Ethics and Corruption in the Federal Public Service: Civil Servants' Perspectives The World Bank Ethics and Corruption in the Federal Public Service Acknowledgements This report was prepared by a World Bank team led by Daniel Ortega Nieto (Senior Public Sector Specialist, World Bank) and formed by Galileu Kim, Joana Isabel Sousa Lourenço, Kerenssa Mayo Kay, Eric Braian Arias, Juliana Brescianini, Maria Elisa Diniz, and Paulo Antonacci. The team worked under the guidance of Paloma Anos Casero (Country Director for Brazil, World Bank), Robert Taliercio (Economic Director for Latin America, World Bank), and Adrian Fozzard (Public Sector Manager for Latin America and the Caribbean, World Bank). A partnership with the Federal Government of Brazil was made possible thanks to the efforts and support of the Office of the Federal Comptroller General (CGU), the Ministry of the Economy, and the National School of Public Administration (ENAP), in particular: Edward Lucio Vieira Borba, Maria Fernanda Colaco Alves, Marcio Denys Pessanha Gonçalves, Jorge Aguiar, Victor Raymond, Thaisis Souza, Gustavo Fleury Souza, Diana Coutinho, Rogerio Meneguim, Mirian Bittencourt, and Rafael Moreira. Their support in the development of this study, their communication with civil servants, and their valuable comments and guidance were essential. Other World Bank experts were also involved in different aspects of the study: Zahid Hasnain, Francesca Recanatini, James Anderson, Kjetil Hansen, and Lida Bteddini. In addition, the team benefited from input received during preparatory consultations. We would like to thank Fernando Filgueiras, Ana Luiza Aranha, Ronaldo Pilati, Gustavo Bozzetti, Guillermo Toral, and Daniel Hidalgo, as well as all CGU staff who participated in several initial meetings. Finally, we are especially grateful to all Brazilian civil servants: this study would not have been possible without the participation of the more than 22,000 civil servants who completed the questionnaire and shared their opinions. 2 The World Bank Ethics and Corruption in the Federal Public Service Executive Summary This Survey on Ethics and Corruption in the Federal Public Service was held online from April 28 to May 28, 2021, in partnership with the Office of the Federal Comptroller General (CGU), the Ministry of the Economy, and the National School of Public Administration (ENAP). All civil servants were represented in the sample, totaling 22,130 respondents. The sample covered all federative units and ministries. Most civil servants report having witnessed some sort of unethical practice during their time in the public sector. Of all respondents, 58.7 percent stated that they witnessed some unethical practice during their career in public service. The most frequent practices were using one's position to help friends or family, and bending the rules under pressure from one’s superiors. Over the past three years, around one third of all civil servants (33.4 percent) witnessed some unethical practice, according to their reports. Almost half of all civil servants believe that their colleagues commit unethical practices, but most of them believe that few civil servants are involved. Among civil servants, 34.5 percent said that not many public agents in their organization were involved in unethical practices;12.8 percent suggested that many were involved; and 0.3 percent believed that all civil servants took part in these practices. Together, they represented 47.6 percent of all respondents. On the other hand, 37.9 percent believed that no-one in their organization engaged in unethical practices. Civil servants report high levels of insecurity when reporting acts of corruption, which could be mitigated by integrity programs. In total, 51.7 percent of all respondents did not feel safe enough to report illegal conducts. Having access to integrity programs seemed to result in lower levels of insecurity, and was seen as an important tool for creating an anticorruption culture. Among the respondents who claimed that they had attended an integrity program, 68 percent felt safer to report. Most civil servants are familiar with integrity programs and consider them an important strategy for fighting corruption. When asked about integrity programs and their opinion about them, most respondents (54.5 percent) stated that they knew about their organization's integrity program, and 69.8 percent generally considered them as a measure that contributed to preventing corruption. Despite considering integrity programs as something relevant, few employees receive training. Only 31.3 percent of all civil servants reported having received integrity training in their organization, which might suggest a potential gap between knowing about such programs and being trained on how to benefit from them. Our survey showed that organizational leaders promoted integrity programs to a limited extent: only 36 percent of civil servants said their leaders regularly promoted their integrity program. Most employees say that their organization's rules and regulations are transparent and strictly complied with. Most employees agreed, totally or partially, that the rules and regulations of their organization were adequately transparent (61.6 percent) and strictly complied with (68.7 percent). It should be noted that 35.8 percent only partially agreed with 3 The World Bank Ethics and Corruption in the Federal Public Service regard to transparency, and 41.6 percent, with respect to compliance. On the other hand, 30.8 percent totally or partially disagreed with the statement on transparency, and 24.4 percent with the statement on compliance. Civil servants view their organization’s promotion and progression scheme as transparent or based on merit, but they express concerns about the influence of political connections and friendship. When asked about their views on their organization's promotion and progression scheme, a significant portion of the employees expressed a positive opinion, considering it transparent (35.1 percent) or merit-based (23.1 percent). However, civil servants raised concerns about the influence of political connections and friendships in the career scheme. Most respondents say that unethical practices increased during the Covid-19 crisis. Most respondents stated that unethical practices such as political interference in their organization's decisions remained the same or increased during the crisis (55.9 percent), or that there had been an increase in conflicts of interest between public and private services (50.6 percent). A number of civil servants stated that procurement and contracting decisions made with little transparency and accountability increased (22.4 percent), and that there was a surge in lobbying activities involving the public and private sectors (22.2 percent). Strengthening the legal framework against corruption must be complemented by its implementation and by measuring its impact among civil servants. Deepening the knowledge of the institutional context in which corruption operates implies listening to the opinion of civil servants, thus generating new evidence to design or improve public policies. Listening to civil servants' perspectives allows us to understand the impacts of rules on specific practices, and thus improve anticorruption policies. This effort complements the review of the anticorruption legal framework, measuring its effective impact. Corruption in the public service is multifaceted, thus requiring granular information about its nature, prevalence, and vulnerable actors. In view of its scope, thematic scope, and representativeness, the data generated by the study could become a valuable source for the development of knowledge about corruption in the federal public service. We hope that this Survey on Ethics and Corruption in the Federal Public Service becomes a tool to complement current and future efforts to fight corruption. 4 The World Bank Ethics and Corruption in the Federal Public Service Table of contents Acknowledgements ............................................................................................................................. 2 Executive Summary.............................................................................................................................. 3 Introduction .......................................................................................................................................... 6 1. Experiences with Corruption and Behavior Assessmentin the Organization ........................... 9 2. Reporting Mechanisms ................................................................................................................ 15 3. Rules and Regulations .................................................................................................................. 21 4. Human Resource Management: ................................................................................................. 24 5. Impact of the Covid-19 Pandemic: ............................................................................................. 28 Conclusion and Recommendations................................................................................................ 30 Annex 1: Sample Characteristics .................................................................................................... 32 Annex 2: Weighting .......................................................................................................................... 36 Annex 3: Behavioral Experiments ................................................................................................... 38 References ......................................................................................................................................... 41 5 The World Bank Ethics and Corruption in the Federal Public Service Introduction Latin America has been facing several challenges related to corruption. Worldwide Governance Indicators (WGI) data reveal that, in the past decade, Latin America as a whole (and Brazil, in particular) saw a drop in indicators such as Control of Corruption and the Rule of Law. 1 Latinobarometer, in turn, measures public opinion about civil servants and perceptions on their involvement in corruption. According to Latinobarometer, in Brazil, almost 75 percent of the population believe that some or no civil servants are involved in corruption. Public opinion about civil servants is relatively positive, but it is not clear whether this perception is shared by civil servants themselves. A comparison with the region shows that the Brazilian population's perceptions on civil servants are better than in other countries (figure 1). However, according to Latinobarometer, more than 25 percent of the Brazilian population believe that all or almost all civil servants are involved in acts of corruption. Figure 1: Public opinion on the prevalence of corruption among civil servants, Latinobarometer 2018. 1 https://info.worldbank.org/governance/wgi/. 6 The World Bank Ethics and Corruption in the Federal Public Service Corruption in the public sector affects people's trust in public institutions, and has negative impacts on a country’s economic development and equality levels, but fighting it requires empirical evidence. 2 Public policies, when subject to corruption, may lead to a reduction in social welfare 3 and negatively affect private sector development. It is critical to understand the incidence of corrupt and unethical practices, and how these are perceived by civil servants themselves. In order to deep our knowledge of the institutional context in which corruption takes place, we must listen to civil servants, and generate new evidence to design or improve public policies. A sophisticated anticorruption legal framework will not be enough if, in practice, civil servants still report or perceive high degrees of unethical behavior. In order to achieve this knowledge objective, the World Bank, in partnership with the Office of the Federal Comptroller General (CGU), the Ministry of the Economy, and the National School of Public Administration (ENAP), seeks to produce empirical evidence to fight corruption in the public sector. The research tool developed under this partnership aims to collect evidence and perspectives from civil servants on corruption in public administration. Applied on a census basis to federal civil servants, the Survey on Ethics and Corruption in the Federal Public Service represents a step forward in the anonymous and safe gathering of information on behavior and perceptions, allowing for greater granularity in both the focus and the quality of data on corruption. By collecting and analyzing civil servants' perspectives, it is possible to generate empirical knowledge about different types of unethical behavior. Evidence on what types of corrupt practices occur, their incidence levels, and the possible motivations behind such acts enable the design of public policies that support the fight against corruption. The survey was conducted as a census at the federal public service level, with a total sample of 22,130 respondents, which ensures a high level of statistical confidence. 4 Data collection took place online from April 28 to May 28, 2021 with the support of the CGU, the Ministry of the Economy, and ENAP. All federal ministries were included in the survey sample. The response rate was higher at the CGU (19.9 percent) and at the Ministry of the Environment (7.0 percent), while the Ministry of Foreign Affairs (1.8 percent) and the Ministry of Regional Development (1.7 percent) presented the lowest response rates. 5 All federative units were included in the sample, and the highest proportion of respondents were based in the Federal District and in Rio de Janeiro. Acre and Roraima had relatively fewer respondents, which was already expected since those states have a smaller presence of federal employees. 2 World Bank 2020; Clausen, Kraay, and Nyiri 2009; Shleifer, and Vishny 1993; Justensen, and Bjornskov 2014. 3 Ferraz, Finan, and Moreira 2012. 4 This study did not include military personnel or retired civil servants. Following CGU recommendations, the survey also excluded those respondents who, when answering the question "What is your link with the public administration", replied “I'm not a public agent” and “I'd rather not answer”, as long as they had less than five years of service. The sample, given these exclusion criteria, totaled 21,356 civil servants. 5 See annex on "Sample Results" for further details. 7 The World Bank Ethics and Corruption in the Federal Public Service Most respondents entered the civil service through a public examination, and had higher education and specialization degrees. In terms of their demographics, 80.5 percent of the respondents were statutory civil servants, 8.3 percent were government employees, 5.2 percent were statutory civil servants on a probation period, and 0.9 percent were political appointees [cargos comissionados], who did not have a permanent employment bond with the State. 6 In terms of gender, 57.4 percent of the respondents identified with the male gender, 38.9 percent identified with the female gender, and 0.2 percent, with another gender. Regarding formal education, most respondents had a specialization degree (31.4 percent) or a higher education degree (24.8 percent); 17.9 percent had a master's degree; and 15.1 percent, a doctoral degree. Most civil servants had some experience in the private sector (about eight years, on average), and their average length of service in the public sector was 16 years. 7 The main findings of the survey are presented in the following sections. Due to the existence of various forms of corruption and unethical behavior, some questions allowed respondents to choose more than one answer so as to capture a wider range of responses. As a result, totals may exceed 100 percent. Responses were weighted to adjust for ministry/department, age, and gender. 8 Section 1 presents data on the different types of experiences with unethical behavior and corruption witnessed by civil servants. Section 2 asked agents about their knowledge of reporting mechanisms. Section 3 focuses on employees' perceptions of compliance with rules and regulations within their organization. Section 4 illustrates insights into human resource management in the public service. Section 5 contextualizes the impact of Covid-19 on corruption. The final section presents the conclusions of this study. 6 During the survey, 0.4 percent of the respondents reported not being public agents, and were therefore excluded from the sample. 7 Further details on the sample are available in annex 1. 8 It is worth noting that response rates vary significantly, both institutionally and demographically. In order to calibrate the sample and make it representative of all federal civil servants, we built a set of weights that make our sample representative of the population, based on the inverse likelihood of the response. More details are provided in annex 2. 8 The World Bank Ethics and Corruption in the Federal Public Service 1. Experiences with Corruption and Behavior Assessment in the Organization This section shows findings related to civil servants' experiences with corruption and unethical practices. More than half of all civil servants have witnessed some type of unethical behavior during their professional career in the public sector, and more than a third have suffered pressure from other civil servants. Nearly half of all respondents believe their peers engage in unethical practices, particularly as a result of personal connections. Most civil servants report having observed some unethical practice during their time in the public sector. When answering the question “In your organization, which of the following practices have you witnessed among civil servants while performing their duties?”, 58.7 percent of respondents said they witnessed some unethical practice, 33 percent said that they did not, and 8.3 percent preferred not to answer. Among the group that responded having witnessed unethical practices (figure 2), the most frequently reported practices included using one's position to help a friend or family member (58.9 percent), or bending rules under pressure from one's supervisor (54.0 percent). The least frequent practices were favoring someone in public procurement or contracting (18 percent), or soliciting money and gifts while performing one's duties (8.1 percent). 9 The World Bank Ethics and Corruption in the Federal Public Service Figure 2: Practices witnessed among civil servants while performing their duties. Respondents were allowed to select more than one option, so the sum of responses exceeds 100 percent. A third of all civil servants (33.4 percent) report having witnessed some unethical practice in the public service in the past three years. They were asked if they had witnessed, in the last three years, other servants engaging in unethical practices in their organization, and the answers focused on different themes from the previous question. Among respondents, 33.4 percent indicated that they had witnessed unethical practices in the last three years, 52.1 percent said that they did not, and 14.5 percent preferred not to answer. Most of those claiming to have witnessed some unethical practice in this period (figure 3) referred to unethical behaviors in the design of policies, projects or programs (37.7 percent), or in procuring and contracting services or works (35.3 percent). The areas where the least unethical practices were observed were the transfer of resources to NGOs (8 percent) or to states and municipalities (6.5 percent). Figure 3: Unethical practices witnessed in the past three years. Respondents were allowed to select more than one option, so the sum of responses exceeds 100 percent. A third of all civil servants reported having suffered some pressure in the past three years to engage in an unethical practice. in the past three years, 9 33.3 percent of all civil servants reported having suffered some type of pressure to engage in unethical practices, 58.3 percent 9 The period referred to here is from April/May 2018 to April/May 2021. 10 The World Bank Ethics and Corruption in the Federal Public Service said they suffered no pressure, and 8.4 percent preferred not to answer. Among those who reported having suffered some type of pressure (figure 4), more than half (62.5 percent) were pressured to bend their organization's rules and procedures, or to ignore inappropriate behaviors (39.9 percent). Some also reported having suffered pressure to share restricted information (13.4 percent). Figure 4: Number of respondents who claimed that they had been pressured to commit a specific unethical practice (vertical axis) in the last three years. Respondents were allowed to select more than one option, so the sum of responses exceeds 100 percent. Most civil servants who suffered some pressure identify a hierarchical superior or a coworker as the agent of such pressure. Of the 7,004 civil servants who reported having been pressured to engage in unethical practices, 65.0 percent indicated that the pressure came from their (direct or indirect) superior, and about 45.6 percent mentioned having been pressured by coworkers. It is noteworthy that 18.3 percent of the respondents preferred not to answer who exerted pressure. A deeper analysis focused on political appointees (figure 5) reveals that DAS officials 10 reported having suffered less pressure from their hierarchical superiors (54.2 percent among DAS level 3 or lower, and 49.8 percent among DAS level 4 or 10 DAS is an acronym that refers to senior management and advisory positions. 11 The World Bank Ethics and Corruption in the Federal Public Service higher), but greater pressure from politicians (14.3 percent among DAS level 3 or lower, and 21 percent among DAS level 4 or higher, respectively). Figure 5: Pressure from different agents among political appointees. Respondents were allowed to select more than one option, so the sum of responses exceeds 100 percent. Junior officers report more experiences with undue pressure; women witness more instances of using one’s position to help personal connections. Civil servants with less time of service in the public sector (low and middle tertiles) reported more cases of pressure from a superior to bend the rules and prioritize illegitimate interests in the organization's strategies and projects (figure 6). 11 There were no statistically significant differences between genders regarding types of corruption, with the exception of reports related to using one’s position to help friends or family, where women reported having witnessed more cases (62.5 percent) than men (55.8 percent). 11 First tertile (0.96 to 14 years of service), second tertile (14 to 27 years of service), and third tertile (27 to 40 years of service). 12 The World Bank Ethics and Corruption in the Federal Public Service Figure 6: Practices witnessed in the behavior of other employees, by length of service in the public sector. Levels indicate terciles, with the highest tercile being 27–40 years of public service, and the lowest, 0–14 years of public service. Respondents were allowed to select more than one option, so the sum of responses exceeds 100 percent. Coworkers and hierarchical superiors are the agents that mostly exert pressure on civil servants to commit unethical practices, and the most frequent practices are bending the rules and overlooking misconduct. For this report, we created a risk matrix in which the vertical axis shows different types of unethical practices that agents were pressured to commit, and the horizontal axis describes the agents who exerted such pressure. 12 Each colored area indicates the frequency (risk) of an action being urged by an agent, where red indicates greater risk, and blue, lower risk. The matrix shows that risks are higher with regard to hierarchical superiors, who generally exert pressure to bend the organization's rules and procedures, or to overlook inappropriate behaviors. Coworkers tend to act in a similar way, pressuring employees to bend the organization's rules and procedures, in addition to ignoring inappropriate behavior. 12 Question on type of pressure: “Over the past three years, have you been under pressure to [commit an unethical practice]”. Pressure agent: “Which of these agents exerted undue pressure?”. 13 The World Bank Ethics and Corruption in the Federal Public Service Figure 7: Risk matrix relating different unethical practices which civil servants were pressured to commit, and the agent who exerted such pressure. Colors indicate the number of respondents for each combination between action and agent, where red indicates higher risk (frequency), and blue indicates lower risk. Almost half of all civil servants indicate that their coworkers engage in unethical practices. On average, 34.5 percent of civil servants report that few public agents in their organization engage in unethical practices,12.8 percent indicate that there are many, and 0.3 percent say that all their colleagues do it (figure 7), adding up to 47.6 percent. A smaller share (37.9 percent) indicate that no-one in their organization engages in unethical practices, with a significant number preferring not to answer (14.4 percent). Assessments vary according to the practice: obtaining professional benefits from personal ties (35 percent said "few", and 18.6 percent, "many"), and bending the rules at the request of a supervisor (40 percent answered "few", and 14.6 percent, many). Accepting money or gifts to fulfill one’s duties is considered less frequent, with 28.6 percent indicating that few civil servants do that, and 5.3 percent, many. 14 The World Bank Ethics and Corruption in the Federal Public Service Figure 8: Perception on the prevalence of unethical practices in the organization. The gray bar indicates the nonresponse rate. 2. Reporting Mechanisms This section focuses on the mechanisms for detecting and reporting unethical practices, as these are a fundamental component for curbing such behaviors. Few civil servants have reported an act of corruption in the past three years. Most claim that they do not feel safe to report, citing lack of protection and fear of retaliation. Although a third of all civil servants have witnessed unethical practices in the past three years, only 12 percent reported corruption in the same period. This low percentage contrasts with the fact that 33.4 percent claimed to have witnessed some unethical practice in the same period. Whistleblowers (2,417 respondents) were asked about the repercussions of their decision, and 41 percent said that the complaint did not proceed, 27.4 percent suffered retaliation, 14.9 percent did not know, and 7.8 percent indicated that the accused party was punished. 13 It is important to point out that not all allegations of corruption necessarily have merit to proceed. Therefore, the low punishment rate does not necessarily indicate impunity. 13 In addition, 2.3 percent of respondents chose not to answer. 15 The World Bank Ethics and Corruption in the Federal Public Service Figure 9: Relationship between training in integrity programs (vertical axis) and number of respondents who say they feel safe enough to report acts of corruption (horizontal axis). Half of all civil servants do not feel safe enough to report wrongdoings, but integrity programs seem to reduce these feelings of insecurity. Among civil servants, 51 percent reported not feeling safe enough to report unlawful conducts in their organization. There seems to be a correlation between integrity programs lower levels of insecurity (as such programs promote an anticorruption culture — figure 9). Among the respondents who fully agreed that they received training in integrity programs, 68 percent said they felt safe to report, while among those who fully disagreed with that statement, only 36.3 percent felt safe to report. The feeling of insecurity about reporting varies depending on the type of contract, gender, and length of service in the public sector. Among statutory civil servants, 51 percent did not feel safe to report misconducts. Political appointees, on the other hand, felt less insecure about reporting cases of corruption, with only 36.8 percent of respondents saying they felt insecure. DAS officials also showed lower feelings of insecurity, with 43 percent (DAS 3 or lower) and 32.6 percent (DAS 5 or higher) reporting that they felt insecure when reporting, in contrast to 52.5 percent for non-DAS servants. Additionally, a greater number of women felt insecure when reporting (59.6 percent) compared to men (44.3 percent). Finally, civil servants with a longer career in public service felt less secure to report (45.9 percent) than their more junior colleagues (54 percent). 16 The World Bank Ethics and Corruption in the Federal Public Service According to the respondents, the top challenges faced by civil servants when reporting is lack of protection for complainants, and fear of conflict with other civil servants. When asked about the main difficulties faced by public agents when reporting such cases (figure 10), most indicated that there was a lack of protection for those who reported (59.7 percent), or that civil servants preferred to avoid conflict (59 percent). Such concerns might indicate that civil servants fear negative repercussions from their decision to report wrongdoings. Others argued that cases of corruption are difficult to prove (46.3 percent). More than a third of respondents reported that they did not know the procedure for reporting (34.4 percent), that the reporting process was lengthy and complex (35.9 percent), or that they preferred to discuss it with their manager or coworkers (12 percent). Finally, 7.1 percent of all participants preferred not to answer the question. Figure 10: Main difficulties reported by public agents when reporting cases of corruption. Respondents were allowed to select more than one option, so the sum of responses exceeds 100 percent. The likelihood of reporting an unethical practice in the past three years is higher when people witness unethical practices in their workplace, or are pressured to engage in unethical practices. 14 Figure 11 illustrates the acts of corruption that were witnessed (vertical 14 Questions about experiences with corruption: “In your organization, which of the following practices have you witnessed among civil servants while performing their duties?" and "In the past three years, have you been under pressure to [commit an unethical practice]?” Likelihood of reporting an unethical practice: “In the past three years, have you reported any wrongdoing?” 17 The World Bank Ethics and Corruption in the Federal Public Service axis) and the percentage of civil servants who reported an unethical practice in the past three years (horizontal axis). In the cases of respondents who witnessed unethical practices, the proportion of those who reported them was higher when the act was “requesting money or gifts to fulfill one’s duties” (34.1 percent), and lower for “using their position to help a friend or family member” (20.3 percent). Figure 12 shows that a higher percentage of civil servants reported having been pressured to “harm a specific individual” (34.1 percent), and a lower percentage were urged to “bend the organization's rules and procedures” (28.9 percent). Figure 11: Percentage of civil servants who reported an unethical practice in the past three years (horizontal axis) by acts of corruption witnessed in the public service (vertical axis). 18 The World Bank Ethics and Corruption in the Federal Public Service Figure 12: Percentage of civil servants who reported an unethical practice in the past three years (horizontal axis) by acts of corruption they were pressured to commit (vertical axis). Most civil servants believe that the most efficient measure to reduce corruption is to reduce impunity. Most respondents indicated that the most effective measures to reduce corruption were reducing impunity (48.3 percent) and introducing more severe penalties for crimes of corruption (47.2 percent). Improving the role of anticorruption agencies was mentioned by 30.8 percent of respondents. To a lesser extent, civil servants indicated that strengthening the monitoring of public policies by civil society (24.3 percent) or increasing civil servants' pay (21.9 percent) would be effective measures, as well. Corruption is mainly attributed to impunity. When asked about the main reasons leading public agents to engage in corruption, 56.6 percent of civil servants suggested “impunity”, 51.2 percent answered “greed”, and 27.7 percent replied “abuse or pressure from authorities” (figure 13). Albeit less frequently, they also cited low wages (15.5 percent), excessive bureaucracy (13.8 percent), and lobbying (5.2 percent). 19 The World Bank Ethics and Corruption in the Federal Public Service Figure 13: Main factors leading public agents to behave in an unethical manner. Opinions expressed by civil servants participating in the online survey reinforce their support for strengthening internal affairs and auditing bodies. When asked for suggestions on how to improve the prevention of and the fight against corruption, one of the respondents replied, “increasing and improving internal controls (audits) and external controls (whistleblowing), including rewards for whistleblowers.” Another civil servant indicated that it was necessary to "offer better conditions to enable auditors and ombudspeople in public organizations to investigate corruption reports more quickly.” Others suggested “improving the correctional system in order to streamline administrative processes for investigating alleged misconduct" and "creating a (virtual) channel where civil servants could feel safer to file and keep track of their complaints without putting their life at risk," a recommendation that mirrors the new FalaBR channel created by the CGU. A behavioral experiment showed that direct peers, immediate superiors, and directors are considered more capable of detecting corruption, in contrast with indirect peers or control bodies. An experiment carried out together with the CGU tested whether the likelihood of detecting corruption varied depending on one's role and hierarchy, as well as the degree of proximity to the civil servant [being accused]. Respondents were asked how likely it was for a particular agent to detect corruption. The findings revealed that, using detection by an indirect colleague as a parameter, direct coworkers, immediate superiors, and directors are 20 The World Bank Ethics and Corruption in the Federal Public Service considered more capable of detecting corruption. Indirect colleagues, on the other hand, have the same likelihood of detecting corruption as a control body. This may indicate that, in the respondents' assessment, there is a lack of institutional monitoring capacity in relation to direct colleagues and hierarchical superiors (that is, immediate superiors or leaders in the organization). 15 3. Rules and Regulations This section presents the main findings on organizations' rules and regulations, trust in the institutions, and integrity programs that aim to strengthen them. Most civil servants indicate that they know about such integrity programs, but few have access to training. Most civil servants are aware of the existence of integrity programs in their organization and see them as measures that contribute to reducing corruption. When asked if they knew about integrity programs in their organizations and their opinion about them, most respondents (54.5 percent) said they knew about their integrity program. More than two thirds of all respondents (69.8 percent) indicated that they considered integrity programs as a measure that contributed to preventing corruption. Training in integrity programs is limited, and is generally little publicized by leaders in the organization. Only 31.3 percent of respondents reported having received training on how to benefit from their organization’s integrity program, indicating a gap between being aware of the existence of such programs and being trained on how to use them. Senior leaders promote integrity programs in a limited way: only 36 percent of civil servants said their leaders regularly promoted their integrity program (figure 14). 15 More details in annex 3. 21 The World Bank Ethics and Corruption in the Federal Public Service Figure 14: Opinions on integrity programs. The gray bar indicates the nonresponse rate. Civil servants indicate that their organizations' rules and regulations are duly transparent and strictly enforced. When asked to assess different statements regarding their organization (figure 15), most respondents agreed, fully or partially, that public standards are duly transparent (61.6 percent) and strictly complied with (68.7 percent). It is noteworthy that a significant number of respondents indicated that they only partially agreed with the same statement (35.8 percent and 41.6 percent, respectively). Regarding the role played by private companies in influencing the opinion of public managers through lobbying efforts, only 19 percent agreed that they had the necessary level of transparency, and 51.3 percent partially or totally agreed that lobbying harms competitiveness. 22 The World Bank Ethics and Corruption in the Federal Public Service Figure 15: Opinions regarding transparency, integrity, and lobbying in one's organization. The gray bar indicates the nonresponse rate. A behavioral experiment showed that most respondents (91.2 percent) believe that acts of corruption are not justifiable in the hypothetical scenarios presented. 16 In the context of this survey, behavioral experiments were conducted together with the CGU (further details in annex 3), including an exercise which sought to test whether the justifications presented for unethical behaviors affected the assessment of how acceptable such behaviors were. Two hypothetical scenarios, each describing different unethical behaviors, were developed: (A) purchasing equipment to favor a specific company; and (B) expediting a bidding process to favor a specific company. The findings revealed that most justifications had a negative effect, with two exceptions. Seeking career promotion is seen as a less acceptable justification for unethical behaviors, but this was not observed in the case of suffering pressure from a superior. Under scenario A, three possible justifications were tested: (a) career promotion; (b) pressure from a superior; (c) no justification (that is, control group). The findings showed that using career promotion as a justification for unethical behavior had a negative effect, thus making 16 It is worth noting that this high percentage may be motivated by a desire for social acceptance. However, differences between alternative justifications are particularly informative, as a controlled experiment allows for a direct comparison (positive or negative) of variations in the assessment of a particular justification. 23 The World Bank Ethics and Corruption in the Federal Public Service such behavior less justifiable. The effect was nil when pressure from above was used as justification, which shows that pressure from above has no effect on how justifiable an unethical practice may be considered. In general, trying to justify an act of corruption makes it less justifiable in the opinion of civil servants, with the exception of NGOs. Under scenario B, respondents randomly received one of six justifications: (a) bribery; (b) bribery in the context of a civil servant's financial needs; (c) bribery in a context where it is common to receive bribes in one's organization; (d) exchanging favors to help a family member in need; (e) favoring an NGO that helps needy families; and (f) no justification (that is, control group). In this scenario, four of the five justifications tested had a negative impact on how justifiable an unethical behavior was considered, with the exception of alternative (e) [favoring an NGO], which had a positive effect, that is, officials considered such act more justifiable. These experiments highlight a number of findings regarding the causal effect of justifications on the assessment of corrupt practices. Three points stand out: firstly, exchanging favors and bribes in a situation of need reduces the magnitude of the negative effect, but does not eliminate it. Secondly, the justification that unethical behavior is common in the organization does not make such behavior more acceptable. Finally, when a licensing process is expedited to favor an NGO, respondents rate such unethical behavior as more justifiable. 4. Human Resource Management: This section presents the main findings on career and promotion systems, and on how witnessing corruption may play a role in civil servants' professional development. Most civil servants believe that their organization's career and promotion system is transparent and based on merit, but they express concerns about the influence of political connections and friendships. Over 40 percent of all civil servants report trusting their peers and leaders in their organization, with DAS officials expressing more positive opinions. When asked about their "degree of confidence in decisions and actions" taken by their peers or leaders in their organization, 43.9 percent of respondents said they had a good or excellent opinion of their leaders, while 44.9 percent said something similar about their peers. A significant number of respondents reported a negative assessment: 25.3 percent in relation to the organization's leaders, and 18.8 percent in relation to their coworkers. DAS officials showed a more positive opinion (good or excellent) of other civil servants: both managers (53.6 percent among DAS 3 or lower, and 61.3 percent among DAS 5 or higher), and peers (52.9 percent among DAS 3 or lower, and 56.5 percent among DAS 5 or higher). Most civil servants believe that their organization's promotion and progression system is transparent, but they express concerns about political or personal influences. When asked about their "opinion on their organization's promotion and progression system", a considerable number of respondents stated that the system is transparent (35.1 percent) or 24 The World Bank Ethics and Corruption in the Federal Public Service meritocratic (23.1 percent). 17 Nonetheless, 24.5 percent of civil servants stated that the system was influenced by political connections, and 22.4 percent believed it was influenced by friendships (figure 16). Figure 16: Opinion about one's organization's promotion and progression system. Respondents were allowed to select more than one option, so the sum of responses exceeds 100 percent. Women in public service consider the promotion and progression system less meritocratic and transparent than men. While 39.5 percent of men considered the promotion and progression system transparent, only 34.0 percent of women agreed with that statement (figure 17). With regard to merit, 20.6 percent of all female civil servants thought that the career system was based on merit, a lower proportion than the 27.4 percent reported by men. Women also believed that promotion and progression decisions were more influenced by friendship (26.6 percent) than men (22.2 percent). Most civil servants indicate that merit-based criteria are more relevant for political appointments ["cargos comissionados"] . When asked what aspects mattered the most to someone being appointed to such a position (which total approximately 22,000 in the federal 17 It is worth noting that respondents were allowed to choose more than one answer for the question “What is your opinion about your organization's promotion and progression system?" Therefore, the sum of responses exceeds 100 percent. 25 The World Bank Ethics and Corruption in the Federal Public Service administration as a whole), respondents mentioned specific skills needed for the activity (49.0 percent), previous professional experience (26.5 percent), or academic qualifications (22.6 percent). On the other hand, 29.5 percent of respondents indicated that having family, friends or personal or political connections in their organization was the most important condition, and 12.8 percent said that directly favoring or rewarding those who decided on the appointment was a key aspect. Figure 17: Opinion about the promotion and progression system, by gender. Respondents were allowed to select more than one option, so the sum of responses exceeds 100 percent. Almost half of all civil servants claim that they have felt harmed due to unethical behaviors related to human resource management in the past three years. When asked “Over the past three years, have you ever felt harmed by unethical behaviors in any of these activities?", 49 percent of respondents indicated that they felt harmed in some way, 47.3 percent did not feel harmed at all, and 3.7 percent preferred not to answer (figure 18). Among the respondents who felt harmed, the majority indicated issues related to the recruitment and political appointments (36.6 percent), or to obtaining permission to travel and benefit from training (34.9 percent). To a lesser extent, unethical behaviors were reported with regard to pay or benefits (25.4 percent), and to disciplinary proceedings (11.3 percent). 26 The World Bank Ethics and Corruption in the Federal Public Service As expected, we found that there is a negative relationship between experiences with corruption and the degree of trust in coworkers, but impacts vary according to the type of infringement. 18 For each of type of corruption practice, we calculated the percentage of civil servants who indicated that they trusted their peers (figure 19). The data show that, as expected, almost all civil servants who said that they had no experience with corruption reported trusting their colleagues (96.4 percent). On the other hand, those who witnessed unethical practices reported lower levels of trust. Among the civil servants who witnessed “bending the rules when there was pressure from a supervisor”, 68.2 percent indicated that they trusted their colleagues. The drop in trust was even higher for practices such as “requesting money or gifts to fulfill one's duties”, with only 52.6 percent saying that they still trusted their coworkers. 18 Question about experiences with corruption: “In your organization, which of the following practices have you witnessed among civil servants while performing their duties?” Trust rating: “What is your degree of confidence in the decisions and actions of the following agents in your organization: other public agents in your organization?” 27 The World Bank Ethics and Corruption in the Federal Public Service Figure 19: Share of employees who trust colleagues in their organization (satisfactory, good, or excellent degree of trust) despite having witnessed specific acts of corruption in their career. 5. Impact of the Covid-19 Pandemic: This section explores the impacts of the Covid-19 pandemic on unethical practices such as lack of transparency in contracting and procurement, and conflicts of interest between the public and private sectors. Most civil servants indicate that unethical practices have increased during the pandemic, but the nonresponse rate is high for all questions. Most civil servants prefer not to answer questions about the impacts of Covid-19. The nonresponse rate ranges from 39 percent to 50 percent for questions about political interference in organizational decisions and increased lobbying between the public and private sectors, respectively. The high nonresponse rate may indicate uncertainty on the part of civil servants regarding the impacts of Covid-19 on their organization's practices. Most respondents believe that unethical practices have increased during the Covid-19 crisis; however, a significant number of participants preferred not to answer. When asked about the evolution of different practices since the beginning of the Covid-19 pandemic in March 2020, most respondents stated that they remained the same, or that they saw an increase in practices such as political interference in their organization's decisions (55.9 percent) or conflicts of interest between public and private services (50.6 percent). Civil servants also stated that more procurement and contracting decisions were made with little 28 The World Bank Ethics and Corruption in the Federal Public Service transparency and accountability (22.4 percent), and that there was an increase in lobbying activities involving the public and private sectors (22.2 percent). Figure 20: Covid-19's impact on unethical practices. The gray bar indicates the nonresponse rate. 29 The World Bank Ethics and Corruption in the Federal Public Service Conclusion and Recommendations The Survey on Ethics and Corruption in the Federal Public Service is the result of a partnership between the World Bank and the Federal Government of Brazil. The World Bank, together with the CGU, the Ministry of the Economy, and ENAP, produced empirical evidence that provides new knowledge about the incidence and perception of unethical practices from the perspective of civil servants themselves. This evidence and data analysis on corruption is expected to contribute not only to enacting future rules on this issue, but also to improving their implementation. The systematic collection of information about unethical practices and corruption will allow the federal government to increase its knowledge about these practices, in order to improve public policies and strengthen mechanisms to mitigate corruption. In this regard, the perception and experience of civil servants at different levels of government is fundamental, as well as those of society and the private sector. The survey included more than 22,000 civil servants across all ministries, federative units, and different levels of leadership, enabling an analysis of corruption from several angles. The survey indicates that, from the point of view of civil servants, corruption in the public service is frequent and multifaceted. Most civil servants said that they had witnessed some inappropriate practice during their professional career (58.7 percent), while a third (33.3 percent) reported having suffered pressure to act in an unethical manner in the last three years. Among those who reported having suffered some type of pressure, more than half (62.5 percent) were pressured to bend their organization's rules and procedures, or to overlook some inappropriate behavior (39.9 percent). Most civil servants who suffered some pressure identified a hierarchical superior or a coworker as the agents of such pressure. 19 Although most civil servants have witnessed some unethical practice, few have reported an act witnessed in the past three years. Although a third of all respondents witnessed some unethical practice in the last three years, only 12 percent filed a report in the same period. Feelings of insecurity are high: 51 percent of respondents said they did not feel safe enough to report it, with a higher proportion of women feeling insecure. The Anticorruption Plan 20 launched by the Interministerial Committee against Corruption (CICC) 21 aims to strengthen whistleblower protection mechanisms. During its implementation, a number of measures must be considered due to the high levels of insecurity expressed by civil servants. Civil servants believe that unethical practices are committed by a minority, and that rules and regulations are generally transparent and respected. Almost half of all civil servants (47.6 percent) believed that their coworkers had committed some unethical practice at some stage in their career. Among respondents. 34.5 percent considered that few coworkers had done it, 12.8 percent indicated that there were many, and 0.3 percent believed that everyone had committed some unethical practice. Most employees agreed, totally or partially, that the 19 Of the 8,866 civil servants who reported having been under pressure, 65.0 percent indicated that the pressure came from their (direct or indirect) superior. 20 https://www.gov.br/cgu/pt-br/anticorrupcao/plano-anticorrupcao.pdf. 21 Reference: https://www.gov.br/cgu/pt-br/anticorrupcao. 30 The World Bank Ethics and Corruption in the Federal Public Service rules and regulations of their organization were transparent (61.6 percent) and strictly complied with (68.7 percent). However, 35.8 percent only partially agreed with regard to transparency, and 41.6 percent, with respect to its compliance. Most civil servants are familiar with their organization's integrity program and see it as an important measure to fight corruption, but few have ever received formal training on it. Over two thirds (68.9 percent) of all civil servants considered integrity programs an important measure to fight corruption, and 54.5 percent of civil servants said they were aware of the existence of an integrity program in their own organization. However, only 31 percent of employees reported having been formally trained on their organization's integrity program. The creation of the Federal Government's Public Integrity System (SIPEF) may contribute to increasing access to integrity programs, given civil servants' positive feelings and their currently limited access. Considering the data collected during the survey, it is important that leaders in different agencies play a proactive role in promoting integrity programs. The Covid-19 crisis coincided with an increase in civil servants' perceptions of corrupt practices, but a significant number of them chose not to answer the questions. The nonresponse rate for questions about how Covid-19 affected unethical practices ranged between 39 percent and 50 percent. Among the respondents, a significant number claimed that there was an increase in practices such as political interference in their organization's decisions (26.3 percent), and reduced transparency and accountability on decisions about contracting and procurement (22.4 percent). Initiatives spearheaded by the CGU, such as monitoring the application of federal funds transferred to municipalities and states, and the creation of a management panel for “Covid-19-related contracts”, are important measures to strengthen transparency and integrity in the context of a crisis. It is crucial to continue strengthening these mechanisms, as well as making greater use of data analytics in all areas — independently of Covid-19 — so as to identify any corrupt practices. Fighting corruption requires a set of actions, mechanisms and rules to improve the prevention, detection, punishment, and remedy of unethical practices. The federal government's 2020–25 Anticorruption Plan shows that, while Brazil has a robust legal framework focused on increasing transparency and fighting corruption, 112 regulatory and non-regulatory improvement needs were identified, as well as 142 actions to address them. The strengthening of the legal framework against corruption must be complemented by measures for its implementation and constant evaluation. Listening to civil servants' perspectives allows us to understand the impacts of rules on practices and improve anticorruption policies. This survey is expected to support the actions of the Anticorruption Plan, given that incorporating civil servants' perspective plays a key role in reducing the weaknesses exposed by them. 31 The World Bank Ethics and Corruption in the Federal Public Service Annex 1: Sample Characteristics The distribution of respondents by different geographic, institutional, and demographic segments is presented in this annex. It is worth noting that the distributions represented herein have not been weighted. Distribution by Ministry Distribution by Federative Unit 32 The World Bank Ethics and Corruption in the Federal Public Service Distribution by Gender Distribution by Age 33 The World Bank Ethics and Corruption in the Federal Public Service Distribution by Educational Level Distribution by Type of Contract 34 The World Bank Ethics and Corruption in the Federal Public Service Length of Experience (Public and Private Sectors) 35 The World Bank Ethics and Corruption in the Federal Public Service Annex 2: Weighting Different biases introduced by nonresponse rates could prevent or influence descriptive inferences from the sample. In order to mitigate such biases, a frequent solution is to add weights to the sample, making it more representative of the general population. 22 There are a number of methods available, ranging from logistic regressions to classification trees. In the context of this study, it was not possible, due to the unavailability of individual identifiers shared between population and sample, to use statistical techniques that rely on individual observations, such as logistic regressions and classification algorithms. As an alternative, we chose to adopt a practice used by the U.S. Census Bureau, that is, estimating stratified weights. 23 Using December 2019 data from SIAPE, different stratifications were developed, based on 3 variables: (1) ministry, (2) age group, and (3) gender. The weighting formula is as follows: −1 = � � Where and respectively indicate the number of respondents and the number of civil servants for ministry i, age group j, and gender k. Ministries that participated in the survey but were not present in the SIAPE database were not assigned a weight. 24 It is worth noting that each stratification's weight is the inverse of the fraction, effectively increasing the weight for each respondent. There was some concern that the results might be sensitive to the specific variables that corresponded to each stratification, but we were able to demonstrate that this was not the case. Our descriptive analyses are robust at different weights, considering the following specifications: 1) ministry, 2) ministry and gender, 3) ministry and age, 4) age and gender, and 5) ministry, gender, and age, our preferred weighting specification. Each chart illustrates the share of respondents per alternative for a set of questions (considering changes in the specification of variables). Each specification is represented by a distinct color. There are no significant variations between different weight choices, which indicates that the results presented do not result from the choice of a specific weight calibration. 22 Groves, and Couper 2012; Kreuter 2012. 23 Groves, and Couper 2012. 24 The only ministry that did not participate in this study was the Ministry of Communications. Other public bodies taking part in the survey but not listed in SIAPE include: Judicial Branch, Legislative Branch, President's Office and its divisions, President's Office Secretariat of Government, Special Secretariat for Strategic Affairs, and President's Office General Secretariat. It is worth noting that 242 respondents belonged to these bodies. Another 2,920 respondents opted for the “other” option, and therefore did not receive any weight. 36 The World Bank Ethics and Corruption in the Federal Public Service 37 The World Bank Ethics and Corruption in the Federal Public Service Annex 3: Behavioral Experiments Under this study, two experiments were conducted. The first sought to test whether the type of justification presented for unethical behavior affected the assessment of how acceptable that behavior might be (on a scale of 1, “not justifiable”, to 5, “justifiable”). Two hypothetical scenarios, each describing different unethical behaviors, were developed: (A) purchasing equipment to favor a specific company; and (B) expediting a bidding process to favor a specific company. In the context of this experiment, the type of justification provided for each of these two scenarios was randomly distributed among respondents in order to test their effect. A third of all respondents were presented with scenario A. These respondents were randomly assigned to see only one of three possible justifications for the behavior described in the scenario: (a) career promotion; (b) pressure from a superior; (c) no justification (that is, control group for comparison). Two thirds of all respondents were presented with scenario B, in which a civil servant is approached by an employee of a private company asking to expedite the issuance of a license, and approve it ahead of other requests. Respondents were randomly assigned to see only one of six possible justifications: (a) bribery; (b) bribery in the context of a civil servant's financial needs; (c) bribery in a context where it is common to receive bribes in one's organization; (d) exchanging favors to help a family member in need; (e) favoring an NGO that helps needy families; and (f) no justification (that is, control group for comparison). In scenario A, with regard to the control group (no justification), the use of career promotion as a justification for unethical behavior had a negative effect, and therefore this behavior was assessed as less justifiable. The effect was nil when pressure from above was used as justification, which shows that pressure from above has no effect on how justifiable an unethical practice may be considered (figure 17.a). 38 The World Bank Ethics and Corruption in the Federal Public Service In scenario B, with regard to the control group (no justification), four of the five justifications tested had a negative impact on the assessment of how justifiable an unethical behavior was (figure 15.b). Relative to these four justifications, two points should be highlighted. Firstly, exchanging favors and bribes in a situation of need reduces the magnitude of the negative effect, but does not eliminate it. Secondly, the justification that unethical behavior is common in the organization (descriptive norm) does not make such behavior more justifiable. However, one of the five justifications had a positive effect. When license approval was expedited to favor an NGO, an effect in the opposite direction was observed, with respondents rating unethical behavior as more justifiable. In summary, we find that, in general, the majority of respondents (91.2 percent) believe that corrupt practices are not justifiable. Most of the justifications examined have an additional negative effect on respondents' opinion of how acceptable such behaviors may be. In other words, when a civil servant presents a reason for an unethical practice, respondents tend to consider the act less justifiable than in the scenario in which no justification is presented. There are only two exceptions: (a) when there is an indication that the civil servant was pressured by a superior, there is no effect; and (b) unethical practices to help an NGO are viewed positively and have a mitigating effect. The second behavioral experiment tested whether the likelihood of detecting corruption varied depending on one's role and hierarchy, as well as one’s degree of proximity with the civil servant [being accused]. In particular, respondents were asked how likely it was for a particular agent to detect corruption (on a scale of 1 ”not likely” to 5 ”very likely”). The role/position of such specified agent varied as part of the experiment. Respondents were randomly assigned to see only one of five possible agents: (a) direct colleague; (b) indirect colleague; (c) immediate superior; (d) senior leader; or (e) control body. 39 The World Bank Ethics and Corruption in the Federal Public Service The findings revealed that, using detection by an indirect colleague as a basis for comparison, direct coworkers, immediate superiors, and directors are considered more capable of detecting corruption. Indirect colleagues, on the other hand, have the same likelihood of detecting corruption as a control body. This may indicate that, in the respondents' assessment, there is a lack of institutional monitoring capacity regarding direct colleagues and hierarchical superiors (that is, immediate superiors or leaders in the organization). 40 The World Bank Ethics and Corruption in the Federal Public Service References Clausen, Bianca, Aart Kraay, and Zsolt Nyiri. 2011. "Corruption and confidence in public institutions: Evidence from a global survey." The World Bank Economic Review 25.2: 212– 249. Ferraz, Claudio, Frederico Finan, and Diana B. Moreira. 2012. "Corrupting learning: Evidence from missing federal education funds in Brazil." Journal of Public Economics 96.9-10: 712– 726. Filgueiras, Fernando, and Ana Luiza Melo Aranha. 2011. “Controle da Corrupção e Burocracia da Linha de Frente: Regras, Discrecionariedade e Reformas no Brasil.” Revista Dados 54: 349–387. Groves, Robert M., and Mick P. Couper. 2002. "Designing surveys acknowledging nonresponse." Studies of Welfare Populations, Data Collection and Research Issues, Washington DC: 13–54. Justesen, Mogens K., and Christian Bjornskov. 2014. "Exploiting the poor: Bureaucratic corruption and poverty in Africa." World Development 58: 106–115. Kreuter, Frauke. 2013. "Facing the nonresponse challenge." The Annals of the American Academy of Political and Social Science 645.1: 23–35. Shleifer, Andrei, and Robert W. Vishny. 1993. "Corruption." The Quarterly Journal of Economics, 108.3: 599–617. World Bank. 2020. Enhancing Government Effectiveness and Transparency: The Fight Against Corruption. 41