (8) (9) FE15 (Baseline) FE (Shadow economy) FE (Sectors) FE (Tax admin- istration) FE (Firm size, employees) FE (Firm size, GVA) FE (CPC) FE (Trade discrepancies) FE (Fiscal prudence) Small–size companies (employees) 0.272*** Medium–size companies (employees) 0.271** Micro–size companies (GVA) 0.059 Small–size companies (GVA) 0.363 Medium–size companies (GVA) –0.161 Fraud proxies Import of risky products 1.006*** 1.047*** 1.312*** 1.007*** 0.413 0.747* 0.973** CPC –0.004* Intra–EU import at risk 0.021 Constant 0.239*** 0.201*** 0.310 0.249*** –0.063 0.145*** 0.238*** 0.24005*** 0.23962*** Observations 468 468 468 468 468 468 468 468 468 R–squared 0.384 0.388 0.429 0.388 0.334 0.316 0.378 0.376 0.384 Number of id 26 26 26 26 26 26 26 26 26 Source: own elaboration, *** p<0.01, ** p<0.05, * p<0.1
- 80. CASE Reports |No. 503 (2020) 80 As a robustness check on the fixed effects specification, we show how the estimates of the model vary across time and countries. Table 5.4 shows the comparison of the base- line estimation with the estimation performed separately across different time periods: 2000–2011 (which were reported in the 2013 Study) and 2006–2017 (which were report- ed across subsequent studies). Columns 4 and 5 report the estimates for low and high VAT Gap countries. The last column shows the model estimated with the full interaction of the time period dummy and explanatory variables. In other words, such a specification allowed to differentiate the value of parameters between low and high VAT Gap Member States. Table 5.4. Robustness Check Source: own elaboration, *** p<0.01, ** p<0.05, * p<0.1 (1) (2) (3) (4) (5) (6) FE (Baseline) F E(2000-2011) FE (2006-2017) FE (LOWVG) FE (HIVG) FE(INTERX_ LOWVG) Macroeconomic variables GDP growth –0.279*** –0.381*** –0.182** 0.359* –0.384*** –0.360*** General gov. surplus (deficit) –0.291*** –0.470*** –0.098 –0.346*** –0.273** –0.299*** Tax administration variables IT expenditure –0.185*** –0.229*** –0.142*** –0.209*** –0.089 –0.123* Economic structure and institutional variables Agriculture share 0.817*** 1.077*** –0.847 –4.191*** 1.006*** 0.867*** Communica- tion share –1.174*** –1.106* –1.395*** –2.181*** –0.847* –0.846* Financial share –0.889*** –0.850*** –0.180 –0.686** –1.101*** –0.968*** Fraud proxies Import of risky products 1.006*** 1.310 0.285 0.247 0.914** 1.209*** Constant 0.240*** 0.229*** 0.237*** 0.330*** 0.265*** 0.277*** Observations 468 312 286 216 252 468 R-squared 0.384 0.333 0.469 0.355 0.479 0.422 Number of id 26 26 26 12 14 26
- 81. CASE Reports |No. 503 (2020) 81 Table 5.4 shows that the baseline model and the model estimated on the 2000–2011 period show very similar results in the values of the estimated effects. In the model estimated on the 2006–2017 time period only (reducing the observations by half), the estimates remain similarly robust. In the equations estimated on different subgroups of countries, general government balances, IT expenditure, communication, and financial sectors, as well as import of risky products remain robust as well. The largest heterogeneity is observed for the share of agricultural sector, which changes sign in the models estimated on the 2006–2017 period and low VAT Gap Member States. Moreover, GDP growth coefficient appeared not to be significant for low VAT Gap counties at the p=0.05 level. Aside from several robustness checks that were performed in order to assess the stability of the coefficients, we also look at the linear predictions for each MS (see Figure 5.5). They show that the model is accurate in predicting trends in VAT Gap changes. As Figure 5.6 shows, the model is able to attribute the majority of shifts in the overall EU VAT Gap to specific factors despite the time-effects used in the model (see Figure 5.6). The results yield an important conclusion – much of the variation in the VAT Gap, especially in periods of economic stress, comes from cyclical factors. The decrease in the VAT Gap in recent years is however only partially related to positive economic tailwinds. Most of the changes are attributed to year effects, which are likely related to efforts of tax administrations not captured by the model.
- 82. CASE Reports |No. 503 (2020) 82 Figure 5.5. Linear Predictions Broken Out by Member State Source: own elaboration. Cyprus and Croatia were not included as the estimates were unavailable for the entire analysed period.
- 83. CASE Reports |No. 503 (2020) 83 Figure 5.6. Contributions to VAT Gap Change Source: own elaboration. VAT Gap in the EU-28 Member States page 72 of 99 Figure 5.6. Contributions to VAT Gap Change Source: own elaboration. -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Year effect Deficit ratio GDP growth IT expenditure Agriculture sector (share) Communication sector (share) Financial sector (share) Import at risk VAT Gap (estimated change)
- 84. CASE Working Paper| No 1 (2015) 84 In this chapter, we examine the potential impact of the coronavirus recession on future VAT collections. The objective is to illustrate that both a decrease in the base as well as an increase in VAT non-compliance will negatively affect VAT revenue over the 2020–21 period. To conduct our forecasts, we operationalise the numerical evidence from the econometric analysis presented in the preceding chapter. We use the coefficients of the interrelations between the VAT Gap and the macroeconomic indicators in the baseline model specification and the Spring Commission’s macroeconomic forecasts as inputs. The predictions are based on the number of assumptions. Not only do we assume that the macroeconomic forecasts will be accurate, but we also assume that the control variables unrelated to the economic situation will not change. For this reason, prediction intervals are relatively large. The results for the EU are reported in the previous section, whereas the indicative results for each EU MS are shown in Annex C. The ongoing COVID-19 recession that will be covered by future VAT Gap Studies is rapidly changing the conditions for collecting VAT, which have remained favourable in recent years. Due to the pandemic, in May 2020, the European Commission significantly revised its forecast of the main economic indicators16 . It was estimated that the EU’s GDP as a whole could contract by 7.4 percent in 2020 and grow by 6.1 percent in 2021 if the following scenario materialises: a) the number of infections in the EU will remain under control even after the loosening of containment measures, b) most of the lockdown measures will be gradually lifted and economic activity will not be affected greatly by the measures that will be kept in place, and c) economic policies put in place by MS governments and the EU will prove to be effective in preventing high unemployment and mass bankruptcies. 16 At the moment of publication of this Study, more up to date (interim) Summer Forecasts became available. However, as they did not include projections of government balances necessary for our projections, they were not included herein. 6. The Potential Impact of the Coronavirus Recession on the Evolution of the VAT Gap
- 85. CASE Reports |No. 503 (2020) 85 As shown in Figure 6.1, the estimates point to a rapid decline in GDP growth and a deterioration of general government balances in 2020. As a result of the recession, the VAT Gap in 2020 is forecasted to increase by 4.1 percentage points up to 13.7 percent (Figure 6.2 and 6.3). The hike in 2020 could be more pronounced than the gradual decrease of the Gap over the three preceding years. This means that the VAT Gap, as a percent of the VTTL, will be higher than in 2016 (Figure 6.3). In nominal terms, the VAT Gap is expected to reach over EUR 164 billion in 2020. A relatively smaller increase of the nominal VAT Gap is related to the sudden decline in the base over the forecasting period. Similarly, to aggregate results, the VAT Gap in most MS will fall rapidly in 2020 and will not fully recover by 2021. The least significant decline in compliance is expected in the EU MS predicted to be least affected by the economic crisis, such as Slovakia and Poland (see Annex B, Table B7 and Annex C)17 . In 2021, the EU economies are expected to recover but only partially. It is expected that despite the stimulus measures introduced, the level of GDP in all EU MS will remain below 2019 nominal values and general government balances will be substantially worse than in 2019. If this scenario materialises, the VAT Gap in the EU would fall in relative terms compared to 2020 but would be unlikely to reach the 9.6 percent estimated for 2019. The scenario for 2021 still poses a number of uncertainties. For this reason, the model forecasts were not visualised herein. 17 The forecasts are presented only for Member States, for which fast estimates for 2019 were available, namely EU28 excluding Cyprus, Luxembourg, Malta and the Netherlands.
- 86. CASE Reports |No. 503 (2020) 86 Figure 6.1. 2020 Spring Forecasts of the European Commission (%) Source: European Commission.
- 87. CASE Reports |No. 503 (2020) 87 Figure 6.2. Change in the VAT Gap and Prediction Intervals (increments, percentage points) Source: own calculations. Figure 6.3. VAT Gap and Prediction Intervals18 (% of the VTTL) Source: own calculations. 18 The prediction intervals were estimated for 95% on the basis of the standard errors of the actual VAT Gap estimates for 2016 and 2017 and the estimates of the model using a 2001–2015 series.
- 88. CASE Working Paper| No 1 (2015) 88 This section of the Annex is based to a large extent on the methodological con- siderations already presented in earlier VAT Gap Reports. More detailed considerations regardingthe approaches to estimate the VAT Gap are presented in the seminal VAT Gap Report (Barbone et al., 2013). a. Source of Revisions of VAT Gap Estimates Every year, the estimates of the VAT Gap are updated and revised backwards. There are three different sources of such revisions: 1) Updates in the underlying national accounts data published by Eurostat: updates in VAT revenues, new supply and use tables, and revised industry-specific growth rates, among others. 2) Updates in the estimated GFCF liability, based on the new information from the own resource submissions (ORS) on taxable shares of GFCF by five sectors: households, government, NPISH, and exempt financial and non-financial enterprises. 3) Revision of the parameters of the VTTL model: effective rates, pro-rata coefficients, and net adjustments, either due to new information from ORS or due to correcting errors in the previous computation. In nominal terms, the most significant revisions in 2018 concerned Malta. The revision of the VTTL in Malta resulted from the availability of data from fiscal registers allowing for a more accurate estimation of the effective rates and propexes for four sectors crucial for the Maltese economy and its output, namely Financial services, except insurance and pension funding (NACE and CPA 64), Insurance, reinsurance and pension funding services, except compulsory social security (NACE and CPA 65), Services auxiliary to financial services and insurance services (NACE and CPA 66), and Gambling and betting services (NACE and CPA 92). Another noteworthy revision concerned Ireland and Germany. The estimates for these two countries were revised backwards due to an improved methodology for imputing missing and confidential values in Eurostat’s SUT. Annex A. Methodological Considerations
- 89. CASE Reports |No. 503 (2020) 89 b. Decomposition of VAT Revenue As VAT Revenue (VR) is the difference between the VTTL and the VAT Gap (VR = VTTL − VAT Gap, and the VTTL is a product of the effective rate and the base (VTTL = effective rate × base VAT Gap), VAT revenue could be decomposed using the following formula: Thus, the year-over-year relative change in revenue is denoted as: where VAT Gap in the EU-28 Member States b. Decomposition of VAT Revenue As VAT Revenue (VR) is the difference between the VTTL and the VAT Gap (𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑇𝑇𝑇𝑇𝑇𝑇 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺), and the VTTL is a product of the effective rate and the base (𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏), VAT revenue could be decomposed using the following formula: 𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 × 𝑉𝑉𝑉𝑉𝑉𝑉 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) Thus, the year-over-year relative change in revenue is denoted as: ∆𝑉𝑉𝑉𝑉 𝑉𝑉𝑉𝑉 = ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ where ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 denotes change in effective rate, ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 denotes change in base, and ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ denotes change in VAT compliance. c. Data Sources and Estimation Method The method used to estimate the VAT Gap in this report uses a “top-down” approach. Top- down approaches rely on national accounts, which cover the full tax base and are an exhaustive description of all productive activities. On the contrary, “bottom-up” approaches use data gathered by tax administrations including audits, surveys, and enquiry programmes. This enables us to estimate non-compliance in VAT for specific taxpayer groups as well as types of non-compliance. Within top-down approaches, VAT liability can be calculated using a “consumption-side” approach focused on the last link in the VAT chain (including intermediate consumption for exempt services) or a “production-side” approach that considers VAT due by each sector of economic activity19 . If the choice of underlying observations is random or if it is possible to estimate selection bias, a “bottom-up” approach might be used to derive the economy- wide tax gap figure. Aside from the different methodologies used, estimates of tax gaps could also be differentiated by the treatment of the tax collected by audit activities and assessed but finally not collected. The estimates presented herein show a “net” gap, meaning that they account for all revenue, including late payments and VAT collected in audit procedures. Estimates of a “gross gap” containing only the liabilities paid on time would be larger. In the “top-down consumption-side” method that is utilised in this Report, the VTTL is estimated as the sum of the liability from six main components: household, government, 19 For more details see IMF (2017). denotes change in effective rate, VAT Gap in the EU-28 Member States b. Decomposition of VAT Revenue As VAT Revenue (VR) is the difference between the VTTL and the VAT Gap (𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑇𝑇𝑇𝑇𝑇𝑇 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺), and the VTTL is a product of the effective rate and the base (𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏), VAT revenue could be decomposed using the following formula: 𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 × 𝑉𝑉𝑉𝑉𝑉𝑉 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) Thus, the year-over-year relative change in revenue is denoted as: ∆𝑉𝑉𝑉𝑉 𝑉𝑉𝑉𝑉 = ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ where ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 denotes change in effective rate, ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 denotes change in base, and ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ denotes change in VAT compliance. c. Data Sources and Estimation Method The method used to estimate the VAT Gap in this report uses a “top-down” approach. Top- down approaches rely on national accounts, which cover the full tax base and are an exhaustive description of all productive activities. On the contrary, “bottom-up” approaches use data gathered by tax administrations including audits, surveys, and enquiry programmes. This enables us to estimate non-compliance in VAT for specific taxpayer groups as well as types of non-compliance. Within top-down approaches, VAT liability can be calculated using a “consumption-side” approach focused on the last link in the VAT chain (including intermediate consumption for exempt services) or a “production-side” approach that considers VAT due by each sector of economic activity19 . If the choice of underlying observations is random or if it is possible to estimate selection bias, a “bottom-up” approach might be used to derive the economy- wide tax gap figure. Aside from the different methodologies used, estimates of tax gaps could also be differentiated by the treatment of the tax collected by audit activities and assessed but finally not collected. The estimates presented herein show a “net” gap, meaning that they account for all revenue, including late payments and VAT collected in audit procedures. Estimates of a “gross gap” containing only the liabilities paid on time would be larger. In the “top-down consumption-side” method that is utilised in this Report, the VTTL is estimated as the sum of the liability from six main components: household, government, 19 For more details see IMF (2017). denotes change in base, and VAT Gap in the EU-28 Member States b. Decomposition of VAT Revenue As VAT Revenue (VR) is the difference between the VTTL and the VAT Gap (𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑇𝑇𝑇𝑇𝑇𝑇 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺), and the VTTL is a product of the effective rate and the base (𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏), VAT revenue could be decomposed using the following formula: 𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 × 𝑉𝑉𝑉𝑉𝑉𝑉 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) Thus, the year-over-year relative change in revenue is denoted as: ∆𝑉𝑉𝑉𝑉 𝑉𝑉𝑉𝑉 = ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ where ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 denotes change in effective rate, ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 denotes change in base, and ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ denotes change in VAT compliance. c. Data Sources and Estimation Method The method used to estimate the VAT Gap in this report uses a “top-down” approach. Top- down approaches rely on national accounts, which cover the full tax base and are an exhaustive description of all productive activities. On the contrary, “bottom-up” approaches use data gathered by tax administrations including audits, surveys, and enquiry programmes. This enables us to estimate non-compliance in VAT for specific taxpayer groups as well as types of non-compliance. Within top-down approaches, VAT liability can be calculated using a “consumption-side” approach focused on the last link in the VAT chain (including intermediate consumption for exempt services) or a “production-side” approach that considers VAT due by each sector of economic activity19 . If the choice of underlying observations is random or if it is possible to estimate selection bias, a “bottom-up” approach might be used to derive the economy- wide tax gap figure. Aside from the different methodologies used, estimates of tax gaps could also be differentiated by the treatment of the tax collected by audit activities and assessed but finally not collected. The estimates presented herein show a “net” gap, meaning that they account for all revenue, including late payments and VAT collected in audit procedures. Estimates of a “gross gap” containing only the liabilities paid on time would be larger. In the “top-down consumption-side” method that is utilised in this Report, the VTTL is estimated as the sum of the liability from six main components: household, government, denotes change in VAT compliance. c. Data Sources and Estimation Method The method used to estimate the VAT Gap in this report uses a “top-down” approach. Top-down approaches rely on national accounts, which cover the full tax base and are an exhaustive description of all productive activities. On the contrary, “bottom-up” approaches use data gathered by tax administrations including audits, surveys, and en- quiry programmes. This enables us to estimate non-compliance in VAT for specific taxpayer groups as well as types of non-compliance. Within top-down approaches, VAT liability can be calculated using a “consumption-side” approach focused on the last link in the VAT chain (including intermediate consumption for exempt services) or a “production-side” approach that considers VAT due by each sector of economic activity19. If the choice of underlying observations is random or if it is possi- ble to estimate selection bias, a “bottom-up” approach might be used to derive the economy- -wide tax gap figure. 19 For more details see IMF (2017). VAT Gap in the EU-28 Member States b. Decomposition of VAT Revenue As VAT Revenue (VR) is the difference between the VTTL and the VAT Gap (𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑇𝑇𝑇𝑇𝑇𝑇 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺), and the VTTL is a product of the effective rate and the base (𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏), VAT revenue could be decomposed using the following formula: 𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 × 𝑉𝑉𝑉𝑉𝑉𝑉 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) Thus, the year-over-year relative change in revenue is denoted as: ∆𝑉𝑉𝑉𝑉 𝑉𝑉𝑉𝑉 = ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ where ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 denotes change in effective rate, ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 denotes change in base, and ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ denotes change in VAT compliance. c. Data Sources and Estimation Method The method used to estimate the VAT Gap in this report uses a “top-down” approach. Top- down approaches rely on national accounts, which cover the full tax base and are an exhaustive description of all productive activities. On the contrary, “bottom-up” approaches use data gathered by tax administrations including audits, surveys, and enquiry programmes. This enables us to estimate non-compliance in VAT for specific taxpayer groups as well as types of non-compliance. Within top-down approaches, VAT liability can be calculated using a “consumption-side” approach focused on the last link in the VAT chain (including intermediate consumption for exempt services) or a “production-side” approach that considers VAT due by each sector of economic activity19 . If the choice of underlying observations is random or if it is possible to estimate selection bias, a “bottom-up” approach might be used to derive the economy- wide tax gap figure. Aside from the different methodologies used, estimates of tax gaps could also be differentiated by the treatment of the tax collected by audit activities and assessed but finally not collected. The estimates presented herein show a “net” gap, meaning that they account for all revenue, including late payments and VAT collected in audit procedures. Estimates of a “gross gap” containing only the liabilities paid on time would be larger. In the “top-down consumption-side” method that is utilised in this Report, the VTTL is estimated as the sum of the liability from six main components: household, government, 19 For more details see IMF (2017). VAT Gap in the EU-28 Member States b. Decomposition of VAT Revenue As VAT Revenue (VR) is the difference between the VTTL and the VAT Gap (𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑇𝑇𝑇𝑇𝑇𝑇 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺), and the VTTL is a product of the effective rate and the base (𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏), VAT revenue could be decomposed using the following formula: 𝑉𝑉𝑉𝑉 = 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 × 𝑉𝑉𝑉𝑉𝑉𝑉 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) Thus, the year-over-year relative change in revenue is denoted as: ∆𝑉𝑉𝑉𝑉 𝑉𝑉𝑉𝑉 = ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 × ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 × ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ where ∆(𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 denotes change in effective rate, ∆𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 denotes change in base, and ∆ (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) (1 − 𝑉𝑉𝑉𝑉𝑉𝑉 𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 ) ⁄ denotes change in VAT compliance. c. Data Sources and Estimation Method The method used to estimate the VAT Gap in this report uses a “top-down” approach. Top- down approaches rely on national accounts, which cover the full tax base and are an exhaustive description of all productive activities. On the contrary, “bottom-up” approaches use data gathered by tax administrations including audits, surveys, and enquiry programmes. This enables us to estimate non-compliance in VAT for specific taxpayer groups as well as types of non-compliance. Within top-down approaches, VAT liability can be calculated using a “consumption-side” approach focused on the last link in the VAT chain (including intermediate consumption for exempt services) or a “production-side” approach that considers VAT due by each sector of economic activity19 . If the choice of underlying observations is random or if it is possible to estimate selection bias, a “bottom-up” approach might be used to derive the economy- wide tax gap figure. Aside from the different methodologies used, estimates of tax gaps could also be differentiated by the treatment of the tax collected by audit activities and assessed but finally not collected. The estimates presented herein show a “net” gap, meaning that they account for all revenue, including late payments and VAT collected in audit procedures. Estimates of a “gross gap” containing only the liabilities paid on time would be larger. In the “top-down consumption-side” method that is utilised in this Report, the VTTL is estimated as the sum of the liability from six main components: household, government, 19 For more details see IMF (2017).
- 90. CASE Reports |No. 503 (2020) 90 Aside from the different methodologies used, estimates of tax gaps could also be differentiated by the treatment of the tax collected by audit activities and assessed but finally not collected. The estimates presented herein show a “net” gap, meaning that they account for all revenue, including late payments and VAT collected in audit procedures. Estimates of a “gross gap” containing only the liabilities paid on time would be larger. In the “top-down consumption-side” method that is utilised in this Report, the VTTL is estimated as the sum of the liability from six main components: household, govern- ment, and NPISH final consumption; intermediate consumption; GFCF; and other, largely country-specific, adjustments. In the “top-down” approach, the VTTL is estimated using the following formula: Where: Rate is the effective rate, Value is the final consumption value, IC Value is the value of intermediate consumption, Propex is the percentage of output in a given sector that is exempt from VAT, GFCF Value is the value of gross fixed capital formation, and index i denotes sectors of the economy. To summarise, the VTTL is a product of the VAT rates and the propexes multiplied by the theoretical values of consumption and investment (plus country-specific net adjustments). For the purpose of VAT Gap estimation, roughly 10,000 parameters are estimated for each year, including the effective rates for each 2-digit CPA (i.e. ratei in the VTTL formula presented above) group of products and services and the percentage of output in a given sector that is exempt from VAT for each type of consumption (i.e. propexi in the VTTL formula presented above). For instance, for Education services (CPA no. 85) in Croatia, like for any other country and group of products and services, we estimated effective VAT Gap in the EU-28 Member States and NPISH final consumption; intermediate consumption; GFCF; and other, largely country- specific, adjustments. In the “top-down” approach, the VTTL is estimated using the following formula: 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 = ∑(𝑟𝑟𝑟𝑟𝑟𝑟𝑒𝑒𝑖𝑖 × 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑒𝑒𝑖𝑖) 𝑁𝑁 𝑖𝑖=1 + ∑(𝑟𝑟𝑟𝑟𝑟𝑟𝑒𝑒𝑖𝑖 × 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑥𝑥𝑖𝑖 × 𝐼𝐼 𝐼𝐼 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑒𝑒𝑖𝑖) 𝑁𝑁 𝑖𝑖=1 + ∑(𝑟𝑟𝑟𝑟𝑟𝑟𝑒𝑒𝑖𝑖 × 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑥𝑥𝑖𝑖 × 𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑒𝑒𝑖𝑖) + 𝑁𝑁 𝑖𝑖=1 𝑛𝑛𝑛𝑛𝑛𝑛 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 Where: Rate is the effective rate, Value is the final consumption value, IC Value is the value of intermediate consumption, Propex is the percentage of output in a given sector that is exempt from VAT, GFCF Value is the value of gross fixed capital formation, and index i denotes sectors of the economy. To summarise, the VTTL is a product of the VAT rates and the propexes multiplied by the theoretical values of consumption and investment (plus country-specific net adjustments). For the purpose of VAT Gap estimation, roughly 10,000 parameters are estimated for each year, including the effective rates for each 2-digit CPA (i.e. 𝑟𝑟𝑟𝑟𝑟𝑟𝑒𝑒𝑖𝑖 in the VTTL formula presented above) group of products and services and the percentage of output in a given sector that is exempt from VAT for each type of consumption (i.e. propexi in the VTTL formula presented above). For instance, for Education services (CPA no. 85) in Croatia, like for any other country and group of products and services, we estimated effective rates in household, government, and NPISH final consumption, as well as the percentage of output that is exempt from VAT. The main source of information is national accounts data and ORS, i.e. VAT statements provided by MS to the European Commission. In a number of specific cases where ORS information was insufficient, additional data provided by MS were used. As these data are not official Eurostat publications, we decline responsibility for
- 91. CASE Reports |No. 503 (2020) 91 rates in household, government, and NPISH final consumption, as well as the percentage of output that is exempt from VAT. The main source of information is national accounts data and ORS, i.e. VAT statements provided by MS to the European Commission. In a number of specific cases where ORS information was insufficient, additional data provided by MS were used. As these data are not official Eurostat publications, we decline responsibility for inaccuracies related to their quality. A complete description of data and sources is shown in Table A1.
- 92. CASE Reports |No. 503 (2020) 92 DESCRIPTION PURPOSE SOURCE COMMENT 1 Household expenditure by CPA/COICOP category. Estimation of effective rates for household final consumption for each 2-digit CPA category. 1ORS / HBS20 … 2 The intermediate consumption of in- dustries for which VAT on inputs can- not be deducted, pro-rata coefficients, alternatively share of exempt output. Estimation of propexes. ORS / assumptions common for all EU MS … 3 Investment (gross fixed capital formation) of exempt sectors. Estimation of VAT liability from investment. ORS / Eurostat Values forecasted two years ahead of available time series. 4 Government expenditure by CPA/COICOP category. Estimation of effective rates for government final consumption for each 2-digit CPA category of products and services. ORS Only individual government consumption and social transfers in kind specifically are a part of the tax base. However, the effective rate is estimated using a broad defi- nition of the base that includes entire government consumption. 5 NPISH expenditure by CPA/COICOP category. Estimation of effective rates for NPISH final consumption for each 2-digit CPA category of products and services. ORS … 6 VTTL adjustment due to small business exemption, business expenditure on cars and fuel, and other country- -specific adjustments. Estimation of net adjustments. ORS In general, adjustments forecasted two years ahead of available time series. 7 Final household consumption, government final consumption, NPISH final consumption, and intermediate consumption. Estimation of VTTL. Eurostat As national accounts figures do not always correspond to the tax base, two corrections to the base are applied: (1) adjustments for the self-supply of food and agricultural products and (2) adjustments for the intermediate consumption of construction work due to the treatment of construction activities abroad. If use tables are not available for a particular year or available use tables include confidential values, use tables are imputed using the RAS method21 . 8 VAT revenue. VAT revenue. Eurostat … 2 Household Budget Survey, Eurostat. 3 The RAS method is an iterative proportional fitting procedure used in a situation when only row and column sums of a desired input-output table are known. Table A1. Data Soures 20 Household Budget Survey, Eurostat. 21 The RAS method is an iterative proportional fitting procedure used in a situation when only row and column sums of a desired input-output table are known. Source: own.
- 93. CASE Reports |No. 503 (2020) 93 d. Fast VAT Gap Estimates The methodology used to estimate the VTTL for 2019 differs markedly from the one employed to estimate the VTTL for 2014–2018. The main simplifications and assumptions include: 1 Structure of household final consumption does not change with respect to 2017. In fact, due to the unavailability of up-to-date figures, it relies in most cases on a three- year lagged series. 2) Non-deductible GFCF liability changes in line with the year-over-year change in govern- ment GFCF published by AMECO 22 . 3) In the vast majority of cases where there are no significant changes in the statuary rates, net adjustments and intermediate consumption liability are rescaled from 2017 using growth rates for the entire tax base. Due to the simplified methodology, uncertainty around the “fast estimates” is sub- stantially larger than for the full estimates. For four MS, namely Cyprus, Luxembourg, the Netherlands, and Sweden, the estimation error was exceptionally large due to the considerable role of country-specific adjustments or to significant changes in the policy structure; hence, we decided not to publish these estimates. The “fast estimates” for 2019 are to be found in the Individual Country Results pages (Tables 3.1 to 3.28) and Annex B. The accuracy of the fast estimates depends on the stability of the structure of the liability components, which results, among others, from economic conditions and tax policies. Regarding the “fast estimates” for 2018 published in the 2019 Report, the direc- tion of year-over-year change was 78 percent in line with the change in sign indicated by the full estimates in the this Report. The mean prediction error was 1.05 percentage points. This relatively small error margin validates our approach and encourages us to continue the publication of the “fast estimates”. 20 Source: https://ec.europa.eu/info/business-economy-euro/indicators-statistics/economic-databases/macro-economic-data- base-ameco_en 21 22 Source: https://ec.europa.eu/info/business-economy-euro/indicators-statistics/economic-databases/macro-economic-data- base-ameco_en
- 94. CASE Reports |No. 503 (2020) 94 e. Derivation of the Policy Gap This section of the Annex defines the concepts used in Chapter 5 for estimating for- egone revenue due to policies introduced and discusses some of the methodological considerations. We begin with the Notional Ideal Revenue that, by definition, should indicate an upper limit of VAT revenue (i.e. the revenue levied at a uniform rate in the environment of per- fect tax compliance). As shown in Figure A1, ideal revenue is larger than the VTTL and sub- sequently larger than VAT collection. However, due to the existence of exemptions, it does not capture the entire VTTL and tax collection. If no exemptions were applied, neither intermediate consumption nor the GFCF of the business sector would be the base for computing the VTTL. The problem arises when deciding whether investment by the non-business sector should be part of the VAT base. According to the OECD (2014), Notional Ideal Revenue is defined as the standard rate of VAT times the aggregate net final consumption. Multiplying the standard rate and final consumption would yield, however, lower liability than in the case where a country applied no exemptions, no reduced rates, and was able to enforce all tax payments. In real life, the VTTL is comprised partially from VAT liability from investment made by households, government, and NPISH. In the case of the non-inclusion of this investment to the base, the VTTL would be partially extended beyond the ideal revenue despite “no exemptions” present in the system (see Figure A1 (c)). Policymakers can see the upper limit of VAT revenue by considering all final use categories of the household, non-profit, and government sectors. Thus, in this Report, Notional Ideal Revenue is defined as the standard rate of VAT times the aggregate net final and net GFCF of the household, non-profit, and government sectors, as recorded in the national accounts (interdependence among the various concepts presented is shown in Figure A1)23 . The Policy Gap is defined as one minus the ratio of the “legal” tax liability (i.e. the chunk of the Notional Ideal Revenue that, in the counterfactual case of perfect tax compliance, is not collected due to the presence of exemptions and reduced rates). The Policy Gap is denoted by the following formula: Policy Gap = (Notional Ideal Revenue – VTTL)/Notional Ideal Revenue 23 National accounts for most countries report final consumption on a gross (i.e. VAT-inclusive) basis. Net consumption is estimated on the basis of the gross consumption recorded in the use tables, from which VAT revenues are subtracted.
- 95. CASE Reports |No. 503 (2020) 95 The Policy Gap could be further decomposed to account for the loss of revenue. Such components are the Rate Gap and the Exemption Gap, which capture the loss in VAT liability due to the application of reduced rates and the loss in liability due to the im- plementation of exemptions. The Rate Gap is defined as the difference between the VTTL and what would be obtained in a counterfactual situation, in which the standard rate, instead of the reduced, parking, and zero rates, is applied to final consumption. Thus, the Rate Gap captures the loss in revenue that a particular country incurs by adopting multiple VAT rates instead of a single standard rate (Barbone et al., 2015). The Exemption Gap is defined as the difference between the VTTL and what would be obtained in a counterfactual situation, in which the standard rate is applied to exempt products and services, and no restriction of the right to deduct applies24 . Thus, the Exemption Gap captures the amount of revenue that might be lost because of exempted goods and services. Note that the Exemption Gap is composed of the loss in the VAT on the value added of exempt sectors, minus the VAT on their inputs, minus the VAT on GFCF inputs for these sectors. Thus, in principle, the Exemption Gap might be positive or negative (if the particular sector had negative value added, or if it had large GFCF expenditures relative to final consumption) (Barbone et al., 2015). In algebraic terms, we have the following: Definitions: VAT Gap in the EU-28 Member States products and services, and no restriction of the right to deduct applies24 . Thus, the Exemption Gap captures the amount of revenue that might be lost because of exempted goods and services. Note that the Exemption Gap is composed of the loss in the VAT on the value added of exempt sectors, minus the VAT on their inputs, minus the VAT on GFCF inputs for these sectors. Thus, in principle, the Exemption Gap might be positive or negative (if the particular sector had negative value added, or if it had large GFCF expenditures relative to final consumption) (Barbone et al., 2015). In algebraic terms, we have the following: Definitions: 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 = 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and the GFCF of non-business activities). 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 – liability from final consumption and GFCF of the non-business activities of group i of products, in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied. Actual liability from intermediate consumption and the GFCF of business activities is assumed. 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 = 𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the event where exempt products within the group are taxed at the standard rate. 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 – liability from the final consumption of group i when exempt products within the group are taxed at the standard rate. Actual liability from final consumption GFCF of non- business activities is assumed. 𝜏𝜏𝑠𝑠 – statutory rate. 𝑖𝑖 ∈ (1; 65) – sectors of the economy. – effective rate for group i of products in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and the GFCF of non-business activities). VAT Gap in the EU-28 Member States products and services, and no restriction of the right to deduct applies24 . Thus, the Exemption Gap captures the amount of revenue that might be lost because of exempted goods and services. Note that the Exemption Gap is composed of the loss in the VAT on the value added of exempt sectors, minus the VAT on their inputs, minus the VAT on GFCF inputs for these sectors. Thus, in principle, the Exemption Gap might be positive or negative (if the particular sector had negative value added, or if it had large GFCF expenditures relative to final consumption) (Barbone et al., 2015). In algebraic terms, we have the following: Definitions: 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 = 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and the GFCF of non-business activities). 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 – liability from final consumption and GFCF of the non-business activities of group i of products, in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied. Actual liability from intermediate consumption and the GFCF of business activities is assumed. 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 = 𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the event where exempt products within the group are taxed at the standard rate. 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 – liability from the final consumption of group i when exempt products within the group are taxed at the standard rate. Actual liability from final consumption GFCF of non- business activities is assumed. 𝜏𝜏𝑠𝑠 – statutory rate. 𝑖𝑖 ∈ (1; 65) – sectors of the economy. – liability from final consumption and GFCF of the non-business activities of group i of products, in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied. Actual liability from intermediate consumption and the GFCF of business activities is assumed. 24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this Report differs from that in Keen (2013). Keen (2013) defines the Rate Gap as the loss from applying reduced and zero rates to the final consumption liability, measured as a percentage of the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumu- lated in the production process as a percentage, on the contrary, of final consumption liability. Due to these definitions, the Policy Gap can be split multiplicatively into gaps attributable to reduced rates and exemptions. Since the numerator of the “ [1 – Rate Gap]” and denominator of the “[1 – Exemption Gap]” are equal, multiplication of these two components yields – VAT revenue as a percentage of Notional Ideal Revenue, which equals “[1 – Policy Gap]” (Barbone et al., 2015).
- 96. CASE Reports |No. 503 (2020) 96 page 83 of 99 In algebraic terms, we have the following: Definitions: 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 = 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and the GFCF of non-business activities). 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 – liability from final consumption and GFCF of the non-business activities of group i of products, in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied. Actual liability from intermediate consumption and the GFCF of business activities is assumed. 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 = 𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the event where exempt products within the group are taxed at the standard rate. 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 – liability from the final consumption of group i when exempt products within the group are taxed at the standard rate. Actual liability from final consumption GFCF of non- business activities is assumed. 𝜏𝜏𝑠𝑠 – statutory rate. 𝑖𝑖 ∈ (1; 65) – sectors of the economy. 24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this Report differs from that in Keen (2013). Keen (2013) defines the Rate Gap as the loss from applying reduced and zero rates to the final consumption liability, measured as a percentage of the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumulated in the production process as a percentage, on the contrary, of final consumption liability. Due to these definitions, the Policy Gap can be split multiplicatively into gaps attributable to reduced rates and exemptions. Since the numerator of the “[1 - Rate Gap]” and denominator of the “[1 - Exemption Gap]” are equal, multiplication of these two components yields – VAT revenue as a percentage of Notional Ideal Revenue, which equals “[1 - Policy Gap]” (Barbone et al., 2015). – effective rate for group i of products in the event where exempt products within the group are taxed at the standard rate. page 83 of 99 Definitions: 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 = 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and the GFCF of non-business activities). 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 – liability from final consumption and GFCF of the non-business activities of group i of products, in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied. Actual liability from intermediate consumption and the GFCF of business activities is assumed. 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 = 𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the event where exempt products within the group are taxed at the standard rate. 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 – liability from the final consumption of group i when exempt products within the group are taxed at the standard rate. Actual liability from final consumption GFCF of non- business activities is assumed. 𝜏𝜏𝑠𝑠 – statutory rate. 𝑖𝑖 ∈ (1; 65) – sectors of the economy. 24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this Report differs from that in Keen (2013). Keen (2013) defines the Rate Gap as the loss from applying reduced and zero rates to the final consumption liability, measured as a percentage of the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumulated in the production process as a percentage, on the contrary, of final consumption liability. Due to these definitions, the Policy Gap can be split multiplicatively into gaps attributable to reduced rates and exemptions. Since the numerator of the “[1 - Rate Gap]” and denominator of the “[1 - Exemption Gap]” are equal, multiplication of these two components yields – VAT revenue as a percentage of Notional Ideal Revenue, which equals “[1 - Policy Gap]” (Barbone et al., 2015). – liability from the final consumption of group i when exempt products within the group are taxed at the standard rate. Actual liability from final consumption GFCF of non-business activities is assumed. page 83 of 99 Definitions: 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 = 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and the GFCF of non-business activities). 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 – liability from final consumption and GFCF of the non-business activities of group i of products, in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied. Actual liability from intermediate consumption and the GFCF of business activities is assumed. 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 = 𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the event where exempt products within the group are taxed at the standard rate. 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 – liability from the final consumption of group i when exempt products within the group are taxed at the standard rate. Actual liability from final consumption GFCF of non- business activities is assumed. 𝜏𝜏𝑠𝑠 – statutory rate. 𝑖𝑖 ∈ (1; 65) – sectors of the economy. 24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this Report differs from that in Keen (2013). Keen (2013) defines the Rate Gap as the loss from applying reduced and zero rates to the final consumption liability, measured as a percentage of the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumulated in the production process as a percentage, on the contrary, of final consumption liability. Due to these definitions, the Policy Gap can be split multiplicatively into gaps attributable to reduced rates and exemptions. Since the numerator of the “[1 - Rate Gap]” and denominator of the “[1 - Exemption Gap]” are equal, multiplication of these two components yields – VAT revenue as a percentage of Notional Ideal Revenue, which equals “[1 - Policy Gap]” (Barbone et al., 2015). – statutory rate. page 83 of 99 Definitions: 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 = 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied (for final consumption and the GFCF of non-business activities). 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝐸𝐸 – liability from final consumption and GFCF of the non-business activities of group i of products, in the case where the standard rate instead of the zero rate, parking rate, or reduced rate is applied. Actual liability from intermediate consumption and the GFCF of business activities is assumed. 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 = 𝑉𝑉𝑇𝑇𝑇𝑇𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 – effective rate for group i of products in the event where exempt products within the group are taxed at the standard rate. 𝑉𝑉𝑉𝑉𝑉𝑉𝐿𝐿𝑖𝑖 ∗,𝑅𝑅 – liability from the final consumption of group i when exempt products within the group are taxed at the standard rate. Actual liability from final consumption GFCF of non- business activities is assumed. 𝜏𝜏𝑠𝑠 – statutory rate. 𝑖𝑖 ∈ (1; 65) – sectors of the economy. 24 The additive decomposition of the Policy Gap into the Exemption and Rate Gap presented in this Report differs from that in Keen (2013). Keen (2013) defines the Rate Gap as the loss from applying reduced and zero rates to the final consumption liability, measured as a percentage of the Notional Ideal Revenue. The Exemption Gap measures unrecovered VAT accumulated in the production process as a percentage, on the contrary, of final consumption liability. Due to these definitions, the Policy Gap can be split multiplicatively into gaps attributable to reduced rates and exemptions. Since the numerator of the “[1 - Rate Gap]” and denominator of the “[1 - Exemption Gap]” are equal, multiplication of these two components yields – VAT revenue as a percentage of Notional Ideal Revenue, which equals “[1 - Policy Gap]” (Barbone et al., 2015). – sectors of the economy. Policy Gap: Exemption Gap: Rate Gap: By definition we have: Thus: VAT Gap in the EU-28 Member States Policy Gap: 1 − 𝑃𝑃 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Exemption Gap: 1 − 𝑃𝑃𝐸𝐸 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Rate Gap: 1 − 𝑃𝑃𝑅𝑅 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) By definition we have: 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Thus: 𝑃𝑃 = 1 − ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸 VAT Gap in the EU-28 Member States Policy Gap: 1 − 𝑃𝑃 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Exemption Gap: 1 − 𝑃𝑃𝐸𝐸 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Rate Gap: 1 − 𝑃𝑃𝑅𝑅 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) By definition we have: 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Thus: 𝑃𝑃 = 1 − ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸 Using the above convention, one can decompose the Rate Gap and the Exemption Gap into components indicating the loss of the Notional Ideal Revenue due to the implementation VAT Gap in the EU-28 Member States Policy Gap: 1 − 𝑃𝑃 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Exemption Gap: 1 − 𝑃𝑃𝐸𝐸 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Rate Gap: 1 − 𝑃𝑃𝑅𝑅 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) By definition we have: 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Thus: 𝑃𝑃 = 1 − ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸 Using the above convention, one can decompose the Rate Gap and the Exemption Gap into components indicating the loss of the Notional Ideal Revenue due to the implementation of reduced rates and exemptions on specific goods and services. Such additive VAT Gap in the EU-28 Member States Policy Gap: 1 − 𝑃𝑃 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Exemption Gap: 1 − 𝑃𝑃𝐸𝐸 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Rate Gap: 1 − 𝑃𝑃𝑅𝑅 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) By definition we have: 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Thus: 𝑃𝑃 = 1 − ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸 Using the above convention, one can decompose the Rate Gap and the Exemption Gap into components indicating the loss of the Notional Ideal Revenue due to the implementation of reduced rates and exemptions on specific goods and services. Such additive decomposition is carried out for the computation of, as defined by Barbone et al. (2015), VAT Gap in the EU-28 Member States Policy Gap: 1 − 𝑃𝑃 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Exemption Gap: 1 − 𝑃𝑃𝐸𝐸 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Rate Gap: 1 − 𝑃𝑃𝑅𝑅 = ( ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ∑ 𝑇𝑇𝑖𝑖 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) By definition we have: 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) + (𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) Thus: 𝑃𝑃 = 1 − ( ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = ( 2𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝐸𝐸 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 − ∑ 𝑇𝑇𝑖𝑖 ∗,𝑅𝑅 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 𝜏𝜏𝑠𝑠 ∑ 𝐶𝐶𝑖𝑖 𝑁𝑁 𝑖𝑖=1 ) = 𝑃𝑃𝑅𝑅 + 𝑃𝑃𝐸𝐸 Using the above convention, one can decompose the Rate Gap and the Exemption Gap into components indicating the loss of the Notional Ideal Revenue due to the implementation of reduced rates and exemptions on specific goods and services. Such additive decomposition is carried out for the computation of, as defined by Barbone et al. (2015), the Actionable Exemption Gap, which excludes the services and notional values that are unlikely to be taxed even in an ideal world.
- 97. CASE Reports |No. 503 (2020) 97 Using the above convention, one can decompose the Rate Gap and the Exemption Gap into components indicating the loss of the Notional Ideal Revenue due to the im- plementation of reduced rates and exemptions on specific goods and services. Such additive decomposition is carried out for the computation of, as defined by Barbone et al. (2015), the Actionable Exemption Gap, which excludes the services and notional values that are unlikely to be taxed even in an ideal world.
- 98. CASE Reports |No. 503 (2020) 98 Figure A1. Components of Ideal Revenue, VTTL, and VAT Collection Source: own. a. b. c.
- 99. CASE Reports |No. 503 (2020) 99 f. Tests of the Econometric Model Within the procedure for selecting exogenous variables aiming at minimising the problems of endogeneity, multicollinearity, and the omitted variables, we created a correlation matrix of pre-selected exogenous variables. As this test proved, there was no case of pairwise correlation of above 0.65 in the specifications presented in Table 5.4. To test whether the data matrix could result in unstable coefficient estimates, we used singular value decomposition method. In all of the data matrices underlying baseline and alternative equations, condition numbers were lower than 30, which is associated with well-behaved data matrices. Several other statistical tests were performed. The appropriateness of including time and country fixed effects was verified through the Hausmann tests. As the tests indicated that in the random effects specification, errors are correlated with the regressors, the fixed effects specification was chosen. Since the model contains time series, we verified that the model does not suffer from the issue of spurious regression. For this purpose, we performed unit root tests – Levin- -Lin-Chu (2002), Harris-Tzavalis (1999), and Im-Pesaran-Shin (2003). All tests indicated that the VAT Gap and explanatory variables included in the specifications are stationary. The tests showed that unemployment is non-stationary and cannot be included in levels in the equation regressing the VAT Gap denoted as a percent of the VTTL. In addition to unit root tests, all model specifications were tested for cointegration using the Pedroni panel-data test (Pedroni, 1999) and the Wald test for groupwise heteroskedasticity. The residuals of all model specifications appeared to be homoscedastic, stationary, and I(0).
- 100. CASE Working Paper| No 1 (2015) 100 Table B1. VTTL (EUR million) Source: own calculations. Annex B. Statistical Appendix 2014 2015 2016 2017 2018 Belgium 30,272 31,416 32,263 33,619 34,670 Bulgaria 4,896 5,045 5,037 5,313 5,711 Czechia 13,948 15,019 15,455 16,694 18,261 Denmark 27,955 28,610 29,308 30,475 31,369 Germany 229,881 232,507 239,911 248,382 257,207 Estonia 1,911 1,986 2,090 2,286 2,458 Ireland 12,406 13,543 14,027 14,652 15,857 Greece 17,287 18,545 20,591 21,898 21,858 Spain 69,824 72,283 74,791 79,003 82,470 France 165,520 167,521 168,611 173,840 180,406 Croatia 6,329 6,440 6,843 7,198 Italy 137,817 139,703 140,400 142,939 144,772 Cyprus 1,761 1,859 2,028 Latvia 2,248 2,348 2,329 2,512 2,705 Lithuania 3,879 3,876 4,015 4,422 4,754 Luxembourg 3,888 3,510 3,736 3,525 3,928 Hungary 11,969 12,693 12,338 13,564 14,140 Malta 935 861 925 984 1,084 Netherlands 47,199 49,756 50,500 52,329 54,897 Austria 27,955 28,736 29,768 30,949 32,231 Poland 38,799 39,922 38,731 42,374 44,862 Portugal 17,020 17,598 17,890 18,872 19,754 Romania 19,347 19,856 17,486 17,727 19,485 Slovenia 3,490 3,491 3,504 3,640 3,913 Slovakia 7,133 7,398 6,866 7,362 7,899 Finland 20,181 20,069 20,679 21,510 22,171 Sweden 40,148 41,709 43,435 44,987 43,739 United Kingdom 177,775 203,309 187,630 184,706 192,126 EU-28, EU-27 (2015), EU-26 (2014) 1,133,681 1,187,640 1,190,518 1,227,266 1,271,953
- 101. CASE Reports |No. 503 (2020) 101 Table B2. Household VAT Liability (EUR million) Source: own calculations. 2014 2015 2016 2017 2018 Belgium 17,326 17,714 18,522 19,230 19,688 Bulgaria 3,533 3,615 3,711 3,977 4,233 Czechia 8,917 9,311 9,776 10,535 11,347 Denmark 16,165 16,604 17,289 17,814 18,438 Germany 142,430 141,011 144,979 149,029 152,971 Estonia 1,338 1,374 1,436 1,530 1,652 Ireland 7,418 7,732 7,815 8,101 8,522 Greece 12,750 13,695 15,673 16,386 16,653 Spain 50,920 52,864 55,178 57,795 59,613 France 98,441 98,826 100,505 102,189 105,477 Croatia 4,555 4,690 4,970 5,241 Italy 97,232 99,621 99,890 100,918 102,246 Cyprus 1,130 1,188 1,245 Latvia 1,748 1,801 1,847 1,965 2,074 Lithuania 3,168 3,164 3,315 3,590 3,839 Luxembourg 1,237 1,289 1,331 1,361 1,469 Hungary 8,297 8,605 9,034 9,471 9,524 Malta 460 488 517 538 582 Netherlands 25,363 25,953 26,218 27,101 28,290 Austria 18,992 19,259 19,885 20,623 21,321 Poland 26,878 27,603 27,432 29,835 31,141 Portugal 12,823 13,190 13,345 13,843 14,397 Romania 11,677 12,086 10,909 11,338 12,846 Slovenia 2,442 2,448 2,573 2,682 2,820 Slovakia 5,303 5,136 5,111 5,421 5,744 Finland 11,074 11,386 11,575 11,830 12,198 Sweden 20,672 21,108 21,539 22,125 21,734 United Kingdom 118,086 133,965 124,855 123,266 127,658 EU-28, EU-27 (2015), EU-26 (2014) 724,690 754,404 760,080 778,654 802,964
- 102. CASE Reports |No. 503 (2020) 102 Table B3. Intermediate Consumption and Government VAT Liability (EUR million) Source: own calculations. 2014 2015 2016 2017 2018 Belgium 7,528 8,110 8,289 8,606 8,878 Bulgaria 722 708 734 794 897 Czechia 3,312 3,530 3,711 3,971 4,372 Denmark 7,795 7,872 7,619 8,043 8,246 Germany 48,657 51,429 53,680 55,605 57,926 Estonia 266 279 326 352 382 Ireland 3,372 3,991 4,022 4,164 4,633 Greece 2,183 2,461 2,681 2,807 2,885 Spain 10,938 10,884 11,046 11,796 12,547 France 28,782 31,790 32,198 33,099 33,955 Croatia 1,095 1,151 1,210 1,255 Italy 23,597 23,556 23,355 24,631 24,748 Cyprus 479 476 514 Latvia 336 366 369 383 405 Lithuania 415 446 448 482 512 Luxembourg 905 1,102 1,171 1,204 1,304 Hungary 1,977 2,102 2,054 2,218 2,320 Malta 410 271 326 356 396 Netherlands 13,409 14,313 14,259 14,642 15,317 Austria 5,050 5,131 5,130 5,276 5,668 Poland 7,180 7,682 7,589 8,242 8,563 Portugal 2,853 2,877 3,218 3,463 3,642 Romania 3,136 3,012 2,522 2,631 2,848 Slovenia 560 544 554 544 612 Slovakia 976 1,067 1,002 1,031 1,158 Finland 5,010 4,754 4,900 5,080 5,160 Sweden 11,981 12,400 12,719 12,962 12,443 United Kingdom 42,476 49,632 44,030 42,253 44,230 EU-28, EU-27 (2015), EU-26 (2014) 233,826 251,403 249,582 256,323 265,817
- 103. CASE Reports |No. 503 (2020) 103 Table B4. GFCF VAT Liability (EUR million) Source: own calculations. 2014 2015 2016 2017 2018 Belgium 4,739 4,957 4,808 5,106 5,440 Bulgaria 600 679 585 534 568 Czechia 1,744 2,192 1,971 2,196 2,502 Denmark 3,276 3,402 3,639 3,826 3,890 Germany 37,176 37,843 39,483 41,458 44,070 Estonia 298 323 318 392 418 Ireland 1,443 1,649 1,995 2,173 2,498 Greece 2,114 2,143 1,948 2,404 2,012 Spain 7,311 7,777 7,891 8,708 9,576 France 32,852 31,667 30,719 33,308 35,550 Croatia 592 567 635 668 Italy 13,305 13,318 13,883 14,005 14,366 Cyprus 134 172 243 Latvia 211 238 175 227 290 Lithuania 442 461 470 505 552 Luxembourg 348 411 626 541 726 Hungary 1,506 1,809 1,092 1,682 2,166 Malta 63 82 58 72 88 Netherlands 7,867 8,962 9,481 10,038 10,744 Austria 2,585 2,890 3,284 3,467 3,676 Poland 4,033 4,072 3,139 3,701 4,552 Portugal 1,017 1,170 941 1,194 1,295 Romania 3,821 4,193 3,638 3,478 3,541 Slovenia 401 419 303 346 406 Slovakia 869 1,206 763 916 992 Finland 3,498 3,316 3,513 3,839 4,096 Sweden 6,861 7,521 8,486 9,166 8,865 United Kingdom 15,202 18,555 17,396 17,022 17,693 EU-28, EU-27 (2015), EU-26 (2014) 153,583 161,849 161,308 171,109 181,482
- 104. CASE Reports |No. 503 (2020) 104 Table B5. VAT Revenues (EUR million) Source: Eurostat. 2014 2015 2016 2017 2018 Belgium 27,518 27,594 28,750 29,763 31,053 Bulgaria 3,810 4,059 4,417 4,664 5,097 Czechia 11,602 12,382 13,101 14,703 16,075 Denmark 24,950 25,672 26,770 27,966 29,121 Germany 203,081 211,616 218,779 226,582 235,130 Estonia 1,711 1,873 1,975 2,149 2,331 Ireland 11,528 11,831 12,603 13,060 14,175 Greece 12,676 12,885 14,333 14,642 15,288 Spain 62,825 67,913 70,214 73,970 77,561 France 148,454 151,680 154,490 162,011 167,618 Croatia 5,699 5,992 6,465 6,946 Italy 96,567 100,345 102,086 107,576 109,333 Cyprus 1,664 1,765 1,951 Latvia 1,787 1,876 2,032 2,164 2,449 Lithuania 2,764 2,889 3,028 3,310 3,522 Luxembourg 3,749 3,420 3,422 3,433 3,729 Hungary 9,754 10,676 10,595 11,729 12,950 Malta 642 673 712 810 920 Netherlands 42,951 44,746 47,849 49,833 52,619 Austria 25,386 26,247 27,301 28,304 29,323 Poland 29,317 30,075 30,838 36,330 40,411 Portugal 14,682 15,368 15,767 16,810 17,865 Romania 11,496 12,939 10,968 11,650 12,890 Slovenia 3,155 3,220 3,319 3,482 3,765 Slovakia 5,021 5,423 5,424 5,919 6,319 Finland 18,948 18,974 19,694 20,404 21,364 Sweden 38,846 40,501 42,770 44,115 43,433 United Kingdom 158,347 183,164 167,827 162,724 168,674 EU-28, EU-27 (2015), EU-26 (2014) 971,566 1,033,741 1,046,721 1,086,332 1,131,912
- 105. CASE Reports |No. 503 (2020) 105 Table B6. VAT Gap (EUR million) Source: own calculations. 2014 2015 2016 2017 2018 Belgium 2,755 3,822 3,513 3,856 3,617 Bulgaria 1,086 985 620 649 614 Czechia 2,345 2,637 2,354 1,991 2,187 Denmark 3,006 2,938 2,539 2,509 2,248 Germany 26,800 20,891 21,132 21,800 22,077 Estonia 200 113 115 137 127 Ireland 878 1,712 1,425 1,592 1,682 Greece 4,611 5,660 6,258 7,256 6,570 Spain 6,999 4,370 4,577 5,033 4,909 France 17,066 15,841 14,121 11,829 12,788 Croatia 630 447 378 252 Italy 41,250 39,358 38,314 35,363 35,439 Cyprus 97 93 77 Latvia 460 472 297 348 256 Lithuania 1,115 987 988 1,111 1,232 Luxembourg 139 90 314 92 199 Hungary 2,215 2,018 1,743 1,835 1,190 Malta 293 188 213 174 164 Netherlands 4,248 5,010 2,651 2,496 2,278 Austria 2,569 2,489 2,466 2,645 2,908 Poland 9,483 9,847 7,893 6,044 4,451 Portugal 2,338 2,230 2,123 2,062 1,889 Romania 7,850 6,917 6,518 6,077 6,595 Slovenia 335 271 186 159 148 Slovakia 2,112 1,975 1,443 1,443 1,579 Finland 1,233 1,095 985 1,106 807 Sweden 1,302 1,207 665 872 306 United Kingdom 19,427 20,144 19,802 21,982 23,452 EU-28, EU-27 (2015), EU-26 (2014) 162,115 153,899 143,798 140,935 140,042
- 106. CASE Reports |No. 503 (2020) 106 Table B7. VAT Gap (percent of VTTL) Source: own calculations. Backcasted series Full estimates Forecast 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Belgium 6.4% 10.9% 8.7% 11.9% 10.3% 10.0% 10.3% 8.6% 12.3% 13.0% 11.3% 12.6% 14.4% 12.7% 9.1% 12.2% 10.9% 11.5% 10.4% 9.4% 13.9% Bulgaria 35.4% 38.0% 46.0% 34.9% 25.8% 21.7% 18.7% 24.2% 16.1% 27.0% 24.0% 25.7% 21.4% 16.3% 22.2% 19.5% 12.3% 12.2% 10.8% 11.1% 15.5% Czechia 23.6% 22.9% 23.3% 25.5% 6.1% 4.2% 9.7% 13.6% 17.4% 19.0% 21.9% 17.3% 20.4% 19.3% 16.8% 17.6% 15.2% 11.9% 12.0% 10.8% 15.3% Denmark 12.6% 12.1% 11.5% 10.9% 11.0% 10.3% 10.4% 10.0% 12.1% 10.6% 11.0% 11.4% 11.2% 12.2% 10.8% 10.3% 8.7% 8.2% 7.2% 7.8% 13.3% Germany 10.2% 12.6% 12.1% 11.9% 12.2% 12.0% 10.7% 12.4% 11.6% 8.8% 9.0% 10.3% 11.5% 11.8% 11.7% 9.0% 8.8% 8.8% 8.6% 7.7% 12.1% Estonia 9.0% 12.5% 13.3% 14.1% 20.0% 10.4% 6.9% 5.7% 15.7% 9.3% 10.5% 12.4% 12.5% 14.1% 10.4% 5.7% 5.5% 6.0% 5.2% 4.8% 10.3% Ireland 13.8% 5.8% 8.3% 10.3% 7.4% 11.6% 11.6% 13.0% 15.0% 19.4% 16.3% 15.6% 15.6% 10.6% 7.1% 12.6% 10.2% 10.9% 10.6% 5.9% 11.4% Greece 20.5% 17.7% 18.5% 23.0% 23.7% 26.5% 27.5% 27.2% 24.9% 30.7% 27.3% 34.8% 29.6% 33.0% 26.7% 30.5% 30.4% 33.1% 30.1% 31.4% 36.9% Spain 5.4% 7.2% 8.5% 5.7% 4.0% -0.4% 0.2% 8.8% 20.9% 33.4% 10.7% 15.1% 11.5% 13.3% 10.0% 6.0% 6.1% 6.4% 6.0% 3.1% 8.4% France 4.4% 6.3% 7.8% 8.3% 7.1% 7.0% 7.5% 7.5% 9.3% 13.5% 8.7% 7.4% 11.7% 10.0% 10.3% 9.5% 8.4% 6.8% 7.1% 3.9% 8.6% Croatia 10.0% 6.9% 5.5% 3.5% 0.6% 5.2% Italy 26.5% 28.5% 27.8% 31.8% 32.3% 31.2% 27.6% 27.2% 30.1% 35.2% 27.6% 30.7% 30.0% 31.3% 29.9% 28.2% 27.3% 24.7% 24.5% 23.9% 29.4% Cyprus 5.5% 5.0% 3.8% Latvia 11.7% 16.5% 17.5% 17.5% 18.7% 10.9% 7.2% 6.7% 21.6% 37.9% 30.1% 32.0% 23.7% 24.0% 20.5% 20.1% 12.8% 13.9% 9.5% 6.6% 11.3% Lithuania 23.9% 27.1% 26.3% 31.6% 35.8% 29.6% 26.3% 22.1% 22.4% 33.4% 28.1% 28.3% 29.5% 29.5% 28.7% 25.5% 24.6% 25.1% 25.9% 21.6% 27.0% Luxembourg 8.4% 8.1% 6.3% 6.1% 3.9% 2.2% 1.9% 4.1% 6.0% 2.1% 2.2% 2.5% 2.1% 3.3% 3.6% 2.6% 8.4% 2.6% 5.1% Hungary 17.0% 22.9% 25.0% 21.0% 18.5% 22.2% 22.4% 19.5% 21.6% 21.4% 21.7% 21.5% 21.7% 21.1% 18.5% 15.9% 14.1% 13.5% 8.4% 6.6% 10.9% Malta 30.9% 31.5% 29.8% 29.5% 34.3% 23.5% 24.2% 27.2% 26.3% 24.6% 28.7% 29.7% 31.1% 30.2% 31.3% 21.8% 23.0% 17.7% 15.1% 16.8% 21.8% Netherlands 12.8% 11.9% 10.7% 10.1% 7.4% 6.9% 6.4% 4.2% 7.7% 12.8% 5.4% 9.9% 9.3% 10.0% 9.0% 10.1% 5.3% 4.8% 4.2% Austria 7.7% 9.4% 6.5% 9.8% 10.2% 10.3% 12.6% 11.5% 11.5% 7.8% 9.9% 11.7% 8.9% 10.3% 9.2% 8.7% 8.3% 8.5% 9.0% 7.5% 11.4% Poland 25.4% 29.4% 26.8% 26.1% 25.4% 17.8% 13.7% 10.5% 17.1% 23.3% 20.6% 20.8% 27.1% 26.6% 24.4% 24.7% 20.4% 14.3% 9.9% 9.7% 14.6% Portugal -0.7% 1.1% 1.8% 1.9% 2.6% -0.9% 1.5% 3.0% 4.4% 15.3% 12.9% 13.2% 15.4% 15.7% 13.7% 12.7% 11.9% 10.9% 9.6% 7.0% 11.5% Romania 37.7% 45.0% 35.5% 35.4% 40.9% 30.6% 33.4% 32.2% 33.4% 45.4% 40.7% 36.6% 37.9% 38.1% 40.6% 34.8% 37.3% 34.3% 33.8% 33.4% 37.4% Slovenia 3.4% 5.3% 4.8% 5.7% 5.5% 5.1% 4.7% 6.5% 8.8% 10.6% 8.5% 6.3% 9.3% 5.7% 9.6% 7.8% 5.3% 4.4% 3.8% 2.3% 7.2% Slovakia 22.5% 22.4% 23.7% 16.2% 19.1% 15.7% 22.4% 26.3% 25.2% 31.6% 33.0% 27.2% 36.7% 31.4% 29.6% 26.7% 21.0% 19.6% 20.0% 16.6% 21.2% Finland 7.2% 8.4% 7.9% 8.0% 8.7% 6.6% 7.0% 9.6% 10.3% 5.2% 8.9% 5.6% 5.4% 5.9% 6.1% 5.5% 4.8% 5.1% 3.6% 3.2% 7.1% Sweden 7.2% 7.3% 7.1% 6.2% 5.9% 5.6% 6.6% 5.4% 4.2% 3.4% 3.1% 3.8% 6.7% 3.4% 3.2% 2.9% 1.5% 1.9% 0.7%
- 107. CASE Reports |No. 503 (2020) 107 Figure C1. VAT Gap Forecasts for 2020 (increments, pp) Source: own calculations. Annex C. Additional Graphs
- 108. CASE Working Paper| No 1 (2015) 108 Aizenman, J., Y. Jinjarak (2008), The collection efficiency of the Value Added Tax: Theory and international evidence, Journal of International Trade & Economic Development, 17(3): 391–410. Allison, P.D. (2001), Missing Data, Sage, Thousand Oaks. Barbone, L., M. Belkindas, L. Bettendorf, R. Bird, M. Bonch-Osmolovskiy, M. Smart (2013), Study to quantify and analyse the VAT Gap in the EU-27 Member States, Final Report of project TAXUD/2012/DE/316. Barbone, L., M. Bonch-Osmolovskiy, G. Poniatowski (2014), 2012 Update Report to the Study to quantify and analyse the VAT Gap in the EU-27 Member States, Report of project TAXUD/2013/ DE/321. Barbone, L., M. Bonch-Osmolovskiy, G. Poniatowski (2015), 2013 Update Report to the Study to quantify and analyse the VAT Gap in the EU Member States, Report of project TAXUD/2013/ DE/321. Carfora, A., S. Dongiovanni, A. Marabucci, S. Pisani (2020), The Impact of Domestic Factors and Spillover Effects on EU Countries VAT Gap, Agenzia Entrate, Discussion Papers, N. 1/2020. Casey, P., P. Castro (2015), Electronic Fiscal Devices (EFDs) An Empirical Study of their Impact on Taxpayer Compliance and Administrative Efficiency?, IMF Working Paper No. 15–73. European Commission (2016), The Concept of Tax Gaps: Report on VAT Gap Estimations, FISCALIS Tax Gap Project Group (FPG/041). European Commission (2018), The concept of Tax Gaps Report III: MTIC Fraud Gap estimation methodologies. Feinstein, J. S. (1991), An Econometric Analysis of Income Tax Evasion and its Detection, RAND Journal of Economics, 22(1): 14–35. Harris, R. D. F., E. Tzavalis. (1999), Inference for unit roots in dynamic panels where the time dimension is fixed, Journal of Econometrics, 91: 201–226. Im, K. S., M. H. Pesaran, Y. Shin. (2003), Testing for unit roots in heterogeneous panels, Journal of Econometrics, 115: 53–74. References
- 109. CASE Reports |No. 503 (2020) 109 Keen, M. (2013), The Anatomy of the VAT, IMF Working Paper, WP/13/111, May. Keen, M., S. Smith (2007), What Do We Know, and What Can be Done? IMF Working Paper No. 07/31. Kleinke, K., M. Stemmler, J. Reinecke, F. Lösel (2011), Efficient ways to impute incomplete panel data, AStA Advances in Statistical Analysis, 95(4): 351–373. Lešnik, T., T. Jagrič, V. Jagrič (2018), VAT gap dependence and fiscal administration measures, Naše gospodarstvo/Our economy, 64(2): 43–51. Levin, A., C.-F. Lin, C.-S. J. Chu. (2002), Unit root tests in panel data: Asymptotic and finite-sample properties, Journal of Econometrics, 108: 1–24. Little, R. J. A., D. B. Rubin, (1987), Statistical analysis with missing data, New York, NY: Wiley. OECD (2014), VAT Revenue Ratio (VRR), in Consumption Tax Trends 2014: VAT/GST and excise rates, trends and policy issues, OECD Publishing, Paris. Pedroni P. (1999), Critical values for cointegration tests in heterogeneous panels with multiple regressors, Oxford. B. Econ. Stat., 61: 653–670. Poniatowski, G., M. Bonch-Osmolovskiy, A. Śmietanka (2017), Study and Reports on the VAT Gap in the EU-28 Member States: 2017 Final Report, TAXUD/2015/CC/131. Poniatowski, G., M. Bonch-Osmolovskiy, J. Durán-Cabré, A. Esteller-Moré A. Śmietanka (2018), Study and Reports on the VAT Gap in the EU-28 Member States: 2018 Final Report, TAXUD/2015/CC/131. Poniatowski, G., M. Bonch-Osmolovskiy, J. Durán-Cabré, A. Esteller-Moré A. Śmietanka (2019), Study and Reports on the VAT Gap in the EU-28 Member States: 2019 Final Report, TAXUD/2015/CC/131. Poniatowski, G., M. Bonch-Osmolovskiy, M. Belkindas (2016), 2014 Update Report to the Study to quantify and analyse the VAT Gap in the EU Member States, Report of project TAX- UD/2015/CC/131.
- 110. CASE Reports |No. 503 (2020) 110 Sczypińska, A. (2019), What drives the VAT gap in the European Union?, Roczniki Kolegium Analiz Ekonomicznych, 55: 69–72. Slemrod, J., S. Yitzhaki (1987), The optimal size of a tax collection agency, Scandinavian Journal of Economics, 89: 25–34. Zídková, H (2014), Determinants of VAT gap in EU, Prague Economic Papers, 23(4): 514–530.