The Quest for Good Governance becomes and coveuption the un and public sesourees are allocated. It primarily asks what lessons we have learned from historical and contemporary experiences in develop- ing corruption conteol, which car aid policy-makers and civil societ steering and expediting this process. ALINA MUNGIU-PrperDI teaches democratization and policy analysis at the Hertic School of Governance in Berlin. She chairs the European Research Centre for Amti-Corruprion and State Building Researct of the EU FP7 five-year research project, Pippidi has served as an adviser on and anticorruption to the European Commission, UNDP, Freedom House, NORAD, and the World Bank, among others. She i is also a popular op-ed writer and the author of ‘wo film documentaries: Where Exrope Ends and A Tale of Two Villages, sereored by the BBC.The Quest for Good Governance How Societies Develop Control of Corruption ALINA MUNGIU-PIPPIDI CAMBRIDGE UNIVERSITY PRESSCAMBRIDGE UNIVERSITY PRESS ray Printing House, C b2 as, United Kingdom, lie: wiv cambridge ongiy7@ ne7s34sz$ ina Mungio-Pippidi 2005 ‘This publication iin copys 53457-5 (pbk) sdminiiraton-Corcopt 3. Government accounibilty. 4. “Tanspard 25.C66N8 2015 3523 '5-6e25——s0rso10906 ‘sux 978-1-107-13399-3 Hardback Iwan 978-1107-52457-5 Paperback Cambridge University Press as no espn of URLs nr exten rhideputy ier and ‘The vesearch for this book was funded in part by the European Union Seventh Framework Research Project ANTICORRP {Anti-corruption Policies Revisited: Global Trends and European Responses to the Challenge of Corruption grant agreement number 290529). The views expressed are solely those of the author and the Exropean Union is not liable for any use that may be made of the information contained therein. ‘This book is the first from Cambridge University Press related ta the ANTICORRP project (wivtw.anticorrp.cu). Alina Mungiu-Pippidi (2015) The Quest for Good Governance kk x x * * * * Kya This project is co-funded by the Seventh Frantework Programme for Research and Technological Development of the Europeart UnionContents List of figures page viti List of tables List of abbreviations Ackrrowledgments Understanding, control of corruption Diagnosis and measurement The road to Denmark: historical paths to corruption control Structure and agency: determining control of corruption Understanding contemporary achievers Domestic coll sive action capacity a International agency and its ancicorruption impact ‘ness to critical mass: some tentative policy conelusions Appendix 1 Explaining bribery Appendix 2 List of variables and sources Appendix 3 Brief description of databases and surveys used Appendix 4 Impact of anticorruption interventions om control of coruption: bivariate regressions Appendis 5 HOI differentials from Figure 4.1 Appendix 6 Cisssification of countries by governance orders 30 161 186 207 227 228 247 252 253 259 265 289Figures 1.1 Individual tolerance of corruption by education level 1.2 Context attributions by bribe payers F cop public contract winning ment (2004-2009) Evolution of the WGI controi of corruption average by region (1996-2013) Predicted control of corruption scores based on the rach model of control of corcuption Selected achievers’ evolution rath, Governance more stable than polities n beter society and conteol of in poxt-Sovier stares Democracy and governance in Pofand, 1985-2010 page 24 2 38 a $2 93 120 137 176 198 Tables ‘The faces of pa Satisfaction with public service by beibe idence of bribery and perception of equality in heal Level of observation and empirical analyses of corruption across disciplines Governance regimes aad their main feacures Allocation of disaster fund per ruling party mayors in Romania over three elect Wicators components on different corruption ineasurements ‘The development of i espace of anticorruption interventions on control of atiate reuressions Impact of anticorruption interventions on contro! of interaction with social openness rium cross-sectional models Panel regressions smporary achievers at a glance of press freedom on corruption: panel regression Checktist of good governance progress-tracing indicators for borderline countries page & 8 9 16 29 36 43 7 8 103 107 m1 123 126 137 168 ngAbbreviations ACA Anticorruption agency ACI ANTICORRP and European Responses to the Challenge of Corruption BpP Botswana Democratic Party CoC Control of Corruption (World Bank) Pl Corruption Perceptions Index 80s society organizations ppp Democratic Progressive Party, Korea Em Extractive Industeies Transparency EQ! European Quality of Government Index EU European Union EVs European Values Survey FPA Foreign Corrupt Practices Act FDL Foreign direct investment FIFA FOIA cop gor GNI GRECO HDI lace IAACA ICAC New South Wales, Australia ICRG International Country Risk Guide IMF ISD KMT KOF McG Mp NGO NORAD NUTS OBL OECD ppp DRS Q0G UN UNCAC UNDP UNESCO usaID WEF war wys International Monetary Fand Indices of Social Developm Kuomi mn (ETH Zurich) ranjum Challenge Cosporation ‘Member of Parliament Non-governmental organization Norwegian Agency for Development Cooperation Nomenclature of Units for Territorial Statistics Open Budget Index Organization for Economic Cooperation and Development Porchasing power pacity The PRS Group, Inc. Quality of Government Institute, University of Gothenbueg, Sweden United Nations United Nations Convention Against Corruption United Nations Development Program United Nations Educational, Scientific, and Itaral Organization United States Agency for International Developmen World Economic Forum Worldwide Governance Indicators, World Bank ‘World Values SurveyAcknowledgments dook was born our of sheer frustration, I started as one of the argued that ‘cry was only a small part of postcommumist corruption, and that the patticularistic allocation of publ jain governance norm aod the main driver of po 2006a), ized in my country (Ron Bank textbooks and mainstream own private, My. second ruption, It argued that particularisin was the “d governance Acknowledgments wit new democracies was mostly a co Corcuption is thus bound to int mn and can decrease only wie ps that can be sustained over T owe many thanks generous peomotio Corothers, supporters of antcoreup- 1d encouraged me over privileged testing ground of my ideas, and I owe fewers and readers who helped me advance my ideas through sev ‘An International took six years ing on the basis of ethical univeesalism? This became the key ques- my seminar on good governance transitions at Hertie Scho which has always been shore 1con stadents. [had to add bis try “was not there” yer folt the same about my own. In 20) Norwegian Ageacy for Development Cooperation {NORAD} commis: od the Hertic School of Governance to write an analytical report 3 first generation of anticorruption ated! ambitions and modest resules, a Moby Dick accoss Every Sea and Oc fh NORAD, deserves thanl chat be Instead, the report was called “Context rcuption: Lessons Learned, NORAD, Repos stracture’and argument became the core of the present book. Thexiv Acknowledgnents views and opinions expressed in the report did not necessarily corres- pond with those of NORAD. The agency graciously agreed that some ‘of the work I initiated there is used here in a more developed form. also owe thanks to the Trust for Civil Soviet Fund, Balkan Trust for Democeacy, Freedom House, Black Stx-feust, Stanford’s Center on Democracy, Development, and the Rule of Law, University, } International ic Forum in Washington DC for the support they have given me over the last fifteen years to develop anticorruption experiments, test them and report the results. | also want to thank New School Report Series for publishing in. advance ‘would later become chapters of this book. In 2011, the European Union Seventh Framework Project awarded me and my colleagues the largest social science EU Framework Project geant Buros. Its is mani- fest in itself of how 1 saw the cure 0} ase — not by lat medicines, but by empowering natural irimuni Historians Mette Frisk Jonsen (for D: (for intellectual history of ethical universal Pippidi ian medieval city staves) gave me great advice and guidance. My Aram Khaghaghordy Martine Barr cutschka for his dedicated work on geaphs and ‘ables, and Raniin Dadaiey for his relentless updates and improve- ‘mens to our statistics, Researchers and practitioners from all over the world brought great contributions to our three-yearlong seminar “Transitions to Good Governance” which helped document this book: Drago Kos, Chris Walker, Monica Macovei, Francesca Recanatini, Alena Ledeneva, Donatella Della Porta, Alberto Vanucci, Richard Rose, Mark Pieth, Masa Loncaric, Bianca Var Mundo, Ana Carolina Sponza Braga, Michael Weinhardt, Angelica Pulido Solares, Aiste Skardziute, Maira Mactini, Fortune Agbele, Mariam Gabedava, Dainius Velykis, Kristof Kleemann, and academics, civil servants and jour South Korea, Chile, and Uruguay. its who patiently received me in -ganizatior worked with me to develop anticorruprion coatirions and ci ety anticorruption projects in Albania, Romani Serbia, Croatia, Uk and Central Asia are e endless and frequently1 Understanding control of corruption Why corruption is worth studying Tack al-Tayeb Mohamed Bonazizi Tunisian strect vendor who 17, 2010, in pro confiscation of his wares following an accusation by officials ‘was trading illegally and evading taxes. That stceet vendor’ action started the fires of the Tunisian Revolution and then the wider Arab Spring, and he was cast as a hero by the ly, while President Ben Ali and his wife prospered. estate could have argued thar since people such as Bouazizi had resources to offer ‘might tuen our that the ‘was insufficient t0 pro rnged who were either not carning enough to pay taxes or consi words, beyond the paradigm of p ly case actors in this particular circumstance ~ what seems 1¢ problem in the Tanisian situation isthe absence of an agreed soci berween these actors, avoiding both c mand tax evasion. Only such a contract would give de Does such a social contract exist roclay, demise? Are other democracies in the workd, more mature than the Tunisian onc,2 Understanding control of carruption _Rrass-roots protests in India, Ukraine, he feeling that people are enu- fed up wi led by rent seekers 1 enhance erony capi ble development. TI “corrupt dn su poste: he and others saw as “cor Soccer World Cup azil to stage the World Cup, struck by the hard truth from stadiums to hospi e standards demanded by * country, che one without stan: an event, a pat ated corresponding t everyone else are and where" democratizat double stendacds, of priv their political leaders. Now clections and more thei own governments 00d governance remains 10 flourish in new dem at a relative majority of countries than ever before are these widespread 2 of corruption across the world as being bascd 1s" But why then do-we find consisten corruption is freq hor market and polities (Mu lox relationship betwee Why corruption is worth studying 3 actors at all ot only insufficiently, direee payments by as bribes fund such services directly negative consequences of co ther subverting y development 8 2013 suevey of 88,000 Bur ists the mechanism by which cocruptios from admission exams into schools t0 public sector employment; from a biased and high opportunity casts, Perceptions of co sul experiences, though few respondents have any dicect evidence of corruption. OF course, economic hardship enbances sen Dbut the negative consequences of cortuption » equal access to public has rect is controversial in academic debat also goes far beyor trading, stating in Article 14 How people understand corruption 5 eased in the last onal anticorruption +y work. In most countries lieve they enjoy corrup- 1 growing frustration we sce on the streets, from GCB respondents believed that corrupt ism in the area of goveen- sear, and for every nine people who co ance, The UNCAG, together International Coven: strategies inoffective, only one thinks ths 966, signed by 167 government conventions, signal that accountability, and proj property.” This puts an end to moral rel which can eady mentioned, we find jum and corruption also believe that merit has litle i Correlation Firstly, aval his particular public levels of favori governance. The angry crowds to demand corruption control, bu How people understand corruption I are important as they ceflece popular experiences in 1g the mechanism of advance- rent in society and the way the state operates. The data from the GCB 2013 show remarkable consistency withia national percep- ace of bribery are sig- he 2013 Global Corruption Barom the government (see Tale in nearly every survey we find a large gap between cor ation (having been asked for a bribe) and perception officials are corcupt {bribe givers pereci ‘Ninety percene of the GCB 2013 ountry as corrupt ‘order to fill chs gap, we simply ‘thee experiences trast, a (eek 27 percent ceported having paid a bribe in the last twelve months. As with the anticorruption protesters sviewed by respondents al to k that bribery is only # coreuption icebe: eve dence in recent years from mukiple survey sources woul dictory and incomprehen population when aske society's capacity to enforce pul has corruption contcol in general succceds or not in sheir own country ndorse fair compet ich alysis with factors other chan bribery, wit jon ot development being very important in the fi 1000) and circumsta rs very importaat iat‘Table 1.4 The faces of particularism es % of respond-% of respond °% of respond ents who think ents who think ents who thar personal that the governe consider public Perception of contacts are ment isto“a offcialseivil corruption ol importansivery large extent or % of zespond-servantsas public ofiial Impotaretouee “enieh"ran austhathne “eeycerpe? vise SHnasdonein byafewhig paiiatiies: orcocsemlyfeighed VARIABLES thepublcwewor ica lewtone comp greg % ofrespondents Pearson Com. who think that Sig. (2-tailed) personal contacrs N° 104 are importantivery important eo get things done inthe public sector % of respondents Pearson Core. 0,647°* 1 who think thar the Sig. 2-tailed) 0 governmentistoN 98 98 a large extent” o “entirely” un by a few big imecests °% of respondents that Pearson Corr. 0.145 0348+ 1 Ihave paid a bribe at Sig.(2-tailed) 0.141 9 feast once N 104 8 aos %ofrespondens Pearson Cor, 0.474" sae oa72+ 1 who consider Sig. (ailed) 9 0 public officials’ = N 101 96 11 101 civil servanrs as “very corcupe” or “extremely corrupt” Pesception of Pearson Core 0.505*" o.seu" 0.37285 70+ 1 corcupsion of Sig. (2-ailed) 0 0 ° a public olficialscivil 103 7 104 101 104 servants {weighted average) © Correlation is significant atthe 0.01 level [2alled. Source: Tesparensy Inersatonal (2013)4 Understanding ‘Table 1.2 Satisfaction with public service by bribe eee of of respondents respondents % of rot satisfied sansied with responde: Secror/Servicw with service service with SEE ee roa: Education 3 2 38 i 66 Health 2 Ey 7 Police 30 22 64 Jess} in GCB 2013, although edu why more or less bribery is 1 be ance. Employees in the private sector IFemployed ate signi they are the ones who must neg wurdles ~ Tunisia’s Bouazizi a perfect exam with favorieism n between the a bribe in a con- 1¢ game are so shaped. Fi though idered personal connections matter a Data from, conducted in all EU member states and some accession-aspiring How people understand cormuption ° ‘Table 1.3 Incidence of bribery and perception of equal service provision Equal Treatment Unequal Teeatment % of ets who agree thar who t everyone is aretreared —% of bribe %oFbribe unequally in payers wito payers who che public perceive perceive equal unequal seatment treatment 2 36 98 7 2 3 13 3 87 10 a 2% Souracs ANTIGORRP Eaopese Qualiey of Goverornent Index 2013 only 36 percent in the Northern European core Meditcrrancan countries perceive favori against when de nic, or ed\ arge perception of unfair tecatment, which is grounded in the individual experience of respondents Around 90) petception gap in public sector governance unless and performance are indeed accounted for together. The discussion in many w: comprehensive hard to distingz Understanding control of corra a contest whether the Tunisian street vendor was a tax evader and a ‘smuggler and the policewoman accosting him was doing her duty, or if the situation was the other way around, with the street vedor the victim and the policewoman the exploites, Surveys and slogans nevertheless convey a clear pictare of what does not collect taxcs from the many to redistribute them the few, where the state is autonomous from any par- group and able to seek the best solutions, maxim in an uneven or pactal allocation of public benefits, benefiting the gramter of favors or his associates. How the anticorruption community understands corruption both quite different from the described above, At one end of the spectrum we rion of corruption,” at the other “integrity building” of “good governance.” Eradi from hygiene. tt supposes that some invader {bacteria, insect, animal) ereates a disordet in an organic mechanism which would not be theze in the absence of the pathogen. Therefore, eradication of the pathogenic agent guar- antees the restoration of the “good” equilibrium. For instance, the Oxford English Dictionary defines corruption as the “perversion ot destruction of integrity in the discharge of public d and favor” implying chat somehow an ex: perverted or destroyed (*Corruption” 198: considered that Evil cosiied metely of the absence of God, the paradigm of integrity building presumes that building good govern ance is ike building a barn: it just needs a plan and some builders, The How the assticorraption commumity understands corruption, " ‘concept that some people with considerable power may oppose the ‘construction of the barn as they benefit from its absence is seldom dis- cussed. All we need is to find an entry poi societies are changed dy reforms, and governments follow the pla, Both paradigms thus fail order as a spe- to power sta- ate, together with the Glob: tion grass-roots rebellions demanding radical overhauls establishment “Good governance” is definitely an unpopular academic concepe, due to its vagueness and normative implications, although the pol. icy world cannot dispense with it due to its tess political and more 997; World Bank 1994), Academic ly private clique} pecuniary or status gains; of cules against the exercise of certain types of privare-regatding influ ence” (Nye 1967: 417] clarified an essential clement: the blurring of the private-public border. However, the double reference to the “formal duties” and “cules” failed to take account of a frequently coccurting situation whereby laws and policies themselves are cor- es for those connected with the -ge. Kaufmann and jon between legal an is useful di reed formas of corruption, Nye's formula was simplified in most current litecature as “the abuse cof public office for private gain” {Kaufmann 1997; Rose-Ackerman 1999; Tanzi and Davoodi 1997), with variants such as “abuse of power” or “abuse of entrusted authority.” The underlying presumption2 Understanding control of corruption the anticorruption con understands corruption 13 remains that § private should have clear borders and that the act of giving gifts creates a social bond with an obligation on the and private spheres are involved part of she recipient to reciprocate since “the objects are never com (1990: jonor” we pletely separated from the men who excl ibsence of reciprocation, a loss of Roland Mous until the eighteenth far more import as France and Spain. Societies of “estates” or “orders,” as he defines insightfully including the USSR and fase! ranks rather berween two private parties is than wealth in certain outsources quasi-monopolistic government services to private which thus come to dispose of public ind organi lic author ry societies in and decisively (cg. referees personal and objective ha ed as informing modern states is some sort of final un \ge ina long chain of evolution (Weber 1968: 959). ¢ cemarkable that UNCAC demands them, but they are most not guaranteed yy also means that the defaul have the stare on one side and indivi (Scott 1972) because it presumes that a order to promote private gain (for self or bur to promo that the ease? The ive and has evolved consid- today is quite rably across his ferent from modern Western be But is the presu 1g with under parcy, club, fa reported to provide person tions between an incl ties that inform exchanges and trans and the state, crossing the gal, others nor. We shonld have ige scope of of interest defini4 Understanding fof cormuption criminalizing only particular exchanges, which include physical gifts (bribes) and which often occur in the absence of a bond or connection ‘of some other nature bat with the goal of establishing one, But con- sider a government captured by an ethnic group or a clan chat distrib- ‘utes benefits mostly to its own members and forces others to engage in is “appropriated” by a ruler who discretion in the exercise of power at the expense of ment applies to everyone regardless of the geoup to which one belongs {Parsons 1997: $0-82} to particularism where indi 0 particular ties or eri ion of a notion such as cor- society foundation. The problem ve is how to change from a govern- ive action problem {You 2006). x deep in his depiction of how power inequality shapes governance. For instance, he described premodern “status soci~ ‘eties," dominated by certain gcoups and goveraed by convention (what we woud call today informal ins \} rather than law (Weber 1968: 177-80}, Such societies ave not primarily structured by capi in the form of casey, onde, or networks, of public goods cannot help but be par- lar in such contexts: particularism is the rule of the game in such societies, and the standards for the way a person should be treated depend on the “estate” to which the person belongs. These are the How the anticorneption commumity understands corruption societies of “estates,” where access is determined by belonging certain gcoup and the personal connections entailed by this affiliation (Mousnier 1969}. Surveys on corruption and rule of law suggest that such societies are not historical relics. For instance, why do res ‘emis of a survey in five Balkan councries, after stating stro certain people are above the law, rate politicians and policemen most highly in wsing the law to their advantage (Mungiu-Pippidi 2006a)? -¢ patcimonialism with the impersonal and func the modern state based on absteac sonal, wri different soci collectivism and part sonal relations and the relationship between the state and is impartial and, wheu implementing laws and policies, tceats citizens as mete individuals, “not taking anything into consideration about te citizen/case that is not beforehand stipulated in the policy ot ions are possible orlly because we supply and demand mechanism, not sense and market rel sonality and impart: kinship, determines price) ‘The international anticorruption community thus faces a daunting problem ~changing basic governance ~ without a cleac understanding of the dimension and nature of the problem, The attempt to agree on 2 universal definition of “corruption,” as well as explaining variation lack of communication across several of the disciplines dealing with the topic (see Table 1.4), all of which have made significant contributions without managing to provide a unitary theory explaining how corruption develops inorcectly defining the goa! of a public policy sdeed the norm is ethical universalism and cor {context A) ‘egy for adequate control of corrupt whi arsenal), based on suppression of cottuption. If ethical univers not present and nonuniversal teansactions of government ate the most numerous (context B), the right strateay should be conceived instead along, the lines of norm building, with suppression only complemeat-Table 1.4 Level of observation and empirical analyses of corruption Prefersed method of inguicy Discipline Jndividual or proup National Andividual Group and organization Notional or sector Some scholars work hard to find differences beeween v festations of par ism (nepotism, ethnic favor ribery, and a democeacy control of pul feacure of the potitical system. Bur € the government side {eeturning of favors as prefer ‘ovo types of corrup far more useful Concepts and plars of this book vy such preferential returns followed kinship hannel for privat ing preferential advantage of pul lern times, crests. But groups tak- mnumerable, Johnston (24006) attempted the best ion to date by country, whereby he combined power pluralism and accou ies where profiteers of public resources are clusters: cout as cooperative “moguls”; and democratic reat seeking (influencing markets}. Before any further attempts at clas- sification, the essential diagnostic element to consider is whether or not the dominant social allocation mode is ethical univers Concepts and plan of this book eneralizing conclusions of a study (see Table 1.4) fost of the modern anticorruption industry operates rel, but defines corruption at the micro level as devi 1g whether it is the norm or the This is the source of very serious pr ticular transactions are the majori can hardly be answered by repressing de imply because appro" priate behavior conformity with different so4 people nsibly behave differently from how their peers would jn a similar situation,corrupt ‘overall rating ro integrity and fairness in che management and di resources to be a defaul not be correct, 1 resources by one group or individual for is extremely common, and the developnx of a system to defend those resources from spoilage by violence, or corruption is the product of a long further reduced '$(0 emulate some version uf Western ing in the adoption of UNCAC. But how do we feet chere, and what lessons can we learn from countries thas have suc- ceeded If con defend rerests rather than the publ of formal and infor are distributed} Concepts and plan of this book 1” and power disteibution in a society ( 2000), so alter reachit ibrium they vend co prove hard to undermine by elections or even chamnes of pol regime, Such equilibria are litecature as either * live under e: rienced bri UNCAC, whos Explaining control of corruption nary effort, whether by leaders, bureaucrats, oF «i of morality 0 fully autonomous morai agents impervious to the soe them, buc are bound by social contexts where social representations af moral concepts suck as right and wrong are devel- oped and sustained by social norms and everyday behavior (Bandura 20025 Moscoviei 2001), Sacial cogn ality theory thos concur that individieal moral behavior is the product logical approach, which began with Durkheim, can be encounter relation to governance in classic historical sociology (Eisenstadt and0 Understanding control of comeption Roniger 1984; Weber 1968) and more recently in social psychology Sah 2007), There economists that the allocation of ies develop “reward structures” that deterr int between pro and unproduc tence of rent seeking creates a ne} operates by un of par “To say that individual choice is bound! ondents reporting that bribery is unacceptal Dong et a surveys such as the GCB include hetter-phrased ques- tions on corruption, but on the other hand miss the kind of questions ‘on individual values chat would offer a model with good explana: tory power at the individual level. Combining the «wo, find evidence supporting a model of mixed! choice, where indivi who have more dealings with authority are more its if they are in a corrupt envicos iployed or men}, but we find also tha n are less tolerant of corruption and would perhaps be more willing to engage in changing the cules of the game {sce Figure 1.1), ‘urthermore, the mere recognition of norms (“appropriateness”) to reduce eogni- dividuals who rance, the duc: Concepts ard plan of this book 21 200 Bribos ae justiabio (= Noverjustilabio, 10» Always jusiabla) ce of bribery also believe that the pr se who have no such exper This World Values Survey find Hofstede’ ral indicators, for instance the connect ractices and performance excellence cannot coe? corruption, Moreover, measuring the inequality and importance of power status shows that socicties with higher power distance are also more corrupt (Hofstede 2011; Husted 1999). Just as individual behavior las to be placed in a certain gov context, the state too must be understood in the context of a givenns conbobe ommerin 2 2% B 50% 48% a igen ge 22% ott Bg a 7 i 20% bbe i. Elie P. c vo ot pea be se pa ite * mNovertarely mSomatimes a Frequentyvery nequenty Figure 1.2 Context att Shree: ANTICORR? yutions by bribe payers a0 Quality of Govsromen Index 2013 location reflects the power distance in a society and the dominant exchange mode, based on eithee performance (market) ot rent secking (violence, corruption). Particularism is the wider govern ance regime indicating the dominant norm and behavior m, favoritism would be the main social allocation mode, with widespread use of connections of any kind, exchange of favors, and, in their absence, monetary inducements. Particularisin in 1 society operates mostly co the advantage of those with more power resources, but no simple elite-theory type explains it. The weaker have their defenses, resorting to patronage, cheating, bribery, tax evasion, and a variety of other practices to ceduce inequality (Scott 1972} once they perceive that the social contract imposed on them is unfair, Elites witha geeac deal of particuiarism enjoy a culture of impun- ity (famously extended to things such as UN diplomats not paying parking fines in New York, as reported in Fisman and Miguel 2007), 1 culture mirrored by the self-indulgence of ehe rest of the population who see the behavior of elites as the best excuse for their own behav. ior. By ane large, vertical and horizontal exchange networks intersect to create the systemic nature of the problem (Johnston 20005}, which remains the particular, nonuniversal character of allocations, resulting in injustice and diseriminacion (Della Porca and Vannucci 1999} a society Concepts and plan of this book 23 Corruption in society is therefure not conceptualized in this book as an aggregate of individual corcuption. ‘The not countries at the top of Transparency International’s Corruption Perception Index (CPI) do not differ from counttics on the bot tom simply by the number of individuals engaged in coreupr acts, but by their institations: in other words, by the rules of the game influenced by power distribution and the shaping of the on of public resources, The countries at the top of the Control of ruption scale managed to insticutionalize open and nondiscrim- native access at some point in their past, and so their institutions differ substantially from che ones at the bottom (Asmerom and Reis 1996; Norch et af. 2009), Many countries in the middle struggle between two worlds, for ia them both universalistic and particu- laristic practices coexist, more or less competitively (Van de Walle 2001). But regardless of how wide the variation might be, some sort of invisible threshold exists between a society where etbical uaiver- salism is the norm and one where the norm is particularism ~ and one can predict fairly well what treatment and what share of public resources to expect from the state if one knows where one stands in the status ranks ‘An anthropologist might call such institucional arrangements ci ture. However, the most frequent question on corruption ~ whether of not itis “cultural” ~ depends greatly on how we define eutuce. If culture is defined as she rules of the game in a society, corruption is definitely a “cultural” phenomenon, Would it then not be a mystery why people complain about it so much when they have in fact lived with it for so long? More interesting still, why do iminigcants find it so easy to abide by new rufes in noncorrupt countries, scemiagty the most attractive counteies in the World as far as immigration is concerned? But if we give culture a deeper, more primordial mean- ing, it becomes even more difficult to claim that corruption is a cul- tural phenomenon. Individuals may value gifts and tradition more in some cultuces and some cultures may be more collectiviste than others, but no culture on record values dishonesty and unfairness Nobody wants to be discrieninated against, which is what corruption invariably does. Conceptualizing public commuption holistically ira broader govern- ance context thus has important policy consequences. Particulacism “malady,” as corruption is usually described, but is not a soci4 Understanding control ofc rather a defau peycholo, sectarian, parison and s 2004), Humans group ~ what Edward C. Banfield (1958) « ‘Treating the rest of the world fairly seems to be a matter of ex: al learning and su es which have trav leed, as Jarnes Q. Wilson (1 ism, which spread in the become generally ageeed Concepts and plax ofthis book 2s fair and accountable government; economic growth can advance is even helped in the meanwhile by a certain degree of corrup. ind development are two very different somists have come to argue pres lar) allocation of public resources hinder development, which # in a vicious cycle of captive states and poor societies. The precise mechanism is not clear, because many institutional explanations of con the premise pioneered by J and power distribut governance contexts or “syndrories” of corruption. fe thas long been marginalized in the grand development theories, and it has only recently started to be more prominent as the nection in Acemoghn 21 North, 012), and the rule these remarkable books, even with- ‘out sometimes naming corruption explicitly, treat governance as an explanatory variable of presperity. Bot since they all conclude that governance matters grea ance deserve to be exp! attempts. More specifically, this book takes a broad development perspective by addressing three main questions: The development question: What was th he comparative questi contemporary socie there any pol conteol of corruption (Chapters 3 and 4); the second by a review of previous econometric work and the author's own contribution, bothanalysis tools, Chapter 2 fying Understanding contro! of corruy 2 Diagnosis and measurement Operationalizing governance contexts of corruption measurement is as contentious as - and relaced While pressure from the international development community the creation of a frst generation of indices such a8 Transparency’ ism, have been lenges of indicators ro he comparable ae inge and policy intervention persist. Many critics macration governance indicators argue chat absence of a c measure (Rothst if the theory fai the because terms such as governance, good governance, and qual- ad and ambitious, and those he risk ity of government are by nature too he consteuct of “eontral of jon” as it not measurable if we focus simply on the ical universalism and parti nd ascer ina taxonomy of governance less normative an confusing than discussions of “good governance,” which caa be used to justify both Faey Diagnosis and measurement 1 and demoeratically elected yoveraments? Some authors, icism over measurements of government based on pro- cedures versus outcomes, agreeing that “procedures, however defined, may not actually correlate with the positive outcomes expected from governments” (Fukuyama 2013; Rotberg 2014). I would argue, how- ‘eves, that social allocation is far more than just a procedure, and if we can establish what the rales of the game are in such can fairly predict other features of governance as wel law and state capacity), as proven by a vast lierature on governance from Bretton Woods institations. A good example is the association between tax collection and control of corcuption quoted in the previr ‘ous chapter (Mungiu-Pippidi 2013a}, Quite a few recent contributions seem to converge on the cen tality of control of corruption so defined for prosperity. The ideal governance types and Weingast (2009) into problem of containing violence by nomic system to generate rents by “orders” developed over the past 500 years. T social order through political and evonomi latter manage to sustain impetition, rather than s for elites in developing countries, even if by doing so they impoverish broader society, In fact, theic she- ory nearly states that the type of exchange (here, corruption) expl development in Full. Bi teast to what they call bad, “extract would be difficult co reduce their dimensions to som able concept, as some of their critics have argued (Ful An attempt fo integrate these taxonomies resting on Webecian foundations and focusing on the notion of control of corruption és, showa in Table 2.1, tt combines the concepts of open access and power pluralism to ercate two governance order types dividing the continuum: opea access orders and limired access orders, with the k ter divided into three sub-types or -contexts: patrimonial {traditional ‘or neo-}, competitive particularism, and borderline. The first ty of governance ~ open access orders ~ corresponds to Robert Dahl's “Table 2.1 Governance regimes and their main features ‘Open access order Limited access order Universalism, Borderline particularism Pateimonialism Giizcnship ‘Competitive with less Stratified with Hicrarchicat with Power distribution stratification power disputed competitively State captured in monopoly of central power Srate captuced by State autonomous from Archipelago of State autonomy privace interest ilegal lobbying, sutonomy and Eshical universalisn tuen by winners of ete.) captured {islands} Particular and Particolar but Particular and Public resouree (inclusive) universal (00F29 competition} Poor ‘unpredictable extractive} predictable allocation (extractive) No Sharp N Separation private! public Relation herween Informal insti, Competitive and Informal insctutions Informal institutions only complementary to formal ones Individualistic Permanent stbsttntive informal Mixed substcutive of formal Colleetvisic substtutive of Josmal ones Collestivistc| No formalfinformat Mentality Only when no longer in Occasional Government power No accountabilicy Rule of law General; “thick” Elites only. No; sometimes “thin”30 Diagno: surement polyarchy or ;pper’s open society. This type ol I equality, high personal auton universal between formal occurs, is indeed a and public impartiality. Open access orders have been promoted for several decades: first, sl I powers when \d a handful of ing on the basis of kind. As they cover so n iges (unlike open access count effort and appropti the autonomy of the vate actors’ capacity to uphold corresponds roughly to some thin form of cule of law. Nevertheless, the state can still be entirely appropriated by one powerful group Operationatizing governance contexts a ocation of public goods. The es only after social certain bound of another, resulting in an une’ development of corcuption control is achieved and violence is preserved and modern on the one che other. In Weber's patrimonialism, authority rest ised by a royal houschold, whose power is formally ry and under the direct conteol of the rule. Dori cal apparatus staffed by slaves, mercen- not a traditional land-owning, controlling the instruments of power inthis way, the pateimos can extend personal grace and favors at the expense of tr the exercise of author and Modern Neopatrinronialism, Risenstade ced term “neopatrimonialism” to describe a mixed systet lements of patrimonial and rat ie sometimes interwoven. This regime imitates formal insti- kleenity from the West, with most informal in i patti In answer to critiques of the concept 5s (Erdmann and Enget be restricted co mean. only the nontraditional governance order based on a power monopoly {of a rules, tribe, or party) treating the state as its “own” patrimony. Many new democracies ro longer have such monopol 1996: 41). We e need 10 add another specific governance context, “com culacise.” Unlike patrimonialism and its modern vari m, which are authocitarian or semiauthi fe particularism is, ‘of rulers spoiling the state despite some form of in lism, such as regular elections.32 Diagnosis and meastrement ‘Competitive particulacism is chus a governance order that replaces le of law is poor as those in power are above the s perceived as an instrument of spo! which greatly subverts ibed as “compet and Chang 2003 dence show we are lo bout rather an ideal belong to the privileged group rather than « 1e game. A culture of pr ing unequal teeatment the led some basic and necessary conditions of progress toward an open access order, in whieh the two normative orders ld and new} coexist confrontationally without one managing idcal-wypes ant differences th uneven, from the classic configu Diagnosing governance contexts 33 relations structure polities in a vertical and particutaristic manner des- t consicaints and no horizontal 11996). Under pa at best (applied pred ¢ particularism campaigns from power; only in open against pr petitive particularism only when they f access order regimes are prosecution at any tim ays be well defined, and the has been passed from one I to identity, exact moment category to another might prove di Diagnosing governance contexts These governance context ideal-types differ across their defining aspects, somte of which are more like featuses (relations| or informal institutions, autonomy of the state coward private interest, separation of publiciprivate spheres, patterns of allocation of public resources}, and others more like probable causes {power resources tuibution}, Do we iave adequace instruments to diagnose and iden the governance contexts described here as ideal-types? Such a quali- ian protesters at the 2014 FIFA World Gup have thy ators by which they judge poor governance. A systematic assessment would first establish the outcome thor a society manages to con resources? Are all pu a particular manner by default? uuce mostly to itself {associated local governments or regions, favorite companies}? How much of the rotal spending budger is rents? DoesM Diagnosis and measurement country ona ism. This variation explains rerature, all of rough bribes) ~ in between the two, The Halian 1996}, has no separation between the post-Sovier char sa ad businessman: old apparat capitalists — pro- fessional ci the “business mal network does not make today, tomorrow sit in the € board membership of a publ is eather difficult wo calcul the profit goes to more than a sit risks punishing only cartel trying ro enter the market on a more ¢ numerical nore systema 1s to classify countties on the particularism-universalism uum? If the majority of state transacti universal kind, we need a benchmack distribution c particular or Diagnosing governance contexts 38 not taking any beforehand si 2008: 170) weeanure on that i (welfare pork barrel, in other words, ig, with economic or service benefits concentrated spread among all taxpayers few exceptions the difference ify in law. Programmatic alloca~ paymenc)anda government spe in a particular area or group, but cost (Shepsle and Weingast 1981), but not so hard to grasp or actual as people with approval of for the bills sponsor will get emergency flood assistance, instance), opaque (often hidden as maintenance works, budget amend- ey tend to push a community ween budget the legal process thac itself so anywhere ria for allocating, money, apart fides a benchmark (were al procedures respect is possible ro document the progress in the fel spoils in Brazil. Des career advanceme: bureaucratic rules semains high. In 2008, the year when the Supeeme Court prohibited nepotism in the exccutive, legislative, and judicial36 Ditagnosis and moasurensest ‘Table 2.2 Allocation of disaster fund per ruling party mayors in Romania over three electoral cycles 2004 2008 2010 (50. (Democrat Democests) {Liberals} Liberals) er a an 45 a 16.2 28.8 28 22 branches of govern nated twenty-three relatives, inch bis government. By 2010, ated in the judiciary by the Ne his wile, cases of nepotism were al Council of Just nder the headi ¢ Government (Bra and Management 2010), with salaries ra this case tet fund in the ‘The beachmark Diagnosing govemance contexts 9 ntary approval duc to their Id follow natural disaster how a regulae pa ibuting emergency grounds {mayors of the ‘ding pacty get the lion's share, regardiess of who gets flooded) has € different government between the ruling pat ey recetved has the government without parlian From these favored allocations, public funds flow further dow networked companies, which collecc nearly every th politicians sometimes have direct st ve otherwise received electoral camps public contract or from which the random occurrence of natural disasters, but rather follows a clear and regular pattern eelated to the party in government. Another example can be seen in Figure 2.1, where data mis Hungarian proct cation over 10 mulion euros for the period of @ nonrandom pattern which ean only be expl favoritism. The graph shows how one group the area of public infrastructure is replaced by a new group following elections and a change in the ruling government coal ‘a0 sampling) for one public agency or an expected pattern of so ‘of e-government will make such methods increasingly ie and reduce the labor lection, However, such data is specific, object lowing both the moni- toring of change ard incceased sensitivity to an eventual policy inter vention {eeform). ig provides a rich source of informa38 Diagnosis and measurement 8.8 8 of total contacts awarded s Top 90 is (2008-10) (2009-10 agusted} ie contract winning companies Figure 2.1. Change of fortune for top 1, Hungary 3 ‘bur not im hied of Romania’s candidates for parliament government private companies when their pacty ‘was in power, WI n would be possible (Mungiu-Pippidi 20131 our a thied verficainegriatea ly Fazekas, 1s University of Budapest Diagnosing governance contexts 3» sector to serve as comparison (for instan structure, which is notoriously vulnerable hhope invested it to replace perceptions of corruption While surveys of users of public services (as iences and measure only audits bring more obj xpayers who ices, companies who apply for public contracts, banks who are aware of the rules of the game ge the ones in this chapter when offering their procurement} 3s. We cannot know, however, how sensitive to change such opinions are, as people's expericnces are anything but systematic. To monitor the progcess of reform efforts, then, data mining offers an indispensable complement many transactions are particular versus random or are completed according ro legal benchmarks ~ is useful also because it is. not40 Diagnusis and measurement “objective repression data, which cannot sep- tion from corruption itself. For instance, has opened more cases on the basis of the OECD at ress people more commonly offer bribes when ‘or that Germany has been more active in enforcing the conver as compared 10 ‘The same apy re number jons. Notoriously corrupt countries have convicted very few people for corruption, as law enforcement itself is inefficient and per haps corrupt. Cross-country comparisons Objective data is ideal for tracing the impact of reforms in one coun- ys but how can one compare across space and over generation of corruption inc nach methodological progress has been made since. of data can be seen ¢ interviewed var iduals and categories. For public officials and frequently amounts to tai sponse rates and tend (0 correlate poorly with exp general surveys (Donchey and Uj secvices are concerned, but they have less dire in areas such as grand corruption. Expests presum- ike the ones shown in the pre but as expert ratings tend co be rotoriow know what indicators were consider Can surveys be used to measure trends? Although samples differ from year to year and we do not have a panel sample in an i national survey on governance, survey results can be used to assess changes in governance, The problem is rather lo since sure vveys are expensive, carried out irregulaey, with different providers fanding and variations in. the wording of questionnaires, making difficult ro compare across years, The Global Corruption Barometet Cross-country comparisons a es ‘00 i wo ae og 5 ni Bae oo i ai G6 40 8 3 we fa ae 3 10 ig ° 2008-2008. «7005==S OOO of udiial eoruption of eorruption in patament judiciary and parliament quently over che years wich new wording, has high s, and had such frequent changes in counsry ross the years, even as the number of countries surveyed has , however, does not have direct experience h cither MPs or courts, and the remarkable consistency across these two very different arcas iaverage assessments for he fact ‘country’s governance, which spi where their direct expericnce is limited. Can surveys be used to make a comparison across countries? We have xno universal tool co measure corrupt «ties might prove more of bias can be and ate studied in experiments. The Eurobarometera2 Diagnosis and measurement 374 survey (Kuicopean Commission 2012) asked nationals in FU mem- bee countries if their country was moré or less corrupt than other EU countries, Sociologically, the question itself is mistakenly formulated, as respondents cannot know enough about other countries: the survey did aot even ask if they were familiar with another EU country or bad even visited one, Amazingly, the average rate of non-response was only 15 percent, andl the tliree most corrupt countries (Greece, Remania, and Bulgaria} as well as the theee least corrupt by selfperception (Sweden, Denmark, and the Netherlands) came up in similar sequence in both comparison and in national assessment, By and large, the ranking derived from this question comparing themselves with other countries returned remarkably similar results 0 the other questions in they were asked co report on corruption in their own national or local authorities. Some countries might be more self-righteous and others more self-deprecating, but the bias is smaller than we would expect: people are wot easily duped, and their opinions about corrup- tion do not seem to be non-attitudes {i., random answers), but rather quite deeply felt, They also have a normative benchmark against which they compare, even if their knowledge of other countries is only hearsay. Furthermore, when we examine expert sources’ catings of 2 ven country (such sources are frequently forcigners, for example risk assessors in the finance industry} with what citizens think abou their own country, We also find a strong correlation (see Table Citizens’ corruption perceptions are nearly identical co ratings of or- ruption aud quality of government made by banking experts. There seems to be something like a national context of control of corruption that people perceive, which is far less specific than the o ments per sector, but also quite consistent across years and vaciations in experience If sources ate so consistent, then they can be aggregated into onc eor- ruption rating pet country. That is what two of the most widely used measures of corruption do, either in the fori of simpler averages, such ‘as the Corruption Perception Index from Transparency International, or as more refined, unobserved components extracted by the World Bank Institute's experts (Control of Corruption, Rule of Law), which allow us to observe the differences between sources. While those in cators have been severely criticized for their lack of validity reliability, and theoretical basis (Andrews 2008; Thomas 2009), the fact remains, as argued by Kaufmann et al. (1999) and Knack (2006), that as most ‘World Bank’s Worldwide Governance Indicators (2020) Global Corruption Barometer 210-11) Global Corruption Baromecer 2010-11) ‘The PRS Group's Ineeenational Country Risk Guide (2010) Source of public Diversion fonds index 1 29 ~0.5238¢* Quality of among polir- among public government Corruption perception a index 1 97 cd 3 ‘corruption officials 1 ” 9.6081"* 0.1323 0.2332, 5 Perception of Perception of ccorcuption seal parties 1 ogee" o 196 0.0467 0.6497 ” ~4asse ° 97 139 corruption Pearson Cort Sig. tailed) N Pearson Core Sig. (2-railed) N Pearson Cott. Controt af Pearson Core, Sig. Q-tailed} N government Sig, (2-raled) N comuption index Extimate corruption among political corruption among, public officials Quality of parties Perception of ‘Table 2.3 Correlations between different corruption indicators Control of Perception of ‘VariablesTable 2.3 (cont) Perception of Perception of coreuption Diversion of public funds Quality of among polit- among public goverument Corruption perception -05167* the Bales (418) 2731 Former Soviet Union ase aay amp : 0.38 Latin America ozs St (0.44) (1.50 The Catena Cowes ans (33m 254 Midale Fast and “Dove ae North Aken . om (6% (038 Sub-SshacnAfca base oes 0 (130) Constant -2n778 aasoees “pa07e* asioee : weet 1497* D419 Sna2? 30 coun (23) Gein 5330 aa BOM) aH ies 3 3 is 123 § Adiesguwed 0909098 aR vay as Os regres ng dn for 210,08 daw or ols ee 2012, Da on GO nae ad taco wes om DF soe ly. Robust standard errors are used; f statistics im parentheses: *pel.10, **p** QO4B** —U.0467** GOST 0.046" (0-8 best} (03) As} 8) GB 4.2AH BT B88) G17) Freedom of the press 0.006"** — g.0ag*** 0.002 0.004 0.005" (1-100 most free) 0,007" 0,002 0.nas** (369) 403) .2} LATE 2.98), BBB (27) 1.98) Social openness (1-100 9,019"*" Q.01S"** 0.009" 0.006 GOT2t**O.011*** 9.013** O.012** ios open) 17) Gat AB} .S6) ABH) BSE) B14) 2S Economic openness 0,006" 9.007"** 0.006" 0.007" O.005"* —0.006* 0,005" 0.007"" (1-100 most open) 629) G2 Re) 267) SR) SHAS) 2.09) Natural resource rents -0.073"* 0.068" -0.083** 089-0173" 0.064" ~.068—-0.058 (% GDP) (237) C479) 2.03) 1.59) L241) LBNL) 1.08) Rural popalation (% of ~0.005¢ 0.002 e001 ° total) e184 28) wey (0.02) ery shoo 0 enrollment (% of roral 7m ° om popula) ons) eos (0.09) ss) ons tcores “Taniees cascaree 279 tape Sle agains A Be city G47 43 Cha) Gad) sea) CSSD Ge obsenations nese BS ae SS Counts Mea emda Rewuad(owal) 0727) a8) Reswacd(wihn) 0120128 ods 3h dean rae enol fcr an WG Te fang coe of ec yarawaguooag DOC OTT (Clustered standard errors by country are usd statintics in parentheses: * pe, ** pels, >=* dummies. 0.01 All regressions include period128 Structure and agency [As mentioned above, in order t0 t wlel p is with rime secics data, we have to use different indicators compared to those that of avai odels presented, lization, capturing freedom openness. The idea behind the use of this restricted the economic transactions of a co wm there is for admi red tape. The second indi ‘openness that I have already used as a proxy for an as and open society in Table 4.4. The data covers the years 1996-20 and a sample of 148 countries. | use three-year averages to account for the per of inseieutional and apply two standard estimation methods for panel data the rest of the se discretion and, rare, tradition, ete, Vin and education to contro! for the effect of development, because HDI data is not available ht be driven by the (8) repeat the es level of control of corruption in the total data sample, most of , however, result from cross-country rather than time variation mn or change in St is sus- yy it and ose obtained from the cross-sectional analysis. [n particu xretion and dependency on natural resource revenues freedom) positively influence control of corrupt Equitl models 129 thet, supecior equilibrium. Cases such as Uruguay, and Georgia prove that this is possible. A forecast based on this model would imply that change ernance order can occur they seem co result more from domestic agency and broad reforms than from typical anticorruprion efforts focused on repressive agencies. auticorruption efforts unless they are contextual ~ in other words, adjusted in which particular transactions are ¢ f particular transactions are widespeead, need to be comprehensive to affect agency matters. In a context where particularism is the rule of the ‘ame, anticorruption needs co involve both state le.., fiscal teanspar- eney} and society {watchdog NGOs) in order to influence bach sides of the formula, it also needs to be grounded among those who o} the atus quo (genuine “principals, power Wd corcuption} and cannot be top-down “reforms” driven by nobody's interest i5 Understanding contemporary achievers ‘The search for control of corruption determinants across. present sal achievers has $0 far returned sonie pieces of rightful place, wry equilibriam ical achievers? To what extent is the contemporary eq) sesh in he previous chapcer discernible in and relevant to these new evolutions? of recent achievers are not only hard What may appear from the out 4 couatry in controlling corrupt people living there, The susta to judge. How many years, perceived as such «yes 0 favorable coreuption ratings are acded before soy i considered to have passed the point of no return into sustainable con- calbioonepe ta n vis one example jing such a point exists rement)? ¢ of the oldest “contemporary achiever 1s long been perceived that they ar their neighbors. Even in these cases, however, more recent evolutions is hard to suppress p in rela examples of transitions to control of cor it not ro agrve with them when consider World Bark charts to a level superi | elections r has risen a Romania, but whose last president bent with 95 percent of the vote and whose yover companies bidding on public contracts (and winti ned the incu it party owns 1g many of them} 130 Undarstanding comtemporary achievers 31 (Bozzini 201 ance ch, ‘next FIFA, Hed in allegations of bribery to win a tender for the id Cup and which has only re from chat of the royal family ( 18 of Transparency I of corruption, evie some positive developments. The cemain- chapter investigates the transition to control of corrup vuntries with the aim of identifying changes that explain each improvement. The time frame for to the most recent period of democrat infocmation about pre mes provided The choive of ulat group of countries was driven by # first, the case had to be classified in the upper third ries on the World Bank CoC scate or 4 regional achiever cantly better than ry had to be an d to have achieved of corrupt mes, particulacly dur- eaty years overlapping with the control of corruption \dwide Governance Indicators (WGI) Control of ¢ for the score that includes most avai and International Country Risk Guide (ICRG) and Business Index scores ( ‘cases, but longer time series) was used for the th: countries with some degree of the handful of author id by ethical universalism ever if they have managed to control bribery and build well-performing bureaucracies and business-friendly environments, Enlightened despots can exist and have132 Understanding contemporary achievers choice of good governance reforms is clear if they want to emulate “che king, of Denmack.” Policy expertise cannot, however the raze occurrence that makes a despot enlightened: the cl remains in building control of corraption by democratic means, therefore by solving cotlective action dilemmas. ‘The group of “achievers” thus selected is highly varied, encompass- ing Estonia and Georgia in Easter Europe, Chile and Uruguay in Latin America, Botswana in Africa, and South Korea and Taiwan in Southeast Asia, Despite regressing after the econoinic crisis of 2008-09, Slovenia is above the cut-off threshold in Eastern Europe. In Asia, Japan has the best and oldest control of corruption. Apart from these countries, this case selection leaves aside very few cases that have reached conteal of cotsuption through a democratic process over the past two decades {mostly small islands in the Caribbean and Africa, such as Barbados, Cape Verde and. Manritius, and tiny kingdoms in Asia like Bhutan), This group of achievers, however, includes two cases - South Korea and Georgia ~ that are both still below the designed cut-off for the first ctiteria described above (65th percentile in control of corruption}, despite improvements in recent times. These countries qualify due to their superior performance in relation ro their region as a whole, but should be considered borderline cases rather than achievers. The evo- lution of these eight countries over the period 1996 to 2011 is shown in Figure 5.1 ison top with a score better than the United States, wed by Uruguay Estonia (ranked 1 in the Eastern European region), Taiwan (2, after Japan}, South Korea (3), Georgia (1 in (ranked 2 in Eastern Europe, How do these achiever cases compare with the rest of theie contin- cent, and what enabled them to outperform other couneries? This jows a theee-step methodology. I first analyze how we tion model and the model of regi paring the country’s performance against then examine briefly the dynamics of the Sntroduced for example, data feom the Airobarometer or other regi sand case selection peovides another way of controling foe eonsinentalfacrocs. Chile and Uruguay ‘33 Wt contre of comution rank (0-100 best) evolution. Control of corruption reflects which public power is exerciced for private gain, i of corruption as well as “capsure? of Source: Woeld Ben, Work! change process (the pol order to understand th tutional f ei oma ime su famework i general and anticorruption interventions in particular to check to what extent progress was due to special anticor- suption interventions. In the end all eases are compared, America has wo achievers ing the threshold (Cos with the loagest democrati ile and Uruguay} and 2 coun- | Which also match the coun- ion on the continent. Uruguay, i. with Argentina, was among the wealthiest cowatries in the world around the World War but stagnated for most of the ing twentieth century, only later resuming growth to arive at the second:-hest economic performance on the comtinent, Chile and Uruguay perforin better on izcome compared to the rest of Latin134 Understanding contemporary achievers ‘America (Chile with over $11,301 GDP per capi with $9,087, $6,465 by 21 homogeneity, which are factors for conteol a average and Umguay worst 0 ind they are both unitary stat exactly the opy 1 for the x corruption were hree types of zov- iculatism, dependence in 1810 ing duc to a cortuption scand te groups began to concest model was the Fre power hub fi remained also constrained by power, whi fe. The autonomy of the ans adopted an extremely spite common problems (lack of sound pub- lar foreign advice {from the American “money recession-strick strong audit ag lie finance} and Chile and Uregiay 135 man repos hee to parliament nor to government. Previously Chile had separate agencies to cover these functions. No presic ean change the head of the “Controler seventy-five and, eemackably, not ation, The presidents do appoi by the upper chamber. Onk replaced a coi ization of a ple In close celation with a strong autonomous central bureaucracy, Chilean governance acquiced carly on remarkable checks ances, even exaggerated ones. The pi ally constrained in the nineteenth centuey: when a budget was voted down, a reformer president committed 5 his will on Congress. Extension of po! lenge ‘hese early features, Presidents can only be elected for one mandate succession, Congress refused to approve travel abroad in the ts alva (Valenzuela 1989). have lifetime tenure and thus acquired her courts, ‘The autonomy of the bureaucracy and accountability of the presi- dene toward peers were neverthetess achieved in a highly untequal soci- ety where ol we enjoyed political access, not unlike eighteenth: remained poos, and only’: the 1950s and 1960s eliminated most vore buying and pork-bacrel Frei Montalva (1964~70}. Even ar favositisen was sill reigning supreme, however, political leaders were » Author's interview with Controller Genesal head of stall, Santiago, May 8, 2014136 Understanding contemporary achievers not in the game of personal ensichment. The last president to run Chile feom the pres palace rather than his own house was in power in the 1950s (the presidential palace aiso and the coin factory) and none has xeturnes Jast conservative president before the d himself the son of a previous president, by himself from his office in La Moneda palace to his apartment: since then, Chilean presidents have only a summer residence provided by saxpayers.! The elite background of Chile’ culers and tis valuing of ‘austerity in public life seems to have played a role in creating a pattern of presidents behaving moze as chief stakeholders among trustees than of spoilers of common assets. le was already institutional the strong oversight mech- ind other autonomous agencies. of politics was evident in persistent lation scandals related to campaign sl (Valenzuela 1977). To rein in part petitive Congcess, one which hael almost wn ts 10 legis- late benefits for key constituencies (e.g, pension benefits) (Chumacero ct al, 2007: 16},a consticutional reform was passed ar the end of Frei ‘Montalya's administration expanding the budgetary authority of the and effectively depriving Congress of the prerogative of leis- che areas of social security, taxation, wages, and fiscal budget. argeted spending aimed at specific constituencies were also Forbidden, thus consolidating che autonomy of the stare toward pr vate interest (Montecinos 2003). The need to rationalize expenses was, however, older, and the trend had begun already under Jorge Alessandei to replace clientelistic allocations rsalistic welfare: but Frei Montalva gave free rein to the {he was a Christian democrat intellectual, a reader of Jacques Mart (Gazmuri et af. 2000) and enjoyed enough of a majority (*p: to be able to leave behind the old transactions to build a in the Congress and move ahead with reducing inequal and increasing rationalization of expenditure (Valenzuela 1977). He ‘o promoted other reforms, including land reform and strong incea- * Authors imerview wich Alfredo Joignant, conscmporary historian, Santiago. Chile and Urieguay 137 4 \ 1 1 Posty 2 ORG quakty of gavemmnt index PPP PEP PEEP SPOS eacaatyesoonneriion Pay 2 for instance) and community-level collective action. The growth of the state under Communist Salvador Allende and the unprecedented power discretion of some army circles under dictacor Pinochet fed +0 ication of rent opportunities again, but Chile never returned (Valenzuela 1977}, Consequently particular allocations in loca! politics died up, and the tendency to promote universalistic welfare allocation rather than concentrated, client-based territorial benefits was mostly preserved in the subsequent dictatorship of Pinochet and the democra- ‘ization that followed (sce Figure 5.2} ‘The 1980 constitution passed under Pin (of the budgerary 1 het reaffirmed many ‘that were already present la ropose legislation rel d to wages, pub- lic employment, social security, and other en ment programs.138 contemporary achievers “limited opportunities and mechanisms to the executive branch” in the Congress, theref bargain with and extract concessions F the stare, ‘cases of clear market failures {except for the freedom ement oF monetary ancial openness, which world provide both ited sire of the domestic economy ose of granting stability to those ru {kdwards and Lederman 1998). roduced as an Office far Transparency and ilped foster a change of cul ship, though ie cemains exceptional to the previous twansparency did not seem t0 foster cor passed after 1990, han in c 's Rood governance roots. is an exal of control + Author's interview, Office foe Transparency, Santisge, May 7, 2014 Chile and Uruguay did authori 's success seems to be celated « jon of the economy and the need for a sound financial system ler was created to save money and ra expenses more than to control corrupt panies that offer public services, and the coms other words, ro the fact that resources for corruption gradita s that provided rent opportunities all along, so its achievement is ail the more remarkable, Control of corcuption toa great extent seems to have preceded the economic lib- cralizacion under Pinochet — it has been a declared concern of nearly ‘each presidency and the ceuci inst | arrangements ~ for instance by making the Comptroller accountable t parliament or ereating an anticorruption agency — should be regarded with caution. The strengeh of Chile’ Uruguay beginning of the tw zens strongly endorse tend to be more skeptical of the integrity of pe res, public must is very high in Uruguay.’ Although this (on was interrupted by a military dictatorship between 1973 and most of the time major stakeholders in the country, not upstart unanimous opinion is that people have never joined politics to become rick: there have always been better alternatives for that. However business and police in Uruguay needs ans through several penecations 0! modest museum of previous presitlr ent president, Jose 2 former urban fnobarometer Director Mara Lagos, Santiago, May140 Understanding contemporary achievers guerillero who rales from his private apartment and drives his old VW Beetle to the office. h ‘i ‘Uruguay has a semipresidential system with two fune- a lances,and it has many decades” experi- ning system of checks and. ence of consoci the opposition seats of public companies parliamenta re high court, mer 1s from a lack of effi ciency and expedtency against organized crime, Uruguay has a long history of interest representation, org formation, evidenced by the dense network of ditfere a large well-educated inequality than other La ces and a strength- s however, we find to control of corruption is built on an older history and grounded deep in society. Even before the last democratic ugh patronage was widely ac nevertheless conducive to deficits. For nearly fifty years 's growth was far inferior to the growth of pub- 2014). This fed © Author's imezview with David Altman, Santiago, May 3, 2014. Chile and Ornguay © that the traditional ways of gaining support should give way. A reduction in material resources began dut nport s occurred, and 1990s and aio the 2000s, following. an economic criss. Starting in the carly 1970s, Uruguay's economy became more outward-oriented. But Uruguay is no triumph of the Washington consensus like Chile: quite fh high public employ- respected fiscal con- 1p, and creative tailor-made mal economy and tax evasion, The con n of an independent judiciary also took place early, not long, after cedemocratization, with an agrocment to appoint Supreme Court judges with a two-thirds majority of parliament. Berween 1985 and 2004, the Uruguayan party system seemed reach a new equilibrium, which decisively influenced control of ty dictatorship itself was an actempt to prevent what had been perceived as a radical challenger party from shaking ional rules of the game. The later integration of this pro technological moderne during years of dictatorship had ‘considerably changed the old ciety, and ethical universalism triumphed for practical as as ideological reasons, There was political consensus thata ret to previous rules of the game {which had played their pa 9 and then dictators! fer been pervasive 10 ion of benefits and public services had been the nore ‘The WGI CoC had already placed Uruguay in the “green atea” in 198. Later pi ices were inthe field economy, showing that ation beyond Uruguay's evolution was the nved for fiscal consolidation. Tex simplification, smarc inceatives, and commissioning142 Understanding contemporary achievers private sector led to the most spectacular fer 2000, tax collection largely t growth of collection in the Estonia Estonia has been a re the nineteenth contury. But itis also overperformin modernization goodness of fit. As a champion of liberal y.and e-government, Estonia progressed con: cess order excluded from the vote most of its Russian-speaking residents, result- ing in an unusually homogenous and politically cohesive community ss numbering under 900,000. 4 2004 survey carried out on behalf of che Ministry of Ju that about one-half of respondents did not consider officials to be corruption; about the same proport common: 16 percent of the respondents over: bribes, 24 percent of entrepreneurs claimed tha foc a bribe for performance sector employees claim to have beet The country has seen later became very popula bat tax evasion and the Estonia 143 e-government reforms by neighboring Finland, Bur key successful reform was anticommunism. In barely «wo ye 1992 t0 1994, Laat, a former anticommunist historian who des that the only book on economics he had read before deco! fer at the age of thirty-two was Free to often spoke about corruption and choices prevented Estonia from sl from “plan toctan” as many other postcommunist countries had dor, evolving to erony capitalism rather than ethical universalism. As he ‘any reform which incecases the competitiveness of the economy ice cocrupt behavior” (2007. As part of the same poli policy toward judiciary, opting to replace most of the inherited fresh with newly trained magisteates (the cost: courts were practic~ ally closed for many months), Normative constraints are also hi with public opinion intolerant of bribery (though ze toward other forms of particulacism), an active civil s rment to civil society development is also grea ch dedicated EU funds and a pact committing e rule of law {Kasemets and Lepp 2010} ince February 2000, when the Estonian Parliament passed ‘guaramecing the general po} has become one of the mos “gis jemet usage rates are due nor only to the prevalence of household computers, but also t access points countrywide. All schoot the to high rates of internet usage, nearly 90 percent of the population144 Understanding contemporary achievers have mobile telephone subscriptions and more than half have digital cards. Another cornerstone of the control of core Bstonia’s system of public e-procurement (RISO 2012) more recent online public expense tacking system sig a 3s was the exch ‘The cost of Estonia's neolibéral coasensus was af nonckstonian speakers feom the vote — about onesthied of Estonia sopulation ate Russian speakers who settled in the country during the Soviet period. According tothe World Values Sarvey 20 ws greater Support for coflectivism and lower suppor ares yy and democtacy than the Estonian-speaking popu and its by-passing by political decision probably helped Estonia's 15. The country’s achievement is not a minor ¢ same institutional problems as ng the same exi chreats from neighboring Russia. Sloneia the xher prominent EasernEuropern ses 00 re from a far szonger poston (ey early economic integration with neighboring Austria, small uebanized population, far Wigher ns mes and excellent moder goodness of i be a i cued ou in recent years its new capitalism is far more based on rents than prev ously acknowledged (Al Jazeera 2013; Cerni 2013). Botswana considered in the “green area” S RAES or ancora 198, Manparar jnternational also has consistently rated Botswana as the least corrupt fat asean countries included ints Corruption Perceptions Index ‘The county frequent pointed to asa classic asiconruption sess son largely due co its antcorruption agency ~ although th me been recent suggestions that its role might be overrated (Acemo ‘Nonetheless, Botswana's achievements arted out as one of the poorest coum independence from Britain in 1966 its modernization goodness of fit. Since. Borswana, by the Wor and is now overperforming Botswana as that time Botswana seems to have traveled a virruous path in compari son with its neighbor, Zimbabwe, which has descended into social and sconomic collapse. Its structural determinants - geography, urbaniza- sion, and health ~ were no better than those of its neighbors. Bur Votswana’s achievement is quite different from the ethical uni- versatism ideal and che resourcevconstraints model. For one thing, ‘material resources for corruption are high, given the countey’s large Potential rents. Diamond reserves are the most important resouce in the country (successfully exploited through a monopoly joindly oper- ated by the state-run Debswana and the privare company De Beers), although cases of corruption in this sector are rare (Bertelsmana Seiffung 2003). Borswana’s public sector is a heavy weight in the economy. Government consumption spending as a share of GDP has remained above 20 percent since 1980 and reached 24 percent in 2009 ~ the highest level among the six contemporary achievers exam- for azound 45 percent wee the early years following independence, hhas been mosily merit-based (Von though some public positions have been subject to patronage {Johnston 2005), Borswana’s road to positive evolution was started by Seretse Khama, the country’s first president, who pursued a tough stance against corruption (¢g., Adamolekun and Morgan 1999; 592} that ‘quickly carned the country a reputation for “clean” Public resources. Although its pol of competition, among the cases studied here it is the country with the longest uninterrupted democratic regime, Throughout forty-five years of democracy, power concentration and material resources have remained high. Nevertheless, rwo positive aspects can be traced back the foundation of the state: the establishment of an autonomous service, which has been protected from widespread politicization; and the development of an auronomous judiciary. Policy formulation and public spending are, however, some of the ‘most transparent aspects of Botswana's governance system, Bud development plaaning, and prioritizing peocesses are extensively consultative, involving communities, local government structures, non-state actors, and political institutions. The social allocation pros cess is targeted but universalist, as there are specific programs for Poorer sections ofthe population, such as inhabitants of remote areas,146 Understanding contemporary achievers including indigenous San people. All allocations are included in a saving litle room for discretion. Borswana’s [Von Soest 200 and has the power impeached by it (Freedom House 2010}, Botswana is also ane among the countries examined jon: the same political party and ps jower since independence. Although the country has 2 mulkipacty 1 presidents in forcy-five years). Irs cur fan Khama, s-born son of Sir Seretse Khaz, who was the country’s foremost independence leader and who served as 966 to 1980, himself the male descendent of the tradi ‘of corruption, Borswana has nal mone mandates as a be all the features of a berween 2008 and 2013 by former servants. The cabinet is regulacty fi blood relations connections of the president and the Members of the BDP government s ot directors of commercial companies and farming ent the BDP has consistently refused to pass le rad cabinet ministers to declare their assets and econom ined the relationship berwe we that De Beers has fina wing the presidency of Ke tswana is a rare case where ethnic particu and elite capture are formally organized and coexist with the rule of Botswana a7 for instance, only has rep. privace interests. remains as car dealers, m ‘The auconomy therefore relative, Big businesses construction antecprises as Ww fh the culing party and to ing, with party executives directly idu 2014), High-level corrupt ‘One area particularly prone to irregu of state Jand: cases have heen reported of a ess persons and members of che 1993, when newspa- ians (and other members of th the state-owned contribute to in business iSebud erupt periodic: Borswana’s largest public lending ins 1996: 602}, Then-president Masire was a ‘These cases were documented in reports various presid to control achievements and rep- ntations, however, ar president appoints the heads of both offices “The judiciary, on independent and actively reviews has become the main accountability insteu- nts the presi to appoint the highest authorities {Transparency International 2007), Except Botswana's first president, under whose ister committed suicide when suspected ty appointed his brother over more g and the cabinet has far too many blood relatives, again gi impression of an enlightened monarchy rather than of a cepublic. By and large, i¢ sorms that the first generation of postcolonial Icaders put148 Understancding contemporary achievers the country a the right path, allowing legal consteaints to develop, mice should nat be raken for granted under new, less jessive and might become jop-down manner as pre in Botswana has grown. Ich sector and depend- 's not active in a watchdog role (Transparency sedom of the press has shown signs of deteri to “partly free” press in 2005, “Amiong the reasons cit ver state-owned media and the governme ¢s, both through formal and informal channels. and private media 0 August 2012 the 1 of information bill a problem for journal sing services, manage- ies and local government services, mining, aod mn of student loans and the are spared, Party mem- party have become an importat loans, opportunities, and government ey of media reports on corruption-rclated offi- but also a genvine perception, ‘Acemogla, devclopment succes Hes of 7.7 percent. ic prosperity made many things easier, and presently overperforming on governance. However, it would be diffi Taiwan and South Korea 49 that Botswana's equilibsium is based on ethical universalism, even if the choice of the pol live after independence was to organize the state as an autonomous defender of property rights, cather than as a source of continuous spoils (Von Soest 200: is thus close 10 the historical Danish path, al authority and top-down reforms by ‘Taiwan and South Korea Corruption scholars have for mam in sub-Saharan Africa is so dettimer it scems to have fewer effects. The obvious answes, disc while ago by Barrington Moore Jr (1966) ina book less quoted today, is that at different levels of societal development predatory behavior by clites can meet differenc degrees of resistance. Where there is no from and society ean fight back, rulers will posit a commission on general economic activ yy for their selfenrichment, with fewer detrimental effects on ety ay a whole. This seems ro have been the Asian situation for the most part, but uperior to the world’s most disenf poor countries was nor really a measure of Asian performance. Only recently has demand for good governance stepped up in importance ig Japan, was 1s, South Korea and Taiwan, and countries ‘such as Indonesia and India have started more important anticorrup- jon activities. The average across the region is personified by China, © poorly, though in comparison. bly well. The ever, do twi f them are atypi therefore will aot be discussed here (Hong Kong, Singapor and Brunei Darussala atypical, we are left with only two tecent performers, control of corruption status: Tai and South Korea.150 Understanding contemporary achievers rt raiwan's goodness of fit co the modernization model is very good; whore i ble perfoemer: a well-developed, edue: a only less than ideal. The country underwent a successful and peace! democracy, although iis somewhat controversial whether is transition also led toa better governance Ine bee ndieated athe os pressing political issue in many opin 5 on party finance and pol = period of Kuorsinzang (KMT} rule, corruption was pr y institutionalized, and the party was able to build real-estate an business empire (Hsueh 2007: 11). Companies bribed the govern- contracts. During wed the factor in teeminating, by the KMT for over the 1990s were a -party authoritarian regime cont fifty years (Yu ef af. 2003 In May 2000, when s KMT peacefully curned over the presi an was regarded as a textbook case of sue the short to an increase in corrupt exchanges, as pol nd more expensive and there was greater dk Despite swings in govern semipreside dent, who apy significant power discretion 1 minister without approval by th appoint the members: idicial Yuan [Taiwan's Taiwan and South Korea 41 constituted as an independent government branch, has auditing and investigative powers and aiso exercises the function of an ombudsman Taiwan are a class in theit cd, and paid. Unlike other young own right ~ sell educated, pro democracies, in mittee, established in 1925 ‘when China became a republic. Taiwanese civil servants have among the most secuse jobs in the world, with a monthiy pens cent preferes bank interest, giving them the world’s income replacement ratio (the percentage of the working income needed to maintain a desited standard of living in ceticement (Cho-shui 2011; Khan 1998). Taiwan began bui 1990s, The 1993 jon framework in the Disclosure Act required y high-level officials and elected representa- ifeung 2009}; tater, President Chen (2000-08) pushed the issue to the forefront of the political ayenda and, shortly after winning che 2000 elections, mapped out an “Action the ‘The main factor motivating the key reforms was the chang ruling party after the DPP' electoral success. AS the previous regime ‘was based on engrained political corcuption the suzvivai of the DPP depended to a large extenc on a change to the rules of the game: anti= core 5 were the best way to enfeeble their political oppo- nents (GBbel 201 4), The ceackdowns on bribery, embezzlement, money Jhundering, and vote-buying removed many top bureaucrats who had beet loyal to the previous regime. Popular support to remove what was perceived as a corrupt regime preceded che DPP's anticorruption, agenda and indead co But_ democratization neously brought more resources for corruption and geearer normative constraints. The latter are appar. gat in the popular movement against President Chen and the low trust in government institutions even alter the implementation ofstanding contemporary achievers 152 ui country. Legal ‘wich harsh pen- larism ay a regime is finished and implemented. Asia for economists studying Latin America. Like Uruguay, Kore: in all four dimensions of the equi South Korea development, as, well as for external support for ies improverents in governan economists rend ta neglect (You 2012). ; South Korea gradually achieved important milestones, including a ‘major land redistribution (1950s}, the creation of a meritocratic and Kim Dae-Jung (1998-2002), the firse opposition leader to be el also contributed co the reduction of ma ry has proved its independence since at fans and businessmen on il received light sentences president, whose pol feast 2003, an corruption charges, so co even presi Bordertiners Georgia 183 Woile Kocea’s control of corruption ratings are inferior to the other countries described above, it is the most populous countey among contemporary achievers, which makes its progress even more signif cant. Th " 2008 of former President Roh Moo-Hynn legations shows how strong normative constraints have outh Korea is doing significantly better than regional aver- components of the moderaization model: es better educated by half, and more ethni ‘neous than many other countries in the region. Si ease in educated people and a rise in the middle class; second, protest movernents in the ci especial jumping over iow global rankings in just a ced this label $0 percent and Azerbaijan to surpass Greece and Hung, EU members, although it remains well below the oth presented here, Georgia’s improvement st the 2003 Rose Revolution, and since then fat fe county has success managed to reduce petty corruption through several icorcupt h-profile campaigns, including the prosecution of senior coreupt154 Understanding contemporary achievers rs: Georgia « officials, police reform, dere business enviconment, as well as public sector reform. If Geor cher than six en our cecoded 1~ moderniza tactics did, howeves, allow the government to quickly mobilize enough revenue to repay the outstanding arrears in pensions and wage sents for pablic sector employ of the top executives who impler ¢ reforms were se for power abuse by their successors, who had accused Saakas author fot managed to cl Bank scale despite overs 1a determinants 5 tert 1990s fired and eeplaced by a well-equipped and newly trained pateol police low USS4,000 per capil Anathema achiren san aloe thoes cor 5 ‘he the mademization model pos kupt system oF university admissions, which became centralized an ee pe TE ee given corruption’: deep roots standardized for the fst time in 2008, replacing the previous policy inthe country’s Soviet past: In Sovier Georgia there was no distibu- of individual university admissions, Safeguards were putin place to tion of firancial and material resources without bribery and nepo- snake corcuption or fav possible (Anderson and Gray sm, According to non-market rules of life and business, personal that were introduced were mosely aimed reducing excessive red tape. These strides forward are the greatest. What fa oduction of “one-stop shops” rudulent parliamentary eles cratic procedures, but also dean reduced the opportunities for corruption. Authorities also promoted the principle of onscat,” forcing time limits for approvals from the administration on the penalty of automatic granting of requests in case of delays, only four procedures and four days necessary, Tbilisi is the most expos 8 city in the world for foreigners Critics: argue that rampan replaced wi explains its success? siya ee 1 2003, mass protests prompted by alted in ent Eduard Shevardnadze, a cevamp of the govern members. Those indicted were summoned to the prowsutors ffs for quesioning, charged srruption and embezzlement, aad arrested. They were aske " dollars, but were then released without ‘ownership and the censorship/scl-censorship of journalists. Howe’ reveled in the spectacle (these proceedings were often broadcast live on the yielding of power by reform champion President Saakash national television) and most of the public felt vindicated, 1 cates that he was a genuine reformer who might have cut some corners ‘was actually in brcach of Georgian legislation. These heavy to yet things done and not a new dictator in the making,156 Understanding contemporary achievers (On the legal constraints fron, Georgia has everything ~ a Freedom of Information Act, an ombudsman, and whistleblower protection legislation ~ and has ratified the UNCAC. Following Saakashvil rion as president, the anticorruption council created in 2001 undee former President Shevardnadze disbanded itself, and no new national steategy has replaced it. The eountcy had made its improvements with- out the usual plan cherished by the anticorruption industry, one that so frequently remains only on paper. Goorgia’s improvements can be attribuced to changes in all four dimensions of the equilibrium model. Resources were deastically cur tailed by reductions in red tape, and policies promoting economic lib- eralization were even steonger than those enacted in Estonia. Legal and seated a new equilibrium through a "big bang” change ~ the 2003 Rose Revolution ~ followed by top-down reforms, Civil society played a large role in bringing down the Shevardnadze regime, but was later weakened by its participation in Saakashvili's government (Tsitsishvili 2010). Independent moniters of the govera- ‘ment existed under all regimes, although the third sector overall is still quite weale in fighting corruption in Georgia, ‘The few civil soci- ety organizations dedicated to fighting corruption engage in heroic deeds with litle volunteer support from the grass roots and pract ro funding from Georgia's business sector, The judiciary remains the most problematic part. As it showed in particular after the second change of regime the interference of the judiciary is more to be feared han welcomed. Saakashvili himself has been harassed after his peace ful departure from power in November 2013 for the accidental death of an ally nearly ten years earlier. The judiciary is not independent I contsol co exercise impartial legal constraints normative const ‘enough from pol Discussion ‘The contemporary achievers differ from one another on many dimten- sions (see Table 5.1 ng the number of years since independence and current trends in control of cozcuption (e.g., the most advanced among them have been stagnating or even regressing slightly}. The two borderline cases — South Korea and Georgia ~ are more advanced than thee neighbors and show a positive trend toward control of cor- ruption, bur some challenges remain, in particular for Georgia die to its less-than-ideal fis to the model. Borswana is a remarkable achiever ‘Table 5.1 Contemporary achievers ata glance “Goodness of fie" to mode! é Sequence Evolution Type Country ‘Control of corruption Stagnamt Southern African Cuscoms Union Foor Gradual Borwana preceded genie pluralism Conzrot of coreuption Economic evisis prior frst good Stagrane Gradual Chile governance reforms; Amictican, model emulation EV accession; preceded pluralism Pluralism preceded control of Big-bang, Good Estonia Scandinavien emulation comruption Pluralism preceded control of revolution, Big-bang, Some diffusion sexoss former Stagnant Poor Georgia republics; American emvlation corruption Negative Austrian emulation, investment Control of corruption Excelent Gradual Slovenia preceded pluratism Patial conteal of corruption, ‘American and IMF conditionaity; Japanese eeaulation| Positive Good South Korea Gradual, preceded pluralism Partial control of corruption confrontational Gradual, Important American assistance in Seagnane Good Taiwan, the past European influence alays preceded pluralisn: Con:rol of corruption eonfronrational Gradual Positive Good highest, Uruguay important, European descent preceded pluralism literacy rate in Ladin America)158 Understarnding contemporary achievers ha more narrowly defined ‘considering the sub-Sabarai ism, which remains far off control of corruption than et ‘each step forward foughe back, leading to av, curve = though many individuals contributed & no one elear leader having, lec the evolution co open acces Uruguay transitioned gradually as well, but with less confrontation, consolidation, External factors are strong in other cases 28 wi Korea and Taiwan were both oa the front lines of che Cold War and ites in charge of the economy were frequé States. Uruguay had 2 Swiss model and its population descends from mmigeaats from vacious European counties, a far more diverse back- mn other Latin Ama in South Korea, where a ci Finnish control of cocrup: Scandinavian countries. Across the contemporary achievers, control of corruption was reached by changes in no less than thee dimensions of the model. The nstraints developed gradually and of power and other goo placed achiever peo} trol of corruption and do no tutional equipment in these co fas an agency which prosecutes anticorruption, but no Freedom of 1 Act; all other countries have FOIAs. Some, however, were late and therefore have had a teol of corruption measues thus fa institutions carry some weight , Unnguay, Taiwan, and Sout Discussion 159 Korea. Having an autonomous, merit service prior to democratization was a pc cases that had to do wsed, and prosperous ci factor in al “Taiwan case. The seven cases examined here have presi residential, oF parliamentary systems (Estonia, Uruguay) lctoral systems and legislature organization, dif- jon (party financing is actually and varying types of juslicial mostly due to strong account system of checks and balances, but these states are noi federal Press freedom and col n by civil society played a con- siderable role in all these cow eept Botswana but includ- ieorgia, where civil society's roots are South Korea and Estonia, the role of ci zreater than is its former civil Media pressure is a key factor, inability problems in some -al leaders played a role everywhere, except in has been so gradual ~ how: anticorruption, one Contemporary achievers have managed their performance ac differ- speeds and by different paths, but all cases confirm that progress cf involving all four control of corruption over decades of gradual build-up, others mixed160 Understanding contemporary achievers of the selected adual evolution ¥ of law and sound control of corruption on el ‘a previous authori sm and competition for power and then tu ca reasonable ft to and Uruguay, ia — quite remarkable indeed. A stake in and commit Lor professional, was le crises and Asian cases) occasionally provide a move to another norm of governance, and popular demand hel snance order, the sustai ‘a whole book in themselves. But the eases do teach us important les- sons even at a more Su reading. All these cases outperform theit 1 governance modef, sometimes fanged the governance in their cous imacy and prosperity. Other reasons a the history of these up of 6 Domestic collective action capacity ‘What are normative constraints? society, and critical citizens are strong enough.’ We also f jons in and of themselves do not offer sufficient empowerment type of governance context has developed and stal rested or unable to change this many of chem contribute to it. jve constraints come from? In his book Development as Freedom, Amartya iduals, in order to achieve cheir potent public resources eq) rty docs not necessarily rest based on ethical universalism. Elites, cla and jroups of every \ A reduced version of this, Chapter 6, was published as Mungiv-Pippidi (20136). 16162 Domestic colleetive action capacity ‘moral values,” “the media," years, awareness o} portance of the development community, yet many 0 lective action remain disconnected both another. In an attempt and later proposed 2002 the United Ns Governance Center. Yet only dhe United States Agency for International Development (USAID) and later the US Millennium Challenge Corporatio jons such as George Soros’ ‘Open Society iery groups and, co a lesser exten good governance prolif indred countries have helped tives media as main prior «of Transparene; nity, though resul Vacious streams ponents of normati scademic fiterature mention four distinct com- e constraints Values: a prev I universalism based on ess and honesty idespread habit of engaging in formal ind shared interests, purposes, and values; tetwork of voluntary assoc understanding of the term, but Some virtuous com mn of these four factors seems to enable soci- ies to overcome competing tendencies toward violence, cronyist, and social domination and to generate normative constraints that empower ethical universalism. All four elements are necessary and no smaller What are normative constrains? 163 individuals are also ief to make ita reality. This does not necessarily require an absolute majority, but rather a majority of active public opinion, including a fraction of the elite. ‘The only source for examining the development of normative consteaints rem quot and worth, France at the turn the ewentieth century, and the extension of the franchise in early ‘eenth-century Britain. The story of Captain Alfred Dreyfus, the Jewish scapegoat anjustly sentenced by a military court to life imprisonment for treason is well known, as is F ¢ denunciation of the trial, “J*Accuse, the front page of the newspaper L’Aurore. Less extensive publi writers such as Marcel Proust ng 1,482 eeleb- rity signatures, in slaude Monet, against che Dreyfus verdiet. Zola’s manifesto on the “ci iginal trial and eight years ia was perfectly aware that fersalism, not a simple miscatciage of (0 poison the small and the stice should be universal and impart jotic appeals o¢ the background of the defendant, Many new democracies have not yet had their Dreyfus affair.164 capacity While the Dreyfus ease is often cited as an example of the power of iss and was sometimes ible and unauchor- igled out for blame those Moreover, there was continuous advocacy for fu hise to the middle class and for abolishing royal extending the fra sinecures. __ For she reform t0 succeed, the Crown and the existing voters (the mandate for ¢ a change that I development driven by an elite steategie | revolution was actually far more bottom-up and top-down elements were present. action at the grass-roots level ion avross the corruption ~ hence the need t0 about, or not, ethical universalism as the main governance norm. What are normative constraints? 165 In his depiction of American society Alexis de Tocqueville was the mptoves assoriations for ei He worried that as men become more eq) to be feared and the costs of aggregating interes he saw the newspapers’ main role urposes. ‘And he was perfectly aware of the difficulties of collective action and the need of practice for successful aggregation to occur: ity that have been widely debated in the relationship between engaged becomes easier to use the social capital thexe produce for any purpose of collective action. Tocqueville acknowledged that people n organize for any purpose, good oF bad, but he viewed the capacity166 Domestic collective action capacity eniably positive and as the only possible path to good governance. A society capable of collective action is capable its most violent oF selfish ies for normative resteaints in fact limi corruption, we sh 4 positive relationship beeween their presence and success ling coreuption. Despite the imperfections have enough material to test Tacquevi of normative constraints in a model of control of corruption. Th membership on the to untary association significant at both the national and individuat experience of democracy, education, age, and propensity toward corzuption was fur of i Collective action and control of corruption 167 development {HDI with the other proxies for normative constraints: freedom (explains 67 percent), the number of internet connections (71 percent denomination of the press an nants of control of corruption that for hnman development). Protestantism is one of the oldest references x0 the role of normative constraints in creating an ethical society and government. It has been tested often and has proven significant in foday many nominally Procestant countries are de facto secular, while quite a fee non-Protestant countries tackle corruption qu The inthe lence of Protestantism seems to stem from its egé thos, which may have worked indirectly to support a general orientation The test of ci wunber of cases 2007: 67-125; Grimes 2008}. & panel regression using eo data fom the [CRG as the dependent variable (see Table 6.1) also confirms the importance of freedom 0}Random effec: =p0216"** (0.00238), 1346"** (0.356) 27age (0.289) 1077 433 047 om earn 0.00952" (0.00339) os.ag7 #338055 4 candard eecors in parenthess}. Sigaficant pe0.01. Fixed effeers 0014s (0.00356), a2agere 156) 1077 133 oat 0.02 2 ~.0239"** (0.00286) Do6ter 0.354) 2381" (0.294) 1077 133 O48 Pooled OLS -oon7* (0.00345) 4.190¢°* (0.194) 1,077 33 oat 4} and 2 show poled OLS eestlis with clistr sobs sasndsed errors in parentheses. Models 3 and report Exed effects estat eesults, wile model shows random effects estimates Hausesan tes suggests random eects to be biased, Significance lees: * pe.1, ** pe 2 Valoes eange betwee (and 6 (higher values indicat les isk of eoeuption) Values zange From 0 bese) 0 400 (worst) = Overall Within ‘Hausman test (Chit based) ‘Time series crosssertion estimations for yearly data herween 2000 and 2008, using the corruption measure cf + Human Development Index (UNDP, vales range from 9 (low development) ro 1 ¢high development. “Table 6.1 Effect of press freedom on cormeption: panel regression Risk Guide as the dependent variable, Models Freedom of the press (Freedom House)* Number of observations Number of countries ICRG tisk of corruption Dependent variable HD Constant Civil society as watchdog 169 freedom and corruption using ethnic fractionalization as an instcu- ‘ment seems to confirm earlier findings that causality runs from press freedom to corruption, but when controlling for GDP the instrument loses significance, The normative constraints proved also fairly robust in more compiex variants of the model, when tested alongside plucal- ism, natural resources, ethnic fractionalization, independence of the judiciary, and various economic policies, as shown in Chapter 4 of this book. Altogether, this group of variables seem to measure one latent variable, normative constraints: the collective action capability of a society to censor the behavior of its powerful elites. “The data shows that in counteies where control of corruption works we find aumerous political and civic associations as well as an active media, and an increase in the freedom of the media in time leads to {greater control of corruption, To understand the mechauiism of their influence, however, and in particular how they could he made to contribute to betier governance in other count in-depth look at both ci countries. 8, we need a more society and the media in the problematic Gi society as watchdog Alongside statistics, the historical examples provide support for the theory, but introduce co | London in the 1830s, Paris at the end of the nineteenth century, and New York after the Civ War likely had higher capacities for collective action in support of eth. ical universalism a8 a main governance norm than many developing countries do today. A modern-day Tocqueville visiting capitals such as Accra, Kiev, Manila, New Delhi, or Moscow, where anticorraption efforts have been raging for many years now, would not remark on the deasity of voluntary associations, but rather on the isolation, exhaus- tion, and underfunding of a few dedicated groups and individuals ‘who have not managed to reach a critical mass. Prague and Sao Paolo might do better, But the widespread picture shows insufficient collect- ive action alongside occasional outbursts and over-reliance on exter- nnal donors due to insufficient support (cash and volunteer work} from a country’s ovn society. People from Brazil to Sofia who cook to the streets in 2012-13 in a remarkable wave of new anticortuption “revo- lusions” cefused to be represented by established NGOs and political ‘oppositions, considering them all part of a coxtupt esta170 Domestic collective action capacity ‘without any organization such spontaneous movernents cannot bring about sustainable change. Tn recent years, atcempts were made co document the role of civil soci- ety in promoting good governance {Armstrong 2007; Boussard 2003; irk 2003; White 1995). Sucvessful case and it is too early to assess them (although the number of Facebook users per country does associate with control of corruption). There are some commons .s democratization in South Korea, Central Europe, and the Bal inated ‘ety movements that had started as grass-roots protests and loosely organized social movements and evolved after transition into more institutionalized and specialized NGO communities, These anticorruption actions span many years and are extremely diverse. Two comparable short-term campaigns allow us trate the mechanism triggering the change in norms. ism, Alist was researched aad compiled with 114 natacs of politicians ‘who were found unsuitable for nomination. AA similar situation happened in Romania in 2004, when the EU ‘country t0 show commitment to an ‘be invited to sign the EU accession treaty. A small but representative muroalists created the coalition comprised of NGOs and individual face a naming and shaming campaign, The criteria were not very different between the two countries, despite the lack of unaccountable personal fortune. Based on the findings on these points, some candidates were blacklisted. Media not under government con trol closely followed the negotiations and the parties’ responses, society as watchdog am escalating public attention and creating a snowball effect. Pofitical parties that scood ro gain {¢om the campaign cooperated put society in Romania, increasing, the feryone pledge pul cempts t0 obtain co Romania and Kore. cain publicity of any form and legal ba tles during and after the campaigns, as some candidates sued for However, with slightly different sequences between the two countries wursued. In Korea, the parties? reac lacklisting of candidates and forty-six candidates on the blacklist were nominated nonetheless. The lliance kept up the movement and campaigns were made in every dis blacklisted} los- ing elections. Forty-two percent of incumbent parliamer voted out, and 106 new candidates were elected uation went equally affocd to be seen as the only corrupt one, The blackl 200 names ~ more than one-third of the two chambers. Opposition parties had far fewer names on the list than those then in government. ‘The campaign pushed corruption as the main agenda topic and within a few weeks caused the governing, party to Jose its original rwenty escent lead in opinion polls. The party leader who fully cleansed his party of controversial people won the presidential elections, for which he had started as an outsider. About a hundred candidates did not ‘make it to parliament: half were retired by their own parties and halé were tejected by voters. One of the civil society leaders became mi Despite these “big bang” successes, preceded by a general feeling personal frustration due to corruption from the general population — perception of r the next years. The two “big bangs” worked in similar ways, helped by the media and extensive mobilization due to elections, By 2011, how- eves, the feeling of sustainable accomplishment was somewhat geeater in Korea than in Romania, despite the latcer’s successful EU admission. igh in the shore run the two campaigns look nearly identical, the172 Domestic collective action capacity level of civil society in the two vountries is very different. According to and Korea's is 48 ‘The success of these ewo campaigns, once significant co managed to cooperate and of media, is worth stad) to sustain once results are achieved 1997} offered the formula that successful a corruption programs are dependent on 18 a significant mumber of projects. Few resemble the geass-roots-inten: society progcams dedicated to good governance in the twenty he Berlin Wall. A research team in an Oper Civil sucioty as watchdog 173 projects compese the rest. About hal ect cooperation with governments at various level spiny esnigal tony over 10 peraatot wanes, Toes tt other half was not made of projects exposing vorrupt activities or prac- tices: less than 10 percent of “engaged” projects dealt with diselosi corruption, and in some countries such proj Most proj and advocacy; and their mose frequent the adoption of legislation, the writing of a repoct, ot the organization of training or conferences, Such projects can only be judged by output indicators (the number of investigative journal ‘was not sought from the design stage (such as, broader than just NGOs. Invol in NGO projects, not just as dissemination agents but as full-lledged participants, exponentially increases the effectiveness of such pro- jects. The more diverse the coalition, the moze elfective the project becomes: those coalitions including NGOs that each have a specific The webpage wow: goremance projets 2008-10, corruption eu and its repository of civil society good product ofthis esearch condacted by the author incapacity 174 Domestic collective act task {not just formal, nominal coalitions} as well as unions and other society actors sev to work dest. Thirdly, the most successful coalitions were not set up by donors, but were grass-roots endeae- ors sponsored by donors, and al thar managed to unearth best, far managed to create an integrity “ma 1 had to compete ro brand themselves as noncorrupt exposure, Hastern Europe had a few suc all with a 3 impact or whose impact cannot be « insufficient concreteness of anticorru projects dealt with corr id only a handful circetly attacked corruption i a specific orga f goverament (see also Spector 2005; Tisne and basic assump s to be that corruption is, ike a voleanie ery ‘group's ~ best interest. Ii for instance, the « Jet and the risk would magically far feom what donor-sponsored anticorra wrong. To challenge corcuption one must understand how it works in ity specific environment. Importing objectives from anticorrup- ion efforts in developed countries that do not suit postcommuni Civil society as watchdog 1s countries was 2 common ercos In almost every country projects existed inspired by the d only a marginal problem in posteommuni reat problems such as favoritism of specific firms, and capture of admi by the governing pasty. Both che bal sucvey, and Furobarometer, an FU survey, ask questi perceived efficiency of anticorruption government are a8 « rule viewed critically in countries where co: Wve the The third problem with many anticorruption projects stems from the confusion between the «wo roles of civil society. This is a widespeead h as Transparency fncernational le consultancy for governments and companies, with the increasing number of Laws and regulations (adopted after their own lobbying). Having the same organization working as a watchdog of government and business and as a deliverer of services for the same government raises conflict of interest issues. If ci ir is supposed tom role and a model is copied by many smaller organizations who se icorrup- tion a sousce of funding and does noc contribute to the development of normative constraints. If we take Rwanda, for instance, which has progressed considerably on paper due to top-down reforms in recent years, but still has a president elected with over 90 percent of the vote, becomes clear that civil society needs to play the role of a real watch: dog, not merely an adviser to govern ‘The fourth caregory of problems is the insufficient domestic sup- port enjoyed by engaged civil society, Watchdog NGOs are often an isolated vanguacd ina society with low association capacity that does really follow behind. Most types of watchdog activities in Eastern Europe were supported by external donors. EU funds do not target ation with the government: hence the backslide on corruption scores tea few new Eastern European member countries after EU acces- when most external funds disappeared and comtribution fom176 Domestic collective action capact 10 best) ‘Conta of couuption (2012) (wooded 1 ciation between civil soviety and contro! of cotruption in Sunree- Worldwide Goversance Indicators and Freedon Hoxse repress. This is a vicious circle; if the main profiters of good gov ‘enance (competitive companies) do not fund such acti unclear whence the money should come. Volunteer work is alsa searce tries of the former Suvict Union fare much worse, They have almost ‘no associations of any kind and mostly authoritarian governments, Media as watchdog of ethical iversalism Ww re 6.1 shows. As watchdog ties, we might see ovcasional outbursts of campaign. ists harasyed by the same 1s because such vanguards have strayed too far from the passive society behind them. Media as watchdog of ethical universalism interests to manifest them ‘This alone is insufficient, however, because such could simply promote private inte ‘ims of autonomy, is connected with the idea that journalists serve a public interest that transcends the interests of particular political parties, owners, and social . (2002; 185-86). Therefore, the media mast be mor. task, Why, in a society where par ttle of the game, would the media be iniversalism, of, in media jargo newspapers sided with the establishment, Some were openly nation- and anti-Semitic: opinions are diverse in a sociery, and most are private and cater to groups of different persuasions. If178. Domestic collective action capacity American journal wajority even in Europe, Quite the ity” could once be defined as a moral ideal, a set of ices, and an observable pattern of news w is of newspapers and news broadcasts. The few media systesns ity is the dominant norm have developed it very grad- jecoment or partisan subst ng opponents and promoting propaganda offered by one of New York City the “Ring” of corruy local Democratic Party and who controlled the 1¢ help of the media, it was tle media long. Most New York newspapers media to become professional, through a mixture of economic incen- tives and legal repression. Media as watebdog of ethical universalism 179 fown papers (Miller 1996) This PR disguised wi co many me nany countries of the world, Alena Ledeneva sketches lar picture of the post-1989 Russian media. She describes a landscape where the main goal of most media is not information, fbut public relations, particularly “black and gray” PR, ce quently amounts t0 disinformation. This is the source of a whole array of pathologies that she de mation campaigns, “kompron and so on, (Ledeneva 2006) tare” (Mungiu:Pippidi 201 vate interests and not objectivity shape the content. The di grounded in the American model of media, relying on the presump- in nead some clear features. For instance, we would find far more media sourees than the advertising market would prediee in such coun- mostly unprofitable but heavily subsidized by € ‘The reason has a business model of ther cypes We would not find Fally functioning. m private sources would not a, for instance; and rax breaks, monopolies, adv ing, or other forms of subsidy distorting the media market would ty of sources exist, as o separation would be found berween the business of media, politics, and business more generally, and the most powerful would own me et directly or through intermediates. In this land- scape, one would be hard pressed to find an investor who makes180 Domestic collect om capacity 4 profit from media alone. More frequencdy, one finds owning media in order to promote their other business or polit- body, which is wea to the presence agents accountabl to oppose very effect mercenaries or be organized to cover Jeremy Pope ack nalism professi proportions: such as Indonesia, are also powerful disincentive to explore misdecds in bigh variovs other countries, pay-olis. This creates places. Pope 2000: 120} ‘The existence of systemic corruption greatly perverts the media sys- sem, and its anal asic terms of the repressive action of govern- 10 media capture either . State subsidies, debt bai is: traded in exchang 2 The gov emment once control resorts to less overt powerful actors, pervert as well as repress try stands on corruption we can predict how its media wo ate, not only the other way around. Again in the words of Hallin and Papathanassopoulos (2002: 18 Media as watchdog of ethical universalism 1 mn tends 10 break down the autonomy of so 4s no exception. It forces the logic logics ~ of parcy 3t breaks down the horizontal solid social groups. institutions, and to merge with -ge for inscance. And iss as it does of other The exi dence shows the reciprocal impact of it report A Free Press fs Bad News for ‘Weder (2003} found evidence nificant relationship between more press freedom and less cor- ruption in a large cross-section of cou 25) and suggested that the direction of causation runs from ‘ionship was weak in authoritarian countries and nonexistent in mediate categories: only consolidated democracy displayed a 1 and strong association {Norris 2010}, A large survey on media ssalso warned thac the mesia the status quo in govern- Jess and less the main cause aff L 2001}, By and large, how can a media landscape be any better th the g ate of marker relations in a society? In their classic For Theories of the Press, Sicbect, Peterson, and Schramm (1956: 8 may try to combat itt lect the state of affair does, that state of affairs, although182 Domestic callective indeed more widespread than media scholars acknowledge, a form of pluralism in its own right which should not be uurnalists would only report cocruprion ir own, and objectivity would result from an ia can be the main account. ‘comes to countering cor- wd rule of sgative social capital civil free media, and corr ruption. As the c Jaw has to impo: ‘The unilateral perpetrator model, with the government as culprie the victim, relies on the prestimy media is by about the world past the fourth wave of tion, when the number of plain auchoritarian counteies ng influence and favors is suboptim: um of eoreupt press ce the virtuous «i ly across countries; s Can collective action be fostered? 183, Postcommunist by external fac pean media, for instance, was greatly in three ways: firstly, by receiving an acvessible bout they are an asset. The problem is the lack of rronomous media outlets to pobli ruption campaigns. Onl seem increasing ers fight ommanized crime more than cstal they also need support. Very few donor programs cater for for ensure the survival of such outlets, which cannot be econor viable in countries dominated by particularism, Training inves ‘own daca gathering problems. or outsiders how to obcain and check theie data. What is solely miss- is support for dissemination and widespread watchdog activity, by ‘own mobile phones and a developmental perspective helps us to understand why the odds184 Domestic collective action capacity ive noted, clicntelism is fre- tm associations docs noe the development of advance from par incerests sted by particularism, i is frequently more con- ry t0 accede ¢0 the privileged group oF to versalism would threaten the existing patrons who would fight against such has passed able chiefly because they are rare; they have become routine 2 handful of y to reach the optimal eq that favors governance reach the permanent capability to exercise normative constraints, and not be forced t0 rely solely accountability provided by s. Otherwise, ever: when successful “big hangs” © end up behaving be closed again. Orange ~ phifanchropists, journalists, a few barcisters. Thus, good governance was brought about not by the most disenfranchised and Can collective action be fostered? 185 powerless, but by merchants, politicians, and intellectuals who engi in collective action against privilege and in favor of faie compe and a ievel playing field, an ¢ of inducing a sense of efficacy that will mol favor of changing the rules of the game scale communities have aleeady been mergect into unwieldy cities and large states, an overextension Adam Fergus saw as a main cause of weakening civil society and rising corruption I mass — helped grea become very visible all around the world 2010 in both erisis-hit ie demand for good governance g-term collective action in favor of new rules of the game is bound to vary greatly acto counties, by7 International agency and its anticorruption impact ins of gener: chapter explored the domestic mech: Bur international factors conditionality in this regacd. Is the create serious constr ing, the national level? The eq fhis book is reached ~ or not reached ~ there ont het on cesources oF constraints, needs to be defined as such. Can exten 1m ~ for the worst, but es influence on governance o' rade or development ‘onal community to support nal actors cepresented by the inter 1989 were well earned and seemed to ind pressure (in the form of economic a ight 3 far less ambitious side goal. But while evidence existed that externals can play some role ay a country (Huntington 1993), 186 International agency and its anticorruption impact 187 accounts sounded the alarm, however, that arcas such as justice and I donors are artempting to change governance orders after having tried ~ and, to some extent, succeeded ~ with political regimes. Bur can governance contexts, understood in verms of chis interplay of formal and infor tutions balancing one another to dctecmin resontces, be affected at alf from the out enced by donoes? What specific strategies have donors developed to push the agenda of anticorruption to recipient countries (for instance, tying foreign assistance to improvements in governance) and impact did they manage t0 achieve? The fourth cl hhas already examined the performance of af 0 across countries, finding tha themselves: this chapter presents evidence of ext corruption, focusing on thrce aggregated secs of transforma leniaim Challenge Corpora ct offering countries grants to improve their control of ing {among others); the Furopean Neighbourhood Policy, leditecranean and Eastern European coune a states, international organizations, and a range of non-state actors, such as multinational companies and civil society organizati agreements have co address coraption in three wa over lective action problems shat might efforts to inercase international constraints to cornu developing and codifyin by promoting governance norins at the nat vel (Johnson et al. 2012)."The key incernational ageeements with regard to anticorruption are:national agency anu its anticarreption impact ro improve transparency and reduce countries; ‘© the 2003 United Nations Convention Ags ook Islands, the Stare of Palestine, and the European and extended where the all greatly across theve vidual agency {2s felony had been national ac the case of the Seeing the geeat effort thar went into legislating, rt in assessing results. However, very h work undertakea on th legal framework or extern: cratly (Johnsan et af. 2012; Mungiu- {Cate and Oouthwaite 2008; Cuervo-Cazurrs 2006, 2008; ull 2003-2004} seems ro suggest that intecnational bribery aers ial couper- es, with Picci (2014) found US cours st cases are filed from more corrupt counties. But forcement, ev ibutes to control of corruption firms are based than ia the situation when enforcement exists, mote in the developed country in International agency and its anticorruption impact 189 proach, researchers also of strong anticbribery fhe OECD Aati-Bribery Conver to convert wrance game, thus forecasting that enforcen bbe weak (Tarullo 2003-2004). Signatories of the OECD Conve on Combating Bribery of Foreign I Is in business transactions p: evider that enforcement of laws against bribery abroad mi by discouraging business engagement in notoriously corrupt cour : prudat invescors have always invested more in developed a noncorrupt countries. In shor ay reduce resources for corruption, atemnational firms ace in any way affected. Seeing thar a country like China with low ci trol of corruption is also a major foreign investor and can ea domestic resources to ceplace any affected Western companies and that recent years, control of eo ‘The impact of other conventions where ensured only. by peer-review mechanisms (UNCAC, GRECO) is also seldom meas- ured. In a recent paper, David-Barrett and Okamura (2013) found some evidence that jes that implemented the EITL standards antracted additional aid, and they argue that governments see FITI membership as a way of buikling a reputation for seeking é work. te \s Convention Against Corruption (UNCAC} con which most hopes are pinned, as it is the world’s first global, com- of a specific internation bor amidst nego ‘Transnational Organized Crime, UNCAC includes provisi itive measures inatization and law enforcement (Chaprei numbers 149) adopted in 201 mechanism for UNCAC implementati or the United Nations Convention Agaijernational agency and its anticorruption impact 190 biggest chal the provisions of UNCA in practive. Considering the reputati for a country that does not adopt evidence that countries adopting UNCAC progressed that had ot (Mungiu-Pippidi ef al, 2011), Once every spted the UNCAC— a moment fast approaching ~ var iminating the value of because too ns had taken place, the before and after test showed no significant improvement, and the regression tests showed no significant association between adopting nd progress toward better control of carruy an effect of poverty, aid should have a postive impac trol of corruption, Conditioning aid on reforms or effec Tor UNCAC would also make a lot of sense. By the end however, academic studies increasingly supported the s had lieele control over how a recipient government jon was created across donor strategies 10 perform in the field of good k 2004; Knack 2001). However, evich ‘exists, showing that forciga aid general ion (Tavares 2003), and its reduct is effect varies by donor country onality is also perplex nality i a set of requirements that must be implemented ‘or to disbursement of a loan or grant (Johnson and Zajone 2006). donors know a country is corrupt, do they also know the recipe tor do they rely on the aforemen tional prescriptions might nor only be too poorly to the local context, but risk being adopted only sulting void of zesponsibility. The conditionality of the internation tutions’ strectural adjustment programs of the 1980s and 1990s for instance, were highly criticized for poor impas and economic growth and blamed fora worseni However, authors such a5 Nicholas van der Walle (2001) have argued that the formal conditions contained in financing agreements were sel- ddom implemented in practice, and this was, in fact, why such reforms were inef main reason y makes governments 1999} argues, is blamed on -apacity to make credible threats to stop. aid in the case of noncompliance (Biedsall 2004}, f donors w breach of promises was rarely tioned by donors such as the World Bank {Mosley et af, 2004; Rade 2005; Svensson 2003), Con: regard the conditions as binding and most donors are reluctant «0 stop the flow of aid even Recause aid continues flowing, poor governatice is perpetuated. Against this backdrop, performance-based aid, based "See Collier snd Dollar (2002) for a synthesis ofthe debate