Economics of Corruption: Demand Side
Case of Western Balkan Countries
EDMIRA CAKRANI1, DIMITRIOS A. KARRAS1,2, GJERGJI SHQAU1,3
1Business Administration Department
Canadian Institute of Technology
Xhanfize Keko Street 12, Tirana 1004
ALBANIA
2National and Kapodistrian University of Athens (NKUA), GREECE
and EPOKA university, Computer Eng, Dept., Tirana, ALBANIA
3Business Administration Department
University “Aleksander Xhuvani” of Elbasan
Ismail Zyma Street, Elbasan 3000
ALBANIA
Abstract: - Corruption is a very negative phenomenon, which distorts markets and harms economic growth.
Corruption has its side of supply and demand. Both supply and demand for corruption are influenced by many
factors. The purpose of this article is to identify some macroeconomic and institutional factors that lead to the
demand for corruption in the Western Balkans. The Corruption Perception Index and Control of Corruption index
are used as measures of corruption, therefore in this paper two models are built, where independent variables are
real income per capita, inequality gap, unemployment rate, rule of law, and government efficiency. A panel
model, with data for the period 2012-2022 is used to identify the most important variables affecting corruption
in the Western Balkans. The results show that the index used to measure corruption affects the statistical
significance of the variables, with inequality gap and rule of law being significant in both models. The
identification of the factors can serve the governments of these countries to design policies and adopt strategies
that will reduce the involvement of people in corrupt practices.
Key-Words: - Corruption, Demand, Macroeconomic factors, Government Policies, Panel data, Fixed Effects,
Random Effects.
Received: April 11, 2024. Revised: September 5, 2024. Accepted: October 9, 2024. Published: November 4, 2024.
1 Introduction
There are different definitions of corruption. [1]
defined corruption as an extra-legal institution used
by individuals or groups to gain influence over the
actions of the bureaucracy. A similar definition is
given by [2]1 who stated that corruption is a behavior
which deviates from accepted norms in order to serve
private ends. The [3]2 defines corruption as "the
abuse of public office for private gains”. This
definition gives the idea that corruption is present
only in the public sector. Corruptive behaviors have
been seen in the private sector too. Another more
complete definition is provided by the Transparency
International, which defines corruption as the misuse
of entrusted power for private gains.
Despite various definitions, it is widely accepted
that corruption implies the use of authority for private
gain. Researchers have been studying corruption for
years. This is because it is widely accepted that
corruption has negative effects on the economy.
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
351
Volume 2, 2024
Corruption distorts markets and weakens the role of
government as a market regulator and guarantor of
rights. If government regulation can be overcome
through corruption, or if property rights become a
"market commodity" by the public servant, then
markets can fail, investments, especially foreign
ones, will fall [4]
3
and there will be negative effects
on economic growth [5]. On the other hand, corrupt
payments increase business costs [6], and increase
the burden especially on the poor, who must pay to
receive services or to have their rights respected, thus
leading to a further deterioration of their economic
situation. In addition to the negative impact on the
economy, corruption also has a negative impact on
democracy. Corruption compromises the effects of
government policies, as well as undermines public
confidence in democratic institutions. It is considered
as an important indicator of the performance of a
political system [7].
But what is considered a corrupted behavior?
According to the United Nations Convention Against
Corruption UNCAC
4
the most common actions that
are considered corrupted are: bribery, embezzlement,
trading in influence.
- Bribery in the public and private sector, whether
briber-initiated or bribe-initiated is the amount given
in a corrupt relationship. The goal is to facilitate
arrangements, to get things done with less effort. The
bribe is usually considered grease money because it
enables the bureaucratic apparatus of the state to
move faster. This is the case when individuals or
businesses evade taxes, when they do not comply
with legal restrictions, etc.
- Embezzlement is the theft and misuse of public
funds by state officials. However, the private sector
also suffers from this form of corruption.
-Trading in influence occurs when a state official is
promised or offered a reward in order to exert his
influence in the state administration to create
advantages for the interested party. This also means
access to take advantage of state resources.
The effectiveness of the fight against corruption
depends on a number of factors of economic, social
and political nature. The purpose of this article is to
identify some macroeconomic and institutional
factors that influence the demand for corrupted acts.
The second section presents different definitions of
corruption and the factors that generally affect
corruption in a country. The third section is dedicated
3
https://www.nber.org/papers/w6030
to the literature review. The fourth section provides
an analysis of the level of corruption and some
macroeconomic factors in Western Balkans, and the
analysis of the econometric models, which serve to
identify the factors that affect the demand for
corruption in these countries. The article concludes
with recommendations towards the situation
improvement.
2 Corruption: Definition and
Determinants
Corruption is not a variable that can be measured
directly in the economy. By its very nature, it is not
possible to have accurate measurements of it. For this
reason, different indexes try to measure the
perception of individuals about the degree of
corruption in a country. These indexes are built on
the basis of questionnaires. Among the various
indexes, some that can be mentioned are:
-International Country Risk Guide Index tries to
capture the extent to which "high government
officials are likely to demand special payments" and
to which "illegal payments are generally expected
throughout lower levels of government" in the form
of "bribes connected with import and export licenses,
exchange controls, tax assessments, police
protection, or loans." The index score is a weighted
average of three sub index scores: The Political Risk
Index (calculated with a maximum of 100 points), the
Financial Risk Index (calculated with a maximum of
50 points), and the Economic Risk Index (calculated
with a maximum of 50 points). In calculating the total
points of the index, the Political Risk accounts for
50% of the total points, while the Financial and
Economic risk account for 25% respectively. The
range of the total points is zero to 100. Countries are
ranked based on the total points of the ICRG index:
up to 49.9 points means Very High risk, and from 50
to 100 points means Very Low Risk.
-Global Competitiveness Report Index suggests a
measure of perception of corruption from the
business perspective. Through a survey with firm’s
managers the perception on "irregular, additional
payments connected with import and export permits,
business licenses, exchange controls, tax
assessments, police protection or loan applications"
is quantified. The respondents are asked to rate the
perceived level of corruption in a 1 to 7 scale.
4
https://www.unodc.org/documents/brussels/UN_Convention_Ag
ainst_Corruption.pdf
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
352
Volume 2, 2024
Countries are ranked using the average scale of
respondents in the corresponding country. The
composite score ranges from 0 to 100, with 0 being
the “highly corrupt”, and 100 “very clean”.
-Business International Index tries to measure the
degree to which business transactions involve
corruption or questionable payments. Respondents
choose a value from 0 to 10, with 10 being the “very
clean” to rate the perceived corruption in a country.
-Corruption Perception Index ranks countries by
their perceived level of public sector corruption.
Data from 13 different datasets are used to calculate
the index. For each country at least three assessments
are made with different data in order to calculate the
index. The CPI reflects the corruption perception of
businesspeople and business analysts. Even though
there exists corruption in the private sector as well,
the CPI measures the perception of corruption only in
the public sector. Countries are ranked on a scale
from 0 (very corrupt) to 100 (very clean).
--Control of Corruption measures perceptions of
the degree to which public power is used for private
benefit, including both small-scale and large-scale
corruption, as well as the "capture" of the state by
elites and special interest. According to the World
Bank definition, corruption is a failure of governance
because it frequently results from a lack of respect for
the laws that regulate the interactions between the
corrupted (usually a public official or politician) and
the corrupter (usually a private citizen or firm) [8].
The values of the index vary from -2.5 to 2.5, where
the lowest values indicate a perception of high
corruption, while high positive values indicate a
perception of low corruption.
Regardless of what index they have used, various
researchers have tried to identify the factors that
affect the level of corruption in a country.
Researchers have identified several factors, that
directly or indirectly affect it. Among many factors,
[8]
5
identified as direct factors:
-Regulations and authorizations: the need to obtain a
license or permit to conduct business activity gives
the civil servant a kind of monopoly power. These
clerks can put pressure on those interested with the
goal to get bribes for themselves.
-Taxation: when the tax laws are not clear or when
the civil servant has discretion over important
decisions (the provision of tax incentives, selection
5
https://www.imf.org/en/Publications/WP/Issues/2016/12/30/Corruption-
Around-the-World-Causes-Consequences-Scope-and-Cures-2583
of audits, etc.) the possibility for the official to be
involved in a corrupted practice increase.
-Spending decisions: investment projects,
procurement spending are decisions that government
officials often use to secure benefits for themselves.
Public projects are often used to favor certain parties
over the bribe.
As indirect factors, among others [9] identified:
-Quality of the bureaucracy: corrupt acts are
committed mainly by state officials. If these officials
are not hired and promoted on the basis of merit, that
will result in a higher level of corruption in the
country [10]. Employment for political reasons,
nepotism and unclear rules of employment or
promotion result in corrupted behavior of the public
servant.
-Level of public sector wages: low public sector
wages can encourage employees to engage in
corrupted acts [11].
-Penalty systems: the lighter the penalties, the more
widespread the corruption.
3 Literature Review
[12] in a study of 41 developing countries, analyze
the determinants of corruption, dividing them into
two groups of factors: economic and non-economic.
As a measure of corruption, the authors use the CPI.
Cross-sectional analysis, with data for 2006, shows
that among the economic factors, the level of income
and globalization, economic freedom, the level of
education result in influencing the level of
corruption, while the distribution of income does not
result in this impact. The study reaches the
conclusion that non-economic factors, such as
freedom of the press, religion, and democracy have
no influence on the level of corruption in these
countries.
Using panel data analysis for the period 2004-2010
for several developing countries, [13] concludes that
foreign direct investments, the export of natural
resources and the level of economic development are
among the most important factors that influence
corruption in these countries. The study also suggests
that other factors, such as country size, state of
democracy, and colonial legacy, have an impact on
the perception of corruption in developing countries.
[14], in a study for 92 countries, with data for 2014,
suggests that different economic, political and social
factors, such as the level of economic development,
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
353
Volume 2, 2024
political stability, religion, level of education, state of
democracy, economic freedom have an impact on the
perception of corruption in these countries, although
the magnitude of this impact differs between
developing and developed countries.
Using a panel data model, with data for the period
1996-2019 for the Visegrad countries, [15] suggest
that different economic and political factors affect the
perception of corruption in these countries, such as:
level of economic development, degree of
globalization, government consumption, degree of
urbanization, share of women in the labor force,
regulatory quality, income inequality.
Using the CPI as a measure of corruption, with
data for the period 2003-2021, [16] use a panel data
model to identify the determinants of corruption in
Developing-Eight (D-8) countries. The authors
suggest that human development index, economic
freedom and taxes as % of GDP have an important
impact on the corruption index, while government
spending, GDP and inflation are statistically
insignificant.
While there are many studies that analyze the
impact of corruption on the economy in the countries
of the Western Balkans, there are relatively few that
identify some individual, mainly demographic
factors that determine corruption in these countries.
In a study for Albania, [17] using data from a
questionnaire conducted in the period January-
February 2016, using cross-sectional data, where the
level of corruption was measured on a scale from 1 to
10, reached the conclusion that the level of income,
area of living and political orientation, all have an
impact on corruption, while gender, age, capital,
previous experience with corruption do not appear to
have an impact on corruption.
With data from the National Survey of Citizens'
Perceptions in Bosnia-Herzegovina, [18] use a
logistic regression to analyze the likelihood that
people engage in corrupt behavior, offering bribes to
employees in the medical, judicial, police, public
service and education sectors. The results of the study
show that individuals with high level of income, who
live in urban areas and are educated are more inclined
to offer bribes. The study also shows that the impact
of these factors is different in different sectors.
This work will complement the existing literature
on this topic, identifying some macroeconomic and
institutional factors that influence the demand for
6
https://data.worldbank.org/
7
https://www.transparency.org/en
corruption in the Western Balkan countries. Another
contribution of this article is the methodology used:
panel data has more information, allows for more
variability, and provides more robust estimates than
the cross-sectional method.
4 Empirical Analysis
4.1. Model Specification
Various researchers have tried to analyze the
factors that motivate people to engage in corrupted
practices. These factors are of economic, social,
political nature, but also cultural factors. The purpose
of this research paper is to identify the factors that
can affect the level of corruption in the countries of
the Western Balkans. Two models will be built for
this purpose. In the first model, the Control of
Corruption index will be used as a measure of
corruption, while in the second model, the Corruption
Perception Index will be used. Independent variables
in both models will be real income per capita,
unemployment rate, government efficiency, rule of
law and inequality gap:
CC/CPI = f (RGDPC, UNEMPL, GE, RL, IG) (1)
where:
-CC represent the control of corruption index. Data is
taken from the World Bank database
6
.
-CPI represents the corruption perception index. Data
is taken from the Transparency International
Database
7
.
-RGDPC represents real GDP per capita, PPP
(constant 2017 international $), which is considered
as a proxy for the economic development of a
country. The higher the economic development, the
lower the incentive for people to engage in corrupted
practices, so a negative relationship between RGDPC
and corruption is expected [19], meaning that an
increase in the value of the RGDPC will improve the
score of the corruption index. Data for this variable is
taken from the World Bank database.
-UNEMPL represents the unemployment rate of a
country. The higher the unemployment rate, the more
people are expected to be involved in corruption, so
a positive impact of this variable on the level of
corruption is expected [20]
8
, resulting in a decrease
of the corruption index. Data for this variable is taken
from the World Bank database.
-GE is a variable that reflects opinions on the
standard of public services, the credibility of the
8
https://openknowledge.worldbank.org/handle/10986/25158?show
=full
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
354
Volume 2, 2024
government's commitment to implementing
appropriate programs, the standard of the civil
service and the extent of its independence from
political control, and the quality of policy
formulation and execution. Values for this variable
varies from -2.5 to +2.5, with low values showing
weak governance performance and high values
showing strong governance performance. Data for
this variable is taken from the World Bank database.
-RL is a variable that reflects opinions about how
much agents trust and follow social norms, especially
regarding the reliability of the police, courts, property
rights, and contract enforcement, as well as the
probability of crime and violence. Values for this
variable range from -2.5 to 2.5 with low values
showing weak governance performance and high
values showing strong governance performance.
Data for this variable is taken from the World Bank
database.
-IG represents the inequality gap. The rich people are
likely to have both greater motivation and
opportunities to engage in bribery and fraud as one
means to preserve and advance their status,
privileges, and interests while the poor are more
vulnerable to extortion at higher levels of inequality
[21]
9
. Data for this variable shows the pre-tax
national income that goes to the bottom 50% of the
adult population. Data is taken from the World
Inequality Database
10
for the years 2013-2021. Data
for the 2022 is not available, so for this year the
average of the two previous years is used as an
estimate for it.
First, the data is tested through pooled regression.
This model suggests that all countries have the same
characteristics and analyzes panel data as time-series
data:
𝒀𝒕= 𝜷𝟎 + 𝜷𝟏𝑿𝟏,𝒕 + 𝜷𝟐𝑿𝟐,𝒕 + ⋯ + 𝒗𝒕 (2)
The model results with a single intercept and
coefficients for all countries, ignoring heterogeneity.
However, since the study includes different
countries, there is the possibility of heterogeneity,
which are specific characteristics of different
countries. In pooled regression, these characteristics
are included in the error term, which may be
correlated with one or several of the independent
variables:
𝑪𝑶𝑽( 𝑿𝒕𝒗𝒕) ≠ 𝟎 (3)
The pooled model can result in coefficients that are
not BLUE, i.e. biased and inconsistent. Therefore, in
the second stage, the data will be tested through the
Fixed effect and random effect models, which
9
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=489823
consider the existence of unobserved heterogeneity.
The fixed effect model allows for different intercepts
for each country, although these intercepts are time-
invariant:
𝒀𝒊𝒕 = 𝜷𝟎𝒊 + 𝜷𝟏𝑿𝟏,𝒊𝒕 + 𝜷𝟐𝑿𝟐,𝒊𝒕 + ⋯ + 𝜷𝒌 𝑿𝒌,𝒊𝒕 +
𝜽𝒊+ 𝜺𝒊𝒕 (4)
where
𝜷𝟎𝒊 shows the intercepts of the equations in each
country;
𝜽𝒊 is a country-dependent error term, which is time-
invariant, but different for different countries.
The random effect model includes heterogeneity in
the error term, unlike the fixed effect model, which
includes it in the intercept:
𝒀𝒊𝒕 = 𝜷𝟎+ 𝜷𝟏𝑿𝟏,𝒊𝒕 + 𝜷𝟐𝑿𝟐,𝒊𝒕 + ⋯ + 𝜷𝒌𝑿𝒌,𝒊𝒕 +
𝝑𝒊𝒕 (5)
where
𝜷𝟎 is the average value of all intercepts of the Fixed
effects model.
𝝑𝒊𝒕 is the error term, which is composed of two
components: country-specific error term and
idiosyncratic error term, which shows the effect of
unobserved variables.
To determine which model between pooled
regression and fixed effects is more appropriate, the
Likelihood ratio test is used. If p-value < 0.05 then
the fixed effects in more appropriate than pooled
model.
To choose between fixed effects and random effects,
we use the Hausman test. If p-value < 0.05 then
random effects may be correlated with the
independent variables, therefore the Fixed effect
model is more appropriate.
This study covers the period 2012-2022, and data
were processed with the statistical package EViews
12.
4.2. Variables analysis 2012-2022
During the period 2012-2022, regardless of what
index is used to measure corruption, the Western
Balkans are positioned in the group of countries with
a high level of corruption. The corruption control
index is negative for all countries, indicating a
ranking below their average level. The best
performance is that of Montenegro, which has values
very close to 0, even for 2018 there is a positive
assessment of corruption control. During the period
under study, the index has improved in Albania and
Kosovo, while in Bosnia and Herzegovina, North
Macedonia and Serbia, a deterioration is observed.
10
https://wid.world/data/
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
355
Volume 2, 2024
Fig.1. Control of Corruption Index WB, 2012-2022
@Authors calculations with data from WB database
If the CPI is used as a measure for corruption, it is
noted that all countries, with the exception of
Montenegro, have an estimate between 30-40 points
for the entire period, showing no strong improvement
in the indicator. Montenegro has an assessment of
over 40 points throughout the entire period, an
assessment which has always been increasing,
reaching 46 points in 2021 and with an average
assessment of around 45 for the entire period. The
country with the lowest rating is Albania, which is
rated between 31 and 39 points with an average rating
of around 35 for the entire period, followed by
Kosovo with an average rating of 36 points.
Fig.2. Corruption Perception Index WB, 2012-2022
@Authors calculations with data from WB database
The level of GDP per capita in Western Balkans has
been constantly increasing. At the beginning of the
period the per capita income is between about $8,100
(Kosovo) and $16,800 (Montenegro). In the
following years, the growth of GDP per capita has
been stable, with the exception of 2020, where GDP
per capita has fallen in all countries, due to the
pandemic. In 2022 GDP per capita has increased, and
the level has exceeded that of the period before the
pandemic.
Fig. 3. Real GDP/c in WB, 2012-2022.
@Authors calculations with data from WB database
Although GDP per capita has increased in all the
countries of the Western Balkans, the distribution of
income has not been fair. The country with the
highest inequality is Serbia, where for the entire
period, only about 15% of the income goes to the
bottom 50% of the population. The country with the
best performance is North Macedonia, where on
average about 20% of the income goes to the bottom
50% of the population.
Fig. 4. Inequality Gap in WB, 2012-2022.
@Authors calculations with data from WB database
Regarding the level of unemployment, during the
period under study, unemployment has been
-0,9
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0
0,1
ALB BIH KS
MNE MKD SRB
30
32
34
36
38
40
42
44
46
48
2012 2014 2016 2018 2020 2022
ALB BIH KS
MNE MKD SRB
0
5
10
15
20
25
ALB BIH KS MNE MKD SRB
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
356
Volume 2, 2024
consistently above the level of 9% for all countries,
although the level has been in continuous decline.
Albania and Serbia have the best performance, where
the average unemployment rate for the entire period
is around 14%, while in Bosnia & Herzegovina, and
North Macedonia the average is around 22%. The
worst performer country is Kosovo, with the average
unemployment rate of 27.5%.
Fig. 5. Unemployment Rate in WB, 2012-2022.
@Authors calculations with data from WB database
4.3. Model Analysis
Pooled regression model shows that regardless of
how the dependent variable is measured, the model is
statistically significant, because p-value F-statistics <
5%. In the case where corruption is measured through
CC, the model explains about 78% of the variation,
while in the case of the CPI index, the model explains
about 70% of the variation in the values of the
corruption index. Regardless of the index used for
corruption, RGDPC, UNEMPL, RL have positive
signs and are statistically significant. GE is
statistically significant at 1% in the case of CC and
10% in the case of CPI. IG has a negative sign and is
statistically significant in the case of CPI, while it is
statistically insignificant in the case of CC.
Table 1. Pooled Regression
CC
CPI
Var
Coeff
p-
value
Coeff
p-
value
RGDPC
0.435265
0.0002
6.126814
0.0161
GE
0.226747
0.0000
1.753908
0.0952
IG
-0.008997
0.2212
-0.556036
0.0012
11
https://www.nber.org/system/files/working_papers/w19483/w19483.pdf
UNEMPL
0.019498
0.0000
0.261202
0.0000
RL
0.47157
0.0018
11.57457
0.0007
R-sqr
0.782519
0.705805
Adj R-sqr
0.764396
0.681289
Prof F-st
0.000000
0.000000
@Authors calculations with data from WB database
The fixed effects model is statistically significant for
both indices, CC and CPI. The CC model explains
about 84% of the variation, while the CPI model
explains about 76% of the variation. These values are
higher than the respective Pooled regression values.
In the case of the CC index, all variables are
statistically significant at 5%, while in the case of the
CPI index, only IG and RL are statistically
significant. Regarding the real GDP per capita
variable, the positive sign suggests a direct link
between it and the corruption indexes: an increase in
the level of income will be accompanied by an
increase in the corruption rating. The same
conclusion was reached by [19]
11
in their study. The
negative sign of the Inequality Gap variable suggests
an inverse relationship between this variable and the
corruption indexes, so an improvement in the
distribution of income in the economy is
accompanied by a decrease in the corruption rating
points. This result is consistent with the conclusion
reached by [21]. The Unemployment variable has a
positive sign, contrary to the expectation of a
negative impact on the value of the index.
Table 2. Fixed effects Model
CC
CPI
Var
Coeff
p-value
Coeff
p-value
RGDPC
0.419478
0.0256
5.369763
0.2189
GE
0.220218
0.0226
3.673714
0.1030
IG
-0.040262
0.0033
-1.212303
0.0002
UNEMPL
0.013181
0.0036
0.115779
0.2621
RL
0.484427
0.0053
8.926768
0.0275
R-sqr
0.842325
0.764205
Adj R-sqr
0.813657
0.721334
Prof F-st
0.000000
0.000000
@Authors calculations with data from WB database
The Likelihood ratio Test shows that for both indices,
the most appropriate model is the Fixed Effects
model, because p-value < 5%.
Table 3. Likelihood Ratio Test
7
12
17
22
27
32
37
2012 2014 2016 2018 2020 2022
ALB BIH KS
MNE MKD SRB
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
357
Volume 2, 2024
CC
CPI
Effects
Test
statistic
p-value
statistic
p-value
Cross-
section F
4.17231
0.0028
2.72440
0.0286
Cross-
section
Chi-
square
21.2240
0.0007
14.6045
0.0122
@Authors calculations with data from WB database
The analysis of the random effects model shows that
the model is statistically significant, with the highest
R-squared value in the case of the CC index.
Regarding the variables, RGDPC, UNEMPL and RL
are statistically significant, while GE is statistically
significant at 10% in the case of CPI. The IG variable
is significant only in the case of the CPI.
Table 4. Random Effect Model
CC
CPI
Var
Coeff
p-
value
Coeff
p-value
RGDPC
0.435265
0.0000
6.126814
0.0103
GE
0.226747
0.0000
1.753908
0.0748
IG
-0.008997
0.1697
-0.556036
0.0006
UNEMPL
0.019498
0.0000
0.261202
0.0000
RL
0.47157
0.0005
11.57457
0.0003
R-sqr
0.782519
0.705805
Adj R-sqr
0.764396
0.681289
Prof F-st
0.000000
0.000000
@Authors calculations with data from WB database
The p-value of the Hausman test is less than 5% and
this suggests that between fixed effects and random
effects, the most appropriate model to explain the
long-term relationship between independent
variables and corruption is the fixed effects model,
regardless of which index is used for measure
corruption.
Table 5. Hausman test results
CC
CPI
Test
Summary
Chi-Sq.
Statistic
p-
value
Chi-Sq.
Statistic
p-
value
Cross-
section
random
20.861563
0.0009
13.62203
0.0182
@Authors calculations with data from WB database
5. Conclusion and Future Work
Corruption is considered a harmful phenomenon,
which negatively affects not only the economy, but
also the rule of law and democracy. The design of
strategies and the implementation of measures
against this phenomenon is a central topic of public
and political debate. The success of these strategies
depends on political, economic, social, cultural
factors, etc.
The purpose of this article is to identify some
macroeconomic and institutional factors that affect
the demand for corruption in Western Balkan
countries. Among the various factors suggested by
previous studies on this phenomenon, this article
includes Real GDP per capita, Unemployment Rate,
Inequality Gap, Government Efficiency, and Rule of
Law.
Between Pooled, Fixed effects and Random effects
models, the most appropriate one to explain the long-
term relationship between these variables and
corruption is the Fixed effects model. However, the
model's conclusions depend on how corruption is
measured, because in the case where it is measured
through the Control of Corruption index, the model
suggests that all variables are statistically significant,
while in the case where the Corruption Perception
Index is used to measure corruption, only Inequality
gap and Rule of Law are statistically significant.
However, the Control of Corruption model is more
robust, as it explains about 84% of the variation,
while the CPI model explains only about 76.4% of
the corruption variation. This conclusion does not
change even though the coefficients of the CPI model
are greater than those of the CC model, as the purpose
is to identify the factors that affect the corruption
assessment, not the extent of this impact.
As expected, economic development is
accompanied by an increase in the public's perception
of keeping corruption under control, while the
positive sign of the variable Unemployment should
be further investigated to discover the reasons for
such an impact.
The model suggests that policymakers design
strategies and implement policies that lead to a better
distribution of income in the economy. Also, the
strengthening of institutions and the rule of law will
positively affect the perception of corruption control
in the Western Balkans.
In the future, this work will be expanded, to
include other microeconomic or social and cultural
factors, which can influence the demand for
corruption and other corruption indexes. It is the
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
358
Volume 2, 2024
objective of the authors to also study the supply side
of corruption, to give a more complete picture of the
economics of corruption.
References:
[1] Leff, Nathaniel (1964), “Economic development
through bureaucratic corruption”, American
Behavioral Scientist, 8(3): 8–14,
[2] Huntington, Samuel. P. (1968), “Political order
in changing societies”, Yale University Press.
JSTOR.
[3] World Bank, “Helping Countries Combat
corruption: The Role of the World Bank”,
Washington DC, World Bank Group,1997.
[4] Wei, Shang-Jin, (1997a), “How Taxing is
Corruption on International Investors?”
Working Paper 6030, NBER, May.
[5] Mauro, Paolo, (1995), “Corruption and Growth”
The Quarterly Journal of Economics. August:
681-712.
[6] Svensson, J. (2005), “Eight Questions About
Corruption”, Journal of Economic Perspectives
19: 19-42
[7] Anderson, C. J., & Tverdova, Y. V. (2003),
“Corruption, Political Allegiances, and
Attitudes Toward Government in
Contemporary Democracies”, American
Journal of Political Science, 47(1): 91-109.
[8] Kaufmann,D.,Kraay,A.,Mastruzzi,M. (2010).
"The Worldwide Governance Indicators : A
Summary of Methodology, Data and Analytical
Issues". World Bank Policy Research Working
Paper No. 5430
[9] Tanzi, Vito, (1998), “Corruption Around the
World: Causes, Consequences, Scope, and
Cures”, IMF Working Paper 63.
[10] Rauch, E.J., Evans, P.B. (2000), “Bureaucratic
Structure and Bureaucratic Performance in
Less Developed Countries”, Journal of Public
Economics, 75: 49-71.
[11] Van Rijckeghem, C., Weder, B, “Bureaucratic
Corruption and the Rate of Temptation: do
Wages in the Civil Service Affect Corruption,
and by How Much?”, Journal of Development
Economics, Vol 65, Issue 2, August 2001:
307-331
[12] Chaudhry,M.A.,Shabbir,G.(2007).
Determinants of corruption in developing
countries, HWWI Research Paper, No. 2-11,
Hamburgisches WeltWirtschaftsInstitut
(HWWI), Hamburg.
[13] Gani, A. (2017). The Empirical Determinants
of Corruption in Developing Countries.
Perspectives on Global Development and
Technology, 16(4), 392-409.
[14] Nafi Ghaniy, Fithra Faisal Hastiadi Political,
Social and Economic Determinants of
Corruption International Journal of Economics
and Financial Issues, 2017, 7(4), 144-149.
[15] Veronika Linhartová, Martina Halásková
Determinants of corruption: a panel data
analysis of Visegrad countries Equilibrium.
Quarterly Journal of Economics and Economic
Policy, 2022, vol. 17, issue 1, 51-79
[16] Fatih Karasac, Halil Kete Panel Data Analysis
on the Socio-Economic Determinants of
Corruption in the D-8 Countries İstanbul
Üniversitesi Sosyoloji Dergisi, 42(Suppl. 1):
117-131 DOI:
[17] Çera,E., Sinamati,A. (2017). Determinants of
Corruption Perception in a Transition
Country: Case of Albania. Baltic Journal of
Real Estate Economics and Construction
Management, November 2017, 5, 177–185.
doi: 10.1515/bjreecm-2017-0013.
[18] Mangafic, J.,Veselinovic,L.(2020). The
determinants of corruption at the individual
level: evidence from Bosnia-Herzegovina.
Economic Research-Ekonomska Istraživanja
2020, Vol. 33, No. 1, 670–2691.
[19] Bai,J., Jayachandran, S.,Malesky,E.J.,
Olken, B.A. (2013), “Does Economic Growth
Reduce Corruption? Theory and Evidence
from Vietnam”, NBER Working Papers
Series 19483, September 2013.
[20] Bouzid, B.N., (2016), “Dynamic Relationship
Between Corruption and youth Employment”,
Policy Research Working Paper No. 7842,
Washington DC, World Bank.
[21] You, Jong-Sung, and Sanjeev Khagram, “A
Comparative Study of Inequality and
Corruption”, American Sociological Review,
Vol 70, No.1:136-157.
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
359
Volume 2, 2024
APPENDIX
Fig. 1A. Normality Test of Residuals
0
2
4
6
8
10
12
-0.2 -0.1 0.0 0.1 0.2
Series: Standardized Residuals
Sample 2012 2022
Observations 66
Mean 4.21e-19
Median -0.002852
Maximum 0.272035
Minimum -0.196076
Std. Dev. 0.079744
Skewness 0.415143
Kurtosis 4.117506
Jarque-Bera 5.330032
Probability 0.069598
Table 1A. Residual Cross-Section Dependence Test
Test
Statistic
d.f.
Prob.
Breusch-Pagan LM
21.30665
15
0.1273
Pesaran scaled LM
1.151431
0.2496
Bias-corrected scaled
LM
0.851431
0.3945
Pesaran CD
-1.345916
0.1783
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
The authors equally contributed in the present
research, at all stages from the formulation of the
problem to the final findings and solution.
Sources of Funding for Research Presented in a
Scientific Article or Scientific Article Itself
No funding was received for conducting this study.
Conflict of Interest
The authors have no conflicts of interest to declare
that are relevant to the content of this article.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US
Financial Engineering
DOI: 10.37394/232032.2024.2.33
Edmira Cakrani, Dimitrios A. Karras, Gjergji Shqau
E-ISSN: 2945-1140
360
Volume 2, 2024