Institutional Design, Prosecutorial Independence, and Accountability: Lessons from the International Commission against Impunity in Guatemala (CICIG)

: In 2007 the International Commission against Impunity in Guatemala began working to support the work of the Public Prosecutor’s Ofﬁce to investigate and prosecute cases of corruption. In this short article, I address three questions: What was the design of CICIG? What were its results? How did the institutional design contribute to the impact it had in Guatemala? To answer these questions, I ﬁrst discuss the problem of impunity and provide an interdisciplinary review of the factors that impact when a state punishes crime. In the following section, I brieﬂy explain the institutionalist framework that guides my analysis of the work of CICIG and explain the theoretical importance of institutional design in the functioning of prosecutorial organs. In the third section I describe the institutional design of CICIG, its impact, and its limitations. In this section I explain how key characteristics in its design made this international commission a creative solution against impunity in a context where prosecuting state agents for human rights violations or corruption was virtually impossible. I conclude this article highlighting some key lessons from CICIG on matters of institutional design.


Introduction
The problem of impunity in cases of corruption has long been recognized (Transparency International 2001), but the idea of creating a permanent multilateral mechanism to tackle this problem has recently gathered momentum. This is most evident in the United Nations General Assembly Special Session against Corruption that convened in June of 2021 to discuss the challenges and measures to combat corruption and strengthen international cooperation (and also in the call for this special issue). As scholars and policy makers reflect on the merits of a new international mechanism to investigate and prosecute corruption, it is important to reflect on what has been done in the past and what lessons we can learn from those efforts.
On 12 December 2006, an unprecedented commission was created with the signing of the Agreement between the United Nations and the State of Guatemala on the Establishment of an International Commission against Impunity in Guatemala (CICIG). This new commission aimed to provide assistance to the Public Prosecutor's Office in the investigation and prosecution of illegal and clandestine security groups or organizations that were known to be involved in organized crime and corruption, and who, through threats, intimidation, and bribes, contributed to a vicious cycle of impunity. The mandate of CICIG was renewed four times and it ceased operations on 4 September 2019. In its 12 years of existence CICIG participated in the investigation and prosecution of over 120 corruption and human rights cases.
The Anticorruption Protocol to the United Nations Convention Against Corruption (APUNCAC) would create an International Commission Against Corruption (ICAC), a permanent multinational commission to investigate corruption cases. The ICAC would be modeled after CICIG. It would employ UN inspectors to conduct investigations, then refer cases to dedicated domestic anticorruption courts. This recommendation is based on the assumption that doing so could lead to similar results as those that CICIG had in Guatemala. In this short article I aim to contribute to this conversation as a political scientist, addressing the following questions: what was the design of CICIG, what were its results, and how did the institutional design contribute to the impact it had in Guatemala? To answer these questions, I first discuss the problem of impunity and provide an interdisciplinary review of the factors that influence the vigor of state actions to punish, to better situate where institutional design fits in this literature. In the following section I briefly explain the institutionalist framework that guides my analysis of the work of CICIG and explain the role of institutional design and how this can influence prosecutorial independence and accountability. In the third section I describe the institutional design of CICIG, its impact, and its limitations. In this section I explain how key institutional design characteristics made this international commission a creative and unprecedented hybrid (domestic and international) legal space for governance, i.e., a solution against impunity in a context where the prosecution of state agents for human rights violations or corruption was virtually impossible. The last section highlights some key lessons from CICIG on matters of institutional design, and in the conclusion I provide some final thoughts on how APUNCAC draws on these lessons.

Understanding Impunity and the Factors That Sustain It
The United Nations Set of Principles for the Protection and Promotion for Human Rights through Action to Combat Impunity defines impunity as "the impossibility, de jure or de facto, of bringing the perpetrators of violations to account-whether in criminal, civil, administrative or disciplinary proceedings-since they are not subject to any inquiry that might lead to their being accused, arrested, tried and, if found guilty, sentenced to appropriate penalties, and to making reparations to their victims." 1 The impact that impunity can have on a country cannot be overstated. Bounded by criminal and international law, a state has the duty to investigate, seek prosecution, and punish the person responsible for violating a fundamental right. Failure to do so has serious implications for the rule of law, governance, and democratic consolidation (Ansolabehere 2007;Kaufmann 2005;O'Donnell 2004;Thome 1998). Thus, the investigation and prosecution of crimes can strengthen the rule of law, as doing so can provide certainty, equality, and impartiality, through judicial protection (Raz 1979). Punishment also serves as a signaling device that the state can use to show that rights violations and the abuse of power will not be tolerated (Kim and Sikkink 2010). In contrast, lack of punishment or a low probability of punishment has been shown to feed more crime (Chalfin and McCrary 2017;Nadanovsky et al. 2009), and can weaken trust in the police and judicial institutions (Fernandez and Kuenzi 2010;Hernández 2019).
Given the negative impact that impunity can have on the rule of law and democratic governance, what do we know about when and why a state punishes or does not punish crimes? Scholars from disciplines as diverse as political science, economics, sociology, socio-legal studies, and criminology have identified conditions when a judicial system is more or less likely to punish crime, factors that range from the micro level to the macro level of analysis.
At the micro level, various individual-level factors can impact why or when cases remain unpunished. The managerial style of judges (Buscaglia 2006) and the investigative strategies, tactics, and experience of prosecutors and police officers can impact clearance rates (Keel et al. 2009;Pizarro et al. 2018). Discretional factors play an important role into why actors may or may not investigate or prosecute a case (Black 1970). Political will has been highlighted by research on high profile human rights and corruption cases, but politics has also been found to impact when prosecutors punish ordinary crimes (Davies 2007). We know that police officers and prosecutors have enormous discretion in their decisions while investigating or prosecuting a crime (Bibas 2009;Shapiro 2013), but the problem is when those discretion-based decisions are biased. Some researchers have found different sources of bias. The perceived characteristics of the victim, the offender, or the location of the crime can introduce bias in a decision to investigate or prosecute a crime. For example, some have found that perceived race and socio-economic status of the victim can reduce the likelihood for the case to be cleared (Litwin 2004). Others have found that the victim's race and the use of firearms may also play a role in how investigators respond to a crime scene (Litwin and Xu 2007). Research in Latin America has found that cases with female victims are less likely to be investigated (Michel 2020;Walsh and Menjívar 2016). Finally, we also know that cognitive biases, such as tunnel vision and implicit bias, can alter the decision-making process and severely impact the fate of a case (Brants 2013). However, while micro-level approaches can be useful to explain individual behavior and are useful to inform the design of institutional policies (such as education and training of state agents), these approaches do not necessarily help us understand systemic patterns of impunity or corruption.
At the macro level, research has looked at various structural and international factors. For instance, when explaining the variations observed across countries in punishing state agents for human rights violation or corruption, scholars have highlighted the role of past repression and the type of transition to democracy experienced in a country (Elster 2004); the presence of networks between domestic and transnational NGOs (Keck and Sikkink 1998;Murdie and Davis 2012;Risse-Kappen et al. 1999; the influence of international law and foreign/regional courts as tools for societal actors to mobilize the law and defend rights (Davis 2014;Simmons 2009); and the role of regional diffusion (Dancy and Sikkink 2011;Kim and Sikkink 2010). Others have highlighted the role of economic development in police performance and clearance rates (Pare 2014). These macro-level approaches have brought to light the key role that international (and regional) courts and transnational networks play in contexts where domestic institutions are unwilling or incapable of providing justice, but these approaches also tend to ignore the design of institutions, as well as how various actors (domestic or international) respond or relate to each other according to a (domestic) institutional setting.
In between these two levels, at the meso level we find another set of explanations that focus on the role of domestic and institutional factors. The availability of resources has been found to impact impunity in different ways. On the one hand, in human rights and corruption cases the disparity of resources (and power) between victims and the state can ensure that a state agent will not face punishment for committing a crime, when the state's preference is to avoid accountability (Brinks 2008). On the other hand, resources (i.e., budget, personnel, forensic tools) can impact a state's own capacity to investigate and prosecute crime (Borg and Parker 2001;Wellford and Cronin 1999). The context in which judicial institutions operate also matters. Some scholars have stressed the importance of legal culture for state agents to foster the rule of law (Couso et al. 2010;Domingo 1999;Hilbink 2007;Sarat and Scheingold 2008). Others have noted that the amount and severity of crime can play a role in punishment rates (Gottfredson and Hindelang 1979). Related to this, some have found that in contexts of cartel-related violence, punishment of violent crime can diminish (Huebert 2019). This literature has been useful to understand the role of institutions, but at the same time these accounts tend to ignore the relationship between law and society in a globalized world.
In contrast to the preceding literature, the struggle against impunity may be understood as an issue of global governance. I draw on research that focuses on the design of institutions and how such design shapes institutional performance. Through the lens of transnational legal pluralism, I interpret law (and institutions) as a social phenomenon that constitutes a (dynamic) regulatory (legal) space in which the state, national society, and world society are all integral elements (Zumbansen 2012, pp. 324-25). Law constitutes and is constituted by these various interactions. The theoretical framework driving my analysis of CICIG's institutional design is informed by the institutionalist (political science) tradition (Thelen and Steinmo 1992;Mahoney and Thelen 2010) that assumes institutions to be dynamic (not static), within which actors' behavior responds to power-distributional implications of the design of institutions (Mahoney and Thelen 2010, pp. 8-13). In the next section I explain in more detail the theoretical framework and explain how institutional design can improve the capacity of a state's prosecutorial organ to investigate and prosecute its own agents (i.e., for corruption or human rights violations).

Prosecutorial Independence and Accountability
State responses to crime have to be recognized as part of a three-stage process that involves (1) the notification or knowledge of a crime, which triggers (2) the investigation and prosecution of a crime, followed by an (3) adjudication. This three-stage process involves four different institutions within the judicial system: the police (to investigate reports of a crime), the prosecutorial office (to gather evidence, present an indictment, and prosecute), the public defender's office (to represent the accused), and the judiciary (to adjudicate cases). 2 We know that judicial independence has an important role in how courts protect and adjudicate rights (Ansolabehere 2007;Ferejohn 1999;Helmke and Ríos-Figueroa 2011;Lupu 2013;Sadek and Cavalcanti 2003;Shapiro 2013;Skaar 2011). However, adjudication in the courts is always preceded by a criminal investigation and a prosecution led by the Public Prosecutor's Office. If we want to prevent impunity, we have to recognize the key role that prosecutors play in this process. With the unilateral power to decide what, when, and whom to prosecute and bring to trial, a state's prosecutorial organ is a gatekeeper to the judicial system. This power becomes particularly problematic in cases where state agents are corrupt or commit human rights abuses, because in both of these instances the state must be willing (and capable) to keep its own agents accountable for these types of crimes and abuses of power.
Past research has shown how various features of institutional design can impact the work of prosecutors, who should be both independent from political pressures on how they conduct their work, as well as accountable for their prosecutorial decisions (Michel 2017(Michel , 2019. This mix of both independence and oversight is necessary to ensure that prosecutorial discretion is not abused or misused, and to provide accountability when failures of commission or omission occur. Prosecutorial independence improves the likelihood that cases related to human rights or corruption are investigated, prosecuted, and punished (Brinks 2007(Brinks , 2008Michel 2017;Ríos-Figueroa 2012;Van Aaken et al. 2010;Van Aaken et al. 2004). By institutional design, two key factors can greatly impact prosecutorial independence de jure: the location of the prosecutorial organ within the structure of government, and the rules of appointment, tenure, and promotion. Across countries we find that prosecutorial organs are housed in the executive branch (like in the US), the judiciary (like in Italy), or can be designed as an autonomous institution (like in Guatemala). Probably more important for shielding prosecutors from external political pressures are the rules of appointment, tenure, and promotion. If the career of a prosecutor depends on politicians, it is likely that they will be more vulnerable to pressures to use their prosecutorial powers to favor those who can impact their career path. If a prosecutor is elected, like in the US, then they are left vulnerable to the whims of public opinion (Savelsberg 1993, p. 932).
Just like variation in the design of the prosecutorial organ can impact how independent a prosecutor is, recent research has found that by design legislators can also introduce mechanisms of control over the discretionary powers of a prosecutor that can bring some accountability over their prosecutorial decisions. A very obvious way in which prosecutorial powers can be kept "accountable" is through the judicial review of key decisions, such as the decision to prosecute or not, the type of charges made, or the request of a warrant. The existence of a disciplinary procedure for prosecutors is another accountability mechanism, but this must be clearly regulated through guidelines that describe the grounds on which disciplinary action must be taken and the sanctions. Obviously, disciplinary action should be impartially applied. Another important accountability mechanism is to require prosecutorial organs to publish annual reports that are accessible to the public. Furthermore, an arguably less well-known accountability mechanism can be introduced by institutional design with the implementation of strong participation rights for victims, such as private prosecution. It has been found, for instance, that when countries allow for private prosecution rights there is a higher likelihood of initiation of prosecutions against state officials (Collins 2010;Dancy and Michel 2016;Michel 2018;Michel and Sikkink 2013;Dancy and Sikkink 2011).
Very briefly I must explain what private prosecution means in comparative procedural law, as understanding this right greatly helps to understand how CICIG was able to operate in Guatemala. The right to private prosecution allows victims and their lawyers to open a criminal investigation and actively participate throughout every stage of the investigation and prosecution. Private prosecution can be auxiliary, where the private party is only allowed to adhere to the prosecution of the state, or private prosecution can be autonomous, where the private party also has the power to press different charges from those introduced in the indictment of the public prosecutor. Some countries grant this right to NGOs or other institutions who may have a "collective" interest in the case. In Guatemala, for instance, any "aggravated party" can claim standing as private prosecutor; this includes "the victim directly affected by the crime, their relatives, and organizations whose work relates directly to those rights that a crime has affected" (Michel 2018, p. 67). A private prosecutor, thus, can keep a public prosecutor accountable because they have the right to appeal all the decisions the public prosecutor makes.
Institutional design, however, does not operate outside of politics. Design may help redistribute power and influence the incentives that prosecutors face, but there are other factors that can impact prosecutorial independence and prosecutorial accountability de facto. 3 One reason could be, for instance, that disciplinary rules are not applied impartially or based on merit. The context in which prosecutors and judges operate may also impact the faculty of these actors to make decisions independently, for instance when their physical safety and security is not provided nor guaranteed by the state (FJEDD 2017, p. 40). The power of accountability mechanisms can also decrease if victims lack access to private prosecutors (as lawyers are expensive).
In the next section I briefly review the work of CICIG, highlighting how by design this unprecedented hybrid institution supported the work of the Public Prosecutor's Office and contributes to the fight against impunity.

The Design and Impact of CICIG
Despite massive judicial reforms that had begun after the transition to democracy (in 1986) and even after the end of a 36-year-long civil war that concluded with the signing of the Peace Accords (in 1996), Guatemala found itself entering the new millennium with increased insecurity and rampant impunity. Plagued by rising crime, gang violence, and drug trafficking, Guatemala has since then been considered one of the most violent countries in the world (World Health Organization 1997, 2002. In the last 20 years, the percentage of homicide cases that remain unsolved has never gone below 90% (Michel 2018, p. 84). Impunity in cases where the defendant was a state agent or was linked to a criminal organization was virtually guaranteed. By the late 1990s it was well known that one key obstacle for the state to prosecute crime was the intimidation and death threats targeted against criminal justice operators (judges, prosecutors) and human rights activists, lawyers, and witnesses. These threats mostly came from former military personnel and former members of dissolved paramilitary groups who were afraid of prosecution, but even these threats were left without investigation or prosecution (Heasley et al. 2001). It was well understood that a key obstacle for the state to end impunity was the state's own incapacity to deal with these actors that paralyzed justice and undermined the rule of law.
The creation of CICIG was not without domestic political resistance and it required time (and the conjunction of various factors) to materialize. Based on grassroots concerns and suggestions (Call 2019, p. 3), in 2002 a proposal emerged from civil society for the UN to develop a mechanism to help the Guatemalan government combat impunity. The following year an initial agreement was signed between the UN and the Guatemalan government to create a commission that was granted independent prosecutorial powers, which was named the "Commission for the Investigation of Illegal bodies and Clandestine Security Apparatuses". This first agreement, however, was declared unconstitutional by the Guatemalan Constitutional Court in 2004, which ruled that the constitution granted the monopoly to prosecute crime to the Public Prosecutor's Office. In the next two years, an alliance of international and civil society actors successfully lobbied for (and contributed to) the drafting of a new agreement that found support among key political actors in government (Call and Hallock 2020, p. 8).
In December of 2006, the Agreement between the United Nations and the State of Guatemala on the establishment of the International Commission Against Impunity in Guatemala (Comisión Internacional contra la Impunidad en Guatemala, or CICIG) was signed. CICIG was an unprecedented experiment in "internationalized prosecution" (Trejo and Nieto-Matiz 2019): an independent investigative international organ designed to work within a domestic jurisdiction to support a Public Prosecutor's Office (Ministerio Público) in the investigation and prosecution of "illegal security forces and clandestine security organizations, and any other criminal conduct related to these entities operating in the country" (Art. 1 §1a). To achieve its objective, CICIG was given three functions: (1) to determine the existence of these groups and organizations, (2) to collaborate with the state in the investigation, prosecution, and punishment of the members of such groups and organizations, and (3) to recommend to the state of Guatemala policies and institutional reforms necessary to dismantle these illegal security groups and clandestine security organizations (Art. 2).
The drafters of the Agreement were careful in the language used, never mentioned the word "corruption," and provided a very broad definition of what individuals could be investigated for within their mandate, which were those who belong to groups or forces who "(i) commit illegal acts in order to affect the full enjoyment and exercise of civil and political rights, and (ii) are linked directly or indirectly to agents of the State or have the capacity to generate impunity for their illegal actions" (Art 1 §1d). In a nutshell, this commission aimed to tackle the problem of weak prosecutorial independence, knowing that the failure to conduct criminal investigations was the gateway to impunity. This agreement was approved by the Guatemalan Congress on August of 2007.
In contrast to the previous commission that had been proposed, the design of CICIG was clearly buttressed in domestic criminal procedure law. To properly aid in the prosecution of crimes, CICIG was granted standing as a private prosecutor (querellante adhesivo) (Art 3 §1b). Like any other private prosecutor, CICIG could conduct an independent investigation and could participate in the prosecution of cases next to the public prosecutor. Carlos Castresana, a Spanish former prosecutor with vast experience in corruption and economic crimes who would later be appointed as the first commissioner of the CICIG, was in charge of its design (Solano 2019, January 14), and in this process he clearly took into account domestic legal tools and practice. Private prosecution had been used in Guatemala as an effective legal tool by human rights activists seeking criminal accountability for human rights violations, and judges, prosecutors, and the public were quite familiar with the practice of private prosecutors. 4 By then it was known that private prosecution could help keep cases open and improve the chances of cases to reach trial (Michel 2018). Thus, by design CICIG itself relied on private prosecution as a tool to improve the accountability of the Public Prosecutor's Office.
As an "auxiliary prosecutor," CICIG was designed to be an autonomous investigative body with various features that strengthened prosecutorial independence. It had financial autonomy from the Guatemalan government, as it was financed by the international community (Art. 7 §1). In terms of costs, the Guatemalan government was only required to provide office space (Art. 7 §2) and provide security and protection to CICIG staff and anyone who cooperated with CICIG (Art. 8). Although it was indeed limited by criminal procedure law to limited prosecutorial powers, 5 it was also empowered well beyond just being an ordinary private prosecutor. For instance, CICIG could enter and implement cooperation agreements with the Public Prosecutor's Office, the police, and the courts (Art. 3 §f). It had great investigative powers as it could request information that state institutions were obligated to provide (Art 3 §h); it was granted legal personality (Art. 4), and its international personnel was protected with the privileges and immunities granted to diplomats by international law (Art. 10). In addition, as part of the agreement, the state of Guatemala was required to provide assistance in any investigation that CICIG undertook, guarantee access to information, appoint special prosecutors that would work with CICIG, and create a special police unit to aid in the investigations (Art. 6). By design, CICIG also had autonomy in the sources and administration of technical and human resources. The head of CICIG was appointed by the Secretary-General of the UN, who was then responsible to recruit its own (international and domestic) personnel. Therefore, CICIG, as an autonomous investigative body with auxiliary prosecutorial powers, was an original and clever response by domestic and international state and non-state actors, and it constituted a hybrid (domestic and international) legal space for governance without precedents. CICIG was not the first time that the international community was seen as a solution to injustice in the face of a state that was unwilling or incapable to prosecute and punish its own state agents for wrongdoing. By 2007, various ad-hoc courts had been created for specific events or purposes, with a temporary mandate and a defined jurisdiction, like the International Criminal Tribunal of Yugoslavia and of Rwanda. Hybrid or internationalized domestic courts had also been created (i.e., domestic courts that operated with the support of the United Nations, like the Special Court for Sierra Leone or the Khmer Rouge Tribunal). The International Criminal Court (ICC), a court with its own prosecutorial organ, was designed to act when domestic jurisdictions failed to prosecute individuals for international crimes. CICIG, however, was unique in that it did not internationalize domestic courts nor replace them with ad-hoc courts. Instead, with CICIG a new hybrid model of supranational support emerged: an international investigative body designed to serve as an auxiliary of the Guatemalan Public Prosecutor's Office.
By all measures CICIG was quite successful in fulfilling its functions. Between September of 2007 until its mandate ended in September of 2019, CICIG identified over 60 criminal networks, investigated over 120 criminal cases, and during its 12 years of work as private prosecutor CICIG collaborated with the Public Prosecutor's Office to achieve over 400 convictions in 25 cases (CICIG Comisión Internacional Contra la Impunidad en Guatemala, p. 13). Although its job was never smooth or easy, and sometimes has been criticized, the overall success of CICIG has been widely praised (Call 2019;Call and Hallock 2020). Its impact on prosecutions and convictions shows that CICIG effectively improved prosecutorial and judicial independence, providing protection to "investigators, prosecutors, witnesses and judges." (Trejo and Nieto-Matiz 2019, p. 4). The CICIG was led first by Carlos Castresana, followed by Francisco Dall-Anese, and Iván Velásquez. Despite differences between the work of each one, all three commissioners were involved in investigating state agents of all ranks.

•
Carlos Castresana (2007Castresana ( -2010. He began the first investigation against a former head of state that eventually led to the prosecution of former president Alfonso Antonio Portillo Cabrera and two of his cabinet members. Castresana also led the investigation of the Rosenberg case, which increased public confidence in CICIG and cleared the incumbent president Alvaro Colom from allegations that he was responsible for the death of Rodrigo Rosenberg, a prominent lawyer. (2010)(2011)(2012)(2013). During his tenure he led the prosecution of various public officials in Caso Pavón (for extrajudicial killings); the investigation of former head of the National Police, Marlene Blanco Lapola in Caso Blanco (also for extrajudicial killings); and during his tenure CICIG successfully prosecuted immigration officials in Caso Migración (for human trafficking).

•
Ivan Velásquez (2013-2019). The last commissioner of CICIG will be remembered for his increased efforts to push for more prosecutions. More prominently, he led the investigation and prosecution of President Otto Pérez Molina and Vice-President Roxana Baldetti for their involvement in a customs fraud network known as "La Línea", a network well entrenched within the government of Otto Pérez Molina. For the first time in the history of the country, a democratically elected president was arrested and indicted, and was forced to resign from office.
Aside from its obvious impact on prosecutions, CICIG also contributed to key changes within Guatemalan institutions that were fundamental for its success as a private prosecutor and which are likely to have a long-term impact on the criminal justice system of Guatemala. The contribution of CICIG included the "support of training and promotion of legal reforms aimed at improving the efficiency of the administration of justice." (IACHR Inter-American Commission on Human Rights, p. 39). In the first two years of its mandate, CICIG proposed various institutional reforms, many of which were eventually adopted (Call and Hallock 2020, p. 17). For instance, CICIG helped in the creation of a specialized unit against impunity within the Public Prosecutor's Office and helped select and train prosecutors for that unit. It also promoted the creation of courts designed to adjudicate only high-profile cases related to human rights violations, organized crime, and corruption. The Tribunales de Crímenes de Alto Impacto, created in 2009, effectively provided more resources and protection (i.e., more independence) to judges (Trejo and Nieto-Matiz 2019, p. 33). CICIG also proposed key reforms to the rules of appointment and tenure, as well as disciplinary rules for public prosecutors that improved prosecutorial independence and accountability. CICIG also contributed to the adoption of laws that greatly improved witness protection and provided technical support to modernize and systematize criminological and judicial data to better inform policy making. Overall, CICIG's strategy behind these reforms was to improve state capacity and strengthen the rule of law (CICIG Comisión Internacional Contra la Impunidad en Guatemala,. The impact of its work was also perceived in the court of public opinion and for this reason CICIG helped improve public trust in judicial institutions. CICIG enjoyed the trust of Guatemalans throughout its mandate, particularly during the tenure of Iván Velasquez when it enjoyed the trust of over 70% of the population. However, over time CICIG positively impacted citizen's perceptions of corruption and of the judiciary itself. Survey studies found that trust in the Constitutional Court and the Public Prosecutor's Office increased significantly over time (Zechmeister and Azpuru 2017).
The experiment of CICIG also had important defects. If CICIG had a mortal flaw it was its temporal design. CICIG's mandate was for a two-year period, renewable only through the agreement of both parties. As noted earlier, the agreement was renewed four times and terminated when political support during the administration of Jimmy Morales (2016-2020) waned, to the point that Morales declared commissioner Velázquez as persona non grata in August of 2017 as reprisal for an investigation of electoral fraud that involved Morales' political party. Fueled by the lack of external political pressure from the Trump administration, which exhibited high tolerance towards regimes that committed human rights violations, the attacks on CICIG continued. In 2018 Morales announced that the government would not renew the CICIG agreement.

Key Lessons from CICIG
There are many lessons to gather from the 12 years of work of CICIG, but here I will focus only on six lessons related to matters of institutional design. First, domestic legislation and legal culture matter. CICIG's institutional design was buttressed in a longstanding legal institution: private prosecution. As a private prosecutor CICIG had the legal advantage that it was clearly defined within criminal procedure law, and as such, it was not totally "foreign" to criminal justice actors. It was designed to collaborate and support the Public Prosecutor's Office in the investigation and prosecution of crimes in domestic courts, and although it was granted "additional" powers to gather information, it did not aim to replace the public prosecutor with an international prosecutorial organ nor with an international court. An investigative body like this might be more easily replicated in countries with similar private prosecution rights.
Second, impunity can be reduced with efforts focused at the investigation stage. CICIG was designed to take over the investigation of grand corruption cases or organized crime, thus it had the clear, tangible impact of preventing these crimes from being relegated to obscurity. CICIG's investigative work shed light on practices and networks that no other party would have dared to investigate. Just opening an investigation greatly increases the odds of any case reaching a court in the future.
Third, institutional design can be limited by state capacity. A review of the experience of CICIG over time reveals that there were limits to what this commission could do to support the state of Guatemala despite its strong investigative powers. By design CICIG was never intended to replace the state and its ability to build local capacity was constrained. For instance, on a site visit to Guatemala on July 31 to 4 August 2017, the IACHR noted that the persistent problem of impunity derives "from the lack of capacity of the Guatemalan State to provide an adequate and efficient system of administration of justice, and an independent and impartial justice in cases from the past as well as the present." (IACHR Inter-American Commission on Human Rights, p. 10). At the end, capacity, as well as survival, depended more than it was hoped for on political will.
Fourth, temporal mandates have a clear disadvantage. As noted earlier, political will matters, and a fundamental lesson from CICIG is that temporary renewable mandates are evidently more vulnerable to whims of domestic political changes. Even though it was granted the power to independently conduct investigations, at the end it could not "enforce cooperation of authorities or impose penalties for noncompliance" (Yeh 2015, p. 232). The CICIG in part was a victim of its own success, as the decision of President Morales to not renew the agreement was both a backlash for its success and a preventive measure to avoid future prosecutions.
Fifth, temporal mandates can also have advantages. CICIG was not conceived to be a permanent commission because one of its core objectives was also to improve the capacity of the Guatemalan state to deal with corruption cases. For this reason, one of the key functions of CICIG was to suggest reforms that would help improve institutional capacity. As noted earlier, it is likely that the impact of CICIG will be long lived in the reforms that it helped design to improve the strength of criminal investigations (through technical training or even improving the witness protection program) and to improve judicial and prosecutorial independence and accountability.
Finally, politics matter. The temporal mandate of CICIG made survival a key concern for all three commissioners. To secure survival the commission had to demonstrate "success" in performing its functions, at the same time that it tried to build legitimacy as a "neutral" actor, beyond domestic politics. For this reason, it became key that CICIG had collaborative partners in the police, the Public Prosecutor's Office, and among judges, and over time CICIG became involved not only in purges (identifying corrupt state agents within criminal justice institutions), but also in important appointments of prosecutors and judges. To some, CICIG became politicized in this process and probably started to go beyond its mandate. This criticism, however, may not be unique to short-lived institutions, as the ICC has also been criticized for being politicized in its prosecutorial choices (Danner 2003;Schabas 2008). However, had it not been designed as a temporal commission would CICIG have been as bold and aggressive as it was in its investigations? Could the concern of securing relationships for the long-term change the type of cases the CICIG was willing to take? If these questions have impacted the ICC, it is not unreasonable to expect that they could have impacted a permanent investigative commission in a similar way. What CICIG and the ICC have both made clear is that building good relationships with key actors is fundamental for the success of a prosecution. Political will can make or break a case.

Conclusions
In summary, CICIG was a hybrid legal space for governance without precedents. In its investigative role CICIG was designed to be autonomous and powerful, but in its prosecutorial powers, CICIG was designed to be auxiliary to the Public Prosecutor's Office. By design the commissioner's investigative team took the lead in identifying and investigating criminal networks. In the eyes of the public and of defendants, these international actors were responsible for this challenging and technically difficult role (gathering the evidence and identifying witnesses). This indirectly increased prosecutorial independence, shielding public prosecutors and detectives involved in the cases that CICIG initiated. However, at the end the success of any investigative and prosecutorial effort greatly depended on the support the CICIG had from the Public Prosecutor's Office (CICIG Comisión Internacional Contra la Impunidad en Guatemala, p. 4; Call 2019, p. 6) and given its temporal character its capacity to perform and survive ultimately depended on the political environment. Politics, in the case of CICIG, won over institutional design.
APUNCAC draws on these lessons. It requires States Parties to establish specialized prosecutors and dedicated courts for corruption cases. Similar to CICIG, ICAC would be internationally funded and would be operated by international personnel. Its design acknowledges that the first obstacle to accountability is lack of investigation. Thus, the ICAC would concentrate the investigation of corruption cases in the hands of an international investigative body, while relying upon domestic criminal justice institutions to prosecute and adjudicate those cases. The appointment of UN inspectors to lead criminal investigations would likely enhance the legitimacy of the evidence that is gathered and shield domestic police and prosecutors from threats that may arise at the investigation stage. UN inspectors would not have prosecutorial powers (unless domestic criminal procedure allowed UN investigators to claim standing as private prosecutors). However, the permanent status of the ICAC would, in contrast to CICIG's temporary mandate, largely eliminate the threat of discontinuation and institutionalize the presence of independent UN inspectors endowed with strong investigative powers. The presence of independent UN inspectors would shine a bright light on corruption and impunity through criminal investigations that expose individuals involved in corruption schemes and restore accountability.
Three international agreements-the agreement that established CICIG, the Rome Statute establishing the independent prosecutor of the ICC, and the APUNCAC proposalare all based on the premise that international investigators are needed when prosecutorial organs are unwilling, overwhelmed, and/or unable to check high level impunity. Only international investigators whose salaries and careers are insulated from manipulation can act independently to break the grip of criminal forces. The main lesson of CICIG is that this type of international institution can be implemented, and can perform well, in the presence of top-down international pressure from other States in conjunction with bottom-up pressure from unified international and domestic civil society organizations.
Funding: Parts of this manuscript were written based on research that was funded by PSC-CUNY Award #67682-0045, as well as by a Seed Funding Grant from the Office for the Advancement of Research at John Jay College.
Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.