Quantifying corruption at a subnational level

25 2, XVIII, 2015 Economics DOI: 10.15240/tul/001/2015-2-003 Introduction Corruption and its potential reduction is a constant topic not only of economic or social science research, but also an issue that plagues governments and citizens alike. This phenomenon is more or less immanent in every social system, regardless of the size and sophistication of the country or culture of the nation. Despite the fact that corruption is not a new phenomenon and a number of foreign and domestic authors have been dealing with this subject for several years, there are still many questions that remain unanswered. The very defi nition of the term corruption is not yet clear, and different authors defi ne corruption with greater or smaller differences. For more detail, see e.g. [6], [9], [29], [20]. Even the question of whether and how corruption can affect the economic level of a country has never been answered by any literature without controversy. One may thus encounter the view that corruption is “sand in the wheels” of the economy, which impedes economic transactions, as it reduces the security of property rights and contributes to ineffi cient allocation of resources [23], [28], [24], [17], [13]. On the other hand, there are authors who believe that corruption is precisely what “greases the wheels” of the economy, because it allows individuals to avoid administrative and bureaucratic delays [12], [14], [15], [1], [16]. It can be said, however, that with the existence of adequate legislation, the argument about corruption as “greasing the wheels of the economy” is totally unacceptable. The issue of quantifying the degree of corruption also raises fi erce debate. Considering the fact that bribery and other forms of corruption are illegal in most countries, the people involved make every effort to carefully conceal their actions and revealing corruption is often almost impossible. Even so, there are currently a number of exact procedures that attempt to quantify the level of corruption in a country. Among the best known current indicators of corruption is one example, the CPI (Corruption Perception Index), published annually by Transparency International and the Control of Corruption of the World Bank [27]. A common feature of all currently existing indices of corruption, however, is the fact that all without exception quantify the level of corruption in a country and are therefore not applicable for quantifi cation of corruption at a sub-national level. The authors of this paper argue that the socio-economic development in a country is not homogeneous, and that it can be assumed that a difference exists in the extent of corruption in different regions within the same country. Under this assumption, more corrupt sub-national regions are detrimental to the national evaluation of corruption in a country as a whole. The fact that the distribution of corruption in a country is not homogenous was confi rmed by authors [2] and [5] in their studies of Italian regions. The level of corruption in the sub national breakdown as reported by these authors was very diverse and its analysis can help explain the differences in the different economic performance of the regions. It can be noted, however, that the study of the Italian authors is unique and fi nding another study on the quantifi cation of regional levels of corruption, or its impact on the region, is virtually impossible. From this it is clear that the issue of quantifying corruption and its consequences at the regional level is a topic that deserves more attention. There are several reasons to consider these issues. Perhaps the strongest is that if corruption is indeed one of the variables that are degrading the performance of economies, the elimination of corruption in certain regions may just be the key to removing regional economic disparities and thereby increasing the economic performance of the country. Analysing regional corruption may also lead to the creation of direct regional anti-corruption QUANTIFYING CORRUPTION AT A SUBNATIONAL LEVEL Veronika Linhartová, Jolana Volejníková EM_2_2015.indd 25EM_2_2015.indd 25 3.6.2015 13:08:523.6.2015 13:08:52 26 2015, XVIII, 2 Ekonomie initiatives that can bring about reductions in the national level of corruption. In general terms, a sub-national resolution in terms of the degree of corruption could bring a new dimension to traditional theories of regional disparities. The main objective of this article is formulated in the context of the above considerations. It contains a design for a method of quantifying the extent of corruption at the level of regional cohesion. The proposed method is then verifi ed and applied to individual regions of the Member States and candidate States of the European Union. Using the proposed method it is possible to evaluate the current state of corruption in the evaluated regions, to mutually compare the regions, to determine the degree of deviation from the “surface” level of national corruption and simultaneously determine the degree of variability in the extent of corruption within the country. The text of this article presents the proposed method of quantifying the regional level of corruption verifi ed using Kendall’s coeffi cient of concordance for further use. Verifi cation of the proposed method is carried out at a national and regional level. Methods at the national level are verifi ed by comparing the evaluation of the newly proposed method to the evaluation of existing corruption indices. This will determine the level of agreement between the already established indices and the newly proposed method. The methods are verifi ed at the regional level using police statistics on recorded corruption offences. After the method is verifi ed, the level of corruption in the various regions is calculated. Special attention is paid to quantifying the extent of corruption in the regions of the Czech Republic. Calculating the level of corruption in the Czech regions will identify those regions which are more affected by corruption than others and would thus worsen the national evaluation of the Czech Republic within the standard published indices of corruption. 1. Proposal for a Method of Quantifying the Extent of Corruption at a Sub-National Level Due to the absence of any method for determining corruption in a more or less affected sub-national region, the next section will present a method for quantifying corruption at a sub-national level. The design of this method is based on the construction of the European Quality of Government Index developed by the European Commission together with The Quality of Government Institute. Corruption is understood here in accordance with the defi nition of Nye, who describes corruption as “behaviour that deviates from the formal duties of a public role because of private-regarding wealth or status gains“ [20]. This defi nition focuses on the abuse of public power, and somewhat ignores corruption in the private sector, which of course also exists. Most existing studies, however, have focused on corruption in the public sector, as the consequences of misuse of public power impact the broad mass of taxpayers and the country as such. 1.1 The European Quality of Government Index The European Quality of Government Index (EQI) was created to quantify the quality of public administration at a regional level. The index so far been developed twice; in 2010 and 2013. 27 EU Member States were included in the EQI in 2010. In 2013, 28 EU Member States are included as well as the Candidate States Turkey and Serbia; in total 30 countries. The European Commission plans to construct EQI regularly every three years. In addition to the national evaluation of the quality of governance, the resulting EQI also takes note of the evaluation of regional administration using regional data which the European Commission has drawn up for the needs of constructing the EQI. The EQI thus consists of two main parts: The fi rst part of the EQI takes into account the national government level, which is represented by the Worldwide Governance Indicators (WGI) of the World Bank. Of the six pillars of the quality of governance, the European Commission chose four for the construction of the EQI: Voice and Accountability (GM1), Government Effectiveness (GM3), Rule of Law (GM5) Control of Corruption (GM6) [4], [26], [11]. The second part of the EQI, which takes into account the regional level of governance, was compiled by the European Commission on the basis of a unique regional survey, conducted for the sole purpose of creating a Regional indicator of government quality, which would take into account regional aspects in the fi nal construction of the EQI. This unique research registered in the fi rst construction of the EQI was executed in 172 NUTS II regions in 18 countries of the EuEM_2_2015.indd 26EM_2_2015.indd 26 3.6.2015 13:08:533.6.2015 13:08:53 27 2, XVIII, 2015 Economics ropean Union in 2010 (from the remaining 9 countries of the European Union only data at the national level was included). The research includes altogether 181 regional units. Data was obtained by means of surveying more than 33,000 inhabitants. The all-European regional research was conducted from 15th December 2009 to 1st February 2010 by means of telephone interviews with respondents older than 18 years and in the local language. In the second construction of EQI, it was executed in 206 NUTS regions in 24 countries of the European Union in 2013 (from the remaining 7 countries of the European Union only data at the national level was included). The research includes altogether 213 regional units. Data was obtained by means of research of more than 85,000 inhabitants. The resulting regional quality of administration indicator refl ects the actual experience of respondents with the use of individual public services, thus the quality of governance in the region is evaluated as it is perceived by its inhabitants; i.e., the recipients of public administration. The Regional indicator of government quality is composed of 16 separate indicators relating to the quality of administration in a particular region. These 16 indicators were developed based on 16 questions (The list of questions is available at http://www.qog.pol.gu.se/ data/datadownloads/qogeuregionaldata/.) developed in accordance with the pillars arising from the methodology of the WGI: Voice and Accountability, Government Effectiveness, Rule of Law and Control of Corruption. In order to capture the most important sub-national differences, questions were focused on three public services that are often funded or administered at sub-national levels. Each of the four pillars mentioned thus involves issues relating to education, health care and law enforcement in the region. With a focus on these three services, respondents were asked to assess these public services with regard to the three fundamental concepts of quality administration - quality, impartiality and corruption. These three concepts are the pillars of the resulting regional indicator of quality government. Data is aggregated three times using a simple average. First is the creation of the average values of responses to the questions. This will create 16 indicators for each region. Then these 16 values are aggregated into three defi ned pillars - quality, impartiality and corruption. Finally, these three pillars are aggregated into a single numerical Regional quality of administration indicator. A simple diagram of the formation of the Regional indicator of government quality is shown in Figure 1. Fig. 1: Approach to creating a Regional Indicator of Government Quality Source: own work according to [3] EM_2_2015.indd 27EM_2_2015.indd 27 3.6.2015 13:08:533.6.2015 13:08:53 28 2015, XVIII, 2 Ekonomie Thus in its fi nal form, the resulting EU Quality of Government Index enriches the national evaluation of quality of administration created by the World Bank (WGI) on a regional scale (Regional indicator of government quality). The fi nal form of the construction of the EQI is as follows: EQIregionXincountryY = WGIcountryY + + (RqogregionXincountryY – CRqogcountryY ), (1) where EQIregionXincountryY is the fi nal European Quality of Government Index in the region of a given country, WGIcountryY is the national average of the above four Worldwide Governance Indicators for each country, RqogregionXincountryY is the score from a regional survey; thus the Regional indicator of government quality, CRqogcountryY is the regional survey of all regions in the country weighted by the proportion of the population of each region to the national population of the country. The EQI has so far been calculated twice; once in 2010 and in 2013. Member States of the EU-28, Turkey and Serbia, were included in the calculation. 1.2 Proposal for a Regional Index of Corruption It is apparent that the resulting EQI, as it was compiled by the European Commission together with The Quality of Government Institute, provides the opportunity to pursue a quantifi cation of corruption at a sub-national level, which had not previously been practically possible. The primary modifi cations of the already created EQI can create a modifi ed index, which, of all the components of quality government, takes into account only corruption; it therefore takes into account only the indicator Control of Corruption in the national evaluation and the indicator Pillars of Corruption in the regional evaluation. Based on this modifi ed methodology of the EQI composition, the modifi ed method of calculating EQI can then be applied only for the purpose of quantifying corruption in the cohesion regions. The resulting Regional Index of Corruption (RIC) is then calculated based on the formula: RICregionXincountryY = CCcountryY + + (PCqogregionXincountryY – CPCqogcountryY ), (2) where RICregionXincountryY is the resulting Regional Index of Corruption for each region of a given country, CCcountryY is the national indicator value of Control Of Corruption (GM6) from the Worldwide Governance Indicators, PCqogregionXincountryY is the score from a regional survey focused on corruption, thus Pillar of Corruption, CPCqogcountryY is the value for the Pillar of Corruption from the regional survey of all regions in a country weighted by the proportion of the population in each region on the national population of the country. Composite indicators often evoke a number of questions relating to their composition and weighting of the individual indicators entering into a composite indicator. There were created tens of aggregated indicators at the time when composite indicators reached a rapid expansion. Most of them, unfortunately, were not built on correct statistical basis [25]. Because of these errors composite indicators failed to meet expectations, which were inserted into them, sparking concerns among their users and negatively affected trust in composite indicators in general. The credibility of these indicators is mainly related to the accuracy of data, based on which they are constructed and the methodology by which they are constructed. Number of composite indicators is constructed by reputable international institution. Such indicators can get known and respected easier and earlier. However, although composite indicators are designed very carefully and statistical requirements have been met, their acceptation is always dependent on bargaining and how they are accepted by experts and public. Acceptation of aggregated indicators mostly depends on how they meet the original goal, i.e. whether measure what they should and the subsequent acceptance of their users. Gaining legitimacy and trust of users is a gradual process. In the case of constructing indicator EQI was by the European Commission used an equal weighting of entering variables. With respect to author´s proposed RIC is a modifi cation of EQI, authors of the article do not consider important further weighting of variables entering into RIC. 2. Applying the Proposed Regional Index of Corruption The Regional Index of Corruption (hereinafter RIC) is applied and tested fi rst at the national EM_2_2015.indd 28EM_2_2015.indd 28 3.6.2015 13:08:533.6.2015 13:08:53 29 2, XVIII, 2015 Economics level, then at the level of the cohesion regions. From the resulting values, the individual regions can be mutually compared and regions can be identifi ed which are more or less affected by corruption. Table 1 shows the resulting ranking of countries in the newly created RIC for the years 2010 and 2013. Countries in the segmented EQI 2013 are included; thus there are a total of 30 countries. The higher the value of the RIC, the better is the evaluation of the country’s RIC. In the evaluation of the RIC between 2010 and 2013, it was found that the new Member States and candidate States of the European Union are at the very bottom of the list of countries evaluated. Conversely, the Nordic countries were evaluated as the least affected by corruption. NUTS I RIC 2010 Ranking NUTS I RIC 2013 Ranking DK 1.811919 1 DK 1.841393 1 FI 1.740486 2 SE 1.559288 2 SE 1.516722 3 FI 1.555572 3 NL 1.438868 4 LU 1.493145 4 LU 1.261475 5 NL 1.479409 5 AT 1.142543 6 DE 0.932501 6 IE 0.948732 7 UK 0.779821 7 DE 0.917613 8 BE 0.749709 8 UK 0.830591 9 IE 0.726454 9 FR 0.488344 10 FR 0.703595 10 BE 0.415918 11 AT 0.609217 11 CY 0.322032 12 PT 0.168304 12 ES 0.157165 13 ES 0.131936 13 MT 0.083101 14 EE -0.0212 14 PT 0.029269 15 SI -0.05617 15 SI -0.07815 16 CY -0.07266 16 EE -0.12856 17 MT -0.1372 17 LV -0.67118 18 PL -0.56423 18 LT -0.70428 19 HU -0.76712 19 HU -0.71697 20 CZ -0.7947 20 PL -0.76271 21 SK -0.85981 21 SK -0.81496 22 LT -0.86415 22 CZ -0.85541 23 LV -0.92744 23 IT -0.87991 24 IT -1.05754 24 GR -1.06275 25 TR -1.08985 25 TR -1.08395 26 HR -1.14626 26 HR -1.23592 27 GR -1.38318 27 RO -1.37328 28 RO -1.39001 28 RS -1.55004 29 BG -1.43259 29 BG -1.55089 30 RS -1.46287 30 Source: Author’s own work Tab. 1: Regional Index of Corruption for 2010 and 2013 EM_2_2015.indd 29EM_2_2015.indd 29 3.6.2015 13:08:533.6.2015 13:08:53 30 2015, XVIII, 2 Ekonomie By using Statistica 12, graphic models were created of the variability of RIC values in individual countries for the years 2010 and 2013. The box plots use the method of min-max comparison and show the range of RIC values marking the best and the worst of the regions evaluated in the country. On the x-axis are plotted the countries evaluated; on the y-axis are the resulting values of the RIC in a given year. The range of values is complemented by the fi nal value of the RIC of the country, which is represented by an asterisk. Figure 2 shows the range of RIC values for 2010 in the thirty countries evaluated. Defi nitely the greatest variability in the assessment of corruption is to be found in the Italian regions. Italian respondents answered questions regarding the impact of corruption on their area with great differences, and perceived corruption very differently depending on which region they live. The most corrupt Italian region, based on the results of the RIC from 2010, is the Campania region (ITF3), while the best ratings were achieved in the Umbria region (ITE2). A high variability was also observed in Romania, France and the Netherlands. Rating corruption at the national level can be particularly misleading for these countries. In the Czech Republic, a middle variability of RIC values was recorded. The top rated region is Jihozápad (Southwest) (CZ03) with a value of -0.9346 and the worst rating is the capital city of Prague (CZ01) with a value of -1.9878. In the evaluation of RIC in 2010, the NUTS II regions which placed best were the Dutch region of Groningen (NL11) with a value of 2.8867. The best ratings in 2010 were achieved generally by Dutch, Danish, Finnish and Swedish regions. In contrast, at the other end of the ranking were Romanian, Italian and Bulgarian regions. Defi nitely the worst ranking among the NUTS II regions was the Romanian region of Bucharest (RO32) with a value of -2.7491. Fig. 2: Box Graph of Values for the Regional Index of Corruption for 2010 Source: Author’s own work EM_2_2015.indd 30EM_2_2015.indd 30 3.6.2015 13:08:533.6.2015 13:08:53 31 2, XVIII, 2015 Economics Figure 3 shows the range of values of RIC for 2013. In 2013, the region with the lowest level of corruption was the Finnish region of Aland (FI20) with a value of 2.3932. On the other hand, the most corrupt region of the European Union was the Bulgarian region of Yugozapaden (BG41) with a value of -2.5237. A high variability of data in 2013 was found again in Italy, as well as Bulgaria, Turkey and Romania. In these countries, the inhabitants of regions had different opinions on the impact of corruption in their area and the corruption assessment may not refl ect the current situation in some regions. In contrast, in Danish, Swedish, Irish and Croatian regions only very small deviations, were detected in the values of RIC of 2013 and evaluation of the national level of corruption relevantly refl ects the evaluation of the regions. Within the Czech Republic, in 2013 the best region evaluated in terms of corruption was Jihozápad (Southwest) (CZ03) with a value of -0.5694 and the most corrupt region was Severozápad (Northwest) (CZ04) with a value of -1.2304. The resulting RIC values demonstrate that some European Union countries show a very high degree of variability in the regional level of corruption. This confi rms the assumption that existing indices evaluating the national level of corruption can ultimately overestimate the regions more affected by corruption and underestimate the less corrupt. Defi nitely the greatest variability of the data evaluated in both years was demonstrated in the Italian regions. In Italy, as one of the smaller countries, several studies on corruption have been conducted in various Italian regions. The authors [2] and [5] in their studies agree that the variability of the degree of corruption in the Italian regions is very high and in this country there are regions with very high levels of corruption, but also regions with much lower levels of corruption. By applying the proposed Regional Corruption Index (RIC), not only were the conclusions of the authors confi rmed regarding the Italian regions, but this conclusion is demonstrated in the majority of countries surveyed. Fig. 3: Box Graph of Values for the Regional Index of Corruption for 2013 Source: Author’s own work EM_2_2015.indd 31EM_2_2015.indd 31 3.6.2015 13:08:533.6.2015 13:08:53 32 2015, XVIII, 2 Ekonomie 2.1 Verifying the Proposed Method The proposed method of quantifying the degree of corruption at the regional level is subsequently verifi ed at national and regional level. Kendall’s coeffi cient of concordance can be used for mathematical verifi cation of the conformity of the assessment methods for the proposed RIC and existing indexes. This is a non-parametric method of mathematical statistics which is primarily used to assess the conformity of individual evaluators. The value of the coeffi cient varies between 0 (no agreement) and 1 (complete agreement) [10]. 2.1.1 Verifying Method at the National Level In order to compare evaluations at the national level, two presently existing indices are selected which measure the degree of corruption in the country. These are indices that focus exclusively on quantifying the national level of corruption. The selected corruption indicators are the Corruption Perceptions Index (CPI) of Transparency International and the Control of Corruption (CC) of the World Bank [11], [27]. Given that data from the Regional government quality indicator, which was used for the construction of the RIC, has been collected among respondents since 2009 and the data of the World Bank to evaluate the situation at the national level was drawn upon in 2008, it is appropriate, in assessing conformity of the ratings, to take into account not only data for 2010. To compare the resulting values of RIC for 2010, a time range of existing indices were selected for the years 2008–2010, which take into account the entire time period during which the data was collected for the RIC. To verify the agreement of the assessment of RIC for 2013, the time range 2011 to 2013 was chosen. Table 2 presents the resulting calculation of Kendall’s coeffi cient of concordance ranking countries according to the RIC in 2010 and the CPI and CC from 2008 to 2010 and to the RIC 2013 and the CPI and the CPI and CC from 2011 to 2013 as evaluated by the program Statistica 12. Kendall’s coeffi cient of concordance assessing the order of the selected indices reaches around 98%. The RIC itself with each of the chosen indices for each year corellates in all cases at least at a level of 95%. High values of the coeffi cients of concordance in both years indicate that the proposed RIC ranks countries in terms of their corruption very similarly to the currently used indices of corruption. These conclusions of Kendall’s coeffi cient of concordance verify the possibility of using the RIC. Variable Kendall´s coeffi cient for RIC 2010 Kendall´s coeffi cient for RIC 2013 (no. of variables–30, no. of indices-7) (no. of variables–30, no. of indices-6) Avg. value r = 0.97501 Avg. value r = 0.98355 Average (ranking) Total (ranking) Average Deviation Average (ranking) Total (ranking) Average Deviation AT 6.85714 48.0000 6.85714 0.89974 10.33333 62.0000 10.16667 1.602082 BE 10.28571 72.0000 10.28571 0.48795 8.08333 48.5000 8.00000 0.632456 BG 28.71429 201.0000 28.57143 1.13389 28.58333 171.5000 28.50000 0.836660 CY 13.57143 95.0000 13.42857 1.39728 13.66667 82.0000 13.66667 1.861899 CZ 20.50000 143.5000 20.42857 1.51186 21.25000 127.5000 21.00000 1.095445 DE 7.35714 51.5000 7.28571 0.95119 6.00000 36.0000 6.00000 DK 1.14286 8.0000 1.00000 1.16667 7.0000 1.00000 EE 14.71429 103.0000 14.57143 2.22539 13.33333 80.0000 13.33333 1.032796 ES 13.85714 97.0000 13.85714 0.89974 13.66667 82.0000 13.66667 1.211060 FI 2.42857 17.0000 2.28571 0.75593 2.41667 14.5000 2.16667 0.983192 Source: Author’s own work Tab. 2: Kendall’s Coeffi cients of Concordance for Regional Index of Corruption 2010 and 2013 – Part 1 EM_2_2015.indd 32EM_2_2015.indd 32 3.6.2015 13:08:533.6.2015 13:08:53 33 2, XVIII, 2015 Economics 2.1.2 Verifying the Proposed Methods at the Regional Level At present, virtually the only possible way to verify the proposed method at the regional level is to compare RIC with statistics of corruption offences in the regions of the Czech Republic. According to offi cial statistics of the Ministry of the Interior and the of the Czech National Police, however, only recorded cases of corruption can be traced, whose number is based on the activity of the state bodies. The strategy of the government in the fi ght against corruption for the period 2013–2014 indicates that corruption in the Czech Republic has a high degree of latency and only a few cases have been uncovered [18]. According to the Government Programme for Combating Corruption of the Czech Republic, only one percent of corruption offences have been uncovered [19]. The actual number of these crimes that have occurred in recent years is likely to be much higher [21]. For the purposes of distinguishing the regions on the basis of corruption, without the need for a precise quantifi cation, this tool is usable. In order to verify the Regional Corruption Index, the following corruption offences are used, related to corruption in public administration, which is defi ned by the Criminal Variable Kendall´s coeffi cient for RIC 2010 Kendall´s coeffi cient for RIC 2013 (no. of variables–30, no. of indices-7) (no. of variables–30, no. of indices-6) Avg. value r = 0.97501 Avg. value r = 0.98355 Average (ranking) Total (ranking) Average Deviation Average (ranking) Total (ranking) Average Deviation FR 10.71429 75.0000 10.71429 0.48795 10.00000 60.0000 10.00000 0.632456 GR 26.07143 182.5000 25.85714 2.03540 28.33333 170.0000 28.33333 1.366260 HR 26.42857 185.0000 26.42857 0.78679 25.08333 150.5000 24.83333 1.329160 HU 19.28571 135.0000 19.28571 1.11269 19.50000 117.0000 19.50000 0.547723 IE 6.85714 48.0000 6.71429 0.75593 9.33333 56.0000 9.16667 0.408248 IT 23.85714 167.0000 23.85714 1.57359 25.41667 152.5000 25.33333 1.211060 LT 21.78571 152.5000 21.57143 2.99205 19.91667 119.5000 19.83333 1.940790 LU 5,00000 35.0000 5.00000 4.66667 28.0000 4.66667 0.516398 LV 21.78571 152.5000 21.57143 1.90238 22.41667 134.5000 22.16667 0.983192 MT 16.00000 112.0000 16.00000 1.41421 17.16667 103.0000 17.16667 1.329160 NL 3.92857 27.5000 3.85714 0.37796 4.33333 26.0000 4.33333 0.516398 PL 19.71429 138.000 19.57143 2.22539 17.33333 104.0000 17.33333 1.211060 PT 14.57143 102.0000 14.57143 1.81265 14.00000 84.0000 14.00000 1.673320 RO 28.14286 197.0000 28.00000 0.81649 27.41667 164.5000 27.33333 1.211060 RS 29.64286 207.5000 29.57143 0.53452 29.41667 176.5000 29.33333 0.816497 SE 2.50000 17.5000 2.42857 0.78680 2.41667 14.5000 2.33333 0.516398 SI 14.28571 100.0000 14.14286 2.11570 16.08333 96.5000 16.00000 0.894427 SK 21.64286 151.5000 21.42857 1.13389 23.33333 140.0000 23.16667 1.722401 TR 24.42857 171.0000 24.28571 1.25357 23.08333 138.5000 22.83333 1.722401 UK 8.92857 62.5000 8.85714 0.37796 7.25000 43.5000 7.16667 0.408248 Source: Author’s own work Tab. 2: Kendall’s Coeffi cients of Concordance for Regional Index of Corruption 2010 and 2013 – Part 2 EM_2_2015.indd 33EM_2_2015.indd 33 3.6.2015 13:08:543.6.2015 13:08:54