- Email: [email protected]
Income inequality and level of corruption in post-communist European countries between 1995 and 2014
Communist and Post-Communist Studies 52 (2019) 93e104
Contents lists available at ScienceDirect
Communist and Post-Communist Studies journal homepage: www.elsevier.com/locate/postcomstud
Income inequality and level of corruption in post-communist European countries between 1995 and 2014 Kristýna Basna Institute of Sociology, Czech Academy of Sciences, Prague, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history: Available online 11 May 2019
This paper analyses the relationships between income inequality and corruption in Europe and looks specifically at post-communist European countries. The scientific community agrees that there is important relationship between income inequality and corruption and many authors believe that low income inequality is connected to low corruption. According to empirical papers, this is true not only on the European scale, but also on a global scale. In this paper, I test this claim by conducting a multilevel analysis on 39 European countries in the period of 1995e2014. This model ascertains that there are immense differences between post-communist countries and the rest of European countries. The effects of income inequality on the level of corruption are discussed. © 2019 The Regents of the University of California. Published by Elsevier Ltd. All rights reserved.
Keywords: Corruption Europe Income inequality Post-communist countries Quantitative
1. Introduction Corruption is a threat to society and to good governance (Scott, 1972) that decreases the quality of the public sector in many areas and may trigger civil unrest (Brown et al., 2011; Pellegata, 2012). Moreover, as Karklins adds (2005: 4), it involves “the loss of equal access to public power and position”, which leads to “the loss of public trust and belief in the political system” and also has a negative impact on people's life satisfaction (Helliwell, 2006). Corruption is also dangerous from an economic point of view. It can be a barrier to economic growth (World Bank, 1997b), negatively affecting the ratio of investment to GDP (Mauro, 1995; World Bank, 1997a) and the level of foreign investment (Wei, Wu, 2001). Finally, it can also contribute to an uncertain business climate, hold back state reform and nourish organized crime (Rose-Ackerman, 1999: 17). Due to the negative effects of corruption, many social scientists have tried to discover and describe the causes of corruption. This task is complicated by the fact that corruption is a clandestine activity, which makes it very difficult to measure and to detect its true effects, as well as its underlying causes. Moreover, corruption is a very complicated phenomenon, which may work and be understood differently in different cultural contexts (Charron et al., 2013). As most authors have conducted their research on a global level or have done case studies, a cross-country analysis including only European countries, which share a common culture, could show the validity of previous research. This article therefore looks at whether income inequality, which was identified as one of the important variables affecting the level of corruption on a global level, also influences corruption when tested exclusively on a European level. The situation in Europe is unique, because Central and Eastern European countries underwent transition to democracy around 30 years ago. This allows us to test also the effect of
E-mail address: [email protected]. https://doi.org/10.1016/j.postcomstud.2019.05.002 0967-067X/© 2019 The Regents of the University of California. Published by Elsevier Ltd. All rights reserved.
94
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
democratisation on corruption and focus on the differences between post-communist European countries and countries without the history of communist rule. 2. Inequality and corruption: literature review Many authors focused on relations between different economic and cultural indicators and corruption. One of the best overview of the different conclusions based on empirical testing of these connections is provided in Treisman's (2007) article which presents indicators that have been found to be predictors of the level of corruption in a country. One of the important indicators is GDP (Kaufmann et al., 1999; Gupta et al., 2002; Lambsdorff, 2003; Treisman, 2000). Most of the authors found that higher GDP per capita decreases corruption (Gupta et al., 2002; Lambsdorff, 2003; Treisman, 2000). However, as Drury et al. (2006) found out, GDP and corruption are related rather in undemocratic regimes. Another important variable is the level of democracy (Blake, Martin, 2006; Pellegata, 2012; Brown et al., 2011; Treisman, 2000), although it seems that the level of democracy influences corruption ambiguously: countries transitioning from authoritarian regime to democracy seem to have higher levels of corruption than democratic or authoritarian countries. Moreover, according to the number of authors, religious beliefs and specifically the share of practicing Protestants in a country are important in determining the level of corruption in a country. In fact, several authors found that countries with a predominant Protestant population have lower corruption levels than countries, which are predominantly Muslim, Orthodox, or Catholic (Treisman, 2000; La Porta et al., 1999). Moreover, La Porta et al. (1999) also provide evidence that Catholicism, Orthodox Christianity, and Islam are more “hierarchical” and less individualistic, and exhibit inferior government performance that might explain higher levels of perceived corruption. Finally, a substantial body of literature focuses on the relationship between corruption and inequality. Most researchers agree that there is indeed a very important relationship; the results of quantitative analyses suggest that countries which have higher corruption also exhibit higher inequality (Treisman, 2007). Some authors also suggest that containing inequality might be a plausible method for lowering corruption (Rose-Ackerman, Soreide, 2006). However, the relationship is probably not that simple: corruption is likely to be the cause and also the consequence of inequality (Rose-Ackerman, Soreide, 2006: 23; Husted, 1999). For example, Gupta et al. (2002) found a significant correlation between income inequality and corruption on a selection of 37 countries. The authors argue that corruption increases inequality; in fact, an increase of one standard deviation in corruption increases the Gini Coefficient of Income Inequality by 11 points. On the other hand, You and Khagram (2005: 70) argue that inequality increases corruption as well. As they say: “Income inequality increases the level of corruption through material and normative mechanisms. The wealthy have both greater motivation and more opportunity to engage in corruption, whereas the poor are more vulnerable to extortion and less able to monitor and hold the rich and powerful accountable as inequality increases” (2005). You and Khagram suggest that inequality increases corruption more strongly in democratic countries where the powerful are forced to hide their dishonest corrupt activities, whereas the powerful in autocratic regimes can oppress the poor without having to hide it (2005). This implies that the effect of income inequality might be less strong in post-communist countries compared to European countries that have never experienced the communist rule. But on the other hand, Li et al. (2000) found that the higher the level of corruption, the stronger is the correlation between corruption and inequality. This might suggest that post-communist countries, which have high levels of corruption, might have also higher levels of inequality. However, as Charron et al. (2013) show in their model, although corruption and inequality are indeed strongly correlated, a few European countries (including Bulgaria, Slovenia and Slovakia) have low inequality and high corruption e which calls into doubt the relationship between inequality and control of corruption in the specific case of post-communist countries. Uslaner is another author closely focusing on the relation between corruption and inequality, presenting the concept of income inequality trap. Uslaner claims that the roots of corruption lie in unequal distribution of resources in society (2009: 127). He argues that there is an indirect link between economic inequality and corruption through trust: inequality lowers trust, which increases corruption (Uslaner, 2008). However, Uslaner does not include post-communist countries in his model. One can see that there is no agreement on the strength and direction of the relationship between corruption and inequality. Finally, the inequality case of post-communist countries might also be specific due to the fact that communist ideology increased equality in certain countries. After the fall of communism, the majority of citizens were educated and income was distributed relatively equally, which meant that the population was capable of participating in a modern state and economy (Kornai, Rose-Ackerman, 2004). 3. Corruption in European countries The level of corruption in European countries is much lower compared to most of the world; in particular, Scandinavian and Western European countries consistently occupy the top positions as countries with the lowest levels of corruption. On the other hand, countries with a communist history generally have higher levels of corruption (Shleifer, 1997) and political corruption is certainly a serious problem there (Karklins, 2005). Communist regimes provide incentives to bribe either due to scarcity of goods or due to the low regime legitimacy connected to the general belief that it is not entirely wrong to steal from government. As the famous Czech saying illustrates: “He who does not steal from the state, steals from his own family”. It has been suggested (Rose, 2001: 105; Rose-Ackerman, 1999) that corruption is the greatest obstacle to progress and democratization in post-communist societies.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
95
Fig. 1. Control of Corruption. Source: World Bank, Control of Corruption, 2014. Control of Corruption shows how the countries are successful in controlling corruption, the indicator goes from 3 to 3, while 3 indicates that country is successful in controlling the level of corruption.
Fig. 1 shows the situation in European countries concerning the level of corruption in 2014 as measured by the World Bank. Their composite indicator is entitled “Control of Corruption” and is based on 22 underlying data sources, which are rescaled and combined to capture “perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as the ‘capture’ of the state by elites and private interests” (World Bank, n.d.). This indicator ranges from 3 to 3, where 3 means that the country is able to control the level of corruption quite well, and 3 means that there is almost no control of corruption in the country. Fig. 1 shows that there is a difference between postcommunist countries and countries, which have never had communist rule. In general, post-communist countries are stacked into the left, having low control of corruption, and the rest of countries are in the right, having generally higher control over corruption. The division is not perfect: Greece is doing very poorly being eighth from the bottom, followed by Italy which is eleventh from the bottom. On the other hand, Estonia is clearly the winner among post-communist countries and Slovenia and Poland are not doing that bad either. Similar results can be observed for direct experience with corruption (Fig. 2) as measured by the Eurobarometer survey. Eurobarometer surveys asked respondents whether anyone asked them, or expected of them, to pay a bribe for his or her service over the last 12 months. Countries are clearly divided by their communist history, although there are some exceptions e Greece is surrounded by post-communist countries and Slovenia and Estonia rather by countries without communist legacy. As shown in Table 1, there is a significant difference within post-communist countries. In fact, on average, the Control of Corruption remains on a similar level between 1995 and 2014, increasing most significantly in Serbia, Croatia or Macedonia, but decreasing in Slovenia, Hungary or in Moldova. It is therefore important not to only look at the post-communist countries
Fig. 2. Direct experience with corruption. Source: Eurobarometer, 2013.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
96
Table 1 Development of Control of Corruption over time. Country/year Average CC 1995e1999 Average CC 2000e2004 Average CC 2005e2009 Average CC 2010e2014 Difference between 2014 and 1995 AL BG BY CZ EE HR HU LT LV MD ME MK PL RO RS RU SI SK UA Average
1.05 0.51 0.78 0.60 0.26 0.77 0.62 0.01 0.42 0.25 0.82 0.60 0.45 1.06 0.98 1.31 0.30 1.09 0.25
0.78 0.10 0.73 0.32 0.75 0.10 0.62 0.21 0.02 0.82 0.54 0.68 0.34 0.35 0.75 0.82 0.84 0.19 0.96 0.18
0.65 0.17 0.69 0.32 0.93 0.03 0.50 0.10 0.22 0.61 0.29 0.29 0.26 0.19 0.32 0.94 0.97 0.34 0.78 0.08
0.63 0.25 0.56 0.26 1.03 0.05 0.25 0.33 0.21 0.70 0.16 0.01 0.52 0.20 0.26 0.99 0.79 0.14 1.02 0.07
0.42 0.26 0.22 0.34 0.77 0.81 0.36 0.35 0.64 0.45 0.82 0.08 0.25 0.79 0.02 0.52 0.16 0.08 0.19
Source: World Bank. Control of Corruption shows how the countries are successful in controlling corruption, the indicator goes from 3 to 3, while 3 indicates that country is successful in controlling the level of corruption.
as a group, but to also observe the differences within countries. According to some authors (Johnson, 2005), it seems that democratization in Central Europe has not reduced corruption. 4. Income inequality e current state and development in time Income inequality varies significantly across European countries. Fig. 3 shows these differences on the UNU-WIDER, Eurostat, and WB datasets indicating average Gini Coefficients in 2013e2014. One can see that there is no clear division between post-communist countries and the rest of Europe: post-communist countries are both among those with the lowest average inequality levels (Norway and Iceland are followed by Ukraine, Slovenia, and the Czech Republic) and those with the highest income inequality e Macedonia, Bulgaria, and Latvia. Looking at the development of the Gini Coefficient over time (Fig. 4), the data cover the period of 1990e2014, that is, from right after the fall of communism until today. For better clarity, I divided the data into several sub-regions according to the UN Statistics Division (United Nations, n.d.), while graphs for each country separately can be found in Fig. 9. We can see that income inequality has not changed very much over the years in countries, which have not experienced communist rule, but a change in income inequality did take place in many post-communist countries. While Uslaner (2008) claims that inequality is on a sharp rise in post-communist countries, Figs. 4 and 9 show that this is only true for some regions and countries. As Fig. 4 shows, it seems that the sharp rise in Gini coefficient took place right after the transition, but it remained steady afterwards.
Fig. 3. Income inequality. Source: UNU-WIDER, Eurostat, World Bank, 2013e2014 average.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
97
Fig. 4. Development of the Gini coefficient, 1990e2014. Source: UNU-WIDER, Eurostat, World Bank.
Even though there are some countries experiencing an increase in the Gini Coefficient, such as Macedonia, Latvia or Romania, there are also countries without any significant changes such as Slovenia, Hungary, Russia and Lithuania, and finally, Ukraine, Moldova or Serbia have experienced a decline in income inequality. It seems that the relations between corruption and income inequality are not straightforward in the case of postcommunist countries therefore a more in-depth analysis might reveal trends in the development of corruption in connection to income inequality in post-communist countries. The following pages will look into this issue and present analysis of relationships between income inequality and level of corruption with a special focus on post-communist countries. 5. Methodology In my analysis, I focus on 39 European countries (see Table 2); of the countries included, 19 have a communist past and the rest (20) do not. I use Eurostat, UNU-WIDER, Polity IV, ARDA World Religion, and World Bank (WB) data as the sources for my analysis. 5.1. Dependent variable Most authors use composite indicators for measuring corruption; among the most widely used are the Corruption Perception Index by Transparency International and the Control of Corruption measure by the World Bank (Fazekas et al., 2013; Kapoor and Ravi, 2012; Zakaria, 2013). The Control of Corruption index by the World Bank is due to number of
Table 2 Countries included into analysis and their codes. Country
Code
Country
Code
Albania Austria Belgium Bulgaria Belarus Switzerland Cyprus Czech Republic Germany Denmark Estonia Spain Finland France Great Britain Greece Croatia Hungary Ireland Iceland
AL AT BE BG BY CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS
Italy Lithuania Luxemburg Latvia Moldavia Montenegro Macedonia Malta Netherlands Norway Poland Portugal Rumania Serbia Russia Sweden Slovenia Slovakia Ukraine
IT LT LU LV MD ME MK MT NL NO PL PT RO RS RU SE SI SK UA
98
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
reasons more suitable indicator for analysis in time than CPI from Transparency International (Chabova, 2017), especially because Transparency International changed its methodology several times since the beginning of its research (Transparency International, 2012), therefore comparison in time is problematic. The data on Control of Corruption that will be used in this analysis is collected every year and it is comparable over time (World Bank, n.d.); therefore, it is possible to perform pseudopanel analysis in time.
5.2. Independent variables I used World Bank data for GDP per capita (given the highly skewed nature of GDP, I logged the indicator to obtain a more normal distribution appropriate for the regression analysis) and I used the Gini Coefficient measured by UNU Wider, WB and Eurostat for the indicator of inequality. I used the Polity IV index (Centre for Systemic Peace, n.d.) to measure democracy. Polity IV indicates explicitly the level of democracy in each country and it varies from 10 to 10, with 10 indicating the most democratic country. To discover the share of Protestants in a country, I used data from ARDA (n.d.) which shows the percentage of population practicing specific religion in each state for every five years. Due to the fact that data on religions are collected only every five years I used interpolation to fill in the missing years and increase the number of observations, this approach is acceptable as share of religious population does not change significantly from one year to another.
6. Results I fitted four regression models, one for each wave. Firstly, I tested whether the theories explaining the level of corruption on the global scale also work on a European level. I expected that the results would be similar to those on the global scale, that is, that a higher GDP per capita, lower income inequality, higher levels of democracy, and a higher share of Protestants would be connected with higher levels of Control of Corruption. However, as discussed above, communism has a strong influence on the level of corruption in a country. It is therefore possible that the effect of variables could be very different in countries, which have a history of communist rule as opposed to countries, which did not experience communist rule. The equation for our models then follows:
Control of Corruptionit ¼ ai þ b Democracyit þ g GDP per capitait þ d Share of Protestantsit ε Giniit z communist historyit þ uit Control of Corruption should be positively influenced by democracy, GDP per capita, share of Protestants, and negatively by Gini Coefficient and communist past. Table 3 shows the results of the OLS regression for the four waves. It is visible that there is a problem with the Gini Coefficient as it is non-significant contrary to our expectation. Moreover, the level of democracy is not an important predictor. However, this non-existent effect of democracy on corruption can be explained by the fact that the vast majority of countries have been democracies in recent years e hence there is no variability among countries. Interestingly, neither the fact of being a post-communist country is an important determinant of control of corruption in the recent waves, which suggests that there might be important differences within the group of post-communist countries. On the other hand, share of protestant in a country has an important effect on the Control of Corruption e as expected, the higher the share of Protestants, the higher the Control of Corruption in a country. To gain more information about the relationships between Control of Corruption and the independent variables, it would be helpful to look at the effects altogether, not divided by time period. As pooling is not appropriate for panel-type data, I used pseudo-panel data analysis. I used fixed and random effects, yet Hausman test indicated random effects analysis to be more appropriate for these data; Table 4 therefore shows only the results for random effects. Of the 39 countries and 419 observations in total, the lowest number of observations per group was 4 and highest was 12 (average number of observations per group was 10.7).
Table 3 OLS Regression analysis e Determinants of Control of Corruption.
GDP per capita (ln) Gini Coefficient Post-communist country Share of Protestants Democracy Intercept Adjusted R2 Number of cases
1.1. 1995e1999
1.2. 2000e2004
1.3. 2005e2009
1.4. 2010e2014
0.45* (0.149) 0.01 (0.022) 0.56 (0.312) 0.852* (0.283) 0.06* (0.02) 3.31 0.88 36
0.528* (0.111) 0.01 (0.015) 0.439* (0.216) 0.929* (0.211) 0.033 (0.023) 4.05 0.91 39
0.531* (0.102) 0.017 (0.016) 0.396 (0.211) 0.879* (0.249) 0.039 (0.022) 4.09 0.88 39
0.602* (0.122) 0.022 (0.02) 0.267 (0.237) 0.822* (0.303) 0.04 (0.025) 4.75 0.84 39
*p < 0.05. In 1995e1999 wave data for IS, ME, MT is missing.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
99
Table 4 Pseudo-panel data analysis - Determinants of Control of Corruption.
Ln GDP per capita Gini Coefficient Communist history Communist history*Gini Coeff. Share of Protestants Democracy Intercept sigma_u sigma_e Rho N of observations N of groups F (Wald chi) Prob > F corr(u_i, Xb)
Model 2.1
Model 2.2
0.25 (0.05)* 0.003 (0.03) 0.93 (0.15)*
0.26 (0.05)* 0.01 (0.004)* 1.47 (0.25)* 0.07 (0.006)* 1.06 (0.207)* 0.04 (0.007)* 1.379 0.4 0.173 0.795 419 39 385.12 0.000 0 (assumed)
1.12 (0.204)* 0.04 (0.007)* 1.566 0.336 0.174 0.788 419 39 384.2 0.000 0 (assumed)
*p < 0.05.
As Model 2.1. in Table 4 shows, the results confirm almost perfectly the findings of the above-mentioned authors analysing the relations between corruption and different variables (for example, that higher GDP per capita, share of Protestants, and level of democracy would increase Control of Corruption and Communist history would decrease CC) with one exception, namely that the Gini Coefficient is not a significant predictor of Control of Corruption. In order to explore what is behind these results, I included an interaction between the Gini Coefficient and being a post-communist country; the results are presented under Model 2.2. Theories suggest that more equality should be connected to less corruption. However, as we can see in the case of post-communist countries, the relationship is reverse, that is, there is more corruption in countries with more equality. As Uslaner writes: “The connection between inequality and the quality of government is not necessarily so simple: As the former Communist nations of Central and Eastern Europe show, you can have plenty of corruption without economic inequality” (Uslaner, 2009). These results are very surprising; in the case of post-communist countries, inequality does not seem to influence control of corruption as it apparently does in countries that have not experienced communist rule. One explanation could be that there was a development of income inequality in time while corruption remained static. This could, in pooled data, create a negative relationship between income inequality and Control of Corruption. A similar explanation is suggested by Uslaner (2009), who argues that even though post-communist countries have low income inequality and high levels of corruption e which undermines the theory of income inequality increasing corruption e this state is only temporary. According to him, after transition to democracy, income inequality should boost and catch up with the rest of Europe. Taking into account Uslaner's claim that even though inequality is low in post-communist countries, the Gini Coefficient is quickly raising, which influences corruption, we should be able to see a shift in the Gini Coefficient and analogical shifts in Control of Corruption. And when first looking at Fig. 5 it seems that Uslaner (2009) is correct: in fact, the average Gini Coefficient in post-communist countries rose very sharply and caught up with the Gini Coefficient in European countries with no communist history in the early 90s. On average, the difference in the Gini Coefficient between post-communist countries and
Fig. 5. Development of the Gini coefficient in Europe. Source: UNU-WIDER, Eurostat, World Bank.
100
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
Fig. 6. Control of Corruption, development in time. Source: World Bank, Control of Corruption. Control of Corruption shows how the countries are successful in controlling corruption, the indicator goes from 3 to 3, while 3 indicates that country is successful in controlling the level of corruption.
the rest of Europe is not big; in fact, the Gini Coefficient in post-communist countries is higher than in the rest of Europe, except for the financial crisis years. However, as discussed before, it seems that each post-communist country has a different path concerning income inequality: in some countries, income inequality is rising, while it remains constant or it is even decreasing in others. Therefore, if we were to believe Uslaner, we would expect corruption to increase or decrease according to the Gini Coefficient. As this theory expects Control of Corruption not to change in time, it is essential to check the development of Control of Corruption in time in post-communist countries as opposed to countries with no history of communist rule. Fig. 6 shows that Control of Corruption has not changed very much over the years e there is still a very big gap between countries that have experienced communist rule and the rest of European countries. This gap has been closing very slowly over the years, yet the improvement of post-communist countries as a group is very small. It is worth noting that Control of Corruption in European countries without a history of communism has been decreasing very slowly over the years. Of course, this graph shows only the development of Control of Corruption in countries as groups, and there is very likely high variability across countries (see Fig. 10). If the hypothesis that inequality influences corruption (or vice versa) is correct, we should be able to see a shift in corruption in those countries where the Gini Coefficient changed. For this reason, it is essential to look at the changes in Gini Coefficient and corruption over years within the groups of countries. The following graphs (Figs. 7 and 8) show the correlations between change of Gini Coefficient (2010e2014 vs. 1995e2000) and change of Control of Corruption (2010e2014 vs. 1995e2000). Fig. 7 shows the relationships between Control of Corruption and Gini Coefficient in European countries without the legacy of communism: there is negative correlation, meaning that lower income inequality is associated with higher control of corruption. The relationship is not very strong, with a correlation at the level of 0.1964.
Fig. 7. Countries without communist legacy e Control of Corruption vs. Gini Coefficient, differences between 2010 and 2014 vs. 1995e2000. Source: WB, UNU-WIDER, Eurostat.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
101
Fig. 8. Post-communist countries e Control of Corruption vs. Gini Coefficient, differences between the years 2010e2014 vs. 1995e2000. Source: WB, UNU-WIDER, Eurostat.
Fig. 8 on the other hand, shows the relationships between the Gini Coefficient and Control of Corruption in postcommunist countries; in this case, the correlation is much stronger (0.4279) and moreover, positive e it has the opposite direction than in the case of countries without the history of communism. It seems that the bigger the change toward a more unequal society, the bigger the change also toward a less corrupt society. These results go against the theories explaining corruption. Moreover, this relationship remains very strong even when deleting from the analysis the countries, which are the strongest drivers of this relationship, such as Moldova or Macedonia. Furthermore, when we consider not the same time
Fig. 9. Development of Gini coefficient in Europe. Source: UNU-WIDER, Eurostat, World Bank.
102
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
Fig. 10. Development of Control of corruption in Europe. Source: World Bank, Control of Corruption. Control of Corruption shows how the countries are successful in controlling corruption, the indicator goes from 3 to 3, while 3 indicates that country is successful in controlling the level of corruption.
frame as in the case of Control of Corruption but the preceding wave (1990e1995), the relationship weakens (0.3435) but remains positive. In the light of these findings, it would be plausible to adjust the equation explaining corruption and to look at the effect of the Gini Coefficient in the first wave and whether it influences the current level of Control of Corruption. As the Gini Coefficient would precede Control of Corruption, we can get a hint of causality. However, of course, we cannot infer a causal effect with certainty as income inequality could in theory influence corruption during a longer time span than twenty years. The updated equation follows:
Control of Corruptionit ¼ ai þ b Democracyit þ g GDP per capitait þ d Share of Protestantsit ε Giniit3 z communist historyit þ uit Table 5 shows the results of the regression analysis; there is no need for pseudo-panel data analysis as we are using the effect of Gini in the first wave (1995e2000) on Control of Corruption in the latest wave (2010e2014). Unsurprisingly, given the previous analysis, the two models show the same effects of income inequality as before. Model 3.1 shows that income inequality in the past, but also being a post-communist country, does not have a significant effect on Control of Corruption. However, as Model 3.2 shows, the results change after including the interaction between income inequality and being a post-communist country. In fact, the Gini Coefficient in the interaction is positive and significant, showing that in post-communist countries, the level of income inequality in the 90s influenced the control of corruption today in a way that high income inequality predicted high control of corruption. This effect is reversed for countries without a history of communism. There might be more explanations to these effects; one of them is supported by Uslaner (2009), who claims that even though post-communist countries have low income inequality while having high corruption due to the rule of communism, this relationship will reverse quickly after post-communist countries catch up in income inequality with the rest of European
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
103
Table 5 Regression analysis e Determinants of corruption, Gini t1.
Ln GDP per capita Gini Coefficient, t3 Post-communist country Communist history*Gini Coeff., t3 Share of Protestants Democracy Intercept Adjusted R2 Number of cases
3.1.
3.2.
0.58* (0.147) 0.013 (0.02) 0.284 (0.273)
0.613* (0.14) 0.07* (0.034) 2.71* (1.22) 0.79* (0.039) 0.938* (0.314) 0.037 (0.024) 3.37 0.86 37
1.115* (0.317) 0.03 (0.03) 4.79 0.84 37
*p < 0.05. IS and MT are missing in this analysis due to missing data.
countries. However, our analysis shows that this claim is not supported by the development of corruption in Europe. It is true that post-communist countries as a group quickly caught up with income inequality levels of the European countries that did not experience communism, but analysis of changes within the group of post-communist countries actually shows the contrary effect. Post-communist countries that have had the highest change in income inequality towards more unequal societies are today the least corrupt and vice versa. In addition to Uslaner's explanation, this effect can be also accounted for by socio-economic changes in post-communist countries. It seems that the highest increase in income inequality might be connected to the differences in countries' socioeconomic transformation, such as the expansion of private sector, retrenchment of redistributive state or promoting foreign investment in their privatization (Bandelj and Mahutga, 2010). It seems that countries that embraced market exchange and capitalism have the highest increase in income inequality and might have been also the most successful in transforming their other institutions connected to good governance and thus decreasing the level of corruption in their country. However, as the level of corruption in a country changes only very slowly in time, more time is needed to confirm this hypothesis. 7. Conclusion This article shows that theories, which point out to the strong correlation between corruption and the Gini Coefficient are not entirely correct. This correlation holds for European countries, which have not experienced communist rule, but it is not true for post-communist countries. This article takes historical data from 1990 to 2014 and observes the developments of the Gini Coefficient and corruption perceptions. The analysis shows that analysis of data from post-communist countries does not support the theory that high income inequality increases corruption; in the case of post-communist countries, unlike in the rest of Europe, high income inequality actually decreases corruption. This is supported by three types of analysis: firstly, by analysing the relationships between Control of Corruption and the Gini Coefficient by time intervals (waves) since 1995, then by multilevel analysis, and finally by analysing the change in these two indicators from 1995 to 2014. However, as the time period we currently have is only 20 years long, we will have to wait for a longer time series to assess whether the results presented by this article hold. Nevertheless, due to the unexpected results of this analysis, it is important to treat post-communist countries in models including corruption and income inequality with caution, as this group behaves significantly differently than the rest of European countries and also differently from the results on a global level. References ARDA. World religion dataset: national religion dataset. Available at: http://www.thearda.com/Archive/Files/Descriptions/WRDNATL.asp. Bandelj, N., Mahutga, M.C., 2010. How socio-economic change shapes income inequality in post-socialist Europe. Soc. Forces 88 (5), 2133e2161. Blake, C.H., Martin, C.G., 2006. The dynamics of political corruption: Re-examining the influence of democracy. Democratization 13 (1), 1e14. Brown, D.S., Touchton, M., Whitford, A., 2011. Political polarization as a constraint on corruption: a cross-national comparison. World Dev. 39 (9), 1516e1529. Centre for Systemic Peace ed., Polity IV Annual Time-Series, 1800-2015, Available at: http://www.systemicpeace.org/inscrdata.html.n.d. Chabova, K., 2017. Measuring corruption in Europe: public opinion surveys and composite indices. Qual. Quantity 51 (4), 1877e1900. Charron, N., Lapuente, V., Rothstein, B., 2013. Quality of Government and Corruption from a European Perspective: A Comparative Study of Good Government in EU Regions. Edward Elgar Publishing, Northampton. Drury, C.A., Krieckhaus, J., Lusztig, M., 2006. Corruption, democracy, and economic growth. Int. Political Sci. Rev. 27 (2), 121e136. Fazekas, M., Toth, I.J., King, L.P., 2013. Anatomy of Grand Corruption: A Composite Corruption Risk Index Based on Objective Data (No. CRC-Wp/2013:02). Corruption Research Center Budapest, Budapest, pp. 1e46. Gupta, S., Davoodi, H., Alonso-Terme, R., 2002. Does corruption affect income inequality and poverty? Econ. Govern. 3, 23e45. Helliwell, 2006. Well-being, social capital and public policy: what's new? Econ. J. 116, C34eC45. Husted, B.W., 1999. Wealth, culture, and corruption. J. Int. Bus. Stud. 30 (2), 339e359. Johnson, M., 2005. Syndromes of Corruption: Wealth, Power, and Democracy. Cambridge University Press, Cambridge. Kapoor, M., Ravi, S., 2012. Determinants of corruption: government effectiveness vs. cultural norms. B. E. J. Econ. Anal. Policy 12 (1), 1e24. Karklins, R., 2005. The System Made Me Do it: Corruption in Post-communist Societies. M.E. Sharpe, Armonk, NY ; London. n, P., 1999. Governance Matters. WB. Kaufmann, D., Kraay, A., Zoido-Lobato Kornai, J., Rose-Ackerman, S. (Eds.), 2004. Building a Trustworthy State in Post-socialist Transition. Palgrave Macmillan, New York. La Porta, R., et al., 1999. The quality of government. J. Law Econ. Organ. 15 (1), 222e279.
104
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
Lambsdorff, J.G., 2003. How corruption affects productivity. Kyklos 56 (4), 457e474. Li, H., Xu, L.C., Zou, H.-F., 2000. Corruption, income distribution, and growth. Econ. Pol. 12 (2), 155e182. Mauro, P., 1995. Corruption and growth. Q. J. Econ. 110 (3), 681e712. Pellegata, A., 2012. Constraining political corruption: an empirical analysis of the impact of democracy. Democratization 20 (7), 1195e1218. Rose, R., 2001. A diverging Europe. J. Democr. 12 (1), 93e106. Rose-Ackerman, S., 1999. Corruption and Government: Causes, Consequences, and Reform. Cambridge University Press, Cambridge. Rose-Ackerman, S., Soreide, T., 2006. International Handbook on the Economics of Corruption. Edward Elgar Publishing, Cheltenham. Scott, J.C., 1972. Comparative Political Corruption. Prentice-Hall, Englewood Cliffs, N.J. Shleifer, A., 1997. Government in Transition. European Economic Review, pp. 355e410. Transparency International, 2012. Corruption perceptions index 2012: technical methodology note. Available at: https://www.transparency.org/files/ content/pressrelease/2012_CPIUpdatedMethodology_EMBARGO_EN.pdf. Treisman, D., 2000. The causes of corruption: a cross-national study. J. Public Econ. 76, 399e457. Treisman, D., 2007. What have we learned about the causes of corruption from ten years of cross-national empirical research? Annu. Rev. Pol. Sci. 10 (1), 211e244. United Nations ed., Composition of Macro Geographical (Continental) Regions, Geographical Sub-regions, and Selected Economic and Other Groupings, Available at: http://millenniumindicators.un.org/unsd/methods/m49/m49regin.htm#europe.n.d. Uslaner, E.M., 2008. Corruption, Inequality, and the Rule of Law. Cambridge University Press, Cambridge. Uslaner, E.M., 2009. Corruption. In: Svendsen, G.T., Svendsen, G.L.H. (Eds.), The Handbook of Social Capital: the Troika of Sociology, Political Science and Economics. Edward Elgar Publishing, Northampton, pp. 127e142. Wei, S.J., Wu, Y., 2001. Negative Alchemy? Corruption, Composition of Capital Flows, and Currency Crises, NBER Working Paper Series 8187. World Bank. n.d. Control of corruption. Available at: http://info.worldbank.org/governance/wgi/pdf/cc.pdf. World Bank, 1997a. Helping Countries Combat Corruption. The World Bank. World Bank, 1997b. World Development Report: the State in a Changing World. Oxford University Press, New York. You, J.-S., Khagram, S., 2005. A comparative study of inequality and corruption. Am. Sociol. Rev. 70 (1), 136e157. Zakaria, P., 2013. Is corruption an enemy of civil society? The case of Central and Eastern Europe. Int. Political Sci. Rev. 34 (4), 351e371.
Contents lists available at ScienceDirect
Communist and Post-Communist Studies journal homepage: www.elsevier.com/locate/postcomstud
Income inequality and level of corruption in post-communist European countries between 1995 and 2014 Kristýna Basna Institute of Sociology, Czech Academy of Sciences, Prague, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history: Available online 11 May 2019
This paper analyses the relationships between income inequality and corruption in Europe and looks specifically at post-communist European countries. The scientific community agrees that there is important relationship between income inequality and corruption and many authors believe that low income inequality is connected to low corruption. According to empirical papers, this is true not only on the European scale, but also on a global scale. In this paper, I test this claim by conducting a multilevel analysis on 39 European countries in the period of 1995e2014. This model ascertains that there are immense differences between post-communist countries and the rest of European countries. The effects of income inequality on the level of corruption are discussed. © 2019 The Regents of the University of California. Published by Elsevier Ltd. All rights reserved.
Keywords: Corruption Europe Income inequality Post-communist countries Quantitative
1. Introduction Corruption is a threat to society and to good governance (Scott, 1972) that decreases the quality of the public sector in many areas and may trigger civil unrest (Brown et al., 2011; Pellegata, 2012). Moreover, as Karklins adds (2005: 4), it involves “the loss of equal access to public power and position”, which leads to “the loss of public trust and belief in the political system” and also has a negative impact on people's life satisfaction (Helliwell, 2006). Corruption is also dangerous from an economic point of view. It can be a barrier to economic growth (World Bank, 1997b), negatively affecting the ratio of investment to GDP (Mauro, 1995; World Bank, 1997a) and the level of foreign investment (Wei, Wu, 2001). Finally, it can also contribute to an uncertain business climate, hold back state reform and nourish organized crime (Rose-Ackerman, 1999: 17). Due to the negative effects of corruption, many social scientists have tried to discover and describe the causes of corruption. This task is complicated by the fact that corruption is a clandestine activity, which makes it very difficult to measure and to detect its true effects, as well as its underlying causes. Moreover, corruption is a very complicated phenomenon, which may work and be understood differently in different cultural contexts (Charron et al., 2013). As most authors have conducted their research on a global level or have done case studies, a cross-country analysis including only European countries, which share a common culture, could show the validity of previous research. This article therefore looks at whether income inequality, which was identified as one of the important variables affecting the level of corruption on a global level, also influences corruption when tested exclusively on a European level. The situation in Europe is unique, because Central and Eastern European countries underwent transition to democracy around 30 years ago. This allows us to test also the effect of
E-mail address: [email protected]. https://doi.org/10.1016/j.postcomstud.2019.05.002 0967-067X/© 2019 The Regents of the University of California. Published by Elsevier Ltd. All rights reserved.
94
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
democratisation on corruption and focus on the differences between post-communist European countries and countries without the history of communist rule. 2. Inequality and corruption: literature review Many authors focused on relations between different economic and cultural indicators and corruption. One of the best overview of the different conclusions based on empirical testing of these connections is provided in Treisman's (2007) article which presents indicators that have been found to be predictors of the level of corruption in a country. One of the important indicators is GDP (Kaufmann et al., 1999; Gupta et al., 2002; Lambsdorff, 2003; Treisman, 2000). Most of the authors found that higher GDP per capita decreases corruption (Gupta et al., 2002; Lambsdorff, 2003; Treisman, 2000). However, as Drury et al. (2006) found out, GDP and corruption are related rather in undemocratic regimes. Another important variable is the level of democracy (Blake, Martin, 2006; Pellegata, 2012; Brown et al., 2011; Treisman, 2000), although it seems that the level of democracy influences corruption ambiguously: countries transitioning from authoritarian regime to democracy seem to have higher levels of corruption than democratic or authoritarian countries. Moreover, according to the number of authors, religious beliefs and specifically the share of practicing Protestants in a country are important in determining the level of corruption in a country. In fact, several authors found that countries with a predominant Protestant population have lower corruption levels than countries, which are predominantly Muslim, Orthodox, or Catholic (Treisman, 2000; La Porta et al., 1999). Moreover, La Porta et al. (1999) also provide evidence that Catholicism, Orthodox Christianity, and Islam are more “hierarchical” and less individualistic, and exhibit inferior government performance that might explain higher levels of perceived corruption. Finally, a substantial body of literature focuses on the relationship between corruption and inequality. Most researchers agree that there is indeed a very important relationship; the results of quantitative analyses suggest that countries which have higher corruption also exhibit higher inequality (Treisman, 2007). Some authors also suggest that containing inequality might be a plausible method for lowering corruption (Rose-Ackerman, Soreide, 2006). However, the relationship is probably not that simple: corruption is likely to be the cause and also the consequence of inequality (Rose-Ackerman, Soreide, 2006: 23; Husted, 1999). For example, Gupta et al. (2002) found a significant correlation between income inequality and corruption on a selection of 37 countries. The authors argue that corruption increases inequality; in fact, an increase of one standard deviation in corruption increases the Gini Coefficient of Income Inequality by 11 points. On the other hand, You and Khagram (2005: 70) argue that inequality increases corruption as well. As they say: “Income inequality increases the level of corruption through material and normative mechanisms. The wealthy have both greater motivation and more opportunity to engage in corruption, whereas the poor are more vulnerable to extortion and less able to monitor and hold the rich and powerful accountable as inequality increases” (2005). You and Khagram suggest that inequality increases corruption more strongly in democratic countries where the powerful are forced to hide their dishonest corrupt activities, whereas the powerful in autocratic regimes can oppress the poor without having to hide it (2005). This implies that the effect of income inequality might be less strong in post-communist countries compared to European countries that have never experienced the communist rule. But on the other hand, Li et al. (2000) found that the higher the level of corruption, the stronger is the correlation between corruption and inequality. This might suggest that post-communist countries, which have high levels of corruption, might have also higher levels of inequality. However, as Charron et al. (2013) show in their model, although corruption and inequality are indeed strongly correlated, a few European countries (including Bulgaria, Slovenia and Slovakia) have low inequality and high corruption e which calls into doubt the relationship between inequality and control of corruption in the specific case of post-communist countries. Uslaner is another author closely focusing on the relation between corruption and inequality, presenting the concept of income inequality trap. Uslaner claims that the roots of corruption lie in unequal distribution of resources in society (2009: 127). He argues that there is an indirect link between economic inequality and corruption through trust: inequality lowers trust, which increases corruption (Uslaner, 2008). However, Uslaner does not include post-communist countries in his model. One can see that there is no agreement on the strength and direction of the relationship between corruption and inequality. Finally, the inequality case of post-communist countries might also be specific due to the fact that communist ideology increased equality in certain countries. After the fall of communism, the majority of citizens were educated and income was distributed relatively equally, which meant that the population was capable of participating in a modern state and economy (Kornai, Rose-Ackerman, 2004). 3. Corruption in European countries The level of corruption in European countries is much lower compared to most of the world; in particular, Scandinavian and Western European countries consistently occupy the top positions as countries with the lowest levels of corruption. On the other hand, countries with a communist history generally have higher levels of corruption (Shleifer, 1997) and political corruption is certainly a serious problem there (Karklins, 2005). Communist regimes provide incentives to bribe either due to scarcity of goods or due to the low regime legitimacy connected to the general belief that it is not entirely wrong to steal from government. As the famous Czech saying illustrates: “He who does not steal from the state, steals from his own family”. It has been suggested (Rose, 2001: 105; Rose-Ackerman, 1999) that corruption is the greatest obstacle to progress and democratization in post-communist societies.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
95
Fig. 1. Control of Corruption. Source: World Bank, Control of Corruption, 2014. Control of Corruption shows how the countries are successful in controlling corruption, the indicator goes from 3 to 3, while 3 indicates that country is successful in controlling the level of corruption.
Fig. 1 shows the situation in European countries concerning the level of corruption in 2014 as measured by the World Bank. Their composite indicator is entitled “Control of Corruption” and is based on 22 underlying data sources, which are rescaled and combined to capture “perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as the ‘capture’ of the state by elites and private interests” (World Bank, n.d.). This indicator ranges from 3 to 3, where 3 means that the country is able to control the level of corruption quite well, and 3 means that there is almost no control of corruption in the country. Fig. 1 shows that there is a difference between postcommunist countries and countries, which have never had communist rule. In general, post-communist countries are stacked into the left, having low control of corruption, and the rest of countries are in the right, having generally higher control over corruption. The division is not perfect: Greece is doing very poorly being eighth from the bottom, followed by Italy which is eleventh from the bottom. On the other hand, Estonia is clearly the winner among post-communist countries and Slovenia and Poland are not doing that bad either. Similar results can be observed for direct experience with corruption (Fig. 2) as measured by the Eurobarometer survey. Eurobarometer surveys asked respondents whether anyone asked them, or expected of them, to pay a bribe for his or her service over the last 12 months. Countries are clearly divided by their communist history, although there are some exceptions e Greece is surrounded by post-communist countries and Slovenia and Estonia rather by countries without communist legacy. As shown in Table 1, there is a significant difference within post-communist countries. In fact, on average, the Control of Corruption remains on a similar level between 1995 and 2014, increasing most significantly in Serbia, Croatia or Macedonia, but decreasing in Slovenia, Hungary or in Moldova. It is therefore important not to only look at the post-communist countries
Fig. 2. Direct experience with corruption. Source: Eurobarometer, 2013.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
96
Table 1 Development of Control of Corruption over time. Country/year Average CC 1995e1999 Average CC 2000e2004 Average CC 2005e2009 Average CC 2010e2014 Difference between 2014 and 1995 AL BG BY CZ EE HR HU LT LV MD ME MK PL RO RS RU SI SK UA Average
1.05 0.51 0.78 0.60 0.26 0.77 0.62 0.01 0.42 0.25 0.82 0.60 0.45 1.06 0.98 1.31 0.30 1.09 0.25
0.78 0.10 0.73 0.32 0.75 0.10 0.62 0.21 0.02 0.82 0.54 0.68 0.34 0.35 0.75 0.82 0.84 0.19 0.96 0.18
0.65 0.17 0.69 0.32 0.93 0.03 0.50 0.10 0.22 0.61 0.29 0.29 0.26 0.19 0.32 0.94 0.97 0.34 0.78 0.08
0.63 0.25 0.56 0.26 1.03 0.05 0.25 0.33 0.21 0.70 0.16 0.01 0.52 0.20 0.26 0.99 0.79 0.14 1.02 0.07
0.42 0.26 0.22 0.34 0.77 0.81 0.36 0.35 0.64 0.45 0.82 0.08 0.25 0.79 0.02 0.52 0.16 0.08 0.19
Source: World Bank. Control of Corruption shows how the countries are successful in controlling corruption, the indicator goes from 3 to 3, while 3 indicates that country is successful in controlling the level of corruption.
as a group, but to also observe the differences within countries. According to some authors (Johnson, 2005), it seems that democratization in Central Europe has not reduced corruption. 4. Income inequality e current state and development in time Income inequality varies significantly across European countries. Fig. 3 shows these differences on the UNU-WIDER, Eurostat, and WB datasets indicating average Gini Coefficients in 2013e2014. One can see that there is no clear division between post-communist countries and the rest of Europe: post-communist countries are both among those with the lowest average inequality levels (Norway and Iceland are followed by Ukraine, Slovenia, and the Czech Republic) and those with the highest income inequality e Macedonia, Bulgaria, and Latvia. Looking at the development of the Gini Coefficient over time (Fig. 4), the data cover the period of 1990e2014, that is, from right after the fall of communism until today. For better clarity, I divided the data into several sub-regions according to the UN Statistics Division (United Nations, n.d.), while graphs for each country separately can be found in Fig. 9. We can see that income inequality has not changed very much over the years in countries, which have not experienced communist rule, but a change in income inequality did take place in many post-communist countries. While Uslaner (2008) claims that inequality is on a sharp rise in post-communist countries, Figs. 4 and 9 show that this is only true for some regions and countries. As Fig. 4 shows, it seems that the sharp rise in Gini coefficient took place right after the transition, but it remained steady afterwards.
Fig. 3. Income inequality. Source: UNU-WIDER, Eurostat, World Bank, 2013e2014 average.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
97
Fig. 4. Development of the Gini coefficient, 1990e2014. Source: UNU-WIDER, Eurostat, World Bank.
Even though there are some countries experiencing an increase in the Gini Coefficient, such as Macedonia, Latvia or Romania, there are also countries without any significant changes such as Slovenia, Hungary, Russia and Lithuania, and finally, Ukraine, Moldova or Serbia have experienced a decline in income inequality. It seems that the relations between corruption and income inequality are not straightforward in the case of postcommunist countries therefore a more in-depth analysis might reveal trends in the development of corruption in connection to income inequality in post-communist countries. The following pages will look into this issue and present analysis of relationships between income inequality and level of corruption with a special focus on post-communist countries. 5. Methodology In my analysis, I focus on 39 European countries (see Table 2); of the countries included, 19 have a communist past and the rest (20) do not. I use Eurostat, UNU-WIDER, Polity IV, ARDA World Religion, and World Bank (WB) data as the sources for my analysis. 5.1. Dependent variable Most authors use composite indicators for measuring corruption; among the most widely used are the Corruption Perception Index by Transparency International and the Control of Corruption measure by the World Bank (Fazekas et al., 2013; Kapoor and Ravi, 2012; Zakaria, 2013). The Control of Corruption index by the World Bank is due to number of
Table 2 Countries included into analysis and their codes. Country
Code
Country
Code
Albania Austria Belgium Bulgaria Belarus Switzerland Cyprus Czech Republic Germany Denmark Estonia Spain Finland France Great Britain Greece Croatia Hungary Ireland Iceland
AL AT BE BG BY CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS
Italy Lithuania Luxemburg Latvia Moldavia Montenegro Macedonia Malta Netherlands Norway Poland Portugal Rumania Serbia Russia Sweden Slovenia Slovakia Ukraine
IT LT LU LV MD ME MK MT NL NO PL PT RO RS RU SE SI SK UA
98
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
reasons more suitable indicator for analysis in time than CPI from Transparency International (Chabova, 2017), especially because Transparency International changed its methodology several times since the beginning of its research (Transparency International, 2012), therefore comparison in time is problematic. The data on Control of Corruption that will be used in this analysis is collected every year and it is comparable over time (World Bank, n.d.); therefore, it is possible to perform pseudopanel analysis in time.
5.2. Independent variables I used World Bank data for GDP per capita (given the highly skewed nature of GDP, I logged the indicator to obtain a more normal distribution appropriate for the regression analysis) and I used the Gini Coefficient measured by UNU Wider, WB and Eurostat for the indicator of inequality. I used the Polity IV index (Centre for Systemic Peace, n.d.) to measure democracy. Polity IV indicates explicitly the level of democracy in each country and it varies from 10 to 10, with 10 indicating the most democratic country. To discover the share of Protestants in a country, I used data from ARDA (n.d.) which shows the percentage of population practicing specific religion in each state for every five years. Due to the fact that data on religions are collected only every five years I used interpolation to fill in the missing years and increase the number of observations, this approach is acceptable as share of religious population does not change significantly from one year to another.
6. Results I fitted four regression models, one for each wave. Firstly, I tested whether the theories explaining the level of corruption on the global scale also work on a European level. I expected that the results would be similar to those on the global scale, that is, that a higher GDP per capita, lower income inequality, higher levels of democracy, and a higher share of Protestants would be connected with higher levels of Control of Corruption. However, as discussed above, communism has a strong influence on the level of corruption in a country. It is therefore possible that the effect of variables could be very different in countries, which have a history of communist rule as opposed to countries, which did not experience communist rule. The equation for our models then follows:
Control of Corruptionit ¼ ai þ b Democracyit þ g GDP per capitait þ d Share of Protestantsit ε Giniit z communist historyit þ uit Control of Corruption should be positively influenced by democracy, GDP per capita, share of Protestants, and negatively by Gini Coefficient and communist past. Table 3 shows the results of the OLS regression for the four waves. It is visible that there is a problem with the Gini Coefficient as it is non-significant contrary to our expectation. Moreover, the level of democracy is not an important predictor. However, this non-existent effect of democracy on corruption can be explained by the fact that the vast majority of countries have been democracies in recent years e hence there is no variability among countries. Interestingly, neither the fact of being a post-communist country is an important determinant of control of corruption in the recent waves, which suggests that there might be important differences within the group of post-communist countries. On the other hand, share of protestant in a country has an important effect on the Control of Corruption e as expected, the higher the share of Protestants, the higher the Control of Corruption in a country. To gain more information about the relationships between Control of Corruption and the independent variables, it would be helpful to look at the effects altogether, not divided by time period. As pooling is not appropriate for panel-type data, I used pseudo-panel data analysis. I used fixed and random effects, yet Hausman test indicated random effects analysis to be more appropriate for these data; Table 4 therefore shows only the results for random effects. Of the 39 countries and 419 observations in total, the lowest number of observations per group was 4 and highest was 12 (average number of observations per group was 10.7).
Table 3 OLS Regression analysis e Determinants of Control of Corruption.
GDP per capita (ln) Gini Coefficient Post-communist country Share of Protestants Democracy Intercept Adjusted R2 Number of cases
1.1. 1995e1999
1.2. 2000e2004
1.3. 2005e2009
1.4. 2010e2014
0.45* (0.149) 0.01 (0.022) 0.56 (0.312) 0.852* (0.283) 0.06* (0.02) 3.31 0.88 36
0.528* (0.111) 0.01 (0.015) 0.439* (0.216) 0.929* (0.211) 0.033 (0.023) 4.05 0.91 39
0.531* (0.102) 0.017 (0.016) 0.396 (0.211) 0.879* (0.249) 0.039 (0.022) 4.09 0.88 39
0.602* (0.122) 0.022 (0.02) 0.267 (0.237) 0.822* (0.303) 0.04 (0.025) 4.75 0.84 39
*p < 0.05. In 1995e1999 wave data for IS, ME, MT is missing.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
99
Table 4 Pseudo-panel data analysis - Determinants of Control of Corruption.
Ln GDP per capita Gini Coefficient Communist history Communist history*Gini Coeff. Share of Protestants Democracy Intercept sigma_u sigma_e Rho N of observations N of groups F (Wald chi) Prob > F corr(u_i, Xb)
Model 2.1
Model 2.2
0.25 (0.05)* 0.003 (0.03) 0.93 (0.15)*
0.26 (0.05)* 0.01 (0.004)* 1.47 (0.25)* 0.07 (0.006)* 1.06 (0.207)* 0.04 (0.007)* 1.379 0.4 0.173 0.795 419 39 385.12 0.000 0 (assumed)
1.12 (0.204)* 0.04 (0.007)* 1.566 0.336 0.174 0.788 419 39 384.2 0.000 0 (assumed)
*p < 0.05.
As Model 2.1. in Table 4 shows, the results confirm almost perfectly the findings of the above-mentioned authors analysing the relations between corruption and different variables (for example, that higher GDP per capita, share of Protestants, and level of democracy would increase Control of Corruption and Communist history would decrease CC) with one exception, namely that the Gini Coefficient is not a significant predictor of Control of Corruption. In order to explore what is behind these results, I included an interaction between the Gini Coefficient and being a post-communist country; the results are presented under Model 2.2. Theories suggest that more equality should be connected to less corruption. However, as we can see in the case of post-communist countries, the relationship is reverse, that is, there is more corruption in countries with more equality. As Uslaner writes: “The connection between inequality and the quality of government is not necessarily so simple: As the former Communist nations of Central and Eastern Europe show, you can have plenty of corruption without economic inequality” (Uslaner, 2009). These results are very surprising; in the case of post-communist countries, inequality does not seem to influence control of corruption as it apparently does in countries that have not experienced communist rule. One explanation could be that there was a development of income inequality in time while corruption remained static. This could, in pooled data, create a negative relationship between income inequality and Control of Corruption. A similar explanation is suggested by Uslaner (2009), who argues that even though post-communist countries have low income inequality and high levels of corruption e which undermines the theory of income inequality increasing corruption e this state is only temporary. According to him, after transition to democracy, income inequality should boost and catch up with the rest of Europe. Taking into account Uslaner's claim that even though inequality is low in post-communist countries, the Gini Coefficient is quickly raising, which influences corruption, we should be able to see a shift in the Gini Coefficient and analogical shifts in Control of Corruption. And when first looking at Fig. 5 it seems that Uslaner (2009) is correct: in fact, the average Gini Coefficient in post-communist countries rose very sharply and caught up with the Gini Coefficient in European countries with no communist history in the early 90s. On average, the difference in the Gini Coefficient between post-communist countries and
Fig. 5. Development of the Gini coefficient in Europe. Source: UNU-WIDER, Eurostat, World Bank.
100
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
Fig. 6. Control of Corruption, development in time. Source: World Bank, Control of Corruption. Control of Corruption shows how the countries are successful in controlling corruption, the indicator goes from 3 to 3, while 3 indicates that country is successful in controlling the level of corruption.
the rest of Europe is not big; in fact, the Gini Coefficient in post-communist countries is higher than in the rest of Europe, except for the financial crisis years. However, as discussed before, it seems that each post-communist country has a different path concerning income inequality: in some countries, income inequality is rising, while it remains constant or it is even decreasing in others. Therefore, if we were to believe Uslaner, we would expect corruption to increase or decrease according to the Gini Coefficient. As this theory expects Control of Corruption not to change in time, it is essential to check the development of Control of Corruption in time in post-communist countries as opposed to countries with no history of communist rule. Fig. 6 shows that Control of Corruption has not changed very much over the years e there is still a very big gap between countries that have experienced communist rule and the rest of European countries. This gap has been closing very slowly over the years, yet the improvement of post-communist countries as a group is very small. It is worth noting that Control of Corruption in European countries without a history of communism has been decreasing very slowly over the years. Of course, this graph shows only the development of Control of Corruption in countries as groups, and there is very likely high variability across countries (see Fig. 10). If the hypothesis that inequality influences corruption (or vice versa) is correct, we should be able to see a shift in corruption in those countries where the Gini Coefficient changed. For this reason, it is essential to look at the changes in Gini Coefficient and corruption over years within the groups of countries. The following graphs (Figs. 7 and 8) show the correlations between change of Gini Coefficient (2010e2014 vs. 1995e2000) and change of Control of Corruption (2010e2014 vs. 1995e2000). Fig. 7 shows the relationships between Control of Corruption and Gini Coefficient in European countries without the legacy of communism: there is negative correlation, meaning that lower income inequality is associated with higher control of corruption. The relationship is not very strong, with a correlation at the level of 0.1964.
Fig. 7. Countries without communist legacy e Control of Corruption vs. Gini Coefficient, differences between 2010 and 2014 vs. 1995e2000. Source: WB, UNU-WIDER, Eurostat.
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
101
Fig. 8. Post-communist countries e Control of Corruption vs. Gini Coefficient, differences between the years 2010e2014 vs. 1995e2000. Source: WB, UNU-WIDER, Eurostat.
Fig. 8 on the other hand, shows the relationships between the Gini Coefficient and Control of Corruption in postcommunist countries; in this case, the correlation is much stronger (0.4279) and moreover, positive e it has the opposite direction than in the case of countries without the history of communism. It seems that the bigger the change toward a more unequal society, the bigger the change also toward a less corrupt society. These results go against the theories explaining corruption. Moreover, this relationship remains very strong even when deleting from the analysis the countries, which are the strongest drivers of this relationship, such as Moldova or Macedonia. Furthermore, when we consider not the same time
Fig. 9. Development of Gini coefficient in Europe. Source: UNU-WIDER, Eurostat, World Bank.
102
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
Fig. 10. Development of Control of corruption in Europe. Source: World Bank, Control of Corruption. Control of Corruption shows how the countries are successful in controlling corruption, the indicator goes from 3 to 3, while 3 indicates that country is successful in controlling the level of corruption.
frame as in the case of Control of Corruption but the preceding wave (1990e1995), the relationship weakens (0.3435) but remains positive. In the light of these findings, it would be plausible to adjust the equation explaining corruption and to look at the effect of the Gini Coefficient in the first wave and whether it influences the current level of Control of Corruption. As the Gini Coefficient would precede Control of Corruption, we can get a hint of causality. However, of course, we cannot infer a causal effect with certainty as income inequality could in theory influence corruption during a longer time span than twenty years. The updated equation follows:
Control of Corruptionit ¼ ai þ b Democracyit þ g GDP per capitait þ d Share of Protestantsit ε Giniit3 z communist historyit þ uit Table 5 shows the results of the regression analysis; there is no need for pseudo-panel data analysis as we are using the effect of Gini in the first wave (1995e2000) on Control of Corruption in the latest wave (2010e2014). Unsurprisingly, given the previous analysis, the two models show the same effects of income inequality as before. Model 3.1 shows that income inequality in the past, but also being a post-communist country, does not have a significant effect on Control of Corruption. However, as Model 3.2 shows, the results change after including the interaction between income inequality and being a post-communist country. In fact, the Gini Coefficient in the interaction is positive and significant, showing that in post-communist countries, the level of income inequality in the 90s influenced the control of corruption today in a way that high income inequality predicted high control of corruption. This effect is reversed for countries without a history of communism. There might be more explanations to these effects; one of them is supported by Uslaner (2009), who claims that even though post-communist countries have low income inequality while having high corruption due to the rule of communism, this relationship will reverse quickly after post-communist countries catch up in income inequality with the rest of European
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
103
Table 5 Regression analysis e Determinants of corruption, Gini t1.
Ln GDP per capita Gini Coefficient, t3 Post-communist country Communist history*Gini Coeff., t3 Share of Protestants Democracy Intercept Adjusted R2 Number of cases
3.1.
3.2.
0.58* (0.147) 0.013 (0.02) 0.284 (0.273)
0.613* (0.14) 0.07* (0.034) 2.71* (1.22) 0.79* (0.039) 0.938* (0.314) 0.037 (0.024) 3.37 0.86 37
1.115* (0.317) 0.03 (0.03) 4.79 0.84 37
*p < 0.05. IS and MT are missing in this analysis due to missing data.
countries. However, our analysis shows that this claim is not supported by the development of corruption in Europe. It is true that post-communist countries as a group quickly caught up with income inequality levels of the European countries that did not experience communism, but analysis of changes within the group of post-communist countries actually shows the contrary effect. Post-communist countries that have had the highest change in income inequality towards more unequal societies are today the least corrupt and vice versa. In addition to Uslaner's explanation, this effect can be also accounted for by socio-economic changes in post-communist countries. It seems that the highest increase in income inequality might be connected to the differences in countries' socioeconomic transformation, such as the expansion of private sector, retrenchment of redistributive state or promoting foreign investment in their privatization (Bandelj and Mahutga, 2010). It seems that countries that embraced market exchange and capitalism have the highest increase in income inequality and might have been also the most successful in transforming their other institutions connected to good governance and thus decreasing the level of corruption in their country. However, as the level of corruption in a country changes only very slowly in time, more time is needed to confirm this hypothesis. 7. Conclusion This article shows that theories, which point out to the strong correlation between corruption and the Gini Coefficient are not entirely correct. This correlation holds for European countries, which have not experienced communist rule, but it is not true for post-communist countries. This article takes historical data from 1990 to 2014 and observes the developments of the Gini Coefficient and corruption perceptions. The analysis shows that analysis of data from post-communist countries does not support the theory that high income inequality increases corruption; in the case of post-communist countries, unlike in the rest of Europe, high income inequality actually decreases corruption. This is supported by three types of analysis: firstly, by analysing the relationships between Control of Corruption and the Gini Coefficient by time intervals (waves) since 1995, then by multilevel analysis, and finally by analysing the change in these two indicators from 1995 to 2014. However, as the time period we currently have is only 20 years long, we will have to wait for a longer time series to assess whether the results presented by this article hold. Nevertheless, due to the unexpected results of this analysis, it is important to treat post-communist countries in models including corruption and income inequality with caution, as this group behaves significantly differently than the rest of European countries and also differently from the results on a global level. References ARDA. World religion dataset: national religion dataset. Available at: http://www.thearda.com/Archive/Files/Descriptions/WRDNATL.asp. Bandelj, N., Mahutga, M.C., 2010. How socio-economic change shapes income inequality in post-socialist Europe. Soc. Forces 88 (5), 2133e2161. Blake, C.H., Martin, C.G., 2006. The dynamics of political corruption: Re-examining the influence of democracy. Democratization 13 (1), 1e14. Brown, D.S., Touchton, M., Whitford, A., 2011. Political polarization as a constraint on corruption: a cross-national comparison. World Dev. 39 (9), 1516e1529. Centre for Systemic Peace ed., Polity IV Annual Time-Series, 1800-2015, Available at: http://www.systemicpeace.org/inscrdata.html.n.d. Chabova, K., 2017. Measuring corruption in Europe: public opinion surveys and composite indices. Qual. Quantity 51 (4), 1877e1900. Charron, N., Lapuente, V., Rothstein, B., 2013. Quality of Government and Corruption from a European Perspective: A Comparative Study of Good Government in EU Regions. Edward Elgar Publishing, Northampton. Drury, C.A., Krieckhaus, J., Lusztig, M., 2006. Corruption, democracy, and economic growth. Int. Political Sci. Rev. 27 (2), 121e136. Fazekas, M., Toth, I.J., King, L.P., 2013. Anatomy of Grand Corruption: A Composite Corruption Risk Index Based on Objective Data (No. CRC-Wp/2013:02). Corruption Research Center Budapest, Budapest, pp. 1e46. Gupta, S., Davoodi, H., Alonso-Terme, R., 2002. Does corruption affect income inequality and poverty? Econ. Govern. 3, 23e45. Helliwell, 2006. Well-being, social capital and public policy: what's new? Econ. J. 116, C34eC45. Husted, B.W., 1999. Wealth, culture, and corruption. J. Int. Bus. Stud. 30 (2), 339e359. Johnson, M., 2005. Syndromes of Corruption: Wealth, Power, and Democracy. Cambridge University Press, Cambridge. Kapoor, M., Ravi, S., 2012. Determinants of corruption: government effectiveness vs. cultural norms. B. E. J. Econ. Anal. Policy 12 (1), 1e24. Karklins, R., 2005. The System Made Me Do it: Corruption in Post-communist Societies. M.E. Sharpe, Armonk, NY ; London. n, P., 1999. Governance Matters. WB. Kaufmann, D., Kraay, A., Zoido-Lobato Kornai, J., Rose-Ackerman, S. (Eds.), 2004. Building a Trustworthy State in Post-socialist Transition. Palgrave Macmillan, New York. La Porta, R., et al., 1999. The quality of government. J. Law Econ. Organ. 15 (1), 222e279.
104
/ Communist and Post-Communist Studies 52 (2019) 93e104 K. Basna
Lambsdorff, J.G., 2003. How corruption affects productivity. Kyklos 56 (4), 457e474. Li, H., Xu, L.C., Zou, H.-F., 2000. Corruption, income distribution, and growth. Econ. Pol. 12 (2), 155e182. Mauro, P., 1995. Corruption and growth. Q. J. Econ. 110 (3), 681e712. Pellegata, A., 2012. Constraining political corruption: an empirical analysis of the impact of democracy. Democratization 20 (7), 1195e1218. Rose, R., 2001. A diverging Europe. J. Democr. 12 (1), 93e106. Rose-Ackerman, S., 1999. Corruption and Government: Causes, Consequences, and Reform. Cambridge University Press, Cambridge. Rose-Ackerman, S., Soreide, T., 2006. International Handbook on the Economics of Corruption. Edward Elgar Publishing, Cheltenham. Scott, J.C., 1972. Comparative Political Corruption. Prentice-Hall, Englewood Cliffs, N.J. Shleifer, A., 1997. Government in Transition. European Economic Review, pp. 355e410. Transparency International, 2012. Corruption perceptions index 2012: technical methodology note. Available at: https://www.transparency.org/files/ content/pressrelease/2012_CPIUpdatedMethodology_EMBARGO_EN.pdf. Treisman, D., 2000. The causes of corruption: a cross-national study. J. Public Econ. 76, 399e457. Treisman, D., 2007. What have we learned about the causes of corruption from ten years of cross-national empirical research? Annu. Rev. Pol. Sci. 10 (1), 211e244. United Nations ed., Composition of Macro Geographical (Continental) Regions, Geographical Sub-regions, and Selected Economic and Other Groupings, Available at: http://millenniumindicators.un.org/unsd/methods/m49/m49regin.htm#europe.n.d. Uslaner, E.M., 2008. Corruption, Inequality, and the Rule of Law. Cambridge University Press, Cambridge. Uslaner, E.M., 2009. Corruption. In: Svendsen, G.T., Svendsen, G.L.H. (Eds.), The Handbook of Social Capital: the Troika of Sociology, Political Science and Economics. Edward Elgar Publishing, Northampton, pp. 127e142. Wei, S.J., Wu, Y., 2001. Negative Alchemy? Corruption, Composition of Capital Flows, and Currency Crises, NBER Working Paper Series 8187. World Bank. n.d. Control of corruption. Available at: http://info.worldbank.org/governance/wgi/pdf/cc.pdf. World Bank, 1997a. Helping Countries Combat Corruption. The World Bank. World Bank, 1997b. World Development Report: the State in a Changing World. Oxford University Press, New York. You, J.-S., Khagram, S., 2005. A comparative study of inequality and corruption. Am. Sociol. Rev. 70 (1), 136e157. Zakaria, P., 2013. Is corruption an enemy of civil society? The case of Central and Eastern Europe. Int. Political Sci. Rev. 34 (4), 351e371.