The Economics of Corruption: Measurement, Causes and Effects. Fabio Padovano

Transcription

1 The Economics of Corruption: Measurement, Causes and Effects Fabio Padovano 1

2 Quality of Institutions Institutions matter How do we measure their quality? Economists used perceptions-based indicators of corruption, property rights, and the rule of law in cross section growth research in the mid-1990s Paulo Mauro obtained national-level indicators of judicial integrity, corruption, and bureaucratic quality from commercial risk rating agencies in 1995 (Mauro, 1995); Arthur Goldsmith used the Corruption Perceptions Index developed by Transparency International a few years later (Goldsmith, 1995, 1999); Alberto Ades and Rafael DiTella introduced the corruption indicators found in the EMF Foundation s World Competitiveness Report at approximately the same time (Ades and DiTella 1997) 2

3 Corruption - 1 The 2014 Report of the UE Commission on corruption shows that The phenomenon is differently spread in the 28 member states It is concentrated especially in: urban development construction health system More at the local than at the central governemnt level The EU Commission estimates the costs of corruption in Europe as about 120 billion per year The mechanisms of internal controls, the efficiency of the authorities, the norms of the conflicts of interest and incompatibility of offices, as well as the statistics are heterogeneous 3

4 Corruption - 2 Economic theory distinguishes 3 types of corruption (Forti, 2003) : public-private; public-public; private-private We focus on a) and b) (i.e. political and administrative corruption) that have the highest social and economic impact and define corruption as in Rose-Ackerman (1975): crimes committed by public officials for personal gain According to the economists corruption requires the existence of three elements (Aidt, 2003): the discretionary power of a policy-maker or bureaucrat to allocate economic resources, define rules and enforce contracts An economic rent or utility associated to the discretionary power A disutility associated to the probability to be caught and punished 4

5 Measurement issues There are different methods to measure corruption and none of them is devoid of conceptual or statistical problems This because there are difficulties in the definition of corruption, lack of objective data, risk of under-reporting, measurement errors The choice of the measure mainly depends on the type of analysis one would like to perform Subjective indicators based on perceptions Judicial indicators based on Court data Objective measures based on direct observations and experiences 5

6 The subjective indicators: Corruption Perception Index (CPI) The Corruption Perceptions Index (CPI) produced by Transparency International since 1995 for more than 170 countries ranks countries in terms of the degree to which corruption is perceived to exist among public officials and politicians It ranges from 0 (more corrupt) to 10 (less corrupt) The confidence range indicates the reliability of the CPI scores and tells us that we can be 90% confident that the true score for each country lies within the calculated range 6

7 The CPI It expresses the views of business people, top bureaucrats, political analysts through surveys conducted by (on average about 13) research institutes such as Asian Development Bank, African Development Bank, Bertelsmann Foundation, Economist Intelligence Unit, Nations in Transit by Freedom House, IHS Global Insight, Institute for Management Development, Political and Economic Risk Consultancy, Country Policy and Institutional Assessment by the World Bank, Global Competitiveness Report by the World Economic Forum The CPI focuses on corruption in the public sector and defines corruption as the abuse of public office for private gain Generally cross-countries studies on corruption rely on the CPI because of its time (1995 to 2013) and spatial (more than 170 countries) dimensions 7

8 CPI scores in

9 Table 1: Comparison of CPI scores in selected OECD countries (source Transparency International) Denmark 9,3 9,5 9,5 9,3 9,3 9,1 Sweden 9, ,3 9,2 8,9 Norway 8, ,9 8,6 8,6 UK 8,44 8,7 8,7 7,7 7,6 7,6 Germany 8,27 7,3 7,7 7,9 7,9 7,8 USA 7,66 7,7 7,5 7,3 7,1 7,3 France 6,96 6,3 6,9 6,9 6,8 7,1 Belgium 6,84 7,1 7,6 7,3 7,1 7,5 Portugal 6,53 6,3 6,6 6,1 6 6,2 Greece 5,1 4,2 4,3 4,7 3,5 4 Spain 4,31 7,1 6,9 6,5 6,1 5,9 Italy 3,42 5,2 5,3 4,8 3,9 4,3 9

10 Table 2. CPI. Comparison between Italy and some OCSE countries (average ) Valori medi CPI Austria 7,9 Belgium 6,74 Denmark 9,54 Finland 9,48 France 6,9 Germany 7,92 Greece 4,46 Irland 7,74 Italy 4,63 Luxemburg 8,57 Netherlands 8,85 Portugal 6,35 Spain 6,34 Sweden 9,24 UK 8,38 Fonte: Transparency International. 10

11 CPI s scores in some countries In 2014 the CPI for Italy is 43, France 69 Italy is very distant from most of the European countries, closer to Greece and Turkey than to any western European country The Nordic European countries, the Netherlands, Germany, the UK, Canada and the US show scores higher than 8; The Mediterranean and Eastern European countries, many countries in Asia, SouthAmerica and Africa show scores lower than 6 11

12 Other subjective indicators - BPI Bribe Payers Index (BPI) It is an index produced by TI which measures the propensity to corruption of the firms which export abroad by sectors for a sample of 22 industrialized and emerging economies I ranges from 0 to 10 The index is computed through surveys addressed to about 3000 businessmen, leaders in international trade, managers of national and international banks, national and multinational corporations 12

13 Other subjective indicators - GCB The Global Corruption Barometer by TI-Gallup International since 2003 measures the corruption perception of citizens in different sectors such as politics, media, judicial system, private sector, public institutions, religious institutions, military. It ranges from 0 (no corruption) to 5 (high corruption) In 2013 Italian citizens perceive a worsening in the last two years and believe that the most corrupt institutions are the political parties (69%), the parliament (44%), the public administration (29%) and the health system (24%) 62% of the interwieved emphasizes the importance of the cultural factor, 14% positively evaluates the efficiency of the anticorruption measures of the governments and 5% declares to be involved in the payment of bribes 13

14 Other subjective indicators: BEEPS and WBES The Business Environment and Enterprise Performance Survey (BEEPS) produced by the WB since 1999 measures the quality of the governance from the perspective of the private sector in Eastern Europe and Central Asia through surveys addressed to firms The World Business Environment Survey (WBES) is a survey conducted by WB in 80 countries in interviewing about 10,000 firms: 70% of managers who work in Asia, in North Africa, in Middle East, 60% of managers who work in Latin America, and 50% of managers who work in Eastern Europe consider corruption as the major obstacle to business 14

15 The shortcomings of the perception-based indexes The meaning of corruption is subjective and can vary greatly from one country to the other one and year-toyear according to the dominant cultural approach Media may affect the score which may change rapidly as a consequence of a political scandal etc. The types of the corrupt activities could be substantially different in each country making comparative analyses difficult Risk of under-estimation or over-estimation by the interviewed They do not consider the regional variation of corruption in the countries 15

16 The regional perspective of the subjective indicators Recently the Goteborg Report by Charron, Lapuerte and Dijkstra European Commission, 2010) measured the quality of institutions, included corruption, at regional level in the 27 countries of the EU in 2008 Italy and Belgium records the highest variance on regional basis WB conducted regional surveys about perception and experience of corruption in several countries in Latin America and in Africa, addressed to citizens, public official, firms 16

17 Worldwide Governance Indicators WGI By the end of the decade Kaufmann et al. at the World Bank had come to believe that many of these indicators serve as imperfect proxies for one of a smaller number of more fundamental concepts of governance, including the rule of law, government effectiveness, and graft (Kaufmann et al. 1999, pp. 1-2) They developed a series of aggregate governance indicators aimed to capture the fundamental concepts in a more transparent and accurate fashion and aggregated data from a wide array of sources including nongovernmental organizations, and commercial risk rating agencies Six indicators: voice and accountability, political instability and violence, government effectiveness, regulatory burden, rule of law, and control of corruption as capture of the State by pressure groups and private interests They are available for more than 150 countries and cover a ten year period and include standard errors as well as individual country scores 17

18 Are the WGI convincing? The WGI are among the most carefully constructed and widely used (Arndt and Oman 2006, p. 49) indicators available for the perceptionsbased measurement of governance but there are some criticisms (Arndt and Oman 2006; Thomas 2007; Knack 2007; Bhagwati 2007, Kurtz and Schrank 2007) KKM emphasizes that the various aspects of governance they aim to measure tend to be quite highly correlated across countries (Kaufmann et al. 2007a, p. 555) Their measures are highly correlated with each other and with the presumed covariates of good governance (e.g., GDP per capita, school attainment). They, too, add explanatory power to multivariate models While they explicitly claim to derive the definitions of the six aspects of governance covered by the WGI from existing definitions or understandings of the concepts (Kaufmann et al. 2007, p. 24), they occasionally admit that their indicators actually drive their definitions and that the meaning is therefore in the measure 18

19 Objective and judicial measures Di Tella e Schargrodsky (2003) the level of prices of some standard inputs used in public hospitals in Buenos Aires before and after an anti-corruption campaign Reinikka and Svensson (2004) difference between central grants to schools and use of the resources in Uganda Golden e Picci (2005) the ratio of the potential stock of public capital (expenditures for different items of infrastructures) and the physical stock of public capital measured for the Italian regions in 1996 Olken (2007) the difference between the citizens estimated cost of building a road and the actual expenditures in Indonesia Adsera et al. (2003), Glaeser and Saks (2003) prosecutions and convictions for corruption crimes 19

20 Advantages of objective indicators 1. Issue-specific since governance is known to vary across issue areas and sectors within countries as well as between countries more generally (e.g., Johnson, 1982) 2. Focus on issues with measurable outputs: education, health care provision, and postal delivery (Putnam 1992) 3. Incorporate data on inputs as well as outputs 20

21 Further methodological developments Identify the different types of distorsions and measurement errors with respect to each methodology and compare them Capture the relation between institutional contexts and behavioral trends of individuals and firms (for example through the experimental economics which analyzes the corrupt behavioral reactions to differents rules) Take into consideration more systematically the environment conditions in the measurement methods of corruption to guarantee more robust estimates of the causes and effects of corruption in different contexts Develop a multi-angle approach that combines micro and macro corruption data in an interaction process between direct and indirect methods (Sequeira, 2012) 21

22 The determinants of corruption: A review of the literature The various economic theories on the determinants of corruption are driven by the logic of a cost-benefit calculus of the potential criminal (economics of crime by Becker) Political scientists and economists have suggested a variety of characteristics of countries economic, political, and social systems that might affect expected costs, benefits, or both 22

23 Democracy, education and income (Modernization theory known as Lipset hypothesis) The risk of being caught may be higher in more democratic systems where competitors for office have an incentive to disclose and publicize the incumbent s misuse of office during the elections Countries characterized by solid democracies, higher levels of income and education are less corrupt voters with more education and income are both more willing and capable to monitor public officials and to take action when these public officials violate the law Greater civic engagement may lead to closer monitoring (Putnam, 2003) Cross-country analyses support this hypothesis ( i.e. Glaeser et al., 2004; Djankov et al., 2003) 23

24 Political institutions and corruption In mature democracies the degree of competitiveness of the party system, the structure of checks and balances, the type of electoral rules, the diffusion of the independent press play a relevant role in affecting corruption Montinola and Jackman (2002) show that only a solid and long lasting democracy reduce corruption Besley, Burgess e Prat (2002), show that high diffusion and independence of the press reduce corruption Persson, Tabellini and Trebbi (2003), Guerring and Thacker (2004), Kunicova and Rose-Ackerman (2005) show that Presidential systems tend to be less corrupt than Parliamentary systems where the separation of powers is less clearcut and there is more collusion between the legislative and the executive 24

25 Electoral rules and corruption Plurality versus proportional system plurality system strenghten the direct link between the individual competence and quality of the candidate and its possibility to be (re)elected Majoritarian systems are more effective in restricting rent extraction because they are more accountable voters give their consensus to individuals rather than to parties (Persson, Tabellini and Trebbi, 2003) Small dimension versus big dimension of the electoral constituencies small dimension constituiencies inhibit the entry of new candidates and increase the adjustment of the potential candidates to pressures by the electorate (Persson, Tabellini and Trebbi, 2003) Individual vote (open list) versus blocked list vote (where what is relevant is the loyalty to the party rather than other qualities) Chang and Golden (2007) show that these two rules of voting perform differently with respect to the dimension of the constituency Overall the impact of the rule of voting seems to be greater than the dimension of the constituency the plurality system reduces corruption 25

26 Economic freedom and corruption Competitive markets and openness to trade reduce corruption (Paldam, 2002; Treisman, 2000; Lambsdorff, 2006) Tariffs and non-tariffs barriers to the international trade generates corruption (Herzfeld and Weiss, 2003, Treisman, 2000) Endowment of natural resources like oil, minerals, gold increase the illegal rents (Ades and Di Tella, 1999) 26

27 Efficiency wages and corruption Relation between the incentives structure and corruption higher salaries can solve both the problem of moral hazard (ex post opportunistic behavior of the public official reduction of monitoring costs) and adverse selection (ex ante difficuty to distinguish between a honest and dishonest official) But the efficiency wages could not work if the corruct bureaucrat requires a higher bribes to counterbalance the risk of losing his job (Besley and McLaren, 1993) Empirical research is mixed: Van Rijckeghem and Weder (2001) find that corruption is lower in countries where public officials are better paid than people that work in the private sector with similar level of responsibility Di Tella and Schargrodsky (2003) do not find any empirical support to this hypothesis 27

28 Public sector and corruption More regulations (production of law, taxation or public expenditure) and greater government size create a potential for corruption since there are more resources to steal and more rules to subvert (De Soto, 1898; Tanzi, 1994; Glaeser and Shleifer, 2003; Adsera et al., 2003) The bureaucrat may use his power to generate some gains for private individuals and obtain in exchange an illegal rent Many empirical analyses (i.e. Gupta, Davoodi and Tjongson, 2002; Glaeser and Shleifer (2003); Adsera, Boix and Payne (2003) find that an increase in public expenditure is associated to higher corruption especially in sectors characterized by monopolistic supply, high technological content and more discretionary and less transparent management of the expenditure But there is the Scandinavian paradox... 28

29 Ethnic fragmentation and income inequality Effect of ethnic fragmentation on corruption and wasteful redistribution (Mauro, 1995; Fearon and Laitin, 1996; Alesina ela Ferrara, 2005) If an area is worn out by ethnic divisions and leaders tend to allocate resources towards groups of their own ethnicity, then members of one ethnic group might continue to support a leader of their own group, even if he is known to be corrupt because of redistributive interests Other forms of division, such as income inequality, may also reduce voters desire to oppose corruption since as voters become more diverse along the income line, they will focus on the redistribution rather than on the honesty of government officials (Mauro, 1995 and Alesina et al., 2002) 29

30 Degree of decentralization and corruption Two contrasting hypotheses: 1. Decentralized systems may induce more honest and efficient governments because of the horizontal competition between jurisdictions (Weingast, 1995; Breton, 1996) when officials compete to sell the same or substitutable benefits to private actors, the price they can charge is driven down to zero (Besley and Case, 1995; Brueckner, 2003; Shah, 2006) 2. Decentralized systems may induce higher levels of corruption at local level the relationships between private agents and public officials are closer and more frequent and because the potential briber has to influence only a segment of the government (Tanzi, 1995); lack of coordination, overlapping and inefficiency (Franzese, 2001) The empirical evidence is mixed 30

31 Crime and punishment: legal system and legal culture The most relevant cost come from the risk of being caught and punished this depends in turn on: efficiency of the country s legal system and on the degree of protection it provides to private individuals harmed by corrupt acts of public officials efficiency of the internal control mechanisms in the public administration (Manon, 1996) and relevance of the international cooperation and external monitoring mechsanisms (Klitgaard, 1998) efficiency of a legal system is rooted not only in the formulations of the laws but also in the legal culture, that is the expectations and practices that inform the way they are enforced (Treisman, 2000) Different conceptions of the social role of law may imply dissimilar perceptions of the gravity of corruption Besides losing their jobs, corrupt officials face a social stigma that highly depends on the prevailing moral norms and cultural expectations 31

32 Legal systems, colonial tradition, religion and corruption Common law versus civil law (Treisman, 2000; Glaeser and Shleifer, 2002; Djankon et al., 2003; La Porta et al. 2004) common law systems are less formal, more independent from the judiciary power and attribute a minor role to the government in the regulation of the economy higher respect of property rights and rule of law less corruption British tradition engenders less corruption (Treisman, 2000) Hierarchical religion as catholicism and islamism seem to be associated with higher corruption rather than protestantism (La Porta et al, 1999) 32

33 Social factors and corruption Social and cultural norms and corruption as a case of frequency-dependent equilibrium (Andvig and Moene, 1990; Andvig, 1991; Tirole, 1996) since the individual reputation is affected by the group reputation the incentive of the member of the group to remain honest is as higher as the repution of the group Transmission mechanisms of social values depends on the education received by parents which in turn depends by the economic conditions and by the distribution of values within society (Hauk and Saez-Marti, 2002) 33

34 Social factors and corruption Horizontal trust (in interpersonal social relations) and vertical trust (in political institutions), respect of formal and informal rules, civicness and political participation affect corruption Social capital creates a loyal and reliable environment which in turn guarantees good economic transactions and good relations between citizens and institutions (Collier, 1998; Coffé and Geys, 2005; Tavis, 2006; Jordal, 2007) Women show more pro-social behavior than men (Alatas et al., 2009; Swamy et al., 2001) More likely the coexistence of greater participatrion of women to public life and less corruption characterize more democratic institutions which guarantee more gender equality 34

35 Corruption and economic growth The impact of public corruption on the level of economic development of different countries has been a topic of debate over the last decades In a recent survey recente (World Bank, 2010) it is estimated that corruption reduces the GDP growth rate by 0,5-1,0 percentage points per year In the world corruption is about 3-4% of the GDP (World Bank, 2008) for Italy about 60 billions The theoretical and empirical literatures still discuss the key question: does corruption sand or grease the wheels of economic development? Does corruption allow investments that would have been otherwise stalled by regulations and bureaucratic procedures or reduce the incentive to invest in productive activities? 35

36 Corruption and growth: a review of the literature - 1 Many economists consider corruption as a major obstacle to economic growth (see, among others, Myrdal, 1989; Andvig and Moene, 1990; Shleifer and Vishny, 1993; Blackburn et al., 2006) The main argument is that a government official that controls the supply of an individually-demanded service may abuse his arbitrary power to restrict the supply, e.g. denying or delaying permissions. The elimination of these barriers requires an extra-price of the service, i.e. a bribe, that increases the bureaucrats' rent Nonetheless, the bribe removes also any incentive to invest, and defines a sub-optimal rent-seeking equilibrium of human capital that hampers growth Existence of a linear and negative correlation between the level of corruption and the average rate of per capita income growth in cross sections (Myrdal, 1989; Andvig and Moene, 1990; Shleifer and Vishny, 1993; Mauro, 1995, 1998; Tanzi and Davoodi, 1998; Hall and Jones, 1999; La Porta et al.,1999; Gupta et al., 2001; Blackburn et al., 2006) Mauro (1998), Tanzi and Davoodi (1998), Gupta et al. (2001), Baldacci et al. (2004) find that corruption distorts the composition of government expenditure towards less productive activities and creates large public sectors where resources are wasted through rent seeking 36

37 Corruption and growth: a review of the literature - 2 Other studies show that corruption greases the wheels of development by counteracting government failures of various sorts and promoting efficiency and economic growth in the short run (Leff, 1964; Huntinghton, 1968; Lui, 1985; Shleifer and Vishny, 1994) This greasing the wheels hypothesis limits the growth-enhancing effect of corruption to those situations where governance is lacking and/or economic policy is inefficient Lui (1985) formalizes this argument into a model where firms value the time to waste in a queue. The more productive firms want to waste least time as possible in the queue, therefore they are willing to buy the priority of their activities by paying a bribe Along this line of research Shleifer and Vishny (1994) develop a model of bargaining between politicians and firms and show that corruption can facilitate an efficient allocation of resources corruption increases efficiency by allowing private sector agents to buy their way out of the inefficiencies that would otherwise be introduced by the politicians 37

38 Corruption and growth: a review of the literature - 3 The existence of a negative linear relation between corruption and growth is challenged in favour of a non-linear relation which predicts a positive growth effect at low levels of corruption incidence (see, among others, Kurrild-Klitgaard, 1988 and Acemoglu and Verdier, 1998, Méndez and Sepulveda (2006) The greasing the wheels hypothesis also implies that the relationship between corruption and economic growth can be affected by both the quality of the institutions and the size of the public sector A non-linear relationship between corruption and growth is empirically found in democratic regimes by Méndez and Sepulveda (2006) and is not modified by the size of government Ehrlich and Lui (1999) develop an endogenous growth model that analyzes the effect of corruption on economic growth in different politico-institutional settings The balanced growth in a democracy (or competitive regime) and in an autocracy (or monopolistic regime) is the outcome of an interaction between accumulation of human capital (socially productive) which engenders growth, and accumulation of political capital (socially unproductive) which mainly assures bureaucratic power and potential corruption 38