Panel Data Research on Corruption. Russia's perspective
|Author:||Bc. Ksenia Pogodina|
|Year:||2015 - summer|
|Leaders:|| doc. Ing. Tomáš Cahlík CSc.
|Work type:|| Bachelors
|Awards and prizes:|
|Abstract:||The thesis assesses causes and consequences of public sector corruption, using panel
data specification. The model presented in this work extends and updates the existing
model, where new studies are incorporated and new methods for empirical evidence are
used. Moreover, the analysis is expected to be more accurate and comprehensive than
the existing one, since cross-sectional analysis is substituted by panel data analysis,
which captures unobserved heterogeneity and country specific effects. The main
correlations I am interested in are between the level of public corruption in a country
and three other important variables: level of market competitiveness within economy,
level of education in a country and the extent of democracy there. Hence, the topic is
covered from both points of view: theoretical and empirical. Additionally, the model is
applied for different samples of countries (developing and developed) in order to
investigate if global tendencies hold for specific groups of countries or not.
Furthermore, the work includes an example of Russian economy, where it is studied
from theoretical and graphical perspective and only after that the proper inference is
made, applying my general model for sample of developing countries.
Empirical research shows that corruption and competition are negatively related. In
addition, higher secondary education and more political rights (democracy) have
depressing effect on corruption in a country. On the other hand, increase in percentage
of people with tertiary education leads to higher corruption. However, when the full
sample was divided for samples of developing and developed countries, the support for
all above mentioned hypotheses was not found, since some variables were insignificant.
The methods implemented in current work are as follows: G2SLS random-effect IV and
Fixed-effects (within) IV regressions. That is a combination of Fixed effects and
Random effects models with Instrumental Variable technique. Additionally, OLS and
2SLS methods are used.