73807 © 2012 the World Bank. 32(077) Originally published November 2012. Code: CTQG-2013 Second printing, January 2013. 4 Publisher’s Foreword C orruption is now one of the most serious problems that draws the attention of nearly everyone. It is quite difficult to assess corruption in daily life properly, whether in Vietnam or other countries, because people only know of corruption cases once they are found out and addressed. It is like the tip of the iceberg on the sea. Through the corruption cases uncovered over the years and based on the corruption assessments of the functional bodies, the Party and State have re-affirmed that corruption in Vietnam is still quite serious. Corruption is so common among state agencies, state officials, citizens and firms, between employees of public services and customers, and the people are concerned. The fight against corruption in Vietnam has progressed since the Law on Anticorruption was enacted in 2005. However, actions have not been bold and there are few visible successes from anti-corruption. The goal of preventing corruption gradually has not been achieved. Aiming at bringing more references for readers on corruption in Vietnam, the National Political Publishing House issues the book “Corruption from the Perspective of Citizens, Firms, and Public Officials - Results of Sociological Surveysâ€? prepared by the World Bank and the Government Inspectorate of Vietnam and in collaboration with the Office of the Steering Committee on Anticorruption, Transformation and Change Consulting, the Asia- Pacific Institute of Management, DFID and UNDP. The book provides readers with a considerable volume of data and analysis on corruption in Vietnam. This is expected to help readers, especially researchers and policy makers, better understand the prevalence and forms of corruption in Vietnam, the causes of corruption, the factors that limit the effectiveness of anti-corruption, and the approaches and solutions needed to step up the fight against corruption in Vietnam in the coming years. Although the survey was conducted in 10 provinces and cities across the country and may not reflect the opinions of the whole population of citizens, firms and public officials 5 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS in Vietnam, the results are significant and provide formal information resources to help set the appropriate measures to promote anti-corruption in Vietnam. Corruption is a sensitive and complicated problem, though it is not insurmountable as said in this report. We would like to receive your comments to help us publish useful studies on this issue in the future and to contribute to prevent corruption more effectively. Enjoy reading! November 2012 THE NATIONAL POLITICAL PUBLISHING HOUSE 6 Acknowledgements T his report and the Anticorruption Diagnostic Surveys were both commissioned by the Government Inspectorate (GI) and the Office of the Steering Committee on Anticorruption (OSCAC). The overall process was guided by an Advisory Board comprising 10 representatives from GI, OSCAC, the Office of the Government (OOG), Office for Business Sustainable Development (SDforB) of the Vietnam Chamber of Commerce and Industry (VCCI), the Vietnam Fatherland Front (VFF) and the World Bank. The Advisory Board was headed by Tran Duc Luong, Deputy Inspector General of the Government Inspectorate, and Le Van Lan, Deputy Head of the Office of the Steering Committee on Anticorruption. The surveys and this report received financial support from the World Bank and from UK-DFID through the GAPAP and VGEMS trust funds. Technical advice was provided by experts from all of the aforementioned institutions, and by experts from UNDP. The surveys were implemented by Transformation & Change Consulting and the Asia-Pacific Institute of Management (APIM, National Economics University), with support from GI, OSCAC, and the World Bank. Overall leadership for the Anticorruption Diagnostic Surveys was provided by a Task Force of ten people from GI (Anticorruption Bureau) and OSCAC. The Task Force, which was established to support and supervise the research process, was led by Ngo Manh Hung - Deputy Head of the Anticorruption Bureau (GI), and Tran Anh Tuan, Deputy Head of the Department for Research, General and International Affairs, OSCAC. The Task Force provided the essential link between the consultant team and the provincial and ministerial contact persons for data collection, monitoring and supporting the data collection process. The Task Force also provided crucial comments on the research approach, initial data analysis and results. The World Bank team provided technical guidance and participated in the pilot survey, and support for the training of interviewers, monitoring and quality assurance of the data collection process. This team also worked jointly with the Consultant team in data analysis and drafting the present report. The core members of the team from the World Bank were James H. Anderson, Tran Thi Lan Huong, Nguyen Thi Phuong Loan, and Do Thi Phuong Thao. We are grateful for the support and encouragement of Victoria Kwakwa, 7 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Country Director, and Deepak Mishra, Lead Economist, throughout the process. The questionnaires and approach were developed jointly by GI, OSCAC, and several development partners including the World Bank, UK-DFID, and UNDP, and the team from T&C Consulting and APIM which was contracted to carry out the surveys. The contributions of Jairo Acuna-Alfaro (UNDP), Do Thi Thanh Huyen (UNDP), Renwick Irvine (UK-DFID), and Nguyen Thi Kim Lien (UK-DFID) during the survey process are greatly appreciated. The Consultant Team included ten researchers/experts from T&C Consulting and APIM. The core members of the Consultant Team were Nguyen Van Thang (Team Leader), Vu Cuong, Le Quang Canh, Bui Duc Tuan and Vu Dong Hung. The Consultant Team helped develop the questionnaires and took overall responsibility for data collection and quality assurance under close supervision and strong support from the World Bank team and the GI/OSCAC Task Force. The Consultant Team relied on the support of ten provincial team leaders during data collection, and more than 100 interviewers. Data analysis and drafting of the present report was carried out by James H. Anderson and Tran Thi Lan Huong (World Bank), and Nguyen Van Thang, Le Quang Canh and Vu Cuong (T&C and APIM). Technical advice and comments on several drafts were provided by Jairo Acuna-Alfaro (UNDP), Do Thi Thanh Huyen (UNDP), Renwick Irvine (UK-DFID), and Nguyen Thi Kim Lien (UK-DFID). Early findings were presented to the Government Inspectorate and OSCAC in April 2012, and the first draft was discussed in May 2012 at two different two-day workshops with the Advisory Board and with stakeholders from ministries and provinces. The insightful and constructive comments of those who participated in these workshops were instrumental in helping to improve the report. We would like to especially recognize the comments provided by representatives of the Office of the National Assembly, Office of the Central Party Committee, OOG, SDforB of VCCI, VFF, concerned ministries and provinces, and express our appreciation to Tran Duc Luong, Deputy Inspector General of the Government Inspectorate, and Le Van Lan, Deputy Head of the Office of the Steering Committee on Anticorruption, for leadership during these workshops. Finally, we wish to thank the staff of the Anticorruption Bureau (GI) for successfully organizing the workshops. The surveys could not have been implemented without the cooperation of ten provincial inspectorates, five ministries, hundreds of village leaders, and the hands-on support of GI and OSCAC. Special thanks are due to the 5,460 citizens, enterprise managers and public officials who provided their time and candid reflections for the conduct of the survey. We are grateful to the team that prepared the pioneering 2005 Anticorruption Diagnostic on which the 2012 Diagnostic draws heavily. We are especially grateful to Professor Nguyen Dinh Cu, Institute for Population and Social Studies (National Economics University) who led that effort and also provided clarifications to our team that helped us to draw comparisons over time. 8 Contents Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Part I - OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 1.1. Rationale for the survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 1.2. Objective and scope of the surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 1.3. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1.4. Limitations of a survey-based approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 1.5. Roadmap to the rest of the report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Part II - SURVEY FINDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.1. Corruption perceptions and experiences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.1.1. Perceptions of corruption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.1.2. Experiences with corruption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.1.3. Information sources on corruption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 2.1.4. Emerging challenge: Interest groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 2.2. Effectiveness of Anticorruption Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 2.2.1. Progress and challenges in anticorruption implementation . . . . . . . . . . . 62 2.2.2. Causes of corruption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 2.2.3. Responses to corruption-risk situations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 2.2.4. How do firms organize their anticorruption activities? . . . . . . . . . . . . . . 70 2.2.5. Tenders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 2.2.6. Recruitment and promotion of public officials . . . . . . . . . . . . . . . . . . . . . . 73 2.2.7. Roles of the media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 2.2.8. Necessity of anticorruption measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 9 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS 2.2.9. The real evidence: what is working, and what is not working? . . . . . . . . 77 2.3. Similarities and differences with the 2005 Anticorruption Diagnostic Surveys . . . 81 2.4. Similarities and differences with other surveys . . . . . . . . . . . . . . . . . . . . . . . . . .87 2.4.1. Citizen’s perceptions and experiences compared to PAPI . . . . . . . . . . . . .87 2.4.2. Enterprise manager’s perceptions and experiences . . . . . . . . . . . . . . . . . .90 Part III - CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . .93 3.1. Policy development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .93 3.2. Policy implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 3.3. Corruption monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107 3.4. Public knowledge enhancement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .108 3.5. Closing remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .110 Part IV - ANNEXES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .111 Annex 1. Additional facts on methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .111 A.1.1. Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .111 A.1.2. Pilot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 A.1.3. Training the interviewers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 A.1.4. Support from OSCAC and GI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116 A.1.5. Quality assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116 A.1.6. Challenges faced by the Team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Annex 2. Institutional factors associated with corruption . . . . . . . . . . . . . . . . . . . .121 Figures Figure 1. The three most serious issues for Vietnam, according to public officials, enterprises, and citizens (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30 Figure 2. What is “Corruptionâ€?? (% saying “definitely corruptionâ€?) . . . . . . . . . . . . . . . . .32 Figure 3. People’s assessment of corruption prevalence (% saying corruption is prevalent, by income) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34 10 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 4. People’s perceptions on corruption seriousness (% saying corruption is serious, by income) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35 Figure 5. Perceptions of corruption prevalence by central vs. local officials (% officials saying corruption is serious) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36 Figure 6. Perceptions of corruption seriousness by central vs. local officials (% officials saying corruption is prevalent) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36 Figure 7. Perceptions of the prevalence of corruption across sectors according to public officials, enterprises, and citizens (% saying prevalent, among those with opinions) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38 Figure 8. The most corrupt sectors according to public officials, enterprises, and citizens (% selecting sector as among the top 3) . . . . . . . . . . . . . . . . . . . . . . . . .39 Figure 9. Suggestions to firms (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .40 Figure 10. Difficulties created by public officials to firms (among those with interactions with state agencies) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41 Figure 11. Firms’ responses to difficulties created by state organizations (%) . . . . . . . . .42 Figure 12. Why firms pay informal charges (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42 Figure 13. Perceptions of the impacts of informal charge to the firms . . . . . . . . . . . . . . . .43 Figure 14. Agencies making difficulties and three agencies making the most difficulties (% of firms) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44 Figure 15. Agencies making the most difficulties are also those receiving the most informal money/gifts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45 Figure 16. To whom is unofficial money paid, and who suggests? . . . . . . . . . . . . . . . . . .46 Figure 17. Keeping good relations with officials. Enterprises’ experiences in the last 12 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47 Figure 18. Firms that make unofficial payments perform worse . . . . . . . . . . . . . . . . . . . . .48 Figure 19. Provinces with more bribery have firms that are doing worse . . . . . . . . . . . . .49 Figure 20. Percentage of citizens using different services in the previous 12 months . . . . .50 Figure 21. Use of services by income category (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50 Figure 22. Probability of paying a bribe when using the service or dealing with the agency, among citizens with contact (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51 Figure 23. Probability of paying a large bribe when using the service or dealing with the agency, among citizens with contact (%) . . . . . . . . . . . . . . . . . . . . . . .52 11 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 24. Proportion of the entire population that paid a bribe, according to citizens (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53 Figure 25. Reactions and outcomes the last time an unofficial payment was suggested (% of citizens) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54 Figure 26. How do officials suggest or request unofficial payments or gifts? (%, among those who paid at least once in the previous 12 months) . . . . . . . .54 Figure 27. When citizens pay unofficial payments or gifts, is it voluntary or suggested? (%, among those who had given an unofficial payment or gift in the previous 12 months) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55 Figure 28. Why citizens pay unofficial payments even when not asked (%, among those who had made a “voluntaryâ€? unofficial payment) . . . . . . . .56 Figure 29. Behaviors encountered by officials at work in the previous 12 months (% of officials) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57 Figure 30. Sources of information about corruption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58 Figure 31. Enterprise views on interest groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 Figure 32. With whom do interest groups influence for negative purposes? (among firms that see only negative aspects of interest groups - they use bribery or connections, but do not help communicate problems) . . . . . . . . . .61 Figure 33. With whom do interest groups influence for positive purposes? (among firms that see only positive aspects of interest groups-they help communicate problems, but do not use bribery or connections) . . . . . . . . . . .61 Figure 34. Officials’ knowledge and attitude toward anticorruption issues . . . . . . . . . . .62 Figure 35. Central and local officials’ assessments of anticorruption measures’ effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64 Figure 36. Officials’ opinions on current detection and treatment of corruption . . . . . . .65 Figure 37. Perceptions of factors that inhibit anticorruption implementation . . . . . . . . .66 Figure 38. Officials’ trust in effectiveness in corruption detection . . . . . . . . . . . . . . . . . . .66 Figure 39. Opinions on the causes of corruption (% agree or strongly agree) . . . . . . . . . .68 Figure 40. Citizens’ response to a corrupt public official . . . . . . . . . . . . . . . . . . . . . . . . . . .69 Figure 41. Reasons for not denouncing corruption (% agreeing or strongly agreeing) . . . .70 Figure 42. Anticorruption activities implemented by firms . . . . . . . . . . . . . . . . . . . . . . . . .71 Figure 43. Officials’ assessment of the tendering activities in their organizations . . . . . .72 12 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 44. Firms’ assessment of the tendering activities in which they participated . . . .72 Figure 45. Importance of different factors in recruitment of officials . . . . . . . . . . . . . . . . .73 Figure 46. Importance of different factors in promotion of officials . . . . . . . . . . . . . . . . . .74 Figure 47. Role of the media’s in helping to fight corruption (% agree or strongly agree) .75 Figure 48. Who is more positive about the media? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 Figure 49. Consensus on the necessity of anticorruption measures among public officials, enterprises and citizens (% saying needed) . . . . . . . . . . . . . . . . . . . . .77 Figure 50. Which behaviors were observed by officials in 2005 and 2012? (% observing in previous 12 months) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Figure 51. Corruption tricks observed by citizens in 2005 and 2012 (%) . . . . . . . . . . . . . .84 Figure 52. Corruption tricks with firms in 2005 and 2012 . . . . . . . . . . . . . . . . . . . . . . . . . .84 Figure 53. Responses by firms to corruption tricks, 2005 and 2012 . . . . . . . . . . . . . . . . . . .85 Figure 54. Why do firms pay bribes, 2005 and 2012 (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . .86 Figure 55. PAPI and the Anticorruption Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 Figure 56. PCI and the Anticorruption Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 Figure 57. Public official sample by administrative levels . . . . . . . . . . . . . . . . . . . . . . . . .113 Figure 58. Public official sample by profession . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Figure 59. Planned and actual enterprise surveyed sample . . . . . . . . . . . . . . . . . . . . . . . .114 Figure 60. Supervising structure of quality assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Tables Table 1. Composition of public official sample in each province . . . . . . . . . . . . . . . . . . . .23 Table 2. Number of respondents by province and sample group . . . . . . . . . . . . . . . . . . . .24 Table 3. Institutional factors associated with lower levels of corruption . . . . . . . . . . . . . .79 Table 4. The most corrupt organizations in 2005 and 2012 surveys . . . . . . . . . . . . . . . . . .82 Table 5. PAPI and the Anticorruption Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 Table 6. PCI and the Anticorruption Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .91 13 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Table 7. Surveyed citizens in the sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .111 Table 8. Common characteristics of surveyed sample . . . . . . . . . . . . . . . . . . . . . . . . . . .115 Table 9. Province-level analysis of the institutional factors associated with lower levels of corruption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .122 Table 10. District-level analysis of the institutional factors associated with lower levels of corruption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 14 Abbreviations ACD Anticorruption Diagnostics APIM Asia-Pacific Institute of Management CECODES Center for Community Development Studies CIEM Central Institute for Economic Management CPV Communist Party of Vietnam DEPOCEN Development and Policies Research Center GDP Gross Domestic Product GI Government Inspectorate GIRI Government Inspectorate Research Institute LURHOC Land Use Right and House and Land-attached Assets Ownership Certificate M&E Monitoring & Evaluation MOF Ministry of Finance MONRE Ministry of Natural Resources and Environment NA National Assembly OOG Office of the Government OSCAC Office of the Steering Committee on Anticorruption PAPI Vietnam Provincial Governance and Public Administration Performance Index PCI Provincial Competitiveness Index PO Public Official PPC Provincial People’s Committee T&C Consulting Transformation and Change Management Consulting Co., Ltd TT Towards Transparency UK-DFID United Kingdom’s Department for International Development UNDP United Nations Development Program UNODC United Nations Office on Drugs and Crime 15 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS VCCI Vietnam Chamber of Commerce and Industry VDR Vietnam Development Report VFF Vietnam Fatherland Front VHLSS Vietnam Household Living Standards Survey VNCI Vietnam Competitiveness Initiative WB World Bank 16 Part I Overview I n 2005, the Internal Affairs Committee of the Communist Party of Vietnam commissioned a study to provide some empirical understanding of the level, pattern, and nature of corruption in Vietnam. The 2005 Diagnostic was instrumental in providing guidance for the development of the Anticorruption Law of 2005, which introduced new approaches to anticorruption such as assets declarations for civil servants, job rotation, and a stronger focus on transparency. The 2005 Diagnostic and the new Anticorruption Law, which took effect in 2006, presaged a period of greater attention to the consequences of corruption and the challenges of reducing corruption. Seven years later, corruption continues to occupy the public’s attention. At the semi- annual Anticorruption Dialogues, co-hosted by the Government Inspectorate and donor partners1, technical approaches are discussed and ideas exchanged. While views may differ, there is one fact on which all agree: corruption remains an important challenge for Vietnam’s development. Vietnam has changed in many ways in those seven years. The economy has expanded by nearly 50%, and 10% of the labor force has shifted out of agriculture and into industry and services. The government is spending more money and collecting more in taxes, with both having more than doubled. The increasingly prosperous population is gradually shifting from bicycles to motorbikes, and from motorbikes to cars - the number of motorbikes has risen by 60%, and the number of cars by 600%. A family is more than twice as likely to own a computer or a telephone. Social services, such as education and health, have been increasingly “socializedâ€?, shifting the official and unofficial burden to the population. The nature and mechanisms of the interface between the population, firms, 1. Sweden was the lead donor on anticorruption for most of this period. Presently, the United Kingdom is the lead donor on anticorruption. 17 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS and the state is also shifting: One-stop shops, which have been flowering since the 1990s, have grown in number and in scope. The role of the state in the economy has also shifted, first with consolidation of the state-owned enterprises into large Economic Groups, and now with a renewed emphasis on equitization - shifting ownership to the private sector - and greater reliance on markets. While Vietnam remains the same in many ways, an ambitious country striving for peace and prosperity, Vietnam is also a different country than it was seven years ago. As Vietnam moves into the ranks of middle income countries, the time to revisit the nature and causes of corruption, the time to bring new empirical data to bear on these issues, and the time for renewed vigor in the fight against corruption is now. 1.1. Rationale for the survey Survey based approaches to understanding corruption are not new. In the decade and a half since they were introduced in countries around the world, such surveys have proven useful in helping to move the debate about corruption from one based on anecdote and assertion to one based on facts and evidence. Surveys help identify where corruption is encountered most often by the population and the business community and how various groups of the population view corruption. In the seven years since the 2005 Diagnostic, the volume of survey evidence in Vietnam has blossomed. Surveys of firms and citizens2 have increasingly pointed to the need to reinvigorate the fight against corruption. The present study, based on surveys of citizens, firms, and public officials, complements these other surveys and brings important new dimensions to our understanding of corruption. Although surveys of citizens and firms have proliferated in the years since the 2005 Diagnostic, surveys of public officials about their experiences are rare. Such surveys help 2. Surveys of firms include the Provincial Competitiveness Index (PCI) survey (annually since 2005), the World Bank’s Enterprise Survey (2009), the University of Copenhagen and CIEM panel survey of SMEs (2005, 2007, 2009) and the Integrity and Transparency in Business Initiative survey of firms (2011). Surveys of citizens include the Governance Module attached to the VHLSS (2008), the Global Corruption Barometer carried out by Transparency International (2011), and the Vietnam Provincial Governance and Public Administration Performance Index (PAPI) (2011-2). Sector specific studies include those carried out by Towards Transparency on health (2011) and youth (2011), and those commissioned the Government Inspectorate with support from UNDP on education, land, and health care. Related aspects of governance, such as access to information have also been the subject of surveys, for example in the study supported by the World Bank on transparency of land related documents. 18 Part I - OVERVIEW us to understand the public official’s perspective on corruption and, more importantly, help us to identify which aspects of anticorruption policies and institutions seem to be working and not working. They help identify areas where reforms are more likely to meet resistance from within the bureaucracy, and those for which public officials are likely to lend their support. Surveys of public officials can also help to identify which aspects of public administration are most associated with integrity and, conversely, with corruption or other measures of poor performance. The surveys of firms and citizens, similarly, bring in new elements that complement existing surveys. The surveys of citizens and firms on which this report is based ask not only about the types of corruption they face, but about how corruption works. They also delve into the more complex and challenging set of issues surrounding the ways that firms collectively express their influence, in both good ways and in bad ways, over policies. By posing the questions to firms directly, the surveys bring data to bear on the emerging challenge of interest groups as Vietnam’s economy becomes more complex. Finally, a key benefit of querying all three sample groups on the same issues is that it helps to identify areas where ideas of social norms are not equally shared among citizens and public officials. In the murky field of corruption, an understanding of social norms is essential for identifying approaches to corruption that are likely to bear results. 1.2. Objective and scope of the surveys The government of Vietnam is in the process of reviewing the implementation of the Anticorruption Law and the Party’s resolution on anticorruption passed in 2006, as well as preparing to submit to the National Assembly an amendment to the current Anticorruption Law. To inform these reviews, the Government Inspectorate (GI) and the Office of the Steering Committee on Anticorruption (OSCAC) commissioned the present study. Under the leadership of GI and OSCAC, the study was carried out with a view to providing an empirically grounded source of information on the nature, causes and consequences of corruption in Vietnam, from the perspective of public officials, firms and citizens. Specifically, the objectives of the study are as follows: • To understand the prevalence and seriousness3, and forms of corruption in Vietnam. 3. “Prevalenceâ€? here is understood as the degree to which corruption is “widespreadâ€?, the number of cases of corruption, the number of corrupt persons detected and sanctioned. “Seriousnessâ€? here is understood to refer to the economic losses associated with corruption, the complexity of corruption cases and the inability to detect corruption. 19 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS • To identify the causes of corruption and the factors that constrain the effective implementation of anticorruption work. • To help inform new directions for anticorruption efforts of Vietnam in the coming years. Geographic and sectoral coverage: The survey covered ten provinces, including the same seven provinces that were included in the 2005 Diagnostic (Ha Noi, Ho Chi Minh City, Son La, Hai Duong, Dong Thap, Nghe An, and Thua Thien Hue) and three additional cities (Da Nang, Hai Phong, and Can Tho). The selection of these ten provinces was driven by the desire to focus on bigger cities and urban areas in Vietnam where there is more socio-economic activity and where the perception of corruption risks is higher. The sample frames in the 10 selected provinces may not necessarily represent the whole population of citizens, enterprises and public officials of Vietnam, but the results are nevertheless compelling as the ten sampled provinces are home to 30% of the country’s population and produce more than 65% of Vietnam’s GDP. The survey also covered officials of five ministries, including three ministries that were included in the 2005 survey (Transportation, Construction, Industry and Commerce) and two additional ministries (Ministry of Finance and the Ministry of Natural Resources & Environment). MOF and MONRE were added to the sample because these ministries oversee areas (e.g., tax, customs, land, and mining) that had been reported by several previous surveys as areas with high corruption risks. Although including officials from the Ministry of Public Security, Ministry of Justice, or the court system would have produced valuable insights on corruption issues and anticorruption work, it was not possible to include these bodies in the survey due to time and resource constraints. (Other limitations are discussed in Section 1.4, below.) Types of respondents: Three groups of respondents were included in this survey: - Citizens: Respondents were approached as representatives of their households. Citizens who were 18 years old or older were eligible to the survey. Since the goal was to learn of experiences in interacting with state bodies and officials, the sample design emphasized citizens who had recently visited One Stop Shops. - Firms: The survey was aimed at formally registered firms. Respondents from the firms could be members of the Board of Directors or the Head of Planning, Logistics, Administrative, or Accounting Departments. There was a single respondent per firm. 20 Part I - OVERVIEW - Officials4: At the ministry level, respondents were the Head of Directorate or Department and below. Ministers or Vice Ministers were not included in the survey. At the province level, respondents were staff members up to Department Head. Higher level officials included representatives of the provincial People’s Councils. People’s Committee Chairmen or Vice Chairmen were not included in the survey. Types of information covered in the surveys: The surveys focused on corruption-risk behaviors emerging from the interaction between public officials and citizens, between public officials and firms, and among public officials. Corruption within the private sector (i.e., in the interactions among firms) was not covered. As will be clear throughout this report, the surveys sought to understand the level of interest in corruption among all three respondent groups, their perceptions and experiences of corruption at different levels, in different sectors and in using public services. Views on the causes of corruption, on denunciations and public procurement, and measures for improving effectiveness in fighting corruption were also included in the surveys. The surveys also sought to break new ground and provide insights into emerging or contentious issues such as interest groups and the recruitment and promotion practices in the public sector. 1.3. Methodology The methodology on which this report is based has been developed over some fifteen years and implemented in dozens of countries around the world. These countries share one key feature: they all have governments that have expressed a commitment to address corruption and a desire to understand corruption’s nature and causes. The present surveys were conducted based on the following fundamental principles: (i) The surveys focused on the elements of the population with the most experience in dealing with state agencies, not the overall population of Vietnam; (ii) randomness in sampling was guaranteed to the extent possible; (iii) all interviews were conducted face-to-face in a private setting; and (iv) close monitoring and supervision for quality assurance was strictly followed throughout all steps of the survey process, including preparation, during field work, post-field work monitoring, and during the data entry and analysis phase. 4. “Officialsâ€? in the public official survey included also cadres working in the provincial councils and public employees in service delivery units. 21 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Preparation for the surveys The present surveys were carried out under the leadership and with the active support of both the Government Inspectorate and OSCAC. The overall process was guided by an Advisory Board including participation of these two bodies, as well as by the Office of the Government, Vietnam Chamber of Commerce and Industry, the Fatherland Front, and the World Bank. A Task Force headed by the Anticorruption Bureau of GI and OSCAC was established to support and supervise the research process. From the side of development partners, UK-DFID and UNDP joined the World Bank in providing expertise, guidance, and resources. Several technical workshops were held among stakeholders during the preparation phase to make needed organizational arrangements and to finalize the survey objectives, approach and methodology, as described below. The questionnaires and approach were developed jointly by GI, OSCAC, and several development partners including the World Bank, UK-DFID, and UNDP, and the team from T&C Consulting and APIM which was contracted to carry out the surveys. The three sets of questionnaires for public officials, enterprise managers and citizens developed for the 2005 Diagnostic Survey were used as starting points. During the development of the questionnaires, references were also made to other questionnaires such as the similar diagnostic surveys carried out by the World Bank in Europe and Central Asia, and Vietnam’s PAPI, which is implemented by UNDP, CECODES and the VFF. For all the three sets of questionnaires, the team took out or updated questions that do not fit in the current context and added questions on emerging issues such as that of interest groups. The draft questionnaires were revised following the pilot survey in Vinh Phuc province, and during the training of trainers for interviewers, and finalized prior to the field surveys. The sampling approach was designed with several purposes in mind. As the key objective of the surveys was to better understand corruption, both the locations and respondents were chosen to provide the most well-informed opinions and experiences. In all, 2,601 citizens, 1,058 firms, and 1,801 public officials were surveyed. The sample of the citizens survey was drawn from households that had recent experience using public services. For this survey, three districts in every province and three communes from each selected district were randomly selected. In each commune, two villages - one central and one distant - were required to provide a list of those who had used public services. The lists, which were provided by the commune People’s Committee offices or by village heads, were provided without knowledge of the survey purposes, and respondents were not informed that this was the basis for their selection. From those lists, interviewees were 22 Part I - OVERVIEW randomly selected for face-to-face interviews. Replacement respondents for interviewees who could not be reached were chosen following a preset rule. The number of respondents per province varied according to the size of the province, with the largest cities receiving the largest share of respondents. The public officials survey covered a sample of 1,801 public officials, of which 95% (1,711) were interviewed from local government levels, and 5% (90) were from ministerial level bodies. The reason a larger proportion of local government officials was sampled is that this reflects the reality of the population of public officials in Vietnam. Public officials from ministerial to commune levels were pre-selected according to their sector, position and profession, but the actual interviewees were only known at the time of the interview. In each ministry, six departments were selected and in each department, one leader and two staff members were interviewed. The composition of public officials in the sample for each province is presented in Table 1, and details of the public official sample are described in Annex 1. Table 1. Composition of the public official sample in each province The enterprise survey covered 1,058 enterprises across ten provinces and followed stratified random sampling by sector, size, and ownership such that the structure of the enterprise sample is close to that of the actual population. The classification of registered enterprises for the stratification exercise was defined by the System for Economic Sector of Vietnam, enacted with Decision No. 10/2007/QD-TTg dated 23/01/2007 of the Prime Minister. The thresholds for size of enterprises as small, medium and large by either labor or capital followed Decree 56/ND-CP, dated 30/6/2009. Enterprise ownership followed the classifications in the Enterprise Law: 100% state owned, more than 50% state owned, less than 50% state owned, purely domestic and foreign invested enterprises. The number of surveyed enterprises in each province was determined by the actual numbers of currently active enterprises reported in the 2010 Census of Enterprises conducted by the General Statistics Office. Replacements were made only for enterprises that could not be found or for those who refused after three efforts to contact. 23 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS The composition of respondents by provinces and sample groups is presented in Table 2 below. The pilot survey was carried out from December 21-22, 2011 in Vinh Phuc province to test the survey plan and tools in practice. Twenty seven pilot interviews were conducted. A brief retrospective was carried out after each day in the field, and an entire third day was spent to draw lessons learned from the pilot survey. Table 2. Number of respondents by province and sample group Training was conducted for provincial team leaders and interviewers. The former received two training sessions, one right after the pilot and the other before the field work, conducted by the core group. The latter received two day training sessions delivered by the provincial team leaders and supported by the core group. The provincial team leaders were trained in how to conduct interviews, how to train others in conducting these interviews, and how to supervise and monitor the interviews. Special attention was also made during the training of interviewers on how best to make the interviewees feel comfortable in sharing their own experience and perceptions on this sensitive topic. Field work Most of the interviews took place from March 12 to April 10, 2012. Interviews in Hanoi began earlier and took longer to complete, from February 26 to April 10, 2012. Each and every interview was a direct, personal interview. The respondent was given a copy of the questionnaire to help follow the questions. After an introduction to the research and some conversation to put the respondent at ease, the interviewer read each 24 Part I - OVERVIEW question, and marked the answer(s) into his/her copy of the questionnaire in front of the respondent. GI and OSCAC played key roles in connecting the interviewing teams with the local contact agency, either the Provincial Inspectorate or the Provincial OSCAC, or both. The team leaders were responsible for directly conducting interviews at the provincial level and for monitoring the interviews conducted by local interviewers. The interviewers, in turn, were responsible for conducting interviews with district and commune public officials, citizens and enterprise managers. The core group was responsible for all interviews at the ministerial level. Monitoring during and after the field work A multi-layer structure of supervision was organized during the field work, led by a general supervisor and coordinator, under whom were three regional supervisors for the Northern, Central and Southern regions. Each regional supervisor was responsible for monitoring progress and randomly checking the field work in three provinces. In addition to regular monitoring channels, the World Bank, GI and OSCAC, and their provincial counterparts (Provincial Inspectorates) also made independent and random on-site checks. The quality of the response sheets was verified in four ways: (i) self checking; (ii) full checking by the team leader; (iii) random checks by regional supervisors; and (iv) re- verification by World Bank and T&C staff. After the field work was complete, a series of call-backs were made to a random selection of respondent enterprises and households to make sure that the interviews were conducted in a professional and proper fashion. Such call-backs were not conducted for respondent government officials since all meetings with this kind of respondent were scheduled in advance and were strictly followed. Data entry and analysis Data entry was conducted in such a way as to minimize random errors. All data sheets were transcribed into a computer program twice by two different staff, and all discrepancies were identified and resolved. Similarly, to minimize errors during the analysis phase, the analysis was jointly conducted by the World Bank and T&C/APIM experts, with each verifying each other’s calculations and agreeing to the findings prior to drafting the report. The early findings of the surveys were presented to GI and OSCAC in April 2012, and the first draft of this report was consulted in two two-day technical workshops in May 2012 with the Advisory Board, GI, OSCAC, and representatives of Office of the Central 25 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Party Committee, Office of the National Assembly and all interviewed ministries and provinces, in addition to cooperating development partners DFID and UNDP. The constructive comments provided by these stakeholders provided directions for a substantial revision of this report. A revised draft was submitted to the Advisory Board and stakeholders in July for final comments. Annex 1 provides more detailed information on methodology, especially on sampling and on the quality assurance mechanisms. 1.4. Limitations of a survey-based approach The benefits of undertaking a survey-based approach to understanding corruption have been outlined earlier. At the same time, surveys should be considered as elements among the many other sources of information about the phenomena of corruption, rather than as the sole tool which can answer all questions. The numbers that surveys provide can illustrate the magnitude of the problem, but cannot convey the impact that corruption has on human dignity or the opportunities lost. Our understanding of the magnitude of corruption may also be shaped by the methodology being used. • Surveys are best suited for illuminating corruption in the interface between public officials and the populace, both citizens and firms, but they are not as well suited for identifying corruption that does not fall in that interface, such as embezzlement or misuse of public funds. • Since the emphasis in the survey is on interactions between public officials and citizens and firms, the emphasis in the report may also place much emphasis on unofficial payments and “petty corruptionâ€?, but that does not mean that grand scale corruption is not also a problem. Similarly, the survey approach targets the forms of corruption behaviors that could be experienced by a large number of respondents. Special behaviors that only few people have opportunities to know (e.g., land acquisition for big projects, embezzlement of public assets) were not the focus of this survey. While these special corruption behaviors may be very serious, a sample survey as this one, even one with a relatively large sample, is not likely to find many respondents able to report on such experiences. • Resource constraints allowed for samples that are reasonably large at the province level (described in more detail below), but not large enough to allow meaningful analysis of responses of all sub-groups. For example, a more meaningful analysis of corruption in the courts would require interviews with investigators, 26 Part I - OVERVIEW prosecutors, judges and lawyers, as well as larger numbers of actual court users, and similarly for health, education, etc. Such an approach was not possible for this study, but would be a useful avenue for future research. • The province-level average sample sizes (260 for citizens and 106 for enterprises) are similar to those used in other surveys such as PAPI (217 citizens) and PCI (141 enterprises). At the same time, larger sample sizes would have made the estimates provided in this study more precise. As explained in Box 1, larger sample sizes would have also come at a cost in terms of breadth and depth of questions and timeliness of the study. In addition, since the surveys covered ten provinces, the results are best interpreted as reflecting the situation in those ten provinces, rather than for the country as a whole. 1.5. Roadmap to the rest of the report The report is organized as follows: Part II will present the main findings about the level of corruption and the types of corruption reported by respondents to the surveys, exploring the understanding of the term “corruptionâ€?, perceptions of corruption at various levels of government and in various sectors, and actual experiences encountering corruption with various state agencies. Part II will also explore which anticorruption measures seem to be having the most impact, and how the approach to anticorruption can be improved. Comparisons to other surveys such as PAPI and PCI and comparisons between the 2005 survey results and those for 2012 are also included in Part II. Part III offers the key conclusions and recommendations. Part IV includes Annexes with some additional details on methodology and certain analyses. Box 1. How do the sample sizes in the Anticorruption Diagnostic Surveys compare with other surveys? Are they large enough? Any sample survey faces a trade-off. Estimates will be more precise the more observations in the sample, but this comes at a cost. Larger samples take longer to implement and are more expensive. With a limited budget and schedule, then, larger samples come at the expense of depth and breadth of questions. The relationship between sample size and estimate precision can be illustrated with an example. One of the questions on the survey is for whether the respondent knows how to denounce a corrupt act. Over all ten provinces and 2,601 citizens, 47.3% of respondents said “yesâ€?, but this is only an estimate. We are 95% sure that the actual 27 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS number of people who know how to denounce a corrupt act is somewhere between 45.4% and 49.3%. At the province level, however, the estimates are much less precise. In one smaller province with 180 observations, for example, 56.7% of respondents said “yesâ€?, but since the number of observations is smaller, the estimate is also less precise. We can be 95% sure that the actual number of people who know how to denounce a corrupt act in this province is between 49.4% and 64.0%.5 Average observations per province Approximate margin of error of estimates Citizens 260 ±6% Enterprises 106 ±10% Public Officials 180 ±7% All ten provinces Approximate margin of error of estimates Citizens 2,601 ±2% Enterprises 1,058 ±3% Public Officials 1,801 ±2% Since the province level estimates are less precise, we will generally not draw conclusions in this report about the levels of corruption at the province level, but rather focus on the insights about the nature and causes of corruption more generally. 5. The examples illustrated here assume a yes/no type of question with roughly 50% answering yes. These are approximate only for the purposes of illustration. For public officials, for whom the population is not well defined, the estimates of confidence intervals are especially imprecise. 28 Part II Survey Findings 2.1. Corruption perceptions and experiences Understanding the level, pattern and nature of corruption after six years of implementation of the Anticorruption Law is essential if we are to understand the new challenges facing Vietnam. As noted in the overview, the results represent the views of the ten provinces in which the survey was carried out, but not necessarily the whole of Vietnam. Since these ten surveyed provinces account for 30% of the population and produce 65% of the GDP of the country, the results are nevertheless compelling. This section begins by examining how corruption compares with other issues that occupy the public’s attention, and probes the question of just what people mean when they refer to “corruptionâ€?. The discussion then presents basic findings regarding the perceptions of the overall prevalence and seriousness of corruption at various levels and in various sectors. The surveys do not only ask about perceptions of corruption, however, they also ask questions about experiences with corruption, and these will be presented next. The discussion then focuses on the information sources on which people base their perceptions of corruption. The discussion will then explore the assessments of firms about the emerging challenge of interest groups. 2.1.1. Perceptions of corruption 2.1.1.1. How does corruption compare with other issues that concern the population? Corruption exists in all countries, although the levels may vary. It has been shown to be an impediment to economic growth and sustainable development of a country. It reduces the effectiveness of public administration and makes public expenditures inefficient. Most seriously, corruption erodes the rule of law, harms the reputation of the state and weakens citizens’ trust in the institutions of state, as well. 29 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS At the same time, we know that society faces many other problems besides corruption. In order to put concerns about corruption into perspective, all the survey respondents were asked to indicate their level of interest in each of ten problems facing society. The results confirm that corruption attracts much concern from society. Although corruption was not cited quite as frequently as an issue of interest as other problems facing society, in absolute terms the level of interest is high: around 75% of respondents from each of the three sample groups indicted that they are interested or very interested in corruption. Although some citizens expressed greater interest in many of the other issues, such as the quality of education, the cost of living, and food safety, it is clear that corruption is viewed as a serious problem. When asked for the three most serious issues for Vietnam at present, more than one out of three respondents selected corruption as among the top three problems. (Figure 1). More public officials selected corruption as among the top problems for the country than any of the other problems on the list. Among enterprises, corruption was second only to the cost of living as a cause of concern - as recent bouts of inflation cause uncertainty for firms, this result is not surprising. And among citizens, only the cost of living and traffic accidents were selected by more people as among the most serious issues for Vietnam. It is obvious that in the views of the respondents, corruption remains one of the top concerns of society. Figure 1. The three most serious issues for Vietnam, according to public officials, enterprises, and citizens (%) 30 Part II - SURVEY FINDINGS Indeed, it should be pointed out that the various problems that Vietnam faces, those depicted in Figure 1, may all be related to corruption in some sense. If people are concerned about traffic accidents, for example, then they need to have trust that the institutions managing traffic patterns and enforcing traffic laws are making decisions based on safety, not corruption. Later in this report, the perceptions of corruption among the traffic police and in healthcare, education, and the market for jobs with the state will be presented. To the extent that corruption undermines the quality of services such as education and health, or erodes the fairness and merit-orientation of the public sector jobs market, or makes enforcement of traffic laws selective, then corruption contributes also to the other social problems Vietnam faces. 2.1.1.2. What does the word “corruptionâ€? mean to people? While the Law on Anticorruption already defines corruption and 12 corrupt acts, understanding how the term is interpreted by the population is also important, for several reasons. First, the surveys repeatedly query respondents about their perceptions of corruption - understanding how that word is interpreted helps us in interpreting the other questions on the survey. Second, the interpretation of the word “corruptionâ€? implicitly points to what society deems acceptable and unacceptable behavior. By understanding more clearly how this term is understood by Vietnam’s citizens, enterprise managers, and public officials, we will see more clearly how and where public information campaigns are needed to try to shift public perceptions. All three respondent groups were presented with eight different hypothetical situations and asked to assess whether it was “definitely corruptionâ€?, “definitely not corruptionâ€?, or “in some cases corruptionâ€?. The responses suggest a fairly high degree of condemnation of the various scenarios as “corruptionâ€?, even many that are not considered corruption according to the Anticorruption Law of 2005. The others, while perhaps inappropriate, would nevertheless not be enough to constitute corruption. Gifts are not forbidden and those acts would be considered corruption only if the agencies or public officials are found to have twisted the rules and in turn received money for personal gain. That there is such strong condemnation as “corruptionâ€? for many acts which are not defined as such by the law suggest an evolving set of standards of public service. 31 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 2. What is “Corruptionâ€?? (% saying “definitely corruptionâ€?) Several notable features of Figure 2 stand out. First, across the board, citizens are more likely than the other two sample groups to call something “corruptionâ€?. This applies to all of the eight scenarios, and the difference is largest when it comes to the unofficial payments for health services: while only 32% of public officials said such unofficial payments are “corruptionâ€?, 45% of the citizens said so. Second, even in situations that seem 32 Part II - SURVEY FINDINGS to be clearly subverting the rules, some respondents do not condemn such acts as “corruptionâ€?. For example, only two-thirds of public officials and enterprises said that a teacher receiving a gift in order to change a student’s grade is “corruptionâ€?. The remainder either denied it was corruption or said only that “in some ways it is corruption.â€? The responses to this question about the understanding of “corruptionâ€? also highlight the need to build a common understanding. It is perhaps encouraging that those who work in a given field tend to be less forgiving of certain behaviors than for the population as a whole. For example, public officials who work in the health sector are significantly more likely than the rest of public officials to declare that a health worker receiving a 300,000 VND unofficial payment is “definitely corruptionâ€?. Whereas 31% of public officials said this was “definitely corruptionâ€?, 42% of public officials who work in the health sector reported so. The same is true for education, although the effect is less strong: a teacher receiving a gift in exchange for a better grade was called “definitely corruptionâ€? by 72% of public officials who work in the education sector, compared to 69% for the rest of public officials. At the same time, there is some evidence that those who supply the unofficial payments are less likely to call the payments “corruptionâ€?. For example, among citizens who have asked schools for something concerning a child’s education, 69% said that a teacher receiving a gift in exchange for a better grade was “definitely corruptionâ€?, compared to 75% for the rest of the population. (Such differences for the health sector are not statistically significant). Whether or not such acts are officially considered corruption by law, a uniform understanding of what is acceptable and is not acceptable would help to reduce the willingness of the population to feed the supply side of bribery. 2.1.1.3. How do citizens and officials perceive the prevalence and seriousness of corruption? An important objective of the survey was to understand how people perceive corruption prevalence (how frequently corruption occurs) and seriousness (how damaging it is) in the country. Citizens were asked about their perceptions of corruption prevalence and seriousness in the nation, province, district, and commune, respectively. Similarly, officials were asked to indicate their perception of corruption prevalence and seriousness in the nation, in the locality they are working, and in their sectors (e.g., environment and natural resources, finance). The results suggest that people perceive a national problem, but are less willing to identify corruption as a local problem. The vast majority (82.3%) of citizens in the sample believe that corruption is prevalent or very prevalent nation-wide.6 However, they assessed the prevalence of corruption to be 6. For presentation purposes, prevalent or very prevalent will be grouped together. 33 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS lower in their province (61%), district (34.8%), and commune (14%). Somewhat surprisingly, the percentage of citizens who answered “don’t knowâ€? increased from 6.2% when asked about corruption prevalence nation-wide to 9.0% when asked about corruption in the province, to 14.8% when asked about districts, and 16.9% for communes. This may reflect the possibility that some people are less willing to talk about corruption in their locality. Understanding how corruption is perceived by respondents with different characteristics helps identify the groups most affected by corruption. People who consider themselves poor see corruption as less prevalent nation-wide as well as in the provinces, compared to people who consider themselves rich. They, however, saw more corruption prevalence in communes (17%) than average income (15%) and rich people (10%). One possible explanation is that poor people are hit more frequently and more strongly by corruption in communes, whereas decisions made at higher levels of government have more impact on those of average income and the wealthy. People’s assessment of corruption seriousness (emphasizing large scale corruption that is most damaging to society) showed a similar pattern with more people perceiving corruption as serious or very serious7 at the national level (86.5%), and a much lower percentage of people perceiving corruption to be serious at the province (66.7%), district (44.1%), and especially at the commune level (19.8%). Similarly, more people answered “don’t knowâ€? when asked about corruption seriousness at the commune level (17.5%) than at the district (15.0%), province (9.0%), or national levels (6.5%). Figure 3. People’s perceptions of corruption prevalence (% saying corruption is prevalent, by income) 7. For presentation purposes, serious and very serious will be grouped together. 34 Part II - SURVEY FINDINGS Figure 4. People’s perceptions of corruption seriousness (% saying corruption is serious, by income) Again, a comparison between income groups shows striking differences in perceptions of the seriousness of corruption at the national and commune levels. At the national level, rich people tended to be more critical than average income people, who in turn were more critical than poor people on corruption seriousness. (Figure 4). At the commune level, the reverse was true: 23% of poor people perceived corruption to be serious, while this number was only 21% for people of average income, and 12% for rich people. Again, poor people appeared to suffer more than others from corruption at the commune level. Officials were asked a similar question on perceptions of corruption prevalence nation-wide, at the localities they were working in, and in their sectors. Seventy-eight percent of officials perceive that corruption was prevalent nation-wide. These numbers are much lower for corruption prevalence at localities (37.5%), and sectors (20.4%). A comparison between officials at different local management levels shows some interesting results. While a smaller percentage of district and commune officials perceive corruption as prevalent at their localities, central and provincial officials are more critical: 72% of central officials and 56% of provincial officials perceive corruption to be prevalent at their localities. Similarly, more central (39%) and provincial (32%) officials perceive corruption to be prevalent in their sectors than district and commune officials (27%). 35 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 5. Perceptions of corruption prevalence by central vs. local officials (% officials saying corruption is serious) Note: Shows the percentage of all respondents who say corruption is “prevalentâ€?, or “very prevalentâ€?, among a set of choices that also included “totally not prevalentâ€?, “very little prevalentâ€?, “little prevalentâ€?, and “difficult to answerâ€?. Figure 6. Perceptions of corruption seriousness by central vs. local officials (% officials saying corruption is prevalent) Note: Shows the percentage of all respondents who say corruption is “seriousâ€?, “very seriousâ€? or “especially seriousâ€?, among a set of choices that also included “not seriousâ€?, “little seriousâ€?, and “difficult to answerâ€?. A comparison between male and female officials showed some differences in their perceptions of corruption prevalence at the national level (80% for women vs. 77% for men). However, male officials tended to perceive more corruption prevalence at the 36 Part II - SURVEY FINDINGS localities (40% for men vs. 33% for women), and their sectors (22% for men vs. 17% for women). Further study will be needed to identify the reasons for the differences in perceptions. The patterns for corruption seriousness mimic those of corruption prevalence: 86.6% of officials perceive corruption to be serious at the national level, 46.7% for localities, and 28.4% for sectors. A comparison between officials at the central and different local levels showed a more critical view from central and provincial officials. The four groups (commune, district, provincial, and central officials) were somewhat similar in their perception of corruption seriousness at the national level. However, more central (68%) and provincial officials (50%) perceived corruption as serious at the localities than district (35%) and commune (31%) officials. Central officials were also more critical when talking about their sector with 33% of them reporting that corruption is serious in their sector. These numbers for other groups were around 20% only. In short, around 80-86% of both citizens and officials perceived corruption as prevalent and serious at the national level. However, a much smaller percentage of citizens and officials perceived corruption to be prevalent and serious in their localities and sectors (for officials). Even in the best case, however, around 20% of respondents to the citizen’s survey said corruption is serious in their communes, and more than 25% of officials said corruption in their sectors is serious or very serious. Poor people were somewhat less concerned about corruption at the national level than rich people. However, poor people were more critical of corruption prevalence and seriousness at the commune level. One explanation is that, as we shall see in Section 2.1.2.1, poor people are hit more frequently and more strongly by corruption at the commune level, whereas the sorts of services used by wealthier citizens are provided by provinces. 2.1.1.4. In which sectors is corruption perceived to be the most prevalent? Corruption ranges from high-level political graft to low-level bribes. Corruption takes many different forms, and varies across sectors and victims of corruption. In this survey, all three groups of respondents were asked their views on how common corruption is in various sectors. According to respondents, the prevalence of corruption varies across sectors, although there is a high degree of consensus among the three surveyed groups on which sectors are the worst. (Figure 7). The top four most corrupt sectors, according to respondents, are the traffic police, land administration, customs and construction. More than 75% of respondents for all three 37 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS groups say corruption is common in these sectors. According to enterprises, finance and banking seem to be less corrupt than in the views of public officials and citizens. The four least corrupt sectors are post and telecommunication, media, treasury, and the ward/commune police. Figure 7. Perceptions of the prevalence of corruption across sectors according to public officials, enterprises, and citizens (% saying prevalent, among those with opinions) Each group of respondents was also asked to give their perception of the three most corrupt sectors among twenty-two, and the results are shown in Figure 8. The results suggest a widespread belief that corruption in the traffic police and land administration is a serious problem, with construction and customs following. As noted earlier, a limitation of a survey-based approach is that it focuses on the forms of corruption experienced by large numbers of firms or citizens, but not necessarily the most serious forms of corruption. Land acquisition, or mismanagement or embezzlement of state assets, for example, may have very harmful effects even if they are less prominent in surveys. Figure 8 focuses on the perceptions of the most corrupt sectors that have the largest interface with the society, and no doubt would cause discontent among the population if not properly addressed. 38 Part II - SURVEY FINDINGS Figure 8. The most corrupt sectors according to public officials, enterprises, and citizens (% selecting sector as among the top 3) 2.1.2. Experiences with corruption An understanding of how the population - citizens, firms, and public officials - perceive corruption is essential, since it is perceptions on which firms base their investment decisions, on which citizens decide to seek services, etc. At the same time, an understanding of actual experiences is also needed to understand where corruption really is encountered by the population and how corruption works. A better understanding of how corruption works can help point the way toward the solutions to the problem. This section presents the survey findings on how corruption is experienced by enterprises, citizens, and officials, and what each group has to say about how corruption works. 2.1.2.1. What kind of requests do firms receive from officials? What sorts of requests, whether corrupt or otherwise, do firms receive from public officials? Firms were asked whether they received different types of requests from officials in the previous 12 months. The requests included things like firms selling assets or land to officials at low prices, recruiting or promoting relatives or acquaintances of officials, and paying to officials and/or his/her organization’s expenses. Fewer than 5% of firms received 39 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS requests to sell assets at low prices, or lend assets/equipment to individual officials, and only a slightly higher percentage received requests to pay for the costs of officials’ studies, visits or private expenses. Close to 8% of firms received a request to recruit or promote relatives or close acquaintances of officials. More than 15% of firms experienced a situation where officials abused the power, names, or reputation of their agencies to suggest that the firms pay money or give gifts to them. All together, 23% of firms said they had encountered one of these six requests in the previous 12 months. Figure 9. Suggestions to firms (%) 2.1.2.2. Why do firms pay bribes and to whom? The previous section showed that some firms receive requests from officials for a range of benefits. But why would some officials believe that firms will comply with those requests, and why would some firms, in fact, play along? Firms were asked what sorts of difficulties state agencies made for them, and about half of the firms in the sample responded to these questions. The other half either did not face any difficulties by state organizations or did not remember. Among those who responded, 63% said that officials intentionally prolong the time to solve firms’ requests, 58% said that officials do not explain the requirements clearly but try to catch firms’ mistakes to refuse solving firms’ requests, and 28% said that officials try to manipulate vague regulations or conflicts in regulations to put pressure on firms. (Figure 10). Firms that faced difficulties created by state organizations were then asked how they responded to these difficulties. (Figure 11). Seventy eight percent of firms kept on waiting, 40 Part II - SURVEY FINDINGS and 86% of firms argued with the agencies. Notably, about 51% of firms said they sought help from powerful people, and 59% of firms chose to give gifts and/or money to the officials to settle the work. Only about 13% of firms looked to law enforcement agencies for help, and fewer than 6% went to press agencies for help. Figure 10. Difficulties created by public officials to firms (among those with interactions with state agencies) 41 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Firms were asked if they had to pay informal charges to various state organizations, and if they did, why. (Figure 12). Around 32% of firms paying informal charges said that it was the quickest and most easily-implementable way to get things done. About 26% reported that the informal charges were small compared to the gains from it, and other firms did the same. About 18% of the firms believed that without such informal charges, it was impossible to get things done. Figure 11. Firms’ responses to difficulties created by state organizations (%) Figure 12. Why firms pay informal charges (%) 42 Part II - SURVEY FINDINGS As noted above, many firms engage in corruption as an expedient means of getting things done. Respondents in the business sector were asked about the impacts of informal charges on their firms. (Figure 13). About 60% of them believe that informal charges are costly. However, 50% think that the benefits from informal charge are greater than the costs. Close to 63% of respondents believe that informal charges create an “unspoken mechanism to get things done more quicklyâ€?, and 53% believe that informal charges motivate cadres to solve public work more actively. Thus, from the firms’ point of view, albeit a short-term and narrow view, informal charges bring more benefits than costs. Figure 13. Perceptions of the impacts of informal charges to the firms The fact that corruption is viewed by some firms as having some benefits does not mean that corruption is efficient, since such corruption may degrade ethics and makes the administrative system inherently unfair. Many firms recognize this: about 37% said informal charges may ruin their staff’s ethics, and 28% thought informal charges created obstacles to getting things done. Moreover, 57% of firms thought that informal charges generate unfairness. Such an administrative system is not only unfair, it also creates incentives for officials who have power and discretion to create difficulties for firms in order to draw out the unofficial payments. Firms were asked whether they encountered difficulties, whether they had to pay unofficial money/gifts, and how they did it. Figure 14 shows the percentage of firms that interacted with each agency that said the agency 43 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS made difficulties for them, and the percentage of firms that said that agency was among the three worst in terms of making difficulties for enterprises. Figure 14. Agencies making difficulties and three agencies making the most difficulties (% of firms) At least 10% of enterprises reveal that difficulties were made when using services provided by any public agencies, and for some the fraction was much higher. Customs, the traffic police, and tax administration were cited by 30% or more of firms as agencies/bodies that create difficulties. When asked to identify the three agencies most likely to make difficulties, 58% of respondents picked taxation, with sector administration agencies in the second place with 23% of responses, the third and fourth places are traffic police with 21% and natural resources and environment with 20%. When agencies make difficulties for enterprises, this does not necessarily mean there is corruption, but it does open opportunities for corruption. Indeed, as Figure 15 shows, agencies that are more likely to make difficulties for firms are, on average, also the agencies to whom firms make unofficial payments most often. This implies that when public agencies actively make difficulties for enterprises, the chance of corruption increases. Respondents were also asked about their direct experiences with paying bribes. (Figure 16). The percentage of enterprises that had to pay unofficial money/gifts varies 44 Part II - SURVEY FINDINGS across sectors. Enterprises reported paying unofficial money most often when using services provided by taxation agencies (33%). When dealing with sector administration agencies, 20% of enterprises said they make unofficial payments. The next three sectors are banking (17%), traffic police (16%), and customs (16%). Figure 15. Agencies making the most difficulties are also those receiving the most informal money/gifts Among enterprises paying unofficial money, a large portion of enterprises actively initiated giving the money/gifts, and the remainder were asked by the officials when using the services. The results show that in more than 70% of the cases, the unofficial payments were actively proposed by enterprises, and in less than 30% of cases the unofficial payments were demanded. Among sectors, market management agencies top the list of agencies requesting unofficial payments, and the second is traffic police, followed by the economic police, natural resources and environment management, and construction. Corruption is not only initiated by people in charge or service providers but also largely actively proposed by service users. Besides giving unofficial money/gifts when using services, corruption may take other forms, as well. Unnecessary visits, giving money/gifts on occasions of holidays and 45 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS festivals, and paying for banquets and entertainment could present forms of corruption risks. Figure 17 shows how often these practices took place, in terms of enterprises’ experiences, in the previous twelve months. Figure 16. To whom is unofficial money paid, and who suggests? According to the enterprises, about 10% said that they have to welcome unnecessary visits from agencies - planning and investment and inspection and control were cited as the first two agencies most likely to make unnecessary visits in the last twelve months. The surveyed results show that less than 5% of surveyed enterprises have to pay unofficial money for banquets and entertainment of service providers, and this occurs most likely in taxation, sector administration, banking, and inspection and control agencies. Finally, the most prevalent activity is giving money/gifts on occasions of holidays and festivals. This transaction does not take place at the time of supplying/using services, so some may not consider this to be a corrupt behavior officially. The results reveal that taxation is on top with 44% of surveyed enterprises using such an approach. The next three are sector administration agencies, banking and local police. 46 Part II - SURVEY FINDINGS Figure 17. Keeping good relations with officials. Enterprises’ experiences in the last 12 months 2.1.2.3. Is there really a pay-off to corruption? The preceding analysis suggests that many firms make unofficial payments unilaterally in order to get things done faster and to deal with difficulties made by state agencies. It seems that many firms see this as a necessary cost of doing business. Whether or not such a strategy actually has a benefit for the firm, however, is another question. The enterprise survey asks firms how they assess the operational efficiency of the firm compared to two years before. Firms could select one of five choices ranging from much worse than before, to much better than before. Overall, 51% of firms said that they were doing better, compared with 35% which said they were doing worse, numbers that are consistent with the overall growth of the Vietnamese economy. Do firms that pay bribes do better than firms that do not? The answer is “no.â€? On average, firms that had paid a bribe in the previous 12 months were actually growing more slowly than firms that said they did not. (Left panel of Figure 18). The effect is even more striking when probing firms about their approach to dealing with difficult situations. The firms that were growing fastest, on average, were the ones that said they routinely do not react 47 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS to difficulties by state agencies by presenting gifts or presents to the officers in charge. (Right panel of Figure 18). In contrast, firms that said they frequently adopt this tactic were, on average, not seeing improvements in their operational efficiency but actually seeing a worsening.8 Figure 18. Firms that make unofficial payments perform worse If individual firms that opt for bribery are growing more slowly, there are several implications. First, it suggests that paying unofficially may not be a good business strategy after all, despite the fact that many myopic firms believe so. Helping firms to understand their legal options and alternative strategies for dealing with difficult situations could help ease the supply side of corruption. Second, it suggests that, on average, the culture of corruption in which firms readily resort to unofficial payments may be driving down growth of the business sector as a whole. Indeed, there could be a race to the bottom, in which firms that are trying to improve their competitive edge unwittingly, collectively, make the problem much worse as officials become accustomed to the practice of accepting bribes to solve problems. When looking at the average growth performance of individual provinces, the ones in which more firms pay bribes are, on average, the provinces where firms report that they are growing more slowly. (Figure 19). 8. These effects are remain strong when controlling for the size, ownership, sector of the enterprise, and for the province. When controlling for these variables the effect remains significant at the 10% level for whether the firm paid a bribe, and the 5% level for how frequently the firm uses gifts to solve its difficulties. 48 Part II - SURVEY FINDINGS Figure 19. Provinces with more bribery have firms that are doing worse 2.1.2.4. To whom do citizens make unofficial payments? The citizens survey also asked about experiences with unofficial payments. Citizens were first asked whether they used some selected public services, and if so, whether they needed to make unofficial payments. Follow up questions asked their assessment of how large those payments were and whether they were voluntary. As a whole, citizens report a large number of interactions with government bodies and agencies. Nearly three fourths had used health care services and more than one in three had interacted with tax officials to calculate or pay taxes, or applied for a loan from a bank or credit institution. (Figure 20). At the other end of the spectrum, very few respondents had any dealings at all with the courts. We saw earlier that people who consider themselves poor were more likely to perceive corruption at the local level, while those who consider themselves wealthy were more likely to perceive corruption at the national or province level. One reason for this difference in perceptions may be that the services used by poor people tend to be provided at the commune level, while wealthier citizens are more likely to use services at higher levels. Indeed, this seems to be the case. Among the 16 services about which citizens were asked to describe their experiences, there are two each that are clearly province, district 49 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS and commune level services, respectively. As is clear from Figure 21, citizens of all income levels make use of commune-level services, but wealthier citizens are more likely to make use of district- and province-level services. Figure 20. Percentage of citizens using different services in the previous 12 months Figure 21. Use of services by income category (%) 50 Part II - SURVEY FINDINGS The results confirm that for many services or functions, citizens have to pay unofficially in order to use those services or interact with those officials. (Figure 22). Among those using services or interacting with officials, the percentage of citizens paying unofficially is highest for the traffic police, to whom about 47% of citizens said they paid. More than 30% said they paid unofficially for school application and education services. When applying for a job in the state sector, 29% said they made an unofficial payment. Healthcare service and housing construction or repair permit were cited by around 25%. Some users reported paying unofficially even for services for which one would not normally expect to make unofficial payments, such as when the poor use social insurance and welfare and when citizens seek a birth certificate for newly born children. Figure 22. Probability of paying a bribe when using the service or dealing with the agency, among citizens with contact (%) Since many unofficial payments may be small, the survey of citizens also asked those who paid whether they consider the payments to be “largeâ€? or “smallâ€?. (Figure 23). Among those who have to pay unofficial money, large unofficial payments are given when applying for jobs in the state sector, and when applying for land use right certificates (LURHOC). When visiting or interacting with a state agency in search of a job, 12% said 51 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS they made a “largeâ€? unofficial payment. The other sectors with the highest probability of making a large unofficial payment are for schools and education, for dealing with the traffic police, and for obtaining a LURHOC. (These percentages are as a fraction of all of those who interacted with the agencies.) Figure 23. Probability of paying a large bribe when using the service or dealing with the agency, among citizens with contact (%) The data presented in Figure 22 and Figure 23 show the probability of paying a bribe and of paying a large bribe when dealing with state agencies. This type of comparison does not indicate the extent of unofficial payments in the economy as a whole, however, since some services, such as health care services, are very widely used by the population, while others are rarely if ever used, such as the courts. In Figure 24 we show the proportion of the entire sample of the citizen survey that said that they made an unofficial payment to each of the agencies. Thus, while the probability of making an unofficial payment is highest for the traffic police and for education, the total number of unofficial payments in the economy of the sampled provinces is highest for healthcare, since such a large fraction of the population visits health facilities. 52 Part II - SURVEY FINDINGS Figure 24. Proportion of the entire population that paid a bribe, according to citizens (%) 2.1.2.5. Why do citizens make unofficial payments? The survey of citizens probed respondents’ attitudes and experiences about why unofficial payments come to be paid. The citizens survey shows how citizens react when they face any hint of the need for an unofficial payment. Citizens that had some interaction with some agency or official in the previous 12 months were asked what happened the last time there was some suggestion for an unofficial payment or gift. The results, presented in the left panel of Figure 25, show that around half of the citizens said they paid the money. Only 3% said they reported to the authorities. Among those who said that they had paid the unofficial payment, shown in the right panel of Figure 25, it appears the payment was generally effective: 58% said the problem was fully solved, and 24% said the problem was partially solved. 53 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 25. Reactions and outcomes the last time an unofficial payment was suggested (% of citizens) The mechanisms through which unofficial payment are suggested are also of interest. When citizens who had made some unofficial payment in the previous 12 months were asked how officials usually suggest unofficial payments or gifts, around 18% said that the officials suggest directly, and 17% indicated the use of intermediaries. (Figure 26). Citizens were more likely to say that the officials intentionally created difficulties or delayed solving problems, cited by 29% of the citizens that had made an unofficial payments. Figure 26. How do officials suggest or request unofficial payments or gifts? (%, among those who paid at least once in the previous 12 months) 54 Part II - SURVEY FINDINGS The experience described above is not uniform across services or agencies. Citizens that reported having made an extra payment in the form of money, gifts, or favors, in addition to the official fees, often said they did so voluntarily. (Figure 27). This is particularly true for services. Many who made such payments for water connection or repair (88%), taxation services (83%), and healthcare (76%) said the payments were voluntary. In contrast, 50% of citizens that made unofficial payments for LURHOCs, and more than 41% of citizens that made such payments for housing construction or repair permits and when seeking a job from the state said that the payment was suggested, rather than being voluntary. Overall, 21% of citizens said they had made at least one voluntary unofficial payment. There could be many reasons that citizens choose to make unofficial payments without being directly asked to do so. The citizens survey asked those who had paid without being asked why they did so. (Figure 28). Among those who had made at least one voluntary payment, many said that it was either because many people do the same (41%), or that it was only a gift to express appreciation for the official (38%), or it was better than coping with complex procedures (32%). Only 17% of those who made at least one voluntary unofficial payment said that “without paying things could not get doneâ€?. Figure 27. When citizens pay unofficial payments or gifts, is it voluntary or suggested? (%, among those who had given an unofficial payment or gift in the previous 12 months) 55 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 28. Why citizens pay unofficial payments even when not asked (%, among those who had made a “voluntaryâ€? unofficial payment) 2.1.2.6. What corruption do officials observe? The surveys of citizens and firms help to illuminate the forms of corruption at the interface between the state and society. As noted in the introduction, however, some forms of corruption do not fall in this interface at all. The survey of public officials asked if they knew clearly any of 12 corrupt behaviors in the previous 12 months, and the results are presented in Figure 29. Four behaviors that were clearly known by more than 20% of the officials were: the use of the agency’s facilities/equipment for personal purposes (26%), taking money/presents to take some action or decision that gives preferential treatment to the giver (25%), intentionally delaying making decisions when conducting duties in order to elicit bribes (22%), and offering high ranking people vacations, meals, or entertainment for personal benefits (20%). The next two most frequently cited behaviors were promoting incompetent people for personal gain (17%) and making phone calls or writing letters of intervention for personal gain (16%). It is instructive that many of the behaviors depicted in Figure 29 do not fall in the interface between state officials and the public, but are entirely internal to the system of public administration. This suggests that there is still much to do in the area of public administration reform, improving the management of human resources, and in public asset management. 56 Part II - SURVEY FINDINGS Figure 29. Behaviors encountered by officials at work in the previous 12 months (% of officials) 2.1.3. Information sources on corruption There are a number of sources that form the opinions of the respondents on the extent of corruption in different sectors. TV, newspapers and radio are the most popular sources of information with 93% of all respondents selecting them. Public opinions are the second 57 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS most popular source for all three respondent groups. On the contrary, reports and reviews from public agencies or localities where respondents reside seem to provide the least information on corruption. Only 5% of respondents from the citizens group, and 10% and 16% from the enterprises and public officials groups, respectively, form their opinions based on this source of information. Although almost two-thirds of enterprises and public officials consider the internet as a source of information, only 18% of citizens do. This suggests a low penetration of internet among citizens. And while close to half of enterprises witnessed corruption themselves, only 29% and 31% of citizens and public officials groups did so. Figure 30. Sources of information about corruption 2.1.4. Emerging challenge: Interest groups The discussion so far has been about the various forms of administrative corruption, meaning corruption surrounding the implementation of laws, rules, and decrees. The experience of other countries that are making the transition to greater reliance on markets, however, shows how the form of corruption can evolve. As the role of the state shifts more towards defining the rules for competition and exchange, rather than directing the economy directly, firms may increasingly seek to carve out selective advantages for themselves. Indeed, firms may organize themselves into interest groups for this purpose. At the same time, firms may organize themselves into groups for positive purposes, as well. Since an individual firm may have little incentive to undergo the costs of communicating its problems to policy makers when the benefits would go to a broader group of firms, it may be more efficient for firms to organize into business associations so they can collectively voice 58 Part II - SURVEY FINDINGS their concerns. In this way, the formation of interest groups can play a positive role in facilitating communication between the business community and policy makers. The survey of enterprises sought to shed light on how firms view interest groups, seeking views on the trends, and the positive and negative contributions of interest groups. (Figure 31). As a whole, the questions about interest groups elicited a much larger percentage of responses of “I don’t knowâ€? than most of the other questions, reflecting the fact that the concept of “interest groupsâ€? is new in Vietnam. Figure 31. Enterprise views on interest groups More influential Use connections to gain advantages Totally disagree Totally disagree 1% 1% Disagree Disagree I don't know 9% 16% I don't know 21% Neither 22% agree nor disagree 19% Totally Totally agree agree 6% 6% Neither agree nor Agree Agree disagree 44% 34% 21% Help communicate problems of Use corruption to achieve their doing business goals Totally disagree Totally disagree 2% 3% Disagree 13% I don't know I don't know Disagree 26% 31% 25% Neither agree nor disagree Totally 22% agree 4% Neither Totally agree nor agree disagree 3% Agree 22% Agree 33% 16% 59 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Firms in Vietnam reported both positive and negative aspects of interest groups. On the positive side, some 37% said that interest groups in their line of business help expose the problems of doing business to policy makers, compared to only 15% who disagreed. There were also slightly fewer firms that agreed that interest groups in their line of business use bribery to achieve their goals than those that disagreed. On the negative side, the fact that only 28% disagreed with that statement suggests some cause for concern. Even more worrying, when asked whether firms use connections to officials to gain advantages, 40% agreed while only 17% disagreed. Finally, when asked whether interest groups were becoming more influential, 50% of firms agreed, while only 10% disagreed. Considering the responses of sub-groups of enterprises provides insights into the shifting landscape of interest groups in Vietnam. In general, larger firms tend to indicate that interest groups are becoming more active, and larger firms also tend to emphasize the positive aspects of interest groups. Firms with state ownership tend to say that interest groups are becoming more active, and emphasize both the positive and the negative aspects of their formation. Members of business associations, likewise, are more likely than firms that are not in business associations to say that interest groups are becoming more active, that they facilitate positive communications, and that they use connections to public officials to gain advantages. Since many of these factors are related to each other, it is necessary to use more sophisticated analysis to identify which factors are most closely associated with different assessments of interest groups. Everything else being equal, the firms that are most likely to say that interest groups in their line of business use bribery to achieve their goals are firms with many competitors and firms in the construction industry. The survey of enterprises also asked firms to identify which groups of policy makers interest groups usually tried to influence. Restricting the sample to firms that said that interest groups in their lines of business use connections to get advantages or use bribery to achieve their goals, but do not help communicate problems, responses tended to be most negative about leaders of People’s Committees and most positive about members of the National Assembly. (Figure 32). Indeed, the same pattern emerges when restricting the sample to firms that emphasized only positive aspects, that interest groups help communication, but do not use connections or bribery to achieve their goals. This suggests that the patterns observed, with Leaders of People’s Committees being the ones most targeted by interest groups, do not necessarily reflect corruption per se, but rather the degree of power and discretion that they have over matters that are important for firms’ profitability. 60 Part II - SURVEY FINDINGS Figure 32. With whom do interest groups influence for negative purposes? (among firms that see only negative aspects of interest groups - they use bribery or connections, but do not help communicate problems) Figure 33. With whom do interest groups influence for positive purposes? (among firms that see only positive aspects of interest groups - they help communicate problems, but do not use bribery or connections) 2.2. Effectiveness of Anticorruption Measures The previous section outlined the state of corruption in Vietnam after six years of implementation of the Anticorruption Law. While the previous section focused on 61 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS corruption, this section focuses on anticorruption. How have the measures introduced in the Anticorruption Law of 2005 been working? Which ones are being implemented, and what additional measures would Vietnamese citizens, enterprises, and public officials support for taking the fight against corruption to the next level? 2.2.1. Progress and challenges in anticorruption implementation The survey sought to explore progress in anticorruption measures in the last decade. Officials were asked a number of questions on their knowledge and attitude toward anticorruption issues, the legal system, and organizations. This section presents these results. About 85% of surveyed officials believe that officials’ knowledge of corruption issues has improved. (Figure 34). However, 64% believe that some officials are willing to join hands with corrupt people, and 86% think that feeling reluctant to fight against corruption is common among officials. Thus, although knowledge of corruption issues has improved, the willingness to fight against it is not very clear in the minds of many officials. Figure 34. Officials’ knowledge and attitude toward anticorruption issues 100% 75% 50% 84.6 64.3 86.0 25% 0% 1. Officials' knowledge 2. Some officials 3. Feeling reluctant on corruption improved are willing to join to fight against corruption hands with corruption is common No opinion Disagree Agree 62 Part II - SURVEY FINDINGS Officials were asked for their opinions on anticorruption laws and enforcement. A majority believe that the laws are generally sufficient (62%). However, officials also raised their concerns on the quality of anticorruption laws with around 78% believing that the laws are too general, are just formalistic, and some are out of date. Officials were requested to give their opinions on the effectiveness of nine anticorruption measures in their organizations. The vast majority of officials believe that openness and transparency (90%), and the development and implementation of entitlement norms and standards (89%) are effective. Administrative reform and development and enforcement of the code of conduct came next with 86% and 83% of officials¸ respectively, believing in these measures’ effectiveness. Slightly more than half of public officials believe in the effectiveness of current income and asset declaration measures. Only 24% of officials thought regulations on returning gifts by public officials were effective. A comparison between local (commune, district, and province) and central officials showed that officials in central organizations are less positive than officials in local agencies regarding the effectiveness of seven of nine measures. The first two measures, openness & transparency and implementation of entitlement norms and standards, were similar across the two types of organizations. Officials were asked to give opinions on the current system of detection and treatment of corruption. Ninety two percent of officials agreed that fewer cases of corruption only meant that corruption has become more sophisticated and harder to detect. Officials were skeptical of enforcement vigor: 61% agreed that a locality with no prosecutions is a place that lacks will to fight against corruption, and 75% agreed that self-detection of corruption remains weak. Officials were quite critical of the treatment of corruption cases: For example, 91% thought that the lack of strict and timely handling of serious cases had caused cynicism in the public. A further 89% believed that weak whistle blower protection had made people afraid to report corruption cases. Respondents also clearly would like to see tougher sanctions: 81% of officials believed that many cases that were treated as economic and/or administrative violations should have been treated as criminal, and 85% believed that recovery of corrupted assets and reparations to the victim have been weak. When strict approaches to corruption are followed, officials see some benefits: 69% of officials believed that handling of corruption cases has had positive impacts on 63 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS deterrence and prevention of corruption. It is also clear, however, that there is little trust in heads of agencies to seriously tackle corruption: 69% of officials agreed that some heads of agencies treated corruption cases lightly out of concerns for organizational and personal reputation. Figure 35. Central and local officials’ assessments of anticorruption measures’ effectiveness Despite continuous efforts to fight corruption in the last decade, corruption is still pervasive. The survey explored why anticorruption measures have not been as effective as expected. All three groups of respondents were asked for their opinions on a number of factors that inhibit anticorruption implementation. The percentage of respondents in each group who agreed or strongly agreed to each factor is presented in Figure 37. The surveys show that more than 90% of people believe that corrupt people are not severely punished. 80% of all three groups said that not enough attention was paid to improving officials’ ethics, and 76-82% said that anticorruption activities lacked focus. 75% of officials and 85% of the other two groups agreed so. 64 Part II - SURVEY FINDINGS Figure 36. Officials’ opinions on current detection and treatment of corruption Citizens were more critical of officials’ attitudes and competencies than the other two groups: 80% of citizens believe there is a conspiracy between officials and corrupt people, 87% of them believe that some superiors cover up for their subordinates, and 76% think officials lack competence. These numbers were less than 75% in both officials and firms samples. In the organization of Vietnam’s anticorruption bodies, the Chairman of the Provincial People Committee also serves as Chairman of the Anticorruption Committee. The survey asked respondents whether they view this as appropriate. This arrangement was seen as inappropriate by 52% of firms, 59% of citizens, and 62% of officials. Although these numbers were less than for other factors, they were still higher than 50% and this arrangement should be reconsidered in future anticorruption strategies. 65 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 37. Perceptions of factors that inhibit anticorruption implementation Public officials were also asked how much they trust the effectiveness of corruption detection of various institutions. (Figure 38). Most officials said that they trust most institutions. The level of trust was highest for the CPV’s Monitoring Committee, followed by the media. It is notable that the level of distrust of the effectiveness of any of the organizations is very small. Figure 38. Officials’ trust in effectiveness in corruption detection 66 Part II - SURVEY FINDINGS 2.2.2. Causes of corruption The causes of corruption are many. All groups of respondents were asked to share their opinions on 16 commonly cited causes of corruption (Figure 39) that have been mentioned in reports of the government or discussed in the media. In general, between half and two-thirds of respondents from all three groups agree or strongly agree with these causes with the exception of the common perception that “those who have corrupt chances but don’t take them are unwiseâ€?. Less than one third of public officials and fewer than half of citizens and enterprises managers agree that this attitude causes corruption. “No serious sanctionsâ€? for corrupt officials is the cause that all three groups of respondents agree with most. Enterprise managers and citizens were more likely to blame habits than public officials. With regard to the attitudes that “public officials are used to receiving gifts or bribes from citizensâ€? and “citizens have the habit of giving bribes to public officials to get things doneâ€?, public officials tend to agree less with that than citizens and enterprises do. While only 57% of public officials agree or strongly agree to the statement that “it is common that public officials receive gifts of unofficial payments while performing dutiesâ€?, 83% and 76% of enterprises and citizens, respectively, agree or strongly agree with this statement. When it comes to “low salaryâ€? as a cause of corruption, 79% of public officials agree or strongly agree that this is a cause of corruption. Citizens, however, are less certain with only 58% agreeing to this cause-citizens are less forgiving of low salaries as a cause of corruption. It is also interesting to note that more public officials than citizens and enterprises agree with limited checks and balances for high ranking public officials or loopholes in regulations as causes of corruption. 67 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 39. Opinions on the causes of corruption (% agree or strongly agree) 2.2.3. Responses to corruption-risk situations A key goal of the anticorruption infrastructure is to build a system whereby people are willing and able to alert the authorities when identifying corruption. For such a system to work, there is a need for both widespread acceptance of the harmful consequences of corruption and confidence that providing such information to the authorities will not have negative repercussions. The Anticorruption Diagnostic Surveys attempted to identify the constraints limiting a more effective system of denunciations. 68 Part II - SURVEY FINDINGS In general, there seems to be strong support for the system whereby public officials could denounce corruption. When asked what a person would do if she/he was certain about a corrupt act in her/his office, 79% of public officials said they would definitely report to competent persons. Only 4% said they would not denounce. For citizens, about 43% said they would likely denounce to competent persons if they know of a corrupt public official. The same number of citizens said they would not be likely to do so. While 66% of citizens indicated that they would not be likely to notify newspapers, radios or television of a corrupt public official, 63% of citizens said they would be likely to confide with their friends, relatives and neighbors, and 59% of them would not be likely to keep silent. (Figure 40). Figure 40. Citizens’ response to a corrupt public official These findings show that although media is seen as an effective force for detecting corruption, as presented in Section 2.2.7 in this report, more needs to be done for the media to gain trust from citizens so that media and citizens could be allies in anticorruption efforts. The results also show that once a citizen knows about a corrupt public official, most of their relatives, friends and community would also know about it. The population, therefore, is apt to know a great deal about corruption, and it is important, therefore, to count on citizens in anticorruption efforts. 69 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Why might citizens, public officials and enterprises be hesitant to denounce corruption? All three groups of respondents were asked to give their views on the reasons that they do not denounce corruption. Citizens are more likely to think that corrupt persons are related to competent persons with almost two-thirds of citizens agreeing and strongly agreeing so, whereas the figures are 55% and 49% for enterprises and public officials, respectively. While more than half of the respondents were afraid of being retaliated or had no trust in the competent persons, only a little more than 20% of citizens and enterprises, and 30% of public officials, agree or strongly agree that the reason that they do not denounce is because there would be no rewards. The lack of rewards is clearly not a strong reason for citizens, public officials and enterprises not to denounce corruption. Factors such as trust in the competent persons and retaliation have much stronger deterrent effects. Figure 41. Reasons for not denouncing corruption (% agreeing or strongly agreeing) 2.2.4. How do firms organize their anticorruption activities? We saw in Section 2.1.2.2 that firms often say they make unofficial payments to speed processes along. Some firms, however, may also take a proactive role in trying to reduce corruption. Firms were asked if they implemented any anticorruption measures in the previous 12 months. Fifty two percent of firms did implement some anticorruption measures in the last 12 months, while 43% of them did not. Among those who implemented anticorruption measures, 74% said they had codes of conduct, 50% said the firms had a campaign “say no to corruptionâ€?, and 37% of the firms organized events to improve knowledge of anticorruption. These three measures were 70 Part II - SURVEY FINDINGS awareness raising activities. Measures to promote anticorruption activities, such as review and rotation of corruption-risk positions and participation in anticorruption initiatives got fewer responses, with only 25% and 12%, respectively. As a share of all firms, of course, these numbers were smaller, ranging from 45% of firms with codes of business conduct to 7% of firms participating in anticorruption initiatives. Figure 42. Anticorruption activities implemented by firms 2.2.5. Tenders One area at the center of the business-state interface that offers the potential for corruption is in the procurement by state bodies of goods and services. The surveys asked respondents about their perceptions and experiences in tenders. Twenty eight percent of officials and 17% of firms said their organizations participated and/or organized tendering activities in the previous 12 months. Among officials whose organizations had tendering activities, 19% believed the tendering was only formalistic, 11% thought there was some coalition or “back doorâ€? lobby to win the tender(s), and 14% thought the tenders were not transparent. (Figure 43). These numbers were much more negative from firm respondents, with 26% believing the tenders were just formalistic, 27% reporting there was some coalition and “back doorâ€? lobbying to win the tenders, and about 21% saying the tenders were not transparent. (Figure 44). It is 71 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS instructive that among those who said there was a “backdoorâ€? and collusion to win the tender, 32% of public officials and 28% of enterprises also said that the process was transparent and clear. Improving checks and balances in procurement processes clearly is only one part of the reform program needed. It is also essential to strengthen controls over the people organizing the tenders - recognizing the special position of public trust of people in these positions, greater attention to lifestyle checks would also be warranted. Figure 43. Officials’ assessment of the tendering activities in their organizations Figure 44. Firms’ assessment of the tendering activities in which they participated 72 Part II - SURVEY FINDINGS 2.2.6. Recruitment and promotion of public officials Officials were asked how important several criteria were for recruitment and promotion in their organizations. Figure 45 and Figure 46 illustrate the results. While an overwhelming percentage of respondents believe that ability and achievement are the two key criteria for recruitment and promotion, there are still sizable numbers who believe that connections with powerful people and with family/friends are important. About 18 - 19% of respondents believe that connections with powerful people helped in recruitment and promotion; similarly 13 - 16% of officials believe that connections with family/friends are important in promotion and recruitment. Although few officials said that money and gifts were important for recruitment, we saw in Section 2.1.2.4 that many citizens reported making unofficial payments in order to apply for a job with a state organization. One may wonder why citizens would pay bribes to get jobs in the public sector despite the oft-heard lament of low salaries in the public sector. Flourishing opportunities or other incentives to recover their “investmentâ€? through commissions or “envelopesâ€? in the public sector might explain this seeming contradiction. Continued efforts to not only strengthen the merit- orientation of Vietnam’s civil service to reduce the prevalence of paying for positions, but also to minimize rent-seeking opportunities in the sector, must be important parts of Vietnam’s anticorruption agenda going forward. Figure 45. Importance of different factors in recruitment of officials 73 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 46. Importance of different factors in promotion of officials 2.2.7. Roles of the media The media has long been recognized as essential for the fight against corruption. The Anticorruption Law of 2005 devotes Article 86 to the role of the press, calling for a proactive role in reporting on corrupt cases and promoting the public’s understanding and awareness of anticorruption, and also calling for professionalism in reporting objectively and honestly. The respondents to the surveys of enterprises and public officials confirm strong agreement that the media has been successful in uncovering corrupt cases and for keeping pressure to pursue cases that might otherwise have died out. (Figure 47). More than 80% of both sample groups agree that the media successfully uncovers cases and then keeps them alive, while fewer than 10% disagree. At the same time, respondents to the surveys tended to agree, albeit to a lesser extent, that the media can sometimes exaggerate, and may not be playing as strong a role as they could in disseminating positive information and focusing on the sectors and places that are doing well. 74 Part II - SURVEY FINDINGS Figure 47. Role of the media in helping to fight corruption (% agree or strongly agree) Officials at the central level agencies were significantly less likely to be positive about the media, although positive nevertheless: While 84% of commune level officials believe that the media uncovers many corrupt cases, only 77% of officials at the level of ministries or central level agencies believe so. (Figure 48). The pattern is reversed when asked whether the media sometimes exaggerates: 83% of officials at the ministerial or central level said the media sometimes exaggerates, while only around 70% of officials at the commune and district levels said so. Figure 48. Who is more positive about the media? 75 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS While the surveys show that most respondents from both the public officials and the enterprise surveys see both positive and negative contributions for the media, it is important to note that the positive assessments outweigh the negative assessments by a considerable extent. Moreover, when asked about the sectors or fields that have corruption themselves, fewer than one percent of all respondents identified the media as among the most corrupt (Figure 8), and some 93% of all respondents say they get their understanding of corruption from the media (Figure 30). On the whole, respondents see the media as a potentially important ally in the fight against corruption. 2.2.8. Necessity of anticorruption measures When designing anticorruption reforms, it is useful to get an idea of which sorts of reforms would have support and which would have strong opposition. The surveys asked the respondents whether they feel a set of reforms is necessary or not, and the results are presented in Figure 49. The widespread prevalence and seriousness of corruption has created such concern to society that people seem willing to do anything to reduce corruption. For every measure proposed, at least two-thirds of respondents said that the reform was necessary. The responses on necessity of anticorruption measures are highly consistent across sample groups. Citizens support any measures that might help to reduce corruption. Indeed, supportive responses of citizens are higher than those of public officials and enterprises in most cases. Among public officials, support for anticorruption solutions is high, but their support for “declarations of assets and incomeâ€? is lowest among the three sample groups. It should be noted that despite the differences, more than two-thirds of public officials agree with publicizing assets declarations. At the same time, public officials support a solution which allows undetected bribe-takers who return money/gifts to avoid being judged, more so than enterprises and citizens. Finally, the very strong consensus in holding votes of confidence for leaders shows the importance of leadership in generating trust among staff and the public, an especially important element of efforts to combat systemic corruption. 76 Part II - SURVEY FINDINGS Figure 49. Consensus on the necessity of anticorruption measures among public officials, enterprises and citizens (% saying needed) 2.2.9. The real evidence: what is working, and what is not working? Although there is no simple recipe for reducing corruption, there are some measures that are suggested by the survey results. In trying to identify workable recommendations, the survey results can help in two ways: by identifying what seems to be working so that those approaches may be scaled up, and in identifying what does not seem to be working. Approaches that are not working should either be fixed or de-emphasized. In some cases, the surveys may also help contribute to our understanding of why they are not working and how they might be fixed. In order to identify what is working and what is not working, we took advantage of a key strength of the approach of the Anticorruption Diagnostic Surveys. Since the surveys provide the perspective of both public officials, who know about the institutional environment in which they operate, and citizens and firms, who can report on their actual experiences with unofficial payments, the combination of the three surveys can help identify which aspects of the institutional environment in the public sector, and which of the many anticorruption measures being implemented, are associated with lower levels of corruption. 77 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS To carry out this analysis, we calculated measures of how well public officials say various measures are being implemented and averaged them at the province and the district level. These 13 sets of assessments - covering the likes of the degree of meritocracy in the civil service, the degree to which assets declarations are thought to be effective, and the levels of salaries - can be thought of as the “inputsâ€? for reducing corruption. We also calculated measures of the proportion of citizens and firms in each province and district that said they paid a bribe in the previous 12 months, as well as citizens’ perceptions of the level of corruption at the province and district levels. As noted earlier, it is important to understand both actual experiences and perceptions of corruption, since it is perceptions on which people and firms make decisions. These three measures of corruption - two based on actual experiences and one on perceptions - can be thought of as the “outputsâ€? of our corruption equation. Matching up data from different surveys in this way provides an especially powerful tool for identifying the relationship between public sector institutions and levels of corruption, more so than trying to use data from a single survey. Since there are only ten provinces, it would not be surprising if the data showed little correlation between these institutional factors and the levels of corruption. For this reason, we also computed all of the same variables at the district level. With three districts per province, this provides a total of 30 districts, facilitating a somewhat stronger set of data on which to conduct this analysis. Considering that corruption has many causes, however, even 30 observations is not a large number and it would not be surprising to find little correlation. However, this was generally not the case. The results of the analysis show that many of the places that are implementing anticorruption and institutional measures (according to public officials) actually do have lower levels of corruption (according to citizens and firms). While it is possible that a variety of other factors are important for explaining levels of corruption and the institutional environment simultaneously, this analysis based on simple correlations at least provides a prima facie case for whether anticorruption or institutional measures are associated with lower levels of corruption. The results can be grouped broadly into two categories: measures which are associated with lower levels of corruption and seem to be working, and those which are not, or only very weakly, statistically associated with lower levels of corruption and do not seem to be working. Table 3 summarizes the results. (Annex 2 provides additional details on levels of significance and scatter plots for all of the relationships.) 78 Part II - SURVEY FINDINGS Table 3. Institutional factors associated with lower levels of corruption The measures that are most clearly associated with lower levels of corruption are: (i) Entitlements, norms and standards. For both province and district levels, and for both citizens’ and firms’ reports of corruption, places where officials assess the “development and implementation of entitlements, norms and standardsâ€? as 79 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS being effective for reducing corruption in the workplace actually do have lower levels of corruption. This is a clear example of how taking firm control of anticorruption policies within an organization can yield positive results. (ii) Openness and transparency. For both province and district levels, and for both citizens’ and firms’ reports of corruption, places where officials assess the “openness and transparency of activities in the agency/unitâ€? as being effective for reducing corruption in the workplace actually do have lower levels of corruption. This is another example of how better implementation of transparency policies and scaling these up can make a difference. (iii) Administrative reform. For both province and district levels, and most strongly for citizens’ perceptions of corruption, places where officials assess administrative reform as being effective for reducing corruption in the workplace tend to be places with lower levels of corruption. (iv) Transfers of public officials. The policy of “transfers of public officials to other working positionsâ€? (rotation) is associated with lower perceptions of corruption among citizens, and at the district level with less bribery, as well. (v) Merit-oriented personnel policies. At the province level, the degree to which recruitment and promotion are based on the candidate’s capacity and achievements, and not on the support of people of high rank or personal relationships, is associated with lower levels of corruption, according to both firms and citizens. Interestingly, this effect only holds at the province level, not the district level. This may be because the scope for introducing competition is larger in the provinces since the labor markets are larger in provinces than in districts. (vi) Willingness to denounce. Provinces where officials are more likely to say that if they knew of a case of corruption in their work place that they would “surely report to the relevant person in chargeâ€? also tended to have lower levels of corruption. Places where people feel confident to denounce corruption are also those with less corruption. The analysis also suggests several measures that are not working very well, including the system of assets and income declarations, the policy of holding the heads of agencies responsible for corruption within the agencies, the use of codes of conduct and professional ethics, and the payment of salaries via bank accounts. These measures, as currently implemented, are only very weakly associated with lower levels of corruption. Such measures should either be de-emphasized, or revisited to attempt to make them work 80 Part II - SURVEY FINDINGS better. Two other factors were not significant for explaining levels of corruption at either the province or the district level, regardless of the measure of corruption. Regulations on returning gifts appears to have little impact, and the levels of salaries also do not help explain levels of corruption. 2.3. Similarities and differences with the 2005 Anticorruption Diagnostic Surveys As the 2005 Anticorruption Diagnostic Surveys were conducted prior to the adoption of the Anticorruption Law, those surveys provided the baseline for developing the questionnaires for the 2012 surveys. Some questions were preserved affording some possibility of comparing results between the two rounds of the surveys. In order to make the comparisons as robust as possible, we focused attention on the same seven provinces and three ministries that were the subject of the 2005 survey. In this section we briefly summarize the main similarities and difference in findings. In both 2005 and 2012 surveys, respondents were asked to assess their perceptions of the most corrupt sectors. There was a very high consensus among the three sample groups on their perceptions of which agencies have the most prevalent corruption. Four of the five sectors identified by respondents as those with the most common corruption in the 2005 survey remained among the top five in the 2012 survey. Public finance & tax, which was fourth in 2005, fell out of the top five in 2012. Transport was cited by citizens and public officials as in the top five in 2012, and minerals management was cited by enterprises as in the top five in 2012, neither of which was cited directly in the 2005 survey. While we may observe notable improvements in certain sectors, such as customs, the magnitude of these improvements still fall short of people’s expectations and customs remains in the top five in people’s perceptions9. In both the 2005 and 2012 surveys, public officials were asked if they had encountered any of the 12 corruption behaviors in the previous 12 months. The results for 2012 were presented earlier in Figure 29. As this question was drawn from the 2005 questionnaire, it is possible to compare the results with those from 2005. (Figure 50). A notable positive result of this comparison is that for each and every corrupt behavior, a smaller fraction of officials reported observing the behavior in 2012 than in 2005. The patterns of behavior 9. For example, the World Bank Enterprise Surveys show a decline in the prevalence of unofficial payments between 2005 and 2009, and the World Bank’s Logistics Performance Index 2012 similarly showed some improvement with 62% of firms surveyed saying solicitation of unofficial payments had improved since 2009. 81 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS were quite similar in the two surveys: the ranking of behaviors by their frequency is nearly identical10. Table 4. The most corrupt organizations in 2005 and 2012 surveys The 2005 and 2012 surveys of citizen also had several identical questions, allowing us to make comparisons. One question centered on four possible corruption tricks that citizens face in interacting with officials. The results (Figure 51) suggest that citizens’ encounters with these tricks tended to decline between 2005 and 2012, with smaller percentages of citizens facing three out of four of these tricks, and about the same percentage for the other trick. For the survey of enterprises, comparisons are possible for several identical questions. As the 2005 survey only included state-owned firms in the sample, we split the sample of firms for the 2012 surveys into two groups, those with and without state capital, to facilitate the comparison. Again, only seven provinces were included in the analysis in order to make the comparisons as robust as possible. Firms were asked if they faced any of six corruption tricks by state agencies, and the results for both 2005 and 2012 surveys are presented in Figure 52. For five of the six tricks, the proportion of firms facing the tricks was higher in the 2012 survey than in 2005 survey. The only trick that fewer firms faced in the 2012 survey was that officials “subjectively suggested to firms what to doâ€?. Thus, while the comparisons of results from the citizens 10. Since Figure 50 is based only on the same seven provinces and three ministries in the 2005 survey, the numbers for Figure 50 are not identical to those of Figure 29 which was based on all ten provinces and five ministries. 82 Part II - SURVEY FINDINGS and public officials surveys suggested some cause for optimism, the comparisons for the surveys of enterprises does not suggest improvement. Figure 50. Which behaviors were observed by officials in 2005 and 2012? (% observing in previous 12 months) Note: The data for 2012 is restricted to the same seven provinces that were in the 2005 sample. For both years, the bars show the percentage of all respondents, including those who answered “don’t knowâ€?, who said they observed the behavior. 83 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 51. Corruption tricks observed by citizens in 2005 and 2012 (%) Note: This chart depicts the share of all respondents who reported these approaches, since that is the way the numbers were calculated for 2005. The results for 2012, therefore, are not directly comparable with those presented in Figure 26, which focuses on the subset of respondents which had made unofficial payments. Figure 52. Corruption tricks with firms in 2005 and 2012 Note: This chart depicts the share of all respondents who reported these issues, since that is the way the numbers were calculated for 2005. 84 Part II - SURVEY FINDINGS Firms were then asked what they did when they faced corruption tricks by state agencies/officials. A comparison between 2005 and 2012 surveys showed that firms engaged less in most types of responses. They are less likely to ask for interventions by powerful people or the judiciary or the media. However, they are slightly more likely to “directly give a bribe to officials in chargeâ€?. This trend of behavior is consistent with other survey findings that suggests that more effort needs to be placed on getting firms to serve as allies in the fight against corruption. Figure 53. Responses by firms to corruption tricks, 2005 and 2012 Note: Sample is restricted to the same seven provinces as in 2005 and to SOEs as in 2005. Firms were asked why they pay bribes. In 2005, firms were more likely to select many of the possible reasons than in 2012. The only exception is that a similar percentage of firms in both years selected “other firms did the sameâ€?. (Figure 54). 85 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 54. Why do firms pay bribes, 2005 and 2012 (%) As a whole, the comparisons between the 2005 and 2012 surveys provides a mixed picture of progress in the fight against corruption. There is considerable consistency in the sectors identified as those with the most prevalent corruption. Four of the five sectors were in both surveys, with some changes in the order: traffic police, land administration, customs, and construction. While improvements may have been made in these sectors over the years, the magnitude was not sufficient to change the public perception of them as the sectors with the most frequent corruption11. The assessments of citizens and public officials showed some positive signs. Compared to the 2005 survey, a much smaller percentage of public officials in 2012 witnessed the 12 corrupt behaviors. Similarly, a lower percentage of citizens reported facing corruption tricks, compared to 2005. By contrast, the firm sample showed a more negative picture of the trend between the two surveys. More firms in 2012 faced corruption tricks than did in 2005, and a larger share of firms prefer giving bribes as a means of overcoming these tricks. Asking powerful persons to intervene or confronting officials by making arguments, asking the media or judicial agencies to assist are all used less by firms in 2012 than in 2005, with direct bribery taking up the difference. In both 11. See, for example, Footnote 10, which describes evidence of improvements in customs. 86 Part II - SURVEY FINDINGS surveys, about one of four firms voluntarily gave bribes because “other firms did the sameâ€?. This points to the fact that collective action needs to be taken in a more positive direction, rather than just letting firms learn corrupt behaviors from each other. 2.4. Similarities and differences with other surveys The surveys of citizens and firms that comprise these Anticorruption Diagnostics are among a growing set of surveys on perceptions and experiences of corruption. Qualitatively, there are several consistencies with other surveys: The importance of bribery for state employment was identified in PAPI 2011, and the frequency of paying informally for healthcare was pointed out in both VDR 2010, and PAPI. The fact that many firms bribe willingly for convenience was pointed out in the VCCI and DEPOCEN study of business corruption. The sector-specific findings for education and land are broadly consistent with the studies by GIRI and T&C Consulting, and the findings on health and the traffic police mirror the findings of the Towards Transparency studies, including the Global Corruption Barometer. The finding of some improvement in some forms of administrative corruption is similar to those of the Provincial Competitiveness Index 2012, the World Bank Enterprise Surveys (2005-2009), and the VDR 2010. In order to ascertain in a more systematic way how the results from the Anticorruption Diagnostics are similar to or different from other surveys, we sought to identify the questions which are broadly similar on other national surveys. The best tools for such an exercise are the Vietnam Provincial Governance and Public Administration Performance Index (PAPI), a survey of citizens conducted by UNDP, CECODES and the VFF, and the Provincial Competitiveness Index (PCI), a survey of firms conducted by the VNCI and VCCI, both of which cover all 63 provinces in Vietnam. All ten of the provinces that were covered by the Anticorruption Diagnostics were also included in the PAPI and the PCI. 2.4.1. Citizen’s perceptions and experiences compared to PAPI The PAPI questionnaire has questions pertaining to five different sorts of citizen-state interactions that are also covered in the Anticorruption Diagnostics: healthcare, education, land, construction, and getting jobs from the state. The questions are not identical and one would expect there to be some differences due to different focus of the questions, different sampling, etc. Nevertheless, the results found in PAPI and those in the Anticorruption Diagnostics are strikingly similar. The approach to questioning used in PAPI bears elements of both perceptions and experiences. While most of the questions are about interactions in general, rather than 87 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS reporting on actual experiences, the focus on the types of interactions that citizens experience and the use of terms such as “people like meâ€? draw the respondent’s attention to their actual experiences. At the same time, since the questions do not refer to actual experiences and are asked of all respondents, even those who did not have such interactions, the PAPI questions have elements of perceptions, as well. For this reason, we examined the correlation with both the Anticorruption Diagnostics “perceptionsâ€? questions, and those that focus on actual experiences of those who dealt with state agencies. Table 5 summarizes the results. Table 5. PAPI and the Anticorruption Diagnostics 88 Part II - SURVEY FINDINGS At the province level, there are only ten observations and so it would not be unusual to find little statistical correlation, even if there really is a correlation in the population. Nevertheless, for three types of interactions the correlations between PAPI and the Anticorruption Diagnostics are high. For both health and education, the degree of correlation between PAPI and the Anticorruption Diagnostics questions on experiences with corruption are strikingly high, and for seeking jobs the degree of correlation is also very high. Figure 55 depicts scatterplots of the province-level assessments from both the Anticorruption Diagnostic Surveys and the PAPI. Figure 55. PAPI and the Anticorruption Diagnostics Note: As the scales of the questions differed, the numbers on the axes are not relevant. Trend lines are shown for relationships that are significant at the 10% level or higher. 89 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS For questions related to land and construction, there is no statistically significant correlation of province-level assessments. There are several possible reasons that health, education and jobs would exhibit strong correlation, while questions related to land and construction would not. The first is that the province-level estimates of experience with unofficial payments for land and construction provided by the Anticorruption Diagnostics are based on a relatively small number of observations. Since the question focuses on actual experience, the question on LURHOCs, for example, was only posed to the subsample of respondents that had sought an LURHOC in the previous 12 months. Although the number of households in the entire sample that had sought an LURHOC and had answered the question on unofficial payments was sizeable (386), the number per province was in some cases not large. Three of the provinces had fewer than 30 observations each. This difference was even larger in the case of construction permits: only two provinces had more than 30 households that had sought a construction permit in the previous 12 months. In contrast, the estimates for health are much more precise: every province had at least 120 responses. It should be noted however, that the numbers of observations for each province for education and job-seeking are also not large, so this can only be part of the explanation. A second reason for the difference is that the questions and approach used in PAPI and the Anticorruption Diagnostics differ in important ways. While both PAPI and the Anticorruption Diagnostics query respondents about unofficial payments for land titles or LURHOCs, PAPI asked this question of all respondents. As such, it is possible that the questions on bribing for land titles brought in an element of perceptions about larger scale corruption related to land allocation. Indeed, the correlation with the Anticorruption Diagnostic question on perceptions of corruption in land management is positive and significant at the 15% level, even though the question on experiences with corruption is not. As a whole, the two surveys are very highly correlated. With ten provinces and five sectors, there are 50 sector-provinces for which to compare results. The correlation of 0.62 is highly significant. Considering that there are only ten provinces, the degree of correlation between PAPI and the Anticorruption Diagnostics is striking, giving some confidence that differences in sampling approaches and methodology between the two surveys are not likely to substantively drive most results. At the same time, the differences in results serve as a reminder to use care in interpreting the results. More importantly, the differences in questions and approaches suggests that the two surveys can both usefully contribute to our understanding of corruption in Vietnam. 2.4.2. Enterprise manager’s perceptions and experiences The survey of enterprises undertaken as part of the Anticorruption Diagnostic Surveys is not the first attempt to gauge manager’s views on issues of corruption. The most well 90 Part II - SURVEY FINDINGS known such survey is the one that forms the basis for the Provincial Competitiveness Index, put together by the VNCI and VCCI. This survey, which has been carried out annually in some form since 2005, includes several questions that are broadly similar to some questions in the enterprise survey carried out for the Anticorruption Diagnostic. It should be noted, however, that while the survey methodology employed by the Anticorruption Diagnostic is similar to that used by PAPI, there is no such similarity for the PCI. Whereas the Anticorruption Diagnostic used face-to-face interviews, the PCI employs a mail-in survey. Among firms that were contacted for the enterprise survey as part of the Anticorruption Diagnostic Surveys, 80% ultimately participated in the survey, whereas for the PCI the response rate is around 35%. Despite these differences in methodology and response rate, exploring the similarities and differences in results between the two surveys could help put the results from the Anticorruption Diagnostic Surveys in context. Table 6 shows the three questions that are most similar between the PCI and the Anticorruption Diagnostic Surveys, and the degree of correlation in the province-level responses. Among the three questions that are similar, two show statistically significant correlations between the province-level assessments in the PCI and the Anticorruption Diagnostic Surveys: the extent to which enterprises are subject to bribe requests, and the degree to which bribes achieve their desired results. Even with only ten provinces on which to base comparisons, the patterns suggested by the Anticorruption Diagnostic Surveys and the PCI survey are similar. Table 6. PCI and the Anticorruption Diagnostics 91 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 56 shows graphically the relationship between province-level indicators from the Anticorruption Diagnostic Surveys and those of the PCI. The high degree of correlation for two of the three indicators is clear. As with the PAPI, the correlations suggest confidence in the broad similarities between the two surveys, while at the same time making clear the differences between the two that make them complements, rather than substitutes, for each other. Figure 56. PCI and the Anticorruption Diagnostics Note: As the scales of the questions differed, the numbers on the axes are not relevant. Trend lines are shown for relationships that are significant at the 10% level or higher. 92 Part III Conclusions and Recommendations C orruption remains a serious problem in Vietnam. Although respondents identified other problems as even more serious, a large proportion of the population is concerned with corruption, with at least a third of the population identifying corruption as among the most serious problems facing Vietnam. Large proportions of the population experience corruption first-hand in the form of unofficial payments: in the 12 months before the surveys 44% of enterprises and 28% of citizens reported direct experience with paying unofficial payments and 45% of public officials encountered corrupt behavior. The fact that all sample groups seem keen to try just about anything to reduce corruption signals the need to reinvigorate the battle against corruption. Although there does seem to be some progress in reducing low level administrative corruption, there is general agreement that fewer corruption cases only means that corruption is becoming more complex, not that corruption is declining. The recommendations described below draw directly from the findings of the Anticorruption Diagnostic Surveys, both for identification of the problems, and for analysis of the approaches that are working and those that need to be fixed. In the presentation that follows, we start with the recommendations related to policy development, those that require legal or regulatory changes. Next we turn to policy implementation, highlighting key areas where implementation of existing policies can be improved. Monitoring progress in reducing corruption comes next, followed by the need to enhance the public’s knowledge so that all of society can work towards to the same goal. 3.1. Policy development Generate real access to information The analysis in Section 2.2.9 showed clearly that one set of existing policies that is succeeding in reducing corruption is openness and transparency. This pillar of the Anticorruption Law of 2005, the Ordinance on Grassroots Democracy and many other laws 93 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS does have a measurable impact on reducing corruption. Provinces and districts in which policies on transparency and openness are implemented more vigorously do in fact have lower levels of corruption. Although many laws call for access to information, the lack of legislation outlining specific responsibilities and establishing a monitoring and enforcement system leads to very uneven implementation of transparency provisions. Earlier studies12 have noted that although Vietnam has more than 30 laws, rules, and regulations calling for access to information, these are often not implemented seriously. And even with rather progressive clauses on openness and transparency in the Anticorruption Law, the Law only provides for the citizens to request information on the operations and activities of organizations in the place where the citizen works or at the commune People’s Committee where she/he resides. Citizens’ ability to request information on operations and activities of People’s Committee Offices from district level above is not guaranteed in the same way. Improving access to information calls for better implementation of existing laws, but without legislative change, such improvements in implementation are not likely to come to pass. The vast majority of all three sample groups (87%) said that it is necessary to issue the Law on Access to Information. Passing the Law on Access to Information would be a major step forward for Vietnam’s anticorruption battle. A draft of a Law on Access to Information was produced by the Ministry of Justice in 2009, but has yet to be submitted to the National Assembly. Passing such a Law would bring several improvements over the current system. First, it would give the opportunity to clarify responsibilities for providing information. Most of the laws in Vietnam regulating the provision of information are vague on the question of who is responsible for providing the information when requested, although the law is clearer with respect to the commune’s obligation to provide information on request than for other state bodies or levels of government. Second, a Law on Access to Information can establish an enforcement body or system to ensure that the law is implemented, and to provide some recourse when it is not. One approach taken by other countries (such as Thailand and Mexico) is to establish an information board or some other autonomous body whose mission is to support compliance with the law13. Such an approach could bring Vietnam’s institutions into the 12. Vietnam Development Report 2010 - Modern Institutions by the World Bank and others. 13. This section draws on Vietnam Development Report 2010 - Modern Institutions by the World Bank and others, and on “Monitoring implementation of access to information in international experience,â€? by Jairo Acuna-Alfaro. 94 Part III - CONCLUSIONS AND RECOMMENDATIONS ranks of middle income countries. Middle income countries are more than twice as likely as low income countries to have an enforcement body to ensure access to information, and a third more likely to specifically outline contact points for requesting information14. Finally, Vietnam would take a huge step toward modernity by introducing the presumption that all information that is not specifically prohibited, for example on national security grounds, should be considered public. Debating and passing the Law on Access to Information would provide Vietnam the opportunity to modernize its institutions along the same lines. Although the Law on Access to Information would open the possibility for major changes in the system of enforcement and in the presumption that information is public unless designated otherwise, the planned revisions to the Anticorruption Law also afford an opportunity to correct other weaknesses that have impeded implementation of the many good transparency provisions in the Law. Although many articles declare that different documents are public information, Article 31, covering the rights of organizations to request information, affords special status to state agencies, political organizations, socio- political and press agencies, but does not mention enterprises or other organizations. More importantly, Article 32, covering the rights of individuals to access information, affords special status to cadres and civil servants, while citizens are limited in their rights in terms of who they can submit requests to (“Chairman of People’s Committees in the communes, wards, or townships where they resideâ€?), and in terms of content (“about the operations or activities of those Committeesâ€?.) Restricting the right to request information in this way weakens the other positive articles in the Law - revising the Anticorruption Law provides an opportunity to correct this weakness. Passing the Law on Access to Information would be even stronger as it open the possibility for strengthening enforcement of transparency provisions. Fix the system of land management to reduce corruption Land was identified in the surveys as a sector with significant corruption problems, a finding that mirrors media reports and earlier studies15. Together with the traffic police, land management was identified by the largest fraction of the population as among the areas with the most serious corruption. It is clear from the surveys that when people think 14. This is according to the World Bank’s Public Accountability Mechanisms database. https://agidata.org/pam/ 15. For example, Corruption Risks in the Issuance and Transfers of Land Use Rights/ House Ownership in Vietnam, conducted by Government Inspectorate Research Institute (GIRI) and T&C Consulting, 2010. 95 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS of corruption in land management, they are not thinking necessarily of the unofficial payments that are made for land use certificates, but rather of corruption surrounding the way that land is allocated, acquired, and managed. Revising the Land Law, scheduled to go before the National Assembly in 2012 (for discussion) and 2013 (for voting), with careful attention to the opportunities that the land management system generates for corruption, could help reduce corruption and conflicts over land more generally. Earlier research16 has already highlighted that throughout the processes of land use rights certificate issuance, land acquisition and land allocation, compensation, and resettlement, the entire system opens many opportunities for, and increases the profitability of, corruption. Many decisions are made at the discretion of the district and province authorities, without commensurate transparency and accountability to control that discretion. In revising the Land Law, several changes would help to reduce the prevalence of corruption in land management. Foremost among these is to reduce the use of compulsory land acquisition for projects that are essentially private. In other words, when a project is commercial or economic in nature, agreement between the land user and the investor would have to be reached voluntarily, rather than using the discretionary power of the local authorities to compel the land user to give up the land as is the case now. By restricting the use of compulsory land acquisition to cases meeting the “national interest or public benefitâ€? criterion, this would not only reduce the chances for corruption, but reduce the prevalence of conflicts over land and improve economic efficiency, as well. This is consistent with the experience of many other countries, some of which (such as Peru) even have provisions in their constitutions requiring that land can only be taken for projects that serve the public, not for private economic projects. Another form of corruption surrounds the mechanisms for estimating “market pricesâ€? for cases where land needs to be acquired for public projects, such as the building of a road. Firmly establishing the conditions for independent land appraisal will also reduce conflicts and opportunities for corruption. 16. This section draws on Recognizing and Reducing Corruption Risks in Land Management (2011) by the Embassy of Denmark, the World Bank, and the Embassy of Sweden; Compulsory Land Acquisition and Voluntary Land Conversion in Vietnam (2011) by the World Bank; Vietnam Development Report 2010-Modern Institutions by the World Bank and others; Vietnam Development Report 2011-Natural Resources Management by the World Bank and others; World Bank - Food and Agriculture Organization Land Governance Assessment Framework (2010); GIRI and T&C Consulting (2010), Corruption Risks in the Issuance and Transfers of Land Use Rights/ House Ownership in Vietnam; See also Revising the Land Law for Sustainable Development in Vietnam, World Bank Policy Note (2012). 96 Part III - CONCLUSIONS AND RECOMMENDATIONS The discretion of local authorities that helps generate opportunities for corruption could be kept under better control in this way. Some provinces are already experimenting with such systems and meeting with some success - making such approaches the rule rather than the exception would help to reduce corruption. Finally, it should be noted that improving transparency also holds great promise for reducing corruption in land management. Another study last year systematically measured the actual level of transparency of land related documents17. Many that are specifically designated as public information were found to be impossible to obtain in the majority of provinces, districts, and communes in the study. Conversely, some documents which are not specifically required to be made public were provided by some provinces, districts and communes. Improving transparency of land related documents, therefore, is well within the realm of the possible. The study on land transparency identified a range of causes for the lack of access to information that have to do with the way that the law is implemented, from capacity constraints to poor record keeping to poor awareness by officials that it is their duty to provide the information. At the same time, some key pieces of information that should be public, such as the prices actually paid by investors, are not currently mandated to be public information. Legislative changes would be required in order to mandate that such information be public. Improving transparency would lead to lower levels of corruption: In addition to the analysis in Section 2.2.9 showing that provinces and districts that emphasize openness and transparency have lower levels of corruption, another recent study analyzed the land transparency data and found that provinces with more complete access to information also had lower levels of corruption18. Build the institutions for dealing with conflicts of interests Conflict of interest is a new concept in Vietnam. When applied to public officials, the term “conflict of interestâ€? means a situation in which an official may stand to benefit from a decision made in his or her official capacity. Although the Anticorruption Law of 2005 clearly defines that “abuse of position and power to illegally use state assets for personal benefitâ€? is a corrupt act, the Law provides little guidance on how to prevent cases where officials (even honest officials) find themselves in the position of conflict of interest. For example, while it is clear that accepting bribes is corrupt and illegal, it is less clear how the 17. Survey Report on Information Disclosure of Land Management Regulations, by DEPOCEN for the World Bank (2010). GIRI and T&C Consulting research (2010) showed that 69% of households get information from cadastral officials and the average informal payment to these officials for this “legally mandatory publicâ€? information is 811,000 VND. 18. Vietnam Development Report 2012 - Market Economy for a Middle Income Vietnam by the World Bank and others. 97 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS law would treat a situation whereby an official’s family member may benefit from an official’s decision, or what sorts of outside business interests an official may be permitted to have19. A system for preventing and resolving conflicts of interest is a key part of the anticorruption infrastructure in many countries. Improving the system of regulating conflicts of interest would also help address the emerging challenge of interest groups. Interest groups may perform positive functions by providing voice to the business community, as well as negative functions by trying to shape the legal framework to their advantage, using corruption or otherwise. Interest groups are becoming more influential in Vietnam according to 50% of firms, and only 10% of firms disagreed. Many enterprise managers see some positive aspects of interest groups: 37% say that interest groups help the business community communicate their problems to policy makers. A significant fraction, however, also see negative aspects: 40% say that interest groups use connections to gain advantages, and 19% say that interest groups use corruption to achieve their goals. The challenge for Vietnam, then, is to control the degree to which interest groups can twist the playing field to their own advantage, while at the same time providing them the space to voice their legitimate concerns to policy makers. Even the most advanced countries constantly struggle to balance these two seemingly competing objectives. Since interest groups can serve positive functions, the solution is not to stop them from forming, nor to stop them from voicing their concerns. Rather, the solution to the conundrum is to limit their ability to exercise influence in a negative way. Defining and controlling the decision making space for public officials to limit their own conflicts of interest is a key part of the set of reforms that are needed to limit the ability of interest groups to shape policy to their own advantage. Regulating conflicts of interest is important not only with regard to interest groups, it is a fundamental part of building an ethical public administration more generally. Even honest officials can find themselves in positions of conflict of interest20. More clearly 19. For example, a 2010 study by GI and T&C Consulting on Corruption in Education showed that 24% of parents asked school teachers or government officials who were their family members or close friends to help getting their children into non-designated schools. 20. The 2010 GIRI and T&C Consulting study - Corruption Risks in the Issuance and Transfers of Land Use Rights/ House Ownership in Vietnam showed that when households hire intermediaries to help register for their LURHOCs, 51% of these households went to cadastral officials for that service. In the same survey, 16% and 19% of cadastral officials admitted that they helped brokers for “all-in servicesâ€? and for “quick resultsâ€?, respectively. 98 Part III - CONCLUSIONS AND RECOMMENDATIONS establishing procedures for resolving conflicts of interest will be essential for Vietnam to reduce the prevalence of corruption. Once defined, such a system needs an enforcement authority to help resolve conflict of interest situations, and to help advise officials so they may avoid such situations in the first place. Many countries (including, recently, Armenia and Mongolia) do this by establishing ethics boards or some similar enforcement body to handle such cases. Indeed, as countries move to middle income status and then to higher levels of development, the establishment of enforcement bodies increases. According to one database, among countries with legislation regulating conflicts of interest, only 30% of low income countries have enforcement bodies, compared to 40% and 64% for lower and upper middle income countries respectively21. In modernizing its institutions, Vietnam would make a large step forward by establishing mechanisms for regulating conflicts of interest. One related set of reforms is worth mentioning here. In the analysis at the outset of this section, the single set of anticorruption measures that best explained the patterns of corruption across provinces and districts was the degree of implementation of entitlements, norms and standards. These provisions, introduced in Section 2 of the Anticorruption Law of 2005, introduced a form of internal control for high ranking public officials and agencies regarding acceptable uses of the state budget. (For example, a head of department is entitled to a certain amount of money a month for telephone allowance, and a Minister is entitled to have a car and a driver exclusively assigned to him or her while carrying out public duties). These provisions were aimed at reducing discretion and abuse by public officials in agencies and departments regarding the use of the state budget. While it was aimed more at preventing misuse or embezzlement of assets within the state apparatus and less so with lowering levels of bribery at the interface of the state and society, it could present a demonstration effect: when the leaders of agencies are subject to such internal controls themselves, discipline within the agency may be enhanced. As a whole, nearly 90% of officials reported that such measures are effective. As Vietnam revises the Anticorruption Law, the norms and standards themselves will surely be revisited. It will be important, however, to ensure that the positive influence that the entitlements, norms and standards seem to be having is preserved, and possibly scaled up through broader application across positions and categories of entitlements. Similarly, in revising the Anticorruption Law, there will be an opportunity to provide greater clarity on what is permitted and not permitted in terms of the giving and 21. World Bank’s Public Accountability Mechanisms database. https://agidata.org/pam/ 99 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS receiving of gifts. At present, public officials are provided guidance in terms of the value of gifts that are acceptable for them to give, but not so clear guidance is provided on the value of gifts that they are permitted to receive. The Anticorruption Diagnostic Surveys showed that many respondents, even public officials themselves, consider many acts to be “corruptionâ€?, even if not strictly defined as such by the law. Clarifying reasonable thresholds for the receiving of gifts would help remove some of the ambiguity on which corruption feeds. Fix the system of assets and income declarations A key finding of the analysis in Section 2.2.9 is that there appears to be only a relatively weak correlation between the degree of implementation of the system of assets and income declarations and the actual levels of corruption. At the same time, there is some indication that districts where the assets and income declarations are taken more seriously do, in fact, have lower levels of bribery according to citizens. As Vietnam prepares to revise the Anticorruption Law, a key question is how to make this system work better. Declarations of assets and income are a centerpiece of the Anticorruption Law of 2005. The Law mandates regular declarations, and by now formal compliance with the provisions of the law is near 100%. Although the Law does outline procedures for verification, the huge numbers of filers makes it difficult to verify more than a fraction of those declarations. Moreover, until very recently all of the declarations were confidential. In recent months the declarations were made quasi-public, being posted for a short time in the place where the officials worked. The ability of society to help identify cases where policy makers have conflicts of interest can be strengthened by making declarations of assets and incomes more rigorous, and publicly available for high level officials. Such a measure also has some support among the population: 68% of officials and an even higher percentage of enterprise managers and citizens, believe that publicizing the declarations is necessary. Making the declarations publicly available, and strengthening them to cover interests as well as assets and income, would also help make the media more effective at identifying cases where the links between interest groups and policy makers have become too close. International experience22 suggests a number of elements of a good system of income and assets declarations. These are (i) a manageable number of filers; (ii) a good data 22. This section draws on and summarizes Public Office, Private Interest – Accountability through Income and Asset Disclosure. Stolen Asset Recovery Initiative, World Bank and UNODC. 2012 (This publication is also available in Vietnamese. www.worldbank.org/vn/quantrinhanuoc.) 100 Part III - CONCLUSIONS AND RECOMMENDATIONS management system for the filings, (iii) effective measures to sanction non-compliance, and (iv) disclosure of asset and income declarations to the public. Like Vietnam, many countries have opted for an income and assets declaration regime covering a large population of filers. With a large population, however, it has proven difficult to manage the large number of filers and verify the information submitted. Whether it is a simple declaration form with only sparse information required or a more complex form with detailed information, asset and income declaration systems are weak tools to combat corruption if the information filed in those systems overwhelms the system’s capacity to regulate and track them. If Vietnam were to restrict the number of filers to higher ranking officials, such as Ministers and above (as is common in higher middle income and high income countries) or officials above certain salary thresholds (as in Guatemala, for example), this would make asset and income declaration systems more effective. A good system for the management of filings is critical to any system of asset and income declarations as it develops over time. Keeping track of filers lists, reducing the amount of paperwork for public officials themselves and for the agencies managing the income and asset declarations, detecting discrepancies in filers’ information and their actual incomes and assets are all necessary. Using information technology to improve data management of income and asset declarations, as has been done in countries at various levels of development (South Korea, Indonesia, Thailand, Mongolia, and Argentina) can help make the system manageable. As a lower middle-income country, Vietnam could do the same. Any functioning income and asset declaration system needs effective sanctions for those who do not comply and/or file wrongful information. Applying sanctions, be they administrative or criminal ones, for non-compliance deters public officials from shirking their public duty to declare. At the same time, it will help build public trust that governments are serious about fighting corruption and that the income and asset declaration system can be a powerful anticorruption tool. Around the world, the debate on disclosing assets and incomes of public officials to the public has always centered around privacy rights of public officials and information security, thus preventing some countries from taking the bold step of publicizing the declarations. However, income and asset filings, if not disclosed may reduce the effectiveness of income and asset declarations as a corruption prevention measure and it may become formalistic. Some places (Argentina, the United States) require full disclosure of income and asset information and some (SAR Hong Kong - China, Croatia, Indonesia) 101 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS require only partial disclosure. In fact, publicizing the income and assets declarations has enabled the media and civil society organizations in countries like Croatia, Romania and Argentina to serve as important allies in the fight against corruption, filling the resource and capacity gaps of the agencies responsible for implementing the income and asset declaration system, checking lifestyles of public officials, improving corruption detection and even deterring corruption before it occurs. Revising the Anticorruption Law to allow full public disclosure of income and assets declarations for high level officials would be a major step forward for Vietnam. Streamline administrative procedures The analysis in Section 2.2.9 showed that one set of measures that seems to be working best for reducing actual corruption is administrative reform. It is easy to see why this would be important. The Anticorruption Diagnostic Surveys showed that officials with discretion over their decisions are perceived as creating difficulties for citizens and firms in order to elicit unofficial payments. On average, the sectors that are identified as creating the most problems for enterprises and citizens are also the sectors most beset by corruption. Many firms said they make unofficial payments simply to speed things along and to avoid cumbersome procedures: 32% of firms said they pay unofficial payments as a means of speeding through processes. Accelerating administrative procedures reform, which has done well under Project 30, needs to continue. Continuing to clarify the legal framework to remove the discretion that allows officials to abuse their positions would also reduce chances for corruption. This is not a new legislative proposal, but rather many small changes in policies to reduce the opportunities for corruption. Empower the media The Anticorruption Diagnostic Surveys show just how important the media could be for the fight against corruption. Positive assessments of the media outweigh the negative assessments by a considerable extent. Both public officials (82%) and enterprise managers (84%) agreed that the mass media discovered many corrupt cases before the authorities commenced their work. At the same time, 71% of public officials and 62% of enterprise managers agreed that the media sometimes exaggerates the information. In addition, when asked about the sectors or fields that have corruption themselves, fewer than one percent of all respondents identified the media as among the most corrupt, and some 93% of all respondents say they get their understanding of corruption from the media. On balance, respondents see the media as a potentially important ally in the fight against corruption. Improving the quantity and quality of information on which the media can base their stories would strengthen their objectivity in reporting, helping them to assemble the 102 Part III - CONCLUSIONS AND RECOMMENDATIONS facts more quickly and efficiently, and at the same time making it easier for them to put their investigative skills to use uncovering cases of corruption. But access to information is not enough. Reporters need the confidence to know that vigorous reporting will not bring negative consequences. More than 80% of citizens and enterprises agreed that one of the causes of limited performance in preventing and combating corruption is that there are ineffective mechanisms to promote participatory roles of the public and the media in detecting and fighting corruption. Among public officials, 70% agreed, an overwhelming majority of respondents. The population, the business community, and public officials alike are ready for the media to play a more active role. Legal changes to make reporting errors subject to civil rather than criminal sanctions would help empower the media to report vigorously on corruption at all levels. In other countries, the media have proven to be extremely effective at identifying cases where the links between interest groups and policy makers have become too close - strengthening the ability of the media to play this role would also constrain the negative side of interest groups while preserving the positive role they can play in communicating concerns to policy makers. 3.2. Policy implementation The Anticorruption Diagnostic Surveys showed that most officials believe the legal system has several shortcomings, especially that the laws are too vague, but an even larger proportion say that the problem is weak enforcement of existing laws. In this section we present recommendations surrounding policy implementation. Make the civil service meritocratic and less prone to corruption Vietnam’s civil service reforms are already aiming to make the civil service more merit-oriented and professional. The Anticorruption Diagnostic Surveys show that there is still a long way to go. Some 28% of citizens paid unofficial payments when applying for a job from the state, a finding that is consistent with other studies23. Of all the interactions between citizens and the state, applying for a job brought with it the highest chance of paying a large unofficial payment. Although few public officials said that recruitment and promotions are due to bribery, 17% said they had seen the promotion of incompetent people for personal gain in the previous 12 months. More effort is clearly needed make recruitment and promotions in the civil service and public service based on merit rather 23. This finding is consistent with a key finding of the Vietnam Provincial Governance and Public Administration Performance Index (PAPI) 2011. 103 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS than other factors. This effort will pay off. The analysis in Section 2.2.9 showed that the provinces where hiring and promotion decisions are based on merit actually do have lower levels of corruption. Taking further measures to enhance the merit-orientation of personnel policies would help reduce corruption further. This is also consistent with a range of international studies showing the importance of meritocracy in the public service for reducing corruption24. While the analysis in Section 2.2.9 showed the importance of merit oriented personnel policies for reducing corruption, there was little correlation between salary levels and levels of corruption. While 79% of public officials believe that low salaries contribute to corruption, only 58% of citizens believed so - citizens are much less forgiving of low salaries than public officials. Any reform to salary levels would need to be accompanied by broader reforms in the salary system and in the civil service system more generally. The emphasis in Vietnam on allowances to supplement the base salary makes the system more opaque. Indeed, although the analysis in Section 2.2.9 finds no link between total remuneration and corruption outcomes, when broken down between base salary and allowances, there is actually a slight positive correlation between the degree to which remuneration is based on allowances and the levels of corruption - places in which remuneration relies more on allowances have more corruption, not less. This is consistent with international experience which shows that the levels of salaries themselves are not as important for explaining levels of corruption as the degree to which merit, accomplishments and good work translate into rewards. Making Vietnam’s salary scales simpler by targeting increases on the skills profiles for which salaries are most out of line, and making salary increases selective and matched to improved qualifications, skills or responsibilities rather than length of service, would bring rewards to good performers. The analysis in Section 2.2.9 highlighted another set of measures related to the civil service that shows some signs of working and could be scaled up. In provinces and districts where the transfers of public officials through job rotation, covered by Section 3 of the Anticorruption Law of 2005, are implemented more thoroughly, the public’s perception of corruption is lower. Although there is no significant effect with respect to enterprise’s or citizen’s experiences with bribery, the impact on the perception of corruption is nevertheless noteworthy, as job rotation may help thwart systemic corruption that goes beyond bribery by disrupting the networks on which systemic corruption relies. As 24. A summary is provided in Vietnam Development Report 2010-Modern Institutions page 18. 104 Part III - CONCLUSIONS AND RECOMMENDATIONS Vietnam revises the Anticorruption Law, attention to how this policy may be more thoroughly implemented could help reduce corruption further. Give whistle-blowers the trust they need Although there seemed to be some willingness to report corruption, this was clearly hindered by a lack of trust that there would be no repercussions. The analysis in Table 3 showed that places where officials are more confident in denouncing corruption are also the places with lower levels of corruption. The system of denunciations has some support: 42% of citizens say if they encountered corruption they would report to the competent person. At the same time, among those who say they would not denounce corruption, 61% of firms and 69% of citizens say it is because they have no trust in competent persons. Strengthening whistle-blower protections could help. Understanding the effectiveness of the current system for protecting whistle-blowers and the status of retribution will be essential. Establishing efficient reporting channels, ensuring systematic data collection of cases, their follow-up and results are also necessary ingredients for a functioning whistle- blower protection system. Importantly, defining protections for whistle-blowers will only be credible if there are also provisions that ensure sanctions against those who inappropriately take revenge on whistle-blowers, and these will only be believed if the population sees visible evidence that these provisions are taken seriously. More importantly, making investigations and prosecutions of corrupt people more vigorous would give people confidence that the risks they take in publicly denouncing corruption would not be wasted. Strengthen the system of enforcement While corruption has many causes, it is clear that the population perceives sanctions of corrupt people to be weak - around 90% of all three sample groups said that a key cause is that there are no serious sanctions of corrupt people. Many clearly feel that corrupt people are not sanctioned, and this is generating cynicism in the population. The surveys showed that 69% of officials agreed that some heads of agencies treated corruption cases lightly out of concern for organizational and personal reputation. Breaking this conflict of interest by making responsibility for leading the anticorruption fight independent from those who would be investigated would open the space for more vigorous enforcement in Vietnam. Develop sector-specific approaches An important distinction needs to be made between being accountable for detecting and investigating corruption in an agency, and being accountable for preventing 105 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS corruption by building ethics and putting in place the systems to limit corruption in the agency. While entrusting the enforcement of anticorruption laws to the heads of the agencies concerned presents a conflict of interest and could weaken enforcement, as discussed above, that does not mean that the heads of agencies should not be accountable for preventing corruption within the agencies. The Anticorruption Diagnostic Surveys showed modest support for the latter approach: 77% of public officials believe in the effectiveness of imposing penalties for the heads of agencies when there is corruption within the agency, and we saw in the analysis in Section 2.2.9 that this measure is, in fact, associated with lower levels of bribery by enterprises at the district level. At the same time, making agencies responsible for their own policies for preventing corruption would seem to have support. Many of the policies that seem to be working, according to the analysis in Section 2.2.9, are implemented unevenly across provinces and districts, and even across organizations. Indeed, it is this variation in implementation that allows the analysis in the first place. A key determinant of implementation of these policies is the degree of leadership being provided in the agency or province or district. This mirrors findings from international research25 that a key determinant of actual corruption levels is leadership, and that leadership is most important in places where the institutional structures are less advanced. More importantly, while most of the policy measures discussed in this report - personnel management, assets declarations, investigations, etc. - are measures which apply across all organizations, it must be recognized that the factors that open space for corruption are different in each sector. The Anticorruption Diagnostic Surveys found that corruption is perceived to be an especially serious problem in land administration, the traffic police, education, health, customs and construction. Although corruption in some of these sectors may not be seen as “seriousâ€? in terms of the real economic losses to the state or in terms of the size of the average bribe, these so called “pettyâ€? corrupt sectors are the ones that have largest interface with society - their negative impact on society should not be underestimated26. 25. Anderson, Reid, and Ryterman. (2003) Understanding Public Sector Performance - An Empirical Contribution. The World Bank. This report is available in Vietnamese at www.worldbank.org/vn/ quantrinhanuoc. 26. This finding supports those of the GIRI and T&C Consulting studies on corruption in Land sector, as well as with GI and T&C Consulting survey on corruption in Education sector, both in 2010. 106 Part III - CONCLUSIONS AND RECOMMENDATIONS Yet, the systemic weaknesses that drive corruption in each sector are different. For land management, a study last year27 examined the whole process flow of land acquisition and land allocation, as well as for the issuance of LURHOCs, and found the system rich with the ingredients for corruption. In the construction sector, the processes for obtaining the myriad approvals similarly opens up many opportunities for corruption28. For the traffic police, the combination of (necessary) coercive authority and the ability to impose fines or other sanctions is a recipe for corruption everywhere. The nature of interactions for health and education are of an entirely different nature, as these are services coveted by the population29. When the accessibility or quality of the services is perceived to be related to unofficial payments, the drivers of corruption are clear. Leadership is essential for reducing corruption, and having the leaders of those agencies outline their plans to prevent corruption is key to establishing accountability. Detecting, prosecuting and penalizing corrupt officials may be the responsibility of the Government Inspectorate, OSCAC, the police and the courts, but responsibility for preventing corruption from happening in the first place must reside with every agency, province, district, and commune. To ensure that progress is being made, a system of monitoring is essential. 3.3. Corruption monitoring The measures outlined so far in this section center on clarifying and strengthening the legal framework for the fight against corruption and for strengthening the implementation of existing provisions. It is equally important to establish a monitoring and evaluation (M&E) framework for tracking progress over time, allowing periodic re- evaluation of what is working and what needs fixing. The Government Inspectorate has recently developed a comprehensive M&E framework which aims to track corruption and progress on anticorruption work nationwide. The M&E framework was designed with a new approach for Vietnam, combining traditional information taken from self-assessments of government agencies with findings from different sociological surveys. With the endorsement of this M&E 27. Recognizing and Reducing Corruption Risks in Land Management (2011) by the Embassy of Denmark, the World Bank, and the Embassy of Sweden. 28. Davidsen, et al. (2009) “Implementation Assessment of the Anti-Corruption Law: How far has Vietnam come at the Sector Level? A Case-Study of the Construction Sectorâ€?. Ministry of Foreign Affairs, Denmark. 29. See, for example, Towards Transparency (2011). “Forms and effects of corruption on the education sector in Vietnamâ€?. 107 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS system, for the first time from 2012 onwards official reports on anticorruption are expected to be produced based on evidence, not only on guesswork as has heretofore been the case. While setting up the framework and introducing external rather than self-assessments is a laudable achievement of the government, making the framework an operationally useful M&E tool is more challenging. The system relies on indexes of various parameters, ranging from the perceptions of citizens and firms on corruption to more complex indicators of the extent of economic losses to firms and citizens. At the same time, much of this information is not readily available nor is it owned by the government. Making full use of the data of these Anticorruption Diagnostic Surveys will help to begin to make the M&E system operational. Conducting regular updates of the surveys will be necessary to track changes over time. Other available datasets that are repeated over time, such as PAPI and PCI, will also support the richness of the data on which the M&E system relies. Finally, simplifying the M&E system as it is rolled out, and making it fully owned by the government, both in terms of data generation and analysis, will help ensure that it is implemented and is useful. 3.4. Public knowledge enhancement Most of the recommendations in the preceding sections focus on the demand side, reducing the opportunities for corruption and enhancing the ability to detect and deter corruption. This is not entirely inappropriate. The surveys show that officials with discretion over their decisions are perceived as creating difficulties for citizens and firms in order to elicit unofficial payments. On average, the sectors that are identified as creating the most problems for enterprises and citizens are also the sectors most beset by corruption. A recurring theme of the Anticorruption Diagnostic Surveys, however, is that corruption is fed by both the demand and the supply sides30. Most firms (60%) report corruption to be costly, but more than half also say that it brings benefits to the firms. When unofficial payments do occur, 70-90% of firms say that it was initiated by them, depending on the sector. The survey finding that many firms willingly participate in corruption is disconcerting, but there may be an opportunity to make the business community an ally, since some firms are clearly ready to help fight corruption. Around half of the firms said they had initiated their own anticorruption activities, especially codes of conduct. Training firms on making and enforcing codes of conduct that work would seem sensible, as would 30. This is consistent with the 2012 study on Current Status of Corruption in the Business Sector in Vietnam by VCCI and DEPOCEN. 108 Part III - CONCLUSIONS AND RECOMMENDATIONS cooperating with universities’ business programs to introduce ethics in their training on corporate social responsibility. The surveys also showed that although firms that engage in bribery believe it to be an expedient means of moving forward, they actually perform worse, on average, than firms that adopt other strategies. Helping firms to understand the costs of corruption and benefits of collective action can help make them allies in the battle against corruption, rather than participants in the problem. The leadership of the large business associations will be key in this regard. The citizens survey similarly shows the need to educate the public on the costs of corruption and their rights when faced with difficulties. When asked why they paid unofficial payments, only 17% said that without paying things could not get done. Citizens that had paid bribes were much more likely to say it was better than dealing with complex procedures (32%), a gift of thanks for those who helped (38%), or simply because others do the same (41%)31. Educating citizens about how they can avoid feeding the system and make better use of tools available to them, such as transparency provisions in the law, can also help to constrain the supply side of corruption. Education of the business community and citizens needs to be supplemented by education of the public officials themselves. Improving implementation of laws requires that public officials know clearly their obligations. For example, the challenges in implementing the many transparency provisions in the Anticorruption Law of 2005 are driven in part by the lack of awareness that providing the specified information is required by law. In several studies, public officials were reported to be puzzled when they were requested to provide information32. One benefit of passing the Law on Access to Information, as suggested earlier in this section, is that it would raise awareness that access to information is a right of the citizen, and an obligation for those in public office. Similarly, mass education campaigns to build awareness on the part of all elements of society on how to prevent and avoid corruption can help ensure that all are working to move Vietnam forward, rather than exerting energy toward opposing ends. 31. An earlier study on corruption in Land (GIRI and T&C Consulting, 2010) showed that 46% of citizen in the sample believed one should pay informal fees to officials when working with land issues. Another study on corruption in Education (GI and T&C Consulting, 2010) found that 67% of parents believed that paying informal money to get their children to good schools is a normal practice. 32. These include a study by the Ministry of Justice and the National Hanoi University (NHU) on the situation of access to information in four provinces, cited in Vietnam Development Report 2010 - Modern Institutions, and the 2011 study on Land Transparency carried out by DEPOCEN for the World Bank. 109 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS 3.5. Closing remarks Most officials believe that despite the fact that corruption remains a challenge, there have been some successes: 85% believe that officials’ awareness of corruption has improved. But knowledge is not enough. Broad institutional reforms are needed: establishing real transparency with a Law on Access to Information, revising the Land Law to reduce the use of compulsory land acquisition for private projects, amending the Anticorruption Law with less restrictive clauses on the right to information of citizens, better measures to control conflicts of interest, bold actions on asset and income disclosure, and making the civil service merit-based. Implementing these reforms will take time, but as Vietnam moves toward middle-income status, the time for modernizing its institutions is now. But there is also a need to shift attitudes of the citizenry, firms, and public officials alike: 64% of officials say that some officials are willing to abet corruption, allowing it to proceed, and 86% say that the feeling of being afraid to fight corruption is still widespread in the public. Vietnam’s fight against corruption needs to include unequivocal actions by the top leadership to show that the fight against corruption is serious. Some countries have taken drastic measures to bring about the needed change in attitude. In Hong Kong SAR - China and Singapore, fighting corruption started with cleaning the traffic police force first, and in Georgia, the traffic policemen were all fired and replaced. In several transition countries of eastern Europe, public declarations of assets, income and interests were placed on the internet for all the world to see. In Mongolia and Indonesia, an independent anticorruption agency was empowered to conduct its own investigations of high-level officials. In Romania, a special anticorruption prosecutor was established and empowered to focus on high-level investigations. Such measures are not easy, but they send the signal to the population that corruption is not to be tolerated and that no one is immune from accountability. 110 Part IV Annexes ANNEX 1. ADDITIONAL FACTS ON METHODOLOGY This Annex provides additional details on the survey methods to supplement the description in Part I of the report. A.1.1. Sample Citizen sample: The citizen sample in each province/city was as follows: Table 7. Surveyed citizens in the sample Official sample: The composition of public officials survey at the local level was as follows: - Elected bodies: In each province, 23 members of People’s Councils at all three administrative levels were interviewed, of which 5 were from provincial levels, 9 from district level (3 for each district) and 9 from commune level (1 for each commune). 111 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS • At the provincial level, five interviewees in each Provincial People’s Council were selected, consisting of a Council Chair/Vice Chair, a head of Bureau of Economic and Budgetary Affairs, a head of Bureau of Social and Cultural Affairs, a head of Legal Bureau, a head of Bureau of Ethnic Minorities and a staff member of the Council Office. If the Bureau of Ethnic Minorities did not exist, another staff member was selected to replace. • At the district level, the three interviewees were a Council leader, a head of the Bureau of Economic and Social Affairs, and a head of the Legal Bureau. • At the commune level, one interviewee was selected from the People’s Council, and she/he was a Council leader. - Executive agencies: The sample size in each province was 52 interviewees, of which 10 were from the provincial level, 24 from the district level (8 per district) and 18 from the commune level (2 per commune). • At the provincial level: 2 interviewees (one leader and one head of division) came from each counterpart department of 5 aforementioned ministries. In addition, one came from the Provincial Inspectorate and the other from the Provincial OSCAC. • At the district level: 8 interviewees were leaders of District People’s Committee, Committee Office, Finance and Planning Division, Division of Natural Resources and Environment, Division of Industry and Trade, Division of Education and Training, Division of Health, and Division of Agriculture and Rural Development (for agro-based districts) or Division of Construction/ Urban Development (for urbanized districts). • At the commune level: a Chair/Vice Chair of Commune People’s Committee, and a person in charge of legal affairs were included in the interviews. - Public service delivery staff: 96 interviewees of which 15 from the provincial level, 27 from the district level (9 per district) and 54 from the commune level (6 per commune) were included in the sample. • At the provincial level: there were 15 public service delivery staff at provincial level, of which 5 staff members came from each counterpart departments of 5 aforementioned ministries; one from inspectorate department; one from Provincial OSCAC; one from district Bureau of Labor and Social Affairs; one from taxation office; three from a provincial public specialized school; and three from provincial general hospitals. • At the district level: one service deliverer came from the following nine divisions/units: Finance and Planning, Natural Resources and Environment, 112 Part IV - ANNEXES Industry and Trade, Education and Training, Health, Agriculture and Rural Development (for agro-based district) or Construction/Urban Development (for urbanized district), Labor and Social Affairs, a public school, and a district health center. • At the commune level: 6 interviewees were from the communal cadastral office, a principal of a public school, head of communal health center, a person in charge of statistics or socio and culture, head of the Fatherland Front or Veterans Association, and head of Women’s Association. A summary of the public officials sample is presented in Figure 57 and Figure 58. Figure 57. Public official sample by administrative levels Figure 58. Public official sample by profession 113 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Enterprise sample: As the enterprise universe is better defined, a less ad hoc approach to sampling was used. Enterprises were selected based on stratified random sampling by their sector, size, and ownership such that the structure of the surveyed enterprise sample is close to the actual population. Numbers of surveyed enterprises in each province were determined by actual numbers of currently active enterprises. Based on the current list of enterprises and sample frame, surveyed enterprises were randomly selected. The enterprise survey covered 1,058 enterprises across ten provinces based on the sample stratification. The sample description by type of firm is shown in Figure 59. Figure 59. Planned and actual enterprise surveyed sample Sample characteristics: Table 8 shows some basic information about the respondents in each of the three sample groups. The relatively small percentage of female respondents reflects the fact mentioned earlier that the citizen’s sample focused on those who had visited one stop shops. 114 Part IV - ANNEXES Table 8. Common characteristics of surveyed sample A.1.2. Pilot The pilot survey was carried out from December 21-22 in Vinh Phuc. The overall objective of the pilot was to test the survey plan and tools in practice. The lessons drawn from the pilot were used to finalize the survey questionnaires and interview instructions, and the survey plan. The pilot survey team included 21 members: The Consultant team (10 members), GI/OSCAC Task Force (9 members), and World Bank (2 members). The team received support from the provincial Inspectorate and Office of Anticorruption Committee in connecting with potential interviewees. Twenty seven pilot interviews were conducted. A brief retrospective was carried out after each day in the field, and an entire third day was spent to draw lessons learned from the pilot survey. The pilot survey provided useful insights for the team in revising the research tools (questionnaires and instructions), data collection plan, and interview process protocol. A.1.3. Training the interviewers Training provincial team leaders: After the pilot survey, the core team provided a general training to all provincial team leaders. The training had participation from World Bank staff and representatives of the Task Force. The provincial team leaders were trained in how to conduct interviews, how to train others in conducting these interviews, and how to supervise/monitor the interviews. Emerging risks and practical logistics arrangements for data collection were also discussed. The training of provincial team leaders took place on Jan. 6th - 7th, 2012. 115 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Training interviewers: Each provincial team leader conducted a two-day training session for interviewers in the province she or he was responsible for. The format and materials of the training were centrally provided by T&C and APIM, which also supervised the training together with the World Bank team. The training included detailed discussion of each question, preparation and conduct of interviews, and mock interviews, among others. These training sessions took place between Feb. 23rd and March 5th, 2012. A.1.4. Support from OSCAC and GI GI and OSCAC played a very hands-on role in helping the interview teams gain access to officials and citizens for the interviews. Key activities included: - Before the training, GI and OSCAC actively worked with the Consultant team on the interview plan. The interview plan needed to match with Taskforce availability in order to best support the interview teams. - After the training, GI and OSCAC connected the Consultant team with the provincial focal agency (normally provincial Inspectorate or provincial committee of AC), and contact persons. These provincial focal agencies and persons helped the interview teams connect with respected interview officials at the provincial level. They also introduced the team to the District contact persons (and then District contact persons introduced the team to commune level) to facilitate the interviews with officials at these levels as well as with citizens. At the central level, GI/OSCAC helped connect the team with Ministry’s contact persons who then assisted the team for scheduling the interviews with central officials. - During the interview period, the Task Force sent people to different provinces to monitor the interview process and to assist the team with any problem in accessing respondents. This direct and hands-on involvement from GI/OSCAC significantly contributed to the completion of interviews with officials and citizens and enabled a very high response rate. A.1.5. Quality assurance A great deal of attention was paid to quality assurance as this was seen as a key factor for the success of the surveys. Different measures were taken before, during and after undertaking the field work. 116 Part IV - ANNEXES Quality assurance measures prior to the field work A set of survey instruction manuals was developed by the core team, in which quality control principles and procedures were clearly set for the field team leaders. Also, a series of forms and templates were provided as tools for monitoring and supervising the field work as well as for checking quality and consistency of respondents’ answers. All field team leaders were required to attend two training sessions in the T&C office: one right after the pilot survey in Vinh Phuc and the other just before the Tet Holiday. The training sessions aimed at making team leaders master the questionnaires, survey processes and tools of quality control. Since they would act as instructors for enumerators when they came back to their own provinces, the fact that they were able to get used to the instruction manual would guarantee that they could conduct retraining in a proper manner. During retraining sessions in the provinces, the core team members, with participation of World Bank staff, were sent to support the team leaders in selected provinces, including Hanoi, Hai Phong, Ho Chi Minh City and Can Tho. Quality assurance during the field work A multi-layer structure of supervision was organized during the field work, led by the general supervisor and coordinator, under whom were three regional supervisors for Northern, Central and Southern regions. Each regional supervisor was responsible for monitoring the progress and randomly checking the field work in three provinces. In each province, the field team leader was responsible for self-conducting surveys with government officers at the provincial level. At the same time, he was required to supervise the surveys with enterprises, government officers at district and commune level, and households, which were conducted by local enumerators. The local enumerators also were subject to frequent checks by representatives of their line agencies - Provincial Statistics Offices - the contractors of T&C. Every week, team leaders had to send progress reports to the regional supervisors in a preset form. The regional supervisors, in turn, made a more synthesized report for the general supervisor, who had responsibility to report to the World Bank, GI and OSCAC. Apart from regular monitoring channels, the World Bank, GI and OSCAC, and their provincial counterparts (Provincial Inspectorates) also made independent and random on-site checks. The supervising structure of the project is illustrated in the following figure. 117 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Figure 60: Supervising structure of quality assurance During the field work, the team leaders took additional quality assurance measures including attending interviews conducted by enumerators, making on-site visits without notice, and checking the notes taken by enumerators and being reconfirmed by local accompaniers. Independent supervision carried out by GI and OSCAC team members in the second week of the field work, by the World Bank staff in the third week, and by T&C staff in the rest. Remote checks were made by the general and the regional supervisors via telephone. The quality of response sheets was also checked in four rounds: - Round 1: Self-checking. In some provinces, such as Hai Phong, enumerators circulated their completed questionnaires among themselves for cross-checking. - Round 2: Check by the team leaders. The team leaders were required to check all completed questionnaires and verify with their signatures. Questionnaires without team leaders’ verification would be returned to the provinces for completing the process. - Round 3: Random check by the regional supervisors. Each regional supervisor took a sample of completed questionnaires for reviewing consistency and logic among answering options. 118 Part IV - ANNEXES - Round 4: Re-verifying. The World Bank staff re-verified randomly a sample of completed questionnaires during their supervising visits or in the Hanoi office. From T&C side, a group of undergraduate students was hired to check all completed questionnaires sent to Hanoi by mail. All questionnaires for which doubts were raised would be rechecked by the team leaders, who in turn had to contact the enumerators for clarifying. If the quality of the completed questionnaires was unacceptable, the team leaders would be required to make new interviews as replacement. Logging and call backs after the field work Logging and call backs were made mainly with enterprises and households to make sure that the interviews were conducted in a professional and proper fashion. Such call backs were not made with government officers since all meetings with this kind of respondents were scheduled in advance and were strictly followed. To prepare for logging and call backs, all enumerators were required to remind the respondents at the end of the interview that someone in the team may call back to them to get their feedback on quality of the interview and enumerators’ attitude. The respondents were asked to provide their telephone number if they did not mind. Based on the list of respondents and the survey notes made by the enumerators, the team leaders in each province undertook call backs for at least 18 different respondents (two respondents per each enumerator). The team leaders either called back to respondents (if their telephone numbers were available), or checked their address via survey notes or local accompaniers and then made direct visits to their houses. All such call backs and/or direct visits were recorded in the log reports, including contact information of respondents. Based on the reports, the general supervisors and independent supervisors of the World Bank, GI and OSCAC had the option of making additional calls to check. A.1.6. Challenges faced by the Team There were several challenges facing the Team during the course of conducting this survey, besides the common challenges of survey research. These special challenges are documented here for those interested in conducting similar work in the future. Sensitive topic: Corruption is a very sensitive topic for surveys, especially for officials and firms. With the help of GI and OSCAC, getting to officials’ interviews was not challenging. However, sometimes the Agencies (such as Ministry or Department at provincial level) asked to have 5 or 6 interviews at the same hour. This required the team 119 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS to have six interviewers ready. In addition, getting them to openly and honestly respond to the survey questions was not easy, especially at the beginning of the interviews. On the other hand, accessing firms was a great challenge. Besides the common challenge of “ghost firmsâ€? (already went out of business without notice) or wrong addresses, getting firms to agree on the interviews was very challenging. Firms in big cities (Ha Noi, Da Nang, Ho Chi Minh City) tended to be less willing to receive interviewers than those in other provinces. In some cases, the firms requested an individual letter from GI addressing their names (and we have 2,600 firms in the sample). In some other cases, the interviewers had to visit the firm 3 times in order to get an interview. It was not uncommon for an interview with a firm to be interrupted several times. Time pressure: To provide timely inputs for the Government and the Central Party’s Meeting in April, 2012, the whole field work needed to be completed within 4 weeks (March 12th - April 10th, 2012). At the same time, the quality control procedures were not compromised. This, coupled with challenges in accessing firms and scheduling interviews with officials, posed a great challenge for the team. The following measures were used to overcome these challenges: - With officials, a flexible scheduling approach was applied. The team was prepared for situations where five or six interviewers were needed at the same hour. Training interviewers, selection of time and location for interviews were given great care to get officials comfortable answering the questions. - For accessing the firms, the team mobilized support from other agencies, such as tax, research institutes, and universities, when needed. These agencies introduced the interviewers to “get through the doorsâ€? of the firms, but were not involved deeper in the interviews which were in all cases conducted privately by trained interviewers. 120 Part IV - ANNEXES ANNEX 2. INSTITUTIONAL FACTORS ASSOCIATED WITH CORRUPTION As described in the text a series of partial correlations was run on province level assessments of the institutional environment (from the public officials survey) against the average percentage of citizens and firms that had paid bribes in that province, as well as the average perception of corruption in that province by citizens. The same exercise was undertaken at the district level. In Section 2.2.9 a summary of the finding is presented. In this Annex we provide scatter plots of each of the combinations of the two variables. As each is a partial correlation, there could be other factors that are driving the correlation. However, since the assessments come from different surveys, one such source of spurious causation can be ruled out. While the results do not necessarily suggest causation, they do provide evidence that places that are implementing certain measures seem to have lower levels of corruption. Each dot on the scatter plots represent one province or district. The vertical axis is the assessment of the indicated institutional factor and the horizontal axis shows the average assessment of corruption as indicated. For example, the first chart on the top left of Table 9 should be read as shown. This shows that the provinces with the better assessment by officials of the effectiveness of openness and transparency actually do have, on average, lower levels of corruption. 121 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Table 9. Province-level analysis of the institutional factors associated with lower levels of corruption 122 Part IV - ANNEXES 123 CORRUPTION FROM THE PERSPECTIVE OF CITIZENS, FIRMS, AND PUBLIC OFFICIALS Table 10. District-level analysis of the institutional factors associated with lower levels of corruption 124 Part IV - ANNEXES 125 126 Responsible for publishing NGUYEN DUY HUNG Responsible for content DO QUANG DUNG Editor DO QUANG DUNG PHAM THI KIM HUE Cover, Design DUONG HONG MAI HOANG MINH TAM Proof reading THAO NGUYEN MINH CHAU 127 Printing 1,000 copies, size 20,5 x 28 cm, at Printing and Cultural Product Single-member Limited (PCP. Ltd) Publishing Plan Registration No: 61-2013/CXB/22-30/CTQG. Publishing Decision No: 2630-QÄ?/NXBCTQG, dated January 22, 2013. Completed and registered in January 2013. 128