反貪腐機構與貪腐認知：瓜地馬拉的案例研究 (2008-2018) - 政大學術集成

全文

(1)International Master’s Program in International Studies National Chengchi University 國立政治大學國際研究英語碩士學位學程. n i on Naatti o N. e rr ssiitt yy. ‧ ‧. ‧ 國 ‧ 國. 學學. Anti-Corruption Agency and the Perception of Corruption: 治政治大(2008-2018) 政 A Case Study of Guatemala 大立立反貪腐機構與貪腐認知：瓜地馬拉的案例研究 (2008-2018）. aal l iv n CChh U eennggcchhii U. 黎宇誠 Adolfo E. Quintanilla Caxaj Advisor: 蘇彥斌 Professor Yen-Pin Su , Ph. D. July 2020. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(2) Acknowledgments I feel extremely grateful to present this thesis to the public, academic, and international community. It is important to remark that this work would not have been possible without the help of specific individuals and institutions.. First and foremost, I want to express my gratitude to my family that always supports me and advises me unconditionally throughout my investigation. This thesis is for Jimena and Estefanía as an example of academic effort.. To my father and mother, thank you for always help me gain patience and encouragement to. 治政政治大大立 Second, I am also grateful to 立 have studied in Taiwan because this country allowed me to complete the subsequent investigation.. ‧ 國 ‧ 國. 學學. study a master´s degree. To my advisor Dr. Yen-Pin Su, for his support and acceptance to be my mentor, for his guidance, follow-up, and constant knowledge contribution during my. n i on Naatti o N. e rr ssiitt yy. ‧ ‧. investigations. My greatest acknowledgment and gratitude.. aal l iv n CChh U eennggcchhii U. 1. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(3) Abstract This thesis aims to understand why some citizens perceive more corruption than others. Previous studies have argued that personal traits and political ideologies matter for explaining the perception of corruption. However, it is essential to consider contextual factors for analyzing the perception of corruption. In Guatemala, I argue that the International Commission Against Impunity in Guatemala (CICIG) performance is an important contextual factor that influences the perception of corruption. My hypothesis suggests that when CICIG had a better performance, people tended to perceive more corruption in Guatemala. It is because a better performance for fighting against corruption might make people more aware of the corruption problems. My qualitative analyses suggest that the efforts made by CICIG. 治政治大政大 Guatemala. 立立 Keywords: Corruption; Political Attitude; Guatemala; Latin American Public Opinion Project 學學. ‧ 國 ‧ 國. had been efficient in promoting anti-corruption reforms and updating legal instruments in. (LAPOP); International Commission Against Impunity in Guatemala (CICIG). ‧ ‧. 摘要. e rr ssiitt yy. n i on Naatti o N. 本論文試圖理解為何有些公民對於國內貪腐會有較高的認知，而有些人卻較低。先前研究指出，個人特質與政治意識型態對於貪腐認知具有解釋力。然而，本論文主張在解釋貪腐認知時，必須要考量到時空脈絡因素才能完整。就瓜地馬拉的案例而言，本. aal l iv n CChh U 的貪腐認知具有重要的影響。本文假設，當CICIG在反貪行動上有較好的工作表現時 eennggcchhii U 論文主張「瓜地馬拉反有罪不罰國際委員會」（CICIG）的工作表現，對於該國公民. ，會讓人民有較高的貪腐認知。這是因為當CICIG的反貪調查成果愈多，大眾傳媒將會利用這些資訊大幅報導貪腐醜聞，而相關的新聞報導則更加吸引人民對於貪腐問題的重視。本文利用「拉丁美洲民意調查計畫」在瓜地馬拉施測的五波民調資料進行量化分析，實證結果支持上述假設。本文的質化研究則發現，CICIG在推動反貪與相關的司法改革，確實有相當的成效。. 關鍵字：貪腐；政治態度；瓜地馬拉；拉丁美洲民意調查計畫；瓜地馬拉反有罪不罰國際委員會. 2. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(4) Table of Content Abstract…………………………………………………………………………...…………..2 List of Figures……………………………….……... ...……..…………...…...…..…………..4 List of Tables………………………………..……... ………...……………...……..………....5 Chapter 1 Introduction…….……………………………………………………..………….6 1.1 Puzzle………..………..………..………..…..………..………..…..…………..…..6 1.2 Why Study Corruption?.....……..…………………..………..……..…....…..……7 Chapter 2 The Development of the International Commission against Impunity in Guatemala (CICIG)……………………...………………………….……………………….9 2.1 Background of Guatemala's corruption problems………..………………………..9 2.2 The State of Guatemala and the relation with the economic elites…………...….10. 治政治大政 2.4 The CICIG and Legal Reforms against impunity…..…………………..………..17 大立 Chapter 3 Theoretical Explanations and Research Design……...……………….………19 立. 2.3 The Creation of the CICIG……..……....…….……….…….…….………...…....11. ‧ 國 ‧ 國. 學學. 3.1 Explanations of Perception of Corruption and Testable Hypotheses....….....……19 3.2 Data and Measurement………....………………………………………….……..20. ‧ ‧. 3.2.1 Dependent Variable…………………………………………....………………21 3.2.2 Independent Variable.……………………………………….…..…………….22. e rr ssiitt yy. n i on Naatti o N. 3.2.3 Control Variables………………………………………….……..……………22 Chapter 4 Empirical Results…………………………………...…….……………….........27. aal l iv n C C 4.2 Comparative Simulation Analyses…………………………………...….…….…32 U hheennggcchhii U Chapter 5 Case Studies……………………………………...……………………………...38 4.1 Statistical Analyses…………………………………...….……………………....27. 5.1 Illicit Electoral Finance ……………... ...…….…………..….………...………...39 5.2 Controlling over the Election of Magistrates………….……….…………...........44 5.3 Public Response…………………...……...…………………………….........…..45 Chapter 6 Conclusion……..……………………………………………..………………….48 6.1 Research Summary………………………………………………………….……48 6.2 Policy Implications………….…………………………………………………....49 Appendix Timeline of Development and Cases of Investigations of the CICIG….…….50 References…………..……...…………....…………………..…………………..…………..53. 3. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(5) List of Figures Figure 1: Cases of Investigations and Prosecutions Conducted by the International Commission Against Impunity in Guatemala (CICIG).............….…………….…...…..…....21 Figure 2: Comparative Simulation Analyses for Total Number of CICIG’s Investigations, Education Attainment, and Corruption Perception in Model 1... ...……………….....…..…..33 Figure 3: Comparative Simulation Analyses for the Number of Uncharged Investigations, Charged Investigations, and Corruption Perception in Model 2... ...…….……………....…..34. 治政政治大大立立 ‧ ‧. ‧ 國 ‧ 國. 學學 e rr ssiitt yy. n i on Naatti o N. aal l iv n CChh U eennggcchhii U. 4. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(6) List of Tables Table. 1:. Description. of. Variables. Used. for. the. Empirical. Analyses. …………….…………..24 Table 2: Summary Statistics………………..…….…...……………..….………..….….........26 Table 3: Probit and Logit Models on Perception of Corruption in Guatemala ….……..…....29 Table 4: Predicted Probabilities of Corruption Perception for Model 1..................................31 Table 5: Elasticities Intervals for Perception of Corruption ………………………...….........34 Table 6: Predicted Probabilities of Corruption Perception for Individuals with Different Political Ideologies…………………………………………………………………………...36 Table 7: Predicted Probabilities of Corruption Perception for Individuals with Different. 治政政治大大Individuals with Different Table 8: Predicted Probabilities立 of Corruption Perception for 立. Frequency of Watching News …………………………….………………………...….........36. ‧ ‧. ‧ 國 ‧ 國. 學學. Sociotropic Evaluations ……………….…………………………………………...…...........37. e rr ssiitt yy. n i on Naatti o N. aal l iv n CChh U eennggcchhii U. 5. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(7) Chapter 1 Introduction 1.1 Puzzle. Corruption is a critical attribution for governance failure. When the mass media exposed severe corruption scandals, the democratic process and regime legitimacy would be undermined (Ehrlich and Lui 1999; Park 2003; Seligson 2002). Guatemala is one of the countries that suffer corruption the most in the world. In Guatemala, politicians and public servants used to utilize state institutions to commit corruption and seek impunity. As the international community, civil society, and government agencies increasingly make efforts to fight against corruption, Guatemala also made a significant move to establish anti-corruption agencies to address various issues about corruption. The United Nations and the Guatemala. 治政政治大大 Impunity in Guatemala (CICIG). Since 2008, the CICIG had, investigated criminal acts of 立立. government signed an agreement in 2007 to create the International Commission Against. ‧ 國 ‧ 國. 學學. bribery, money laundry, and illegal electoral finances. In 2015, CICIG’s investigation of corruption scandals contribute to the resignation of former President Pérez Molina.. ‧ ‧. Despite remarkable performance on identifying, reporting, and prosecution, CICIG only. n i on Naatti o N. e rr ssiitt yy. lasted for 11 years (2008-2018). CICIG's presence in Guatemala created a controversial debate toward national sovereignty. As a result, a series of political obstacles became evident for the CICIG (Althoff 2019; Brannum; 2018; Maihold 2016; Luiña, 2011). Surrounded by. aal l iv n CChh U ee nefficiency ggcchhiiofU CICIG in January 2019. Still, the efforts andn the CICIG have made this agency a corruption scandals, president Jimmy Morales decided to terminate the UN's agreement for. role model for anti-corruption agencies in other countries (O’Boyle 2019).. Why do some people perceive more corruption than others? Previous studies have argued that personal traits and political ideologies matter for explaining the perception of corruption. However, few studies have taken into account the contextual factors that might affect citizens’ perception of corruption. This thesis aims to join the debate about the perception of corruption by highlighting the importance of anti-corruption agency performance. Focusing on the case of Guatemala, I argue that the performance of CICIG is an important contextual factor that influences the perception of corruption. My hypothesis suggests that when CICIG had a better performance, people tend to perceive more corruption in Guatemala. It is because. 6. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(8) a better performance for fighting against corruption might make people more aware of the corruption problems.. Using the Latin American Public Opinion Project (LAPOP) data for 2010, 2012, 2014, 2016, and 2018, this thesis examines the impact of CICIG’s performances on the perception of corruption in Guatemala. LAPOP data shows that corruption is one of the main concerns for most Guatemalans. As a result, CICIG performance combined with confidence on the Ministerium Público, facilitate Guatemalans’ understanding of corruption issues in the country.. 1.2 Why Study Corruption?. 治政政治大大 Previous studies have found 立 that corruption has negative impacts for a country’s national 立 ‧ 國 ‧ 國. 學學. development. For instance, a higher level of corruption might be correlated with more military spending (Gupta, de Mello, and Sharan 2000). Moreover, more corruption might. ‧ ‧. reduce public spending on health care and education (Mauro 1998; Gupta, Davoodi, and Terme 1998; Gupta, Davoodi, and Tiongson 2001), and reduce investment in human capital (. n i on Naatti o N. e rr ssiitt yy. Ehrlich and Lui 1999). In a political environment where corruption is prevalent, government agencies are likely to perform as “independent rent-seeking, monopolist providers of complementary goods and services because of corrupt and extortionist tax inspectors”. aal l iv n CChh U eennggcGlomm, chhii U manage public sources (Bearse, and Janeba 2000). For instance,. (Gupta, Davoodi, and Tiongson 2001, 133). Corruption undermines the government’s ability to effectively. corruption reduces spending on medicine and textbooks (Tanzi and Davoodi 1997).. Why does the perception of corruption matter? First of all, when corruption problems are widely perceived by the citizens, there will be more distrust of institutions, and thus political legitimacy will be eroded (Canache and Allison 2005; Davis, Camp, and Coleman 2004; Gomez and Vilchis 2012; Sole and Sorribas 2018; Williams 1999). A higher level of corruption perception might lead to less participation in government activities such as tax paying (Birdsall, Kenny, and Diofasi 2018).. 7. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(9) High levels of corruption perception affect political culture. Specifically, when corruption problems are widely perceived, citizens tend to generate strong distrust towards certain political institutions, and interpersonal trust would be undermined (Seligson 2002). Rampant corruption might create a cultural tradition that emphasizes gift giving and ignores accountability, which lead to more corruption (Melgar, Rossi, and Smith 2010). Moreover, the perception of corruption might affect the dynamics of party politics. When people perceive more corruption, new parties with an anti-corruption platform might be more likely to enter the political arena, and existing parties might split because of the scandals (Sole and Sorribas 2018).. Given that studying corruption matters for understanding the political process of a country, it. 治政政治大大 The perception of corruption indicates the level of citizens’ awareness of this issue. A lower 立立. is also essential to study why some citizens perceive more corruption than others in a country.. ‧ 國 ‧ 國. 學學. level of corruption perception does not necessarily mean that the government is less corrupt. In fact, it is possible that citizens might perceive less corruption in a country that is identified. ‧ ‧. with serious corruption by international organizations. When citizens perceive more corruption, it is expected that there will be higher popular pressure for demanding the. n i on Naatti o N. e rr ssiitt yy. government to tackle the issues of corruption. In this sense, it is important to examine what determines various levels of corruption perceptions among citizens.. aal l iv n CChh U U eenngdiscuss gcchhiiimportant development of the CICIG. Chapter 3 will theoretical explanations of. This thesis is organized as follows. In Chapter 2, I will provide an overview of the. corruption perception and present the research design for the empirical analyses. In Chapter 4, I test the hypothesis about the relationship between CICIG’s performance and corruption perception and present the statistical results. Chapter 5 provides qualitative analyses by focusing on important cases of illicit electoral financing and controlling over the magistrate elections. Chapter 6 concludes and provides some policy implications based on the finding of this thesis.. 8. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(10) Chapter 2 The Development of the International Commission against impunity in Guatemala (CICIG) 2.1 Background of Guatemala's corruption problems. During the Cold War, Guatemala experimented with a severe internal conflict that lasted thirty-six years. The UN Commission for Historical Clarification (Comisión para el Esclarecimiento Histórico, CEH) concluded that this conflict register the death of 200,000 humans and 70,000 disappeared, which a majority was part of the Mayan indigenous people (CEH 2). The Commission also states that most of the war crimes can contribute to the State of Guatemala (CEH 17). A set of human rights violations can be easily tracked because of the. 治政治大政大 agency in charge of fighting (Estado Mayor Presidencial, EMP) was a state intelligence 立立 insurgents. During the war, this agency benefited from substantial power quota, and it was institutional frame used to commit torture and acts of violence. The Presidential Guard. ‧ 國 ‧ 國. 學學. usually linked with Guatemala political elites and other relevant actors of Guatemala society.. ‧ ‧. The influence of state intelligence agencies in Guatemala increased rapidly during the. e rr ssiitt yy. n i on Naatti o N. internal war. Judiciary, Legislative branches and Police forces created a network of control over political and business interests. In 1996 the Agreement on the Strengthening of Civilian Power and Role of the Armed Forces in a Democratic Society (Acuerdo Sobre. aal l iv n CChh U sections founded in the State to ii U the necessity of the regulation and restructuring eennggccofhhthose. Fortalecimiento del Poder Civil y Función del Ejército en una Sociedad Democrática) stated eradicate networks of control.. However, in 2002 the UN Verification Mission in Guatemala (MINUGUA) founded that commitments signed during the peace agreements in 1996 related to developing investigatory faculties and strength the police apparatus have not been addressed (CICIG 2010). As a result, the MINGUA also stated in its report that structures linked with the State have reorganized and become active in illegal economic interest and operated with complete impunity with political influence (CICIG 2015).. 9. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(11) Both the lobbying sector from the government and business community members and illegal groups like CIACS have a strong influence over the judiciary. Illegal bodies use the judiciary mainly to main impunity for civil war crimes and maintain strong relations with key actors of each four-year administration.. 2.2 The State of Guatemala and the relation with the economic elites. The relation between land tenancies and economic elites is very close. Historically, economic elites in Guatemala are responsible for maintaining political regimes that contribute to financial benefits and security (Gutierrez 2016). After industrialization took place in the middle 20th century, a small middle-rural class emerge. It is considered that the economic. 治政政治大大 sentiment was a set of ideals 立 that incentive the financial elites to support any action against 立. elites and middle-rural call interest were embraced during the Cold war. The anti-communist. ‧ 國 ‧ 國. 學學. minorities in the rural areas of Guatemala.. ‧ ‧. By the end of the Cold War, Guatemala had a robust centralized State with a high clientele system. This system operated with corruption and racism (Gutierrez 2016). The interactions. n i on Naatti o N. e rr ssiitt yy. of power among Guatemalan State are related to the weakness of the national institutions. As a result, the current Guatemalan State is highly inefficient and failed to provide essential services to society (Nyberg 2015).. aal l iv n CChh U U eennggarmy cchhiiofficials Instead, as part of the internal conflict, heritage started to work with criminal organizations that took advantage of state institutions' fragility. The lack of procedures and weak bureaucracy made the State of Guatemala, a machine of unskilled unemployed individuals. The degradations of State institutions also damage the legitimacy of authority and economic elites.. Under the context described in the below section, after 1985, drug and human trafficking were already taking place in Central America. The formation of criminal activity networks was rising while economic elites emerged under the new capitalist model. A regional criminal organization often recluses unemployed military and elite soldiers that suffer from state programs to provide them with a proper job after internal conflict.. 10. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(12) On the one hand, large parts of the demobilized army in Guatemala became part of Colombian and Mexican criminal organizations. In the other, corrupt officials provided the infrastructure needed to build a network of logistics for smugglers and money laundry.. The consolidation of all these criminal networks and the label of illegal Clandestine Security Apparatuses (Cuerpos Ilegales y Aparatos Clandestinos de Seguridad CIACS) was given at the same time the regional drug trafficking was fully established. Currently, CIACS are responsible for multiple crimes; arms trafficking, drug trade, black-market adoptions, money laundry, and extortion (Nyberg 2015). Consequently, CIACS are using to understand the relationship between the related. 治政政治大大 2016). The peace agreements立 identified and included a section for the restructure and 立 deterioration of the Guatemalan State and corrupt law enforcement institutions (Postema. ‧ 國 ‧ 國. 學學. reformation of those state-agencies. The treaties did not address the dynamics of power between economic elites, the Guatemalan army, and CIACS (Gutierrez 2016). As a result, the. ‧ ‧. contemporary history of Guatemala has vast information about CIACS operations and modus operandi. Thus, these actors continue to ensure impunity and intricate inside the roots of. n i on Naatti o N. e rr ssiitt yy. public institutions (Postema 2016 ).. 2.3 The Creation of the CICIG. aal l iv n CChh U U government has been unable to eenstate, Guatemala has been depicted as a failing hhii the nggccwhere adequately provide essential services for the population because criminal groups operate and parasite the state institutions (Briscoe and Pellecer 2010, 38). Security and justice are among the many problems that civilians face daily in this country (Hampson 1995). As a result, high distrust among citizens toward the political elite grows year by year (Park 2003).. However, even though Guatemala suffers from various issues of poor governance (e.g., high crime rates), many civil society organizations work hand-in-hand with the international community to make the country a better place. In particular, there had been several efforts made to reduce corruption. The first remarkable effort was formed in 2003 when the Alfonso Portillo administration openly recognized that the judicial system is too weak to tackle. 11. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(13) organized crime within state institutions, and later the government requested the intervention of the Department for Political Affairs of the United Nations to investigate and enforce the laws for groups that involving political corruption (Luiña et al. 2014).. After intense negotiation with the UN, an agreement draft was made on January 7, 2004, to create the International Commission to Investigate Illegal Groups and Clandestine Security Forces (Comisión de Investigación de Cuerpos Ilegales y de Aparatos Clandestinos de Seguridad, CICIACS). The so-called “illegal groups and clandestine security forces” (CIACS) are criminal units that have been operated as an illicit or collaterally illicit network. The groups can exert influence by corruption, infiltration, or co-optation of decision-making processes through their informal members within the state apparatus (CICIG 2007). More. 治政政治大大 almost without any punishment or prosecution. They enjoyed significant benefits sponsored 立立. broadly, these groups were part of an intricate criminal network operating since the post-war,. ‧ 國 ‧ 國. Congress members (Lopez 2010; Luiña et al. 2014).. 學學. by the cover of the national security forces, functionaries in public institutions, or even. ‧ ‧. In 2004, Oscar Berger won the presidential election due to its successful campaign platform. n i on Naatti o N. e rr ssiitt yy. for promoting a strong agenda against corruption and organized crime and impunity. Unfortunately, homicides rates increased significantly during Berger’s term. On December 12, 2006, the United Nations and Guatemala signed an agreement to create International. aal l iv n CChh U eennggcchhii U ratification on August 01, 2007 (CICIG 2008).. Commission against impunity in Guatemala (CICIG), and the Congress made the final. Strong support of the international community was essential for creating the CICIG. Different countries, including the United States, provided the commission an annual budget of $15-20 million(OSJI 2016, 8).In the first report, the CICIG describes itself as an independent organization from the political and financial aspects (CICIG 2008). The constitution of this commission is one of its kind by 2008, nor the United Nations or any other international organizations had registered any similar document designated for the promotion of law enforcement (CICIG 2008). The new experience with prosecutorial mission represented bureaucratic challenges for both CICIG and the UN (OSJI 2016).. 12. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(14) CICIG manages to operate with high efficiency and effectivity in its administrative procedures. This represented an advantage, especially during the initial organizational and recruitment processes. However, the installation of this anti-corruption agency also faced several challenges at the national level. An evident debate emerged between the pro-status-quo groups and the reformist group (Brannum 2019). The nationalists, evangelical networks, and military (Althoff 2019) argue that the CICIG would undermine the country’s sovereignty (Maihold 2016). Other opponents contended that the government transferred its capacities and duties to an external actor (Valladares 2010). Not surprisingly, this argument was widely used by those who were later investigated and prosecuted by the CICIG (Brannum, 2019).. 治政政治大大 investigated and identified illegal groups in or out of the state. Second, the CICIG cooperated 立立. Since its establishment, the CICIG aimed to pursue three primary objectives. First, the CICIG. ‧ 國 ‧ 國. 學學. with the Guatemalan government to tackle illegal bodies or clandestine security groups, activities, and financial sources, and these cooperative actions included. Investigations, legal. ‧ ‧. prosecution, and the punishment of members of this organization. Third, the CICIG made recommendations to the Guatemala government for the adopting policies and institutional. n i on Naatti o N. their reappearance.. e rr ssiitt yy. reforms for eradicating clandestine apparatuses and illegal security forces and preventing. In practice, the CICIG mandate serves as a guideline to justify the investigatory work related. aal l iv n CChh U ii U eennggand cchhclandestine tracking illegal groups. to the protection of human rights in Guatemala (OSJI 2016). The CICIG’s initial work was mainly focused on. security forces (CIACS).. Beginning from 2010, CICIG started to work on tracking money laundry actions, fraud, and other crimes related to impunity (CICIG 2016).. Article 6 of the “Agreement between the United Nations and the State of Guatemala on the 1. establishment of an International Commission Against Impunity in Guatemala” stipulates important powers for this organization, including:. “(a) freedom of movement without restriction throughout Guatemalan territory;. 1. See https://www.cicig.org/uploads/documents/mandato/cicig_acuerdo_en.pdf. 13. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(15) (b) freedom of access without restriction to all State locations, establishments and installations, both civilian and military, as well as to all penitentiaries and detention facilities without prior notice, in accordance with and subject to the relevant provisions of Guatemalan Constitution; (c) freedom to meet and interview any individual or group of individuals, including State officials,. military and police personnel, community leaders, non-governmental. organizations, private institutions and any persons whose testimony is deemed necessary for the discharge of its mandate; (d) free access to information and documentary material that has a bearing on its investigations, official archives, databases and public records and any similar report, file, document or information in possession of the relevant persons or entities, whether civilian or military, in accordance with and subject to the applicable provisions of the Guatemalan Constitution.”. 治政治大政 The CICIG had experienced five propagations before it 大 was terminated in 2019. The five 立立 propagations were necessary for the CICIG to build investigations, identify criminal activity, ‧ 國 ‧ 國. 學學. and structure institutional reforms. The first propagation took place on March 24, 2009, the second on December 23, 2010, the third propagation took place between September 4, 2013. ‧ ‧. and September 3, 2015, the fourth took place between September 4, 2015, and September 3,. place from September 04, 2017, to September 03, 2019.. e rr ssiitt yy. n i on Naatti o N. 2017. Requested by the then-president Jimmy Morales to the UN, a final propagation took. aal l iv n CChh U iU eenperiods development, and prosecution require long of The two-year term of each hipreparation. nggcch. Overall, the CICIG experience suggests that the capacity building, investigations. propagation constrained the CICIG to act on specific cases (OSJI 2016). Thus, it was not surprising that those linking with political-criminal networks used the propagation opportunity to detract (Brannum, 2019).. Before 2019 CICIG’s last propagation period, the anti-corruption agency had three different commissioners. According to Article 3 of the “Agreement between the United Nations and the State of Guatemala on the establishment of an International Commission Against 2. Impunity in Guatemala,” the main roles of the commissioners include:. 2. See https://www.cicig.org/uploads/documents/mandato/cicig_acuerdo_en.pdf. 14. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(16) “(a) Collect, evaluate and classify information provided by any person, official or private entity, non-governmental organization, international organization, and other states; (b) Promote criminal prosecutions by filing criminal complaints with the relevant authorities. The commission may also follow the Agreement and the Code of Criminal Procedure and join a criminal proceeding as a private prosecutor (querellante adhesivo) concerning all cases within its jurisdiction; (c) Provide technical advice to the relevant state institutions in the investigation and criminal prosecution of crimes committed by presumed members of illegal groups and clandestine organizations, and advise state institutions in the implementation of administrative procedures that might be required against state officials allegedly involved in such organizations; (d) Report to the relevant administrative authorities the names of civil servants who in the exercise of their duties have allegedly committed administrative offenses so that the proper. 治政政治大大立立. administrative proceedings may be initiated, especially those civil servants or public employees accused of interfering with the commission’s exercise of its functions or powers, without prejudice to any criminal proceedings that may be instituted through the Office of. ‧ 國 ‧ 國. 學學. the Public Prosecutor;. (e) Act as an interested third party in the administrative disciplinary proceedings referred to above;. ‧ ‧. (f) Establish and implement cooperation agreements with the Office of the Public. e rr ssiitt yy. n i on Naatti o N. Prosecutor, the Supreme Court, the Office of the Human Rights Ombudsman, the National Civilian Police and any other state institutions to carry out its mandate;. (g) Guarantee confidentiality to those who assist the commission in discharging its functions, whether as witnesses, victims, experts or collaborators;. aal l iv n C C U hheenorngadministrative cooperation in general from any official gcchhii Uauthority of the state and any (h) Request, under the terms of its mandate, statements, documents, reports, and. decentralized autonomous or semi-autonomous state entity, and such officials or authorities are obligated to comply with such request without delay; (i) Request the Office of the Public Prosecutor and the government to adopt measures necessary to ensure the safety of witnesses, victims and all those who assist in its investigations, offer its good offices and advice to the relevant State authorities for the adoption of such measures, and monitor their implementation; (j) Request and supervise an investigation team made up of national and foreign professionals of proven competence and moral integrity, as well as such administrative staff is required to accomplish its tasks; (k) Take all such measures it may deem necessary for the discharge of its mandate, subject to and following the provisions of the Guatemalan Constitution;. 15. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(17) (l) Publish general and thematic reports on its activities and the result thereof, including recommendations under its mandate.”. To operate effectively CICIG create inter-institutional agreements with the Guatemalan government agencies, particularly the public prosecutor's office, for performing functions such as investigations, criminal prosecution, technical assistance, and institutional coordination. These inter-institutional agreements aimed to help track and dismantle illegal security groups and clandestine security organizations.. Under the first commissioner Carlos Castresana, the first eight inter-institutional cooperation agreements were signed. One important agreement is the agreement with the Superintendency. 治政治大政 most agreement between the Public Attorney Office and the 大CICIG (MP-CICIG), which was 立立 seen as the most important inter-institutional agreement, was signed on February 29, 2008.. of Banks (SIB-CICIG), which was related to administrative offenses (CICIG, 2007). The. ‧ 國 ‧ 國. 學學. The MP-CICIG aimed to tackle illegal security groups and clandestine security organizations.. ‧ ‧. Moreover, this agreement established the figure of Third-Party Prosecutor, or querellante adhesivo. According to the agreement, the Third-Party Prosecutor “may exercise criminal. e rr ssiitt yy. n i on Naatti o N. prosecution or adhere to that initiated by the Public Ministry in the case of crimes allegedly committed by members of those criminal structures that are the object of the commission’s mandate (CICIG 2020).. aal l iv n CChh Uto facilitate investigations related On March 6, 2008, a three-institution agreement ii U eennggwas cchhsigned to violence and crimes against women. The nature of this agreement was the cooperation between the Secretary of Social Works of the President's Wife, Ombudsman for Indigenous Women, Woman Presidential Secretary, and the CICIG (SOSEP-DEMI-SEPREM-CICIG) (CICIG 2008).. On March 25, 2008, an agreement between the CICIG and the Ministry of the Interior was signed (MINIGOB-CICIG) to promote cooperation of the CICIG and the National Civil Police for identifying illegal security groups and clandestine security organizations. To persecute money laundry and crimes related to taxation and customs fraud, crimes or offenses. 16. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(18) against the tax administration, CICIG signed an agreement with Superintendency of Tax Administration (SAT-CICIG) on July 2, 2008 (CICIG 2008).. Under Castresana term as a commissioner, the CICIG also developed three different inter-institutional cooperation agreements with institutions that belonged to the UN. Two of those agreements were signed in 2009. The first is the UNIFEM-CICIG agreement, which aimed to persecute impunity related to violence against women, signed by the CICIG and the United Nations Development Fund for Women.. The second is the UNICEF-CICIG, which aimed to persecute impunity with crimes related to children, signed by the CICIG and the United Nations Children's Fund. A third inter-institutional agreement was signed in 2010 with the United Nations Office on drugs and Crime (UNODC-CICIG) to fight against drug. 治政政治大大立立. crime and terrorism (CICIG 2009).. ‧ 國 ‧ 國. 學學. Under the commissioner Francisco Dall ,two relevant domestic agreements related to the Third-party prosecutor or querellante adhesivo were signed. The first was signed in 2011 with. ‧ ‧. the National Accounts Controller (CGC-CICIG), and the second was signed in 2013 between the CICIG and the General Attorney and General Register of Property (MP-RGP-CICIG). n i on Naatti o N. e rr ssiitt yy. (CICIG 2013). The last commissioner, Ivan Velasquéz, appeared to decrease the activity related to the promotion of inter-institutional cooperation agreements of the CICIG. Still, Velasquez signed an agreement with the Supreme Electoral Court in 2018 to advance an. aal l iv n CChh U eennggcchhii U. electoral reform (CICIG 2018).. 2.4 The CICIG and Legal Reforms against Impunity The first obstacle that the CICIG faced was that the national prosecutorial system was weak and highly politically dependent (Hudson 2010). Therefore, as part of CICIG's ten-year operations, the CICIG pushed forward a series of legal reforms and proposed legal instruments for combating criminal networks, which would meet the urgent need for the upgrade of Guatemala's judiciary system (WOLA 2015).. The first two reforms promoted by CICIG was the law of strengthening criminal prosecution and the law against organized crime. CICIG's comprehension of organized crime in. 17. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(19) Guatemala allow this international organization to promote legal prosecutions. Important proposals and reforms were approved like the law regulating private security services, the act of extinction of domain, the law against corruption, and migration law (CICIG 2019).. Influenced by the CICIG’s efforts for promoting legal reforms, the Guatemalan Congress passed a law against corruption in 2012. This law stipulates crimes from illicit enrichment of public service and the use of power to influence judge which influenced the dynamics of corruption in Guatemala. From 2015 to 2019, CICIG had been an important actor in the promotion of a more independent judiciary power.. While the CICIG had made impressive performance to promote judicial reforms, not all. 治政政治大大 several reforms related to the立 judicial career, prosecutor, police, forensic, and penitentiary 立. changes were approved by the Congress. For example, under the Law against Corruption,. ‧ 國 ‧ 國. 學學. system never were discussed between congressmen (CICIG 2019). Other reforms related to investigatory faculties, criminal prosecution like those related to the amparo, the prejudice,. ‧ ‧. the code of criminal procedure, and the preventive prison had been pending for a long time.. n i on Naatti o N. e rr ssiitt yy. Besides the legal reforms promoted by CICIG, several new investigation tools had been established. The inter-institutional coordination, combined with the new investigatory methodology facilitates the prosecutor to have solid cases in justice tribunals. Empirical. aal l iv n CChh U iU eenngg(Trejo cchhiand rates through these new tools of investigation Matiz 2019). The first example is. studies suggest that the presence of CICIG in Guatemala contributes to reducing homicide. wiretapping, which was often used by CICIG and Public prosecutor's office to track phone calls and information found in modern communications systems. The second example was the witness protection program. The third tool was the confidential informant, which helped investigators provide legal help to organized crime members to improve to a potential dismantling of criminal groups (WOLA, 2015). The second and the third tools eventually proved to be rather successful because they had received support from the international community, which provided technical guidance and help for the CICIG to protect those that contribute with vital information and guarantee their safety. Last, the CICIG also implemented tracking control inside and outside Guatemala related to money laundering,. 18. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(20) drug trafficking, and weapons circulations. However, these controls faced limitations because of domestic laws and the Ministry of the Interior's budget.. 治政政治大大立立 ‧ ‧. ‧ 國 ‧ 國. 學學 e rr ssiitt yy. n i on Naatti o N. aal l iv n CChh U eennggcchhii U. 19. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(21) Chapter 3 Theoretical Explanations and Research Design 3.1 Explanations of Perception of Corruption and Testable Hypotheses At the macro level, Park (2003) finds that economic freedom, socio-political stability, a tradition of law abidance, and national cultures help explain the degree of corruption in a country (Park 2003). At the micro-level, individuals’ perception of corruption is affected by various factors. For instance, Melgar et al. (2010) show that because more educated people tend to access information about corruption news and more able to process the information, they tend to perceive more corruption. Birdsall, Kenny, and Diofasi (2018) contend that an individual’s educational attainment matters in explaining her perception of corruption, but the. 治政治in high-income 政 Diofasi (2018) find that higher educated people countries perceive less 大大立 corruption than higher educated people in low-income countries. 立. effect of education differs depending on specific contexts. Specifically, Birdsall, Kenny, and. ‧ 國 ‧ 國. 學學. Other empirical studies argue that perceptions of corruption among individuals can be. ‧ ‧. explained by political ideology and economic outlook. In general, those individuals that consider themselves with a left-wing ideology compared to others are more likely to perceive. e rr ssiitt yy. n i on Naatti o N. corruption (Navia, Perelló, and Masek 2019). Moreover, the relationship between an individual’s outlook toward national economic performance and the perception of corruption. aal l iv n CChoutlook individuals with a positive economic are less likely to U perceive corruption (Navia, heennggcchhii U Perelló, and Masek 2019). was shaped by the evaluation of this person’s future income. Empirical analysis shows that. Another essential factor that helps explain the perception of corruption is media exposure. In general, mass media in democracies tend to report corruption scandals. Therefore, when an individual has more exposure to media, he or she will be likely to receive information about corruption, and thus this person will perceive more corruption ( Weyland 1998).. Media reports of corruption scandals might be the result of reporters’ investigation. More importantly, when a country has an anti-corruption agency, the investigation of corruption conducted by this agency will be an essential source for the mass media. Specifically, when a. 20. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(22) country has a clear anti-corruption policy agenda, and the anti-corruption agency coordinates with judicial institutions for combating corruption, the reveal of corruption scandals is more likely to attract public attention. In countries where significant efforts to reduce corruption have been successful, it might heighten awareness of corrupt practices and might increase the perception of corruption precisely when actual corruption is declining (Seligson 2006). Sole and Navarro’s (2018) study of corruption scandals in Spain between 1999 and 2009 demonstrates that corruption scandals have remarkable effects on local politicians’ levels of trust and perceptions of corruption (Solé and Sorribas 2018).. Based on the discussion above, I propose two testable hypotheses for the empirical analysis: H1: An individual is more likely to perceive corruption when more individuals are identified. 治政政治大大 H2: An individual is more likely to perceive corruption when more individuals are identified 立立 as involving corruption crimes but not charged with sentences or fines.. ‧ 國 ‧ 國. 學學. as involving corruption crimes charged with sentences or fines. 3.2 Data and Measurement. ‧ ‧. e rr ssiitt yy. n i on Naatti o N. Unlike other government agencies in Guatemala, the CICIG enjoyed a high degree of autonomy and budget for conducting corruption investigations. However, President Jimmy Morales terminated the agreement with the UN for CICIG in January 2019. Table 1 shows. aal l iv n C C 11 years, CICIG had successfully attracted Uthe Guatemalans and international hheenattention ii U nggcchhfrom. the performance of the CICIG for corruption investigation from 2008 to 2018. During these. community, and the CICIG served as a role model for anti-corruption agencies in other countries like Honduras (MACHH) (Call 2019), and El Salvador (CICIES) (Renteria 2019).. 21. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(23) Figure 1 International Commission Against Impunity in Guatemala (CICIG). (2008-2010) Year. Number of case studies conducted by CICIG. Number of individuals under CICIG’s investigation. Number of people identified as involving corruption crimes but not charged with sentences or fines. Number of people charged with sentences or fines because of committing crimes of corruption. 2008. 5. 65. 40. 25. 2009. 8. 152. 95. 57. 2010. 3. 18. 13. 5. 治政治大政 32 21 大立立 120 87. 6. 2013. 2. 105. 55. 2014. 5. 45. 37. 2015. 15. 371. 347. 2016. 10. 178. 165. 2017. 10. 135. 132. 2018. 9. (2013-2019). ‧ 國 ‧ 國. 2012. 11 33. ‧ ‧. 3. 學學. 2011. n i on Naatti o N. e rr ssiitt yy. (2010-2013). aal l 128 127 i v C C Uton2019. Source: Elaboration by the Author, based on CICIGh annual reports from i2008 h eennggcchh i U. 50. 8 24 13 3 1. 3.2.1 Dependent Variable. The dependent variable in this analysis is the perception of corruption based on the data from the LAPOP surveys for Guatemala conducted in 2010, 2012, 2014, 2017, and 2019. The variable was initially coded as an ordinal variable, which takes the value of 1 to 4, where 1 = very widespread, 2= somewhat widespread, 3= little generalized, and 4= nothing generalized. I followed Navia et al. (2019) to transform the corruption perception variable into a dichotomous variable, coded 1 when the original ordinal variable takes the values of 1 or 2; and 0 when the original ordinal variable takes the values of 3 or 4. The change in the wording in the 2017 LAPOP survey presents a challenge for an appropriate intertemporal comparison.. 22. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(24) To address this issue, again, I followed Navia et al. (2019), coding “None” and “less than half” as 0, and “common,”; “very common”; “more than half”; and “all” coded as 1.. 3.2.2 Independent Variable. The main independent variable for the empirical analysis is the level of anti-corruption performance, operationalized as how much efforts that the CICIG had made to fight against corruption in Guatemala. The presence of the government’s anti-corruption efforts suggests specific actions taken against corrupted and corruption dynamics. In Guatemala, the press releases of the CICIG used to be an essential information source for the society to follow the anti-corruption agenda. The information about corruption include 1) the number of. 治政政治大大 and 2) those that are charged with sentences or fines. Most of the individuals under CICIG’s 立立. individuals identified as involving corruption crimes but not charged with sentences or fines,. ‧ 國 ‧ 國. 學學. investigations involved issues of corruption, impunity, and bribery.. ‧ ‧. I coded the data using CICIG’s annual reports on a year basis. The variables are lagged by one year to avoid endogeneity. For instance, all the individual respondents for the 2012. n i on Naatti o N. e rr ssiitt yy. LAPOP survey are assigned the value of the CICIG anti-corruption performance in 2011. I expect that other things being equal, an individual is more likely to perceive corruption when the total number of individuals under CICIG’s investigation of corruption is higher. I also. aal l iv n CChh U U eenngthat hhiiindividual gccan independent variable. Specifically, I expect is more likely to perceive. examine how anti-corruption performance affects corruption perception by decomposing the. corruption when more individuals are identified as involving corruption crimes but not charged with sentences or fines. I also expect that an individual is more likely to perceive corruption when more individuals are identified as involving corruption crimes charged with sentences or fines.. 3.2.3 Control Variables. Education Attainment. The years of education an individual has affected her perception of corruption. For instance, Melgar et al. (2010) and Birdsall, Kenny, and Diofasi (2018) show that because more educated people tend to access information about corruption news and. 23. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(25) more able to process such information, they tend to perceive more corruption. Therefore, education attainment is controlled for its possible impacts on the perception of corruption.. Political Ideology. As Navia et. al (2019) and Julnes and Villoria (2014) show, political ideology can influence the perception of corruption. Specifically, rightist-leaning individuals tend to be more skeptical about the state than those in the left that center their beliefs on the role of the state. It is expected that those in the right-wing will perceived more corruption than the rest. The LAPOP surveys coded political ideology using a 1 to 10 scale. I follow Navia et al. (2019) to construct three dummy variables for right-wing (7-10), left-wing (1-3), and centrist (4-6). I include the dummy variables for right-wing and left-wing in the models, with no response as the reference group.. 治政政治大大 Media exposure. Another critical factor that affects the perception of corruption is the 立立 ‧ 國 ‧ 國. 學學. frequency individuals read the news. In general, mass media in democracies tend to report corruption scandals. Therefore, when an individual has more exposure to media, he or she. ‧ ‧. will be likely to receive information about corruption, and thus this person will perceive more corruption (Weyland 1998). The LAPOP surveys coded the frequency of following news in. n i on Naatti o N. e rr ssiitt yy. mass media as a categorical variable. In my analysis, I generated five dummy variables for every day, every week, every month, and rarely. I include all of the dummy variables except “never read news,” which is excluded as the reference group. I expect that individuals that. aal l iv n CChh U eennggcchhii U. read news more frequently are more likely to perceive corruption.. Economic Outlook. Navia et al. (2019) argue that the relationship between an individual’s outlook toward national economic performance and the perception of corruption was shaped by the evaluation of this person’s future income. It is expected that an individual that considers the national economic situation as better in the past twelve months will perceive less corruption than the rest. Moreover, a person with a positive evaluation of personal economic situation will be more tolerant of corruption, and therefore, perception might not be affected. Sociotropic and egotropic economic outlook are coded as trichotomous variables in my empirical analyses.. 24. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(26) Presidential Approval. Gómez-Vilchiz’s (2012) study shows that as the perception of corruption increases, the presidential approval decrease. In my analysis, I believe that presidential approval might have a negative impact on the corruption perception. Using the LAPOP survey data, I coded presidential evaluation as trichotomous variable. I include presidential approval and presidential disapproval in the models, and leave the regular approval as the reference group. Table 1 displays all of the variables used in the models. Table 2 provides summary statistics for all the variables used in the analyses.. 治政政治大大立立 ‧ ‧. ‧ 國 ‧ 國. 學學 e rr ssiitt yy. n i on Naatti o N. aal l iv n CChh U eennggcchhii U. 25. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(27) Table 1 Description of Variables Used for the Empirical Analyses. Variables. Variables Description coded 1 if the response is very generalized or somewhat generalized; 0 if the response is little generalized or not generalized.. Dependent variable Perception of corruption Independent Variables Total number of CICIG’s investigations. Total number of individuals under CICIG’s investigation of corruption. Number of uncharged investigations. Number of people identified as involving corruption crimes but not charged with sentences or fines. 治 Number of people charged with sentences 政治政大 Number of charged investigations or fines because 大 of committing crimes of 立 corruption 立. Individual economic outlook. e rr ssiitt yy. Country economic outlook. Ideology left. 1 for respondent who lives in rural area; 0 otherwise. n i on Naatti o N. Age. 1 for female respondent; 0 otherwise. ‧ ‧. Rural. ‧ 國 ‧ 國. Female. 學學. Control Variables. Age of the respondent.. Sociotropic evaluation, coded 1 if the response is better; 0.5 if the response is same; 0 if the response is worst.. aal l iv n CChh U eennggEgotropic cchhii U evaluation, coded 1 if the. response is better; 0.5 if the response is same; 0 if the response is worst. Ideology left if the respondent give 1 to 3 answer; 0 otherwise *The values none and 1– 10. Ideology center. Ideology center if the respondent give 4 to 6; 0 otherwise *The values none and 1– 10. 26. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(28) Ideology right if the respondent give 7 to 10 answer; 0 otherwise. Ideology right. *The values none and 1– 10 Education Attainment. Respondent total years for education attainment. Presidential approval. coded 1 if presidential approval; 0 otherwise. Presidential disapproval. coded 1 if presidential disapproval; 0 otherwise. 治政政治大大 coded 1 if respondent follow news every 立立 day ; coded 1 if presidential regular; 0 otherwise. ‧ ‧. News rarely. coded 1 if respondent follow news every week;0 otherwise. n i on Naatti o N. News month. 學學. News week. 0 otherwise. ‧ 國 ‧ 國. News everyday. coded 1 if respondent follow news every month;0 otherwise. e rr ssiitt yy. Presidential regular. coded 1 if respondent follow news rarely;;0 otherwise. aal l iv n CChh U eennggcchhii U. 27. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(29) Table 2 Summary Statistics. Variable. Mean. Standard Deviatio n. Minimu m. Maximu m. Perception of Corruption. 2.345. 1.276. 1. 5. Total number of CICIG’s investigations. 119.4. 49.63. 32. 178. Number of uncharged investigations. 93.35. 50.92. 21. 165. Number of charged investigations. 26.02. 22.6. 1. 57. Female. 0.504. 0.5. 0. 1. Rural. 0.516. 0.5. 0. 1. 0.17. 0.91. 0. 1. 1. 3. Country economic outlook. Ideology right. Presidential approval. n i on Naatti o N. Education Attainment. Presidential disapproval Presidential regular. 0.219. 0.413. 0. 1. 0.388. 0.49. 0. 5. 0.238. 0.426. 0. 1. 0.336. 0.212. 0. 1. 0.247. 0.432. 0. 1. 0. 1. 0. 1. ‧ ‧. Ideology center. 學學. Ideology left. ‧ 國 ‧ 國. Individual economic outlook. 0.385 0.153 治政治政 0.228 大大0.303 立立 2.255 0.682. e rr ssiitt yy. Age. aal l 0.215 0.411 iv n CChh U 0.5 eennggcc0.513 hhii U. News everyday. 0.477. 0.5. 0. 1. News week. 0.314. 0.464. 0. 1. News month. 0.0662. 0.249. 0. 1. News rarely. 0.0721. 0.259. 0. 1. 28. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(30) Chapter 4 Empirical Results 4.1 Statistical Analyses. Because my dependent variable is coded as dichotomous, the main models are estimated using probit. I also estimated logit models for robustness checks. The predicted probabilities were generated using CLARIFY (Tomz, Wittenberg, & King 1999; King, Tomz, & Wittenberg 2000). My data has a pooled cross-sectional structure for the years 2010 to 2019. The results from the models on the perception of corruption can be seen in Table 3.. As expected, in Model 1 and Model 3, the total number of CICIG investigations has a. 治政政治大大 effort is an important 立 Guatemalans. It indicates that the government’s anti-corruption 立. positive and statistically significant effect on the perception of corruption among. ‧ 國 ‧ 國. 學學. determinant of perception of corruption in the country from its begin in 2008 to the end in 2018. In the Model 2 and Model 4, I decompose the total number of CICIG investigations. ‧ ‧. into two variables. Both coefficients for the number of charged investigations and the number of uncharged investigations are positive and statistically significant. This finding suggests. n i on Naatti o N. e rr ssiitt yy. that the number of people identified with corruption with sentences or without them increases the likelihood that an individual perceives corruption in the country. More importantly, the. aal l i v This finding suggests that n CCofhhuncharged investigations. coefficient than that for the number U eennggcchhii U the number of charged investigations makes citizens more likely to perceive corruption.. results demonstrate that the number of charged investigations has a more significant. Table 3 also demonstrates exciting findings for the control variables. Among Guatemalans, older generations are more likely to perceive corruption. Also, those that consider that the country’s national economic performances are negative are more likely to increase the likelihood of perceived corruption. However, the coefficient of egotropic economic outlook does not reach statistical significance. Education attainment has a positive and statistically significant coefficient for the perception of corruption. Specifically, higher educated individuals tend to perceive corruption. This finding suggests that highly educated individuals. 29. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(31) have more access to information about corruption, so they are more likely to perceive corruption.. My results show whether individual life in a rural or urban area does not matter for the likelihood of the perception of corruption. Contrary to expectations, the presidential approval or disapproval do not increase perception. There are also puzzling results regarding political ideology and the frequency of reading news. For instance, both leftist and rightist leaning individuals are less likely to perceive corruption than those who do not have a clear political ideology. Individuals that follow news every day, those that follow news every week, and those that rarely follow news have a higher likelihood of perceiving corruption than those. 治政政治大大立立學學 ‧ ‧. ‧ 國 ‧ 國. who never follow news.. e rr ssiitt yy. n i on Naatti o N. aal l iv n CChh U eennggcchhii U. 30. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(32) Table 3 Probit and Logit Models on Perception of Corruption in Guatemala Variables Total number of CICIG’s investigations. Model 1 Probit 0.00225*** (0.000364). Model 2 Probit. Model 3 Logit 0.00380*** (0.000623). Model 4 Logit. Total number of uncharged investigations. 0.00212*** (0.000376). 0.00356*** (0.000645). Total number of charged investigations. 0.00329*** (0.000870). 0.00559*** (0.00152). Female Rural. Presidential approval Presidential disapproval. News everyday News week News month News rarely Constant. e rr ssiitt yy. ‧ ‧. Education attainment. Presidential regular. -0.173*** (0.0654) -0.0735 (0.0693) 1.552*** (0.242) -0.330*** (0.113) -0.0497 (0.105) -0.390*** (0.109) -0.0349 (0.100) -0.248** (0.109) 1.761*** (0.190) 0.0531 (0.239) 0.184 (0.244) 0.0606 (0.234) 0.653*** (0.133) 0.661*** (0.136) 0.263 (0.169) 0.374** (0.166) -0.562* (0.312). 學學. Ideology right. n i on Naatti o N. Ideology center. ‧ 國 ‧ 國. Sociotropic. Ideology left. -0.100*** (0.0376) -0.0405 (0.0399) 0.888*** (0.138) -0.197*** (0.0652) -0.0201 (0.0601) -0.203*** (0.0635) -0.0101 (0.0577) -0.122* (0.0633) 1.010*** (0.108) 0.0232 (0.141) 0.108 (0.143) 0.0294 (0.138) 0.383*** (0.0809) 0.389*** (0.0826) 0.152 (0.103) 0.220** (0.101) -0.330* (0.184). 治政政治大大立立. Age. Egotropic. -0.102*** (0.0376) -0.0416 (0.0399) 0.884*** (0.137) -0.194*** (0.0651) -0.0244 (0.0600) -0.215*** (0.0628) -0.0165 (0.0574) -0.134** (0.0626) 0.998*** (0.107) 0.0340 (0.140) 0.114 (0.143) 0.0413 (0.138) 0.391*** (0.0807) 0.396*** (0.0823) 0.165 (0.102) 0.229** (0.100) -0.312* (0.183). aal l iv n CChh U eennggcchhii U. 31. -0.170*** (0.0654) -0.0716 (0.0693) 1.558*** (0.242) -0.334*** (0.113) -0.0437 (0.105) -0.369*** (0.110) -0.0233 (0.101) -0.227** (0.110) 1.783*** (0.191) 0.0364 (0.239) 0.173 (0.244) 0.0421 (0.235) 0.641*** (0.133) 0.649*** (0.137) 0.241 (0.169) 0.359** (0.166) -0.593* (0.313). DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(33) Pseudo R squared Observations. 0.0469 6,041. 0.0472 6,041. 0.0469 6,041. 0.0471 6,041. Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1. In Table 4, I calculate the predicted probability of key variables in Model 1 to analyze an individual’s propensity further to perceive corruption. If all the variables are held at their means, a Guatemalan citizen has 79.19% of the probability of perceived corruption. Different situations can increase or decrease the percentage of the probability of perceived corruption. When fewer individuals are under CICIG’s investigations (at one standard deviation lower), the probability of an individual perceiving corruption is 75.85%. However, when there are more individuals under CICIG’s investigations (at one standard deviation higher), the. 治政政治大大立立. probability of perceiving corruption increases by 6.4%.. ‧ ‧. ‧ 國 ‧ 國. 學學 e rr ssiitt yy. n i on Naatti o N. aal l iv n CChh U eennggcchhii U. 32. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(34) Table 4 Predicted Probabilities of Corruption Perception for Model 1 Situation. %Probability of perceiving corruption. All at mean. 79.19%. Fewer individuals under CICIG’s investigations. 75.85%. Higher individuals under CICIG’s investigations. 82.25%. Female. 77.71%. Male. 80.64%. No sociotropic. 80.44%. Sociotrotropic. 74.65%. 治政政治大大立立. Leftist-leaning Rightist. 81.11% 72.63% 84.73%. ‧ ‧. Higher educated. 77.77%. 學學. Lower educated. ‧ 國 ‧ 國. Center. 75.29%. Note 1: Fewer= the number of individuals under CICIG investigation at 1 standard deviation lower. Note 3: Lower educated=education at 1 standard deviation lower Note 4: Higher educated=education at 1 standard deviation higher. e rr ssiitt yy. n i on Naatti o N. Note 2: More= the number of individuals under CICIG investigation at 1 standard deviation higher. aal l iv n C C U resulted to be an important hheeeducation Holding all the variables at their means, ii U nnggcchhattainment. determinant. The breach between those that are highly educated and low educated is significantly large. Higher educated individuals are more likely to perceive corruption (84.73%). However, those that are lower educated are considerably less likely to perceive corruption (72.63%) at the time others do.. Some of the interesting scenarios are related to the sociotropic evaluation variable. Those that consider the national economy is going good have a 74.65% chance of perceiving corruption. In contrast, having a negative sociotropic evaluation increases the probability of 80.44%, holding all the variables at their means. This shows how the perception of corruption is affected by the economic outlook. Another situation is the difference founded by ideology. 33. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(35) preferences. The predicted probabilities suggest that left-leaning has less probability of perceiving corruption (75.29%) compare to those in the right-leaning (77.77%) and center (81.11%). Finally, the simulation results show that a female respondent has less probability (77.71%) than male respondents (80.64%) to perceive corruption. Simulations using key variables of Model 1 different situations can create a difference in the likelihood that an individual perceives corruption. Overall, Table 3 and Table 4 demonstrate that the presence of the government’s anti-corruption efforts, education attainment, and economic outlook serve as important determinants for explaining the likelihood of corruption perception in Guatemala. 4.2 Comparative Simulation Analyses. ‧ 國 ‧ 國. 學學. 治政政治大 Based on the results in Model立 1 and Model 2, I created a大 series of simulations for different 立 scenarios of corruption perception. Figure 1, compares the predicted probabilities of corruption perception for educational attainment and total number of CICIG’s investigations.. ‧ ‧. Figure 2 compares the predicted probabilities of corruption perception for the number of charged investigations and the number of uncharged investigations. In addition to Figure 1. e rr ssiitt yy. n i on Naatti o N. and Figure 2, I also conducted simulations of the likelihood of corruption perception under different scenarios for political ideology, media exposure, and economic outlook.. aal l iv n CChh U eennggcchhii U. 34. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(36) Figure 2 Comparative Simulation Analyses for Total Number of CICIG’s Investigations, Education Attainment, and Corruption Perception in Model 1. 治政政治大大立立 ‧ ‧. ‧ 國 ‧ 國. 學學 e rr ssiitt yy. n i on Naatti o N. Figure 1 shows that both the number of CICIG’s investigations and the educational attainment level have positive effects on increasing the likelihood of perceived corruption. However, the simulation shows that higher educational attainment has a more substantial. aal l iv n CChh U 5 suggest that a change of 1% in iU eennshown investigation. The calculated arc elasticities ggcchinhiTable effect on increasing the likelihood of corruption perception than the total number of CICIG’s. the total number of CICIG’s investigations creates an increase of the perception of corruption ranging from 5.43% to 15.35%. Moreover, a change of 1% in educational attainment creates an increase in the perception of corruption ranging from 6.79% to 19.60%.. 35. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(37) Table 5 Elasticities Intervals for Perception of Corruption. Respect to total number of CICIG’s investigations. Min Max. Respect to education attainment. Respect to. Respect to. number of uncharged investigations. number of charged investigations. 5.43%. 6.79%. 4.28%. 0.66%. 15.35%. 19.60%. 9.44%. 4.67%. Figure 3 Comparative Simulation Analyses for the Number of Uncharged. 治政政治大大立立. Investigations, Charged Investigations, and Corruption Perception in Model 2. ‧ ‧. ‧ 國 ‧ 國. 學學 e rr ssiitt yy. n i on Naatti o N. aal l iv n CChh U eennggcchhii U. The simulation results in Figure 2 show that both the number of uncharged investigations and the number of charged investigations increase the likelihood of perceived corruption. However, the result shows that the number of charged investigations has a more substantial effect on increasing the likelihood of corruption perception than the number of uncharged investigations. The calculated arc elasticities shown in Table 5 suggests that a change of 1% in the number of CICIG’s uncharged investigations creates an increase in the perception of. 36. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(38) corruption ranging from 0.66% to 4.67%. Moreover, a change of 1% in the number of CICIG’s charged investigations creates an increase in the perception of corruption ranging from 4.28% to 9.44%.. Table 6 shows the simulation results for predicted probabilities of perceiving corruption under different scenarios of political ideologies. On average, an individual with left-leaning characteristics has a percentage of perceiving corruption of 75.89% in Model 1 and 76.07% in Model 2. An individual with right-leaning characteristics has a 78.37% percentage of perceived corruption in Model 1 and 78.47% in Model 2. Overall, the simulation shows that leftist leaning individuals perceive less corruption than rightist leaning individuals.. 治政政治大大立立 ‧ ‧. ‧ 國 ‧ 國. 學學 e rr ssiitt yy. n i on Naatti o N. aal l iv n CChh U eennggcchhii U. 37. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(39) Table 6 Predicted Probabilities of Corruption Perception for Individuals with Different Political Ideologies. Variables. Model 1. Model 2. Leftist. 75.89%. 76.07%. Rightist. 78.37%. 78.47%. Table 7 Predicted Probabilities of Corruption Perception for Individuals with Different Frequency of Watching News. 80.95%. Week. 81.07%. 74.02%. 74.00%. 76.28%. 76.05%. n i on Naatti o N. Rarely. 81.13%. ‧ ‧. Month. 80.96%. e rr ssiitt yy. Day. ‧ 國 ‧ 國. Model 1. 學學. Variables. 治政政治大大 Model 2 立立. aal l iv n C C U news daily or weekly are more Table 7. The results show that an individual hheennggthat ii U cchhfollows The predicted possibility of perceiving corruption related to media exposure can be seen in. likely to perceive corruption. Finally, simulations on individuals’ sociotropic evaluation confirmed that a negative economic outlook increases the perception of corruption.. 38. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.

(40) Table 8 Predicted Probabilities of Corruption Perception for Individuals with Different Sociotropic Evaluations. Model 1. Model 2. Worst. 80.97%. 80.98%. Same. 78.21%. 78.24%. Better. 75.23%. 75.26%. 治政政治大大立立 ‧ ‧. ‧ 國 ‧ 國. 學學. n i on Naatti o N. e rr ssiitt yy. Variables. aal l iv n CChh U eennggcchhii U. 39. DOI:10.6814/NCCU202001225 DOI:10.6814/NCCU202001225.