AD ALTA
JOURNAL OF INTERDISCIPLINARY RESEARCH
4 The calculation of the tax gap
To calculate the tax gap, it was first necessary to calculate the
total tax liabilities. In this calculation, we used theoretical VAT
liabilities according to Barbone (2013), included five VAT sub-
aggregates, i.e. the final household consumption, government
expenditures, intermediate consumption, gross fixed capital
formation (GFCF), and the final consumption of non-profit
institutions serving households (NPISH). We considered the
sectoral classification of the economy, the effective tax rate, and
the percentage of exports of goods that are exempt from VAT.
Barbone (2013) classified theoretical VAT liabilities as follows:
=∑(
)+ ∑(
+
=1
=1
+=1 + +
(1)
where:
rate – effective tax rate
Value – the final household consumption NPISH and
government consumption
IC Value – the intermediate consumption
Propex – a percentage of output exempt from VAT in the sector
GFCF Value – gross fixed capital formation
i – economic sectors.
In our calculation of tax liabilities, we used a study CASE (2018)
which the European Commission considers as key research in
assessing VAT evasion. We adjusted the total VAT liability and
added non-sectoral economic classification, i.e. we considered
final consumption of all the products regardless of the goods and
services for which reduced or super-reduced tax rate is applied. In
the formula, the percentage of output that is exempt from taxation
represents the sum of export within the EU (intra-EU export) non-
EU export and the percentage of taxes and duties excluding VAT.
Total tax liabilities are calculated as follows:
=(++)∗+
(+)∗∗
(2)
where:
Gov – the final government consumption, in million EUR
Hous – the final household consumption, in million EUR
NPISH – the final consumption non-profit institutions serving
households, in million EUR
er – effective tax rate, in %
GFCF – gross fixed capital formation, in million EUR
IC – the intermediate consumption, in million EUR
out – a percentage of non-EU export and intra-EU export,
percentage of taxes and duties excluding VAT, in %.
Since one of the input variables is the effective VAT rate, we
used the following formula for its calculation:
=
+
(3)
where:
Hous – the final household consumption, in million EUR
corporation – output for the final consumption of non-financial
corporations, in million EUR.
We included the final household consumption in the tax base as
VAT is the most burdened by it, and also received output for the
final consumption of non-financial corporations retrieved from
Eurostat (2018) which includes all economic sectors based on
NACE classification. After calculation total tax liabilities, we
measured the tax gap using the following formula:
= − (4)
5 The regression analysis
The regression analysis examined the relationship between
individual variables and the evolution of tax gaps. The explained
variable represents the tax gap with VAT tax revenue. The
general panel model for our regression analysis is defined as
follows:
=+
+
(5)
where:
y
it
x
- dependent (response) variable (i.e. tax evasion as a
proportion of tax gap to tax revenues);
it
- independent (explanatory) variables (GDP per capita, import
ratio,
standard VAT rate,
consumption-to-GDP ratio,
intermediate consumption, unemployment rate, corruption index,
value added-to-GDP ratio, shadow economics, gross public debt,
and the amount of population (Tab.1).
The selection of variables for both analyses was determined by
the theoretical basis of the following studies: Aizenmann &
Jinjarek (2008), Ebrill et al. (2001). Agha & Haughton (1996).
Bird et. al. (2004), Barbone et al. (2013), CASE (2018), and
Reckon (2009). In these studies, authors followed many
variables which have either a direct, or an indirect impact on the
volume of tax evasion. The degree of impact of the above factors
varied depending on the intensity of the relationship between the
variables. The determinants themselves were specific and
dynamic, constantly evolving and influencing each other.
Table 1 Independent explanatory variables X
Variable
ij
Abb.
Unit
Reason for inclusion in
the model
Relation to the tax
gap (hypothesis)
Author
Source
GDP per capita
GDPpc
mil. €
wealth level of
development
decrease
Reckon (2009)
Eurostat
unemployment
unemp
% of active
population
economic cycle tax
revenues inequality
increase
Barbone (2013)
Eurostat
import to GDP
IMP
%
economy openness
carousel fraud risk
increase (if there is
VAT carousel)
Aizenmann & Jinjarak
(2008), Ebrill (2001)
Eurostat
VAT
VAT
%
tax burden
increase
Reckon (2009), Ebrill
(2001), Agha (1996)
European
Commission
Corruption
Perceptions
Index
CPI
index
level of corruption
population trust in the
public sector
decrease (the higher
CPI the lower
corruption)
Bird et al. (2004),
Reckon (2009)
Transparency
International
population
pop
mil. €
country size
increase
Barbone (2013)
Eurostat
public debt
debt
%
worse financial condition
increase
Barbone (2013)
Eurostat
shadow
economy
shadow
%
significance of the shadow
economy
increase
Bird et al. (2004)
IMF
added value to
GDP
AV
%
the relative size of
economic sectors
decrease
Reckon (2009)
Eurostat
intermediate
consumption to
GDP
iC
%
incorporating the
corporate sector
increase
the variable we choose
Eurostat
consumption to
GDP
C
%
size of potential tax base
decrease
Reckon (2009)
Eurostat
Source: own calculation based on Zidková (2014)
- 9 -