Document of The World Bank Report No.: ICR00003807 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IBRD-83760) ON A LOAN IN THE AMOUNT OF US$216 MILLION TO THE STATE OF AMAZONAS, BRAZIL FOR A AMAZONAS: MODERNIZING PUBLIC SECTOR MANAGEMENT, CITIZEN SECURITY AND GENDER POLICIES DEVELOPMENT POLICY LOAN January 27, 2017 Governance Global Practice Latin America and the Caribbean Region CURRENCY EQUIVALENTS (Exchange Rate Effective as of April 14, 2014) & (Exchange Rate Effective as of October 12, 2016) Currency Unit = Real (R$) R$2.33 = US$1 (April 2014) & R$3.20 = US$1 (October 2016) FISCAL YEAR January 1 – December 31 ABBREVIATIONS AND ACRONYMS AISC Integrated Citizen Security Area (Áreas Integradas de Segurança Cidadã) CONSEG Community Security Council (Conselhos Interativos Comunitários Segurança Pública) DIP Integrated Police Unit (Distritos Integrados de Policiamento) DPL Development Policy Loan GAGIS Integrated Citizen Security Management Advisory Unit (Gabinete de Assessoramento da Gestão Integrada de Segurança Cidadã) GBV Gender-based Violence GDP Gross Domestic Product GoAM Government of Amazonas IADB Inter-American Development Bank ICMS Brazilian Value-Added Tax (Imposto sobre a Circulação de Mercadorias e Serviços) ICR Implementation Completion and Results Report M&E Monitoring and Evaluation NEPAS Center for Studies and Applied Research in Public Safety (Núcleo de Estudos e Pesquisas Aplicadas em Segurança) NFC-e Electronic Consumer Receipt (Nota Fiscal do Consumidor Eletrônica) PDO Program Development Objective PPA Multiyear Plan (Plano Plurianual) PSIA Poverty and Social Impact Assessment SEFAZ Amazonas’ Secretariat of Finance (Secretaria da Fazenda do Estado de Amazonas) SEJUSC State Secretariat for Justice, Human Rights, and Citizenship (Secretaria de Estado de Justiça, Direitos Humanos e Cidadania) SEPM Executive Secretariat for Gender-related Policies (Secretaria Executiva de Políticas para Mulheres) SSP Secretariat of Citizen Security (Secretaria de Segurança Pública) Vice President: Jorge Familiar Calderon Country Director: Martin Raiser Senior Practice Director: Deborah L. Wetzel Practice Director: James A. Brumby Practice Manager: Arturo Herrera Project Team Leader: Laura De Castro Zoratto ICR Team Leader: Daniel Ortega Nieto Table of Contents A. Basic Information.................................................................................................................... i B. Key Dates ................................................................................................................................ i C. Ratings Summary .................................................................................................................... i D. Sector and Theme Codes ....................................................................................................... ii E. Bank Staff ............................................................................................................................... ii F. Results Framework Analysis .................................................................................................. ii G. Ratings of Program Performance in ISRs .............................................................................. v H. Restructuring (if any) ............................................................................................................ vi ICR Background and Summary .................................................................................................. 7 1. Program Context, Development Objectives, and Design ...................................................... 8 2. Key Factors Affecting Implementation and Outcomes ....................................................... 13 3. Assessment of Outcomes......................................................................................................... 19 4. Assessment of Risk to Development Outcome...................................................................... 28 5. Assessment of Bank and Borrower Performance ................................................................ 29 6. Lessons Learned ...................................................................................................................... 30 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners .................. 32 Annex 1. Bank Lending and Implementation Support/Supervision Processes ..................... 32 Annex 2. Summary of Borrower’s ICR and/or Comments on Draft ICR ............................. 34 Annex 3. List of Supporting Documents ................................................................................... 38 Annex 4. Achievements by Policy Area..................................................................................... 39 Map: State of Amazonas - Brazil ............................................................................................... 44 A. Basic Information Country: Brazil Program Name: BR Amazonas DPL Program ID: P147979 L/C/TF Number(s): IBRD-83760 ICR Date: 01/27/2017 ICR Type: Core ICR Lending Instrument: DPL Borrower: STATE OF AMAZONAS Original Total US$216.00 million Disbursed Amount: US$216.00 million Commitment: Revised Amount: US$216.00 million Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 11/14/2013 Effectiveness: 01/10/2014 01/10/2014 Appraisal: 03/10/2014 Restructuring(s): n.a. Approval: 05/23/2014 Mid-term Review: n.a. Closing: 01/31/2016 01/31/2016 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Satisfactory Risk to Development Outcome: Moderate Bank Performance: Satisfactory Borrower Performance: Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Satisfactory Government: Satisfactory Implementing Quality of Supervision: Satisfactory Satisfactory Agency/Agencies: Overall Bank Overall Borrower Satisfactory Satisfactory Performance: Performance: C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments (if Indicators Rating: Performance any) Potential Problem Quality at Entry Program at any time No None (QEA): (Yes/No): i Problem Program at any Quality of Supervision No None time (Yes/No): (QSA): DO rating before Satisfactory Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) Other social services 13 13 Sub-national government administration 87 87 Theme Code (as % of total Bank financing) Administrative and civil service reform 19 19 Gender 19 19 Public expenditure, financial management and procurement 38 38 Tax policy and administration 12 12 e-Government 12 12 E. Bank Staff Positions At ICR At Approval Vice President: Jorge Familiar Calderon Jorge Familiar Calderon Country Director: Martin Raiser Deborah L. Wetzel Practice Manager/Manager: Arturo Herrera Gutierrez Arturo Herrera Gutierrez Program Team Leader: Laura De Castro Zoratto Laura De Castro Zoratto ICR Team Leader: Daniel Ortega Nieto Daniel Ortega Nieto ICR Primary Authors: Courtney Price Ivins F. Results Framework Analysis Program Development Objective (from Project Document) To assist the Government of Amazonas to (i) improve tax administration and procurement procedures and (ii) strengthen the delivery of citizen security services and women’s access to justice and social services. Revised Program Development Objectives (if any, as approved by original approving authority) ii (a) PDO Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised approval Completion or Target Values documents) Target Years Indicator 1: Number of firms issuing the NFC-e (electronic invoice) Value 0.00 4,000 9,396 (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Target achieved and surpassed significantly. The Government of Amazonas requested to change the description of the indicator from ‘number of taxpayers registered at SEFAZ’ to ‘Number of firms issuing NFC-e’ to correctly reflect what Comments was being measured. The baseline and target values for the indicator were referring to (incl. % the actual number of firms issuing an electronic consumer receipt (Nota Fiscal do achievement) Consumidor Eletrônica, NFC-e) consistent with the reform supported by the Development Project Financing and not to the number of taxpayers registered at SEFAZ, the value of which was not zero in December 2013. Indicator 2: Percentage of procurement processes completed in less than 45 days. Value 29.20% 40% 66.6% (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments (incl. % Target achieved and surpassed significantly. achievement) Indicator 3: Percentage of value of purchases made through Price Registration Value 47.9% 55% 55.3% (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments Target achieved. (incl. % achievement) Indicator 4: Percentage of procurement plans of GoAM published on e-Compras.am Value 0% 80% 100% (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments (incl. % Target achieved and surpassed. achievement) Indicator 5: Number of services for which standard specifications for contracting are obligatory Value 9 19 21 (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 iii Comments (incl. % Target achieved and surpassed. achievement) Indicator 6: Percentage reduction in unitary cost of selected services Value 0% 10% 12.8% (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments Target achieved. Target refers to an average reduction for three of the Secretary of (incl. % Education services (school maintenance, painting, and air conditioning maintenance). achievement) Indicator 7: Number of government institutions adopting the new budget execution practices Value 22 60 49 (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 The target was achieved. The original 60 represented 100% of government Comments institutions. Due to organizational rearrangements that resulted from the economic (incl. % crisis, the total number of government institutions was reduced to 49; therefore, all of achievement) them have adopted the new budget execution practices. Indicator 8: Number of AISCs operating in the state Value 6 18 22 (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments (incl. % Target achieved and surpassed. achievement) Indicator 9: Number of performance indicators implemented Value 0 15 31 (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments (incl. % Target achieved and surpassed substantially. achievement) Indicator 10: Number of police trained Value 1,030 2,000 2,017 (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments (incl. % Target achieved. achievement) Number of disclosures per year of indicators of expenditures of AISCs in periodic Indicator 11: reports Value 0 4 7 (quantitative or iv qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments (incl. % Target achieved and surpassed. achievement) Number of communities reached with mobile units providing judicial and basic Indicator 12: services to women Value 2 60 47 (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments (incl. % About 78% of the target was achieved. achievement) Number of women victims of violence receiving judicial and social services at the Indicator 13: Casa da Mulher Brasileira Value 0 6,000 0 (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 The target was not achieved. The attainment of the indicator depended on funding Comments from the federal government to construct the Casa da Mulher Brasileira. However, (incl. % judicial and social services for women continued to be offered through the existing achievement) state institutions—over 23,000 women received services during this operation. Indicator 14: Number of implementers of the Maria da Penha Law receiving training Value 320 2,750 1,699 (quantitative or qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 About 62% of the target was achieved. The definition of implementer changed over the course of the operation. The definition of implementer was expanded to include both implementers in the traditional sense (that is, service providers such as social Comments workers, psychologists, policymakers, and the judiciary) and ‘multipliers’ (including (incl. % members of the general population, students, and so on). This report only considered achievement) those that fall under the traditional definition of implementer. Nevertheless, it is important to mention that over 24,000 additional people (multipliers) improved their awareness or became aware of the Maria da Penha Law. (b) Intermediate Outcome Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised Target approval Completion or Values documents) Target Years G. Ratings of Program Performance in ISRs v Date ISR Actual Disbursements No. DO IP Archived (US$, millions) 1 08/04/2014 Satisfactory Satisfactory 0.00 2 01/27/2015 Satisfactory Satisfactory 216.00 3 10/14/2015 Satisfactory Satisfactory 216.00 H. Restructuring (if any) Not Applicable vi ICR Background and Summary Brazil is going through the greatest recession in decades (2015–2016) while undergoing a major political crisis. The deterioration in both the external environment and domestic policies resulted in a deep-protracted crisis starting in 2014. Given the tightening fiscal situation, the State of Amazonas decided to embark on a reform program to increase efficiency and improve service delivery, particularly through reforming tax administration and procurement procedures, as well as service delivery in citizen security and women’s access to justice and social services. The single-tranche stand-alone Development Policy Loan (DPL) of US$216 million to the State of Amazonas was a direct and urgent request from both the Federal and State Governments, requiring preparation in a particularly short period of time. The operation complemented one by the Inter-American Development Bank (IADB) and filled a financing gap for which the IADB had insufficient resources. The rapid engagement was essential to ensure the success of the intended reforms and open an important space for policy dialogue. The World Bank expedited the preparation and approval processes (six months from concept review to Board approval) to enable the two operations to move forward simultaneously and to meet the fiscal deadlines of the Government of Amazonas (GoAM). In spite of the short period of time available to design the operation, the reform areas supported by the operation strengthened the state’s systems and yielded notable legacies. The World Bank went well beyond the Results Framework and was instrumental in helping the GoAM incorporate sophisticated information systems that collect and leverage data throughout the state for decision making in tax administration, procurement, and citizen security. The overall outcome of the operation is rated Satisfactory given the substantial results in a particularly challenging environment. Notable results in the tax administration component include increasing tax compliance in the retail sector, reducing tax compliance costs to businesses, and improving citizen engagement, fiscal transparency, and accountability. An illustration of the success of the reforms in this area is that, even though there was a 7.5 percent decrease in retail sales (the basis of most tax revenues), the GoAM increased revenues by 3 percent. Similarly, procurement reforms led to important gains in efficient procurement processes, government savings, and increased transparency. For example, in 2015, as a result of one of the measures under the DPL (that is, standard specifications for contracting), the GoAM saved R$91 million. In citizen security, there were significant results in efficiency and coordination between military and civil police, increased use of evidence for decision making, and improved transparency and community engagement. The community policing approach is now fully embedded in the GoAM’s citizen security policies, and coordinated actions between police forces are becoming the norm. Finally, the World Bank was critical in advocating for the advancement of important reforms in gender policies. The GoAM created the Executive Secretariat for Gender-related Policies (Secretaria Executiva de Políticas para Mulheres, SEPM) and, in spite of institutional rearrangements that affected the entire government structure, SEPM’s budget increased fivefold and SEPM is now also responsible for additional gender-related services that cut across different secretariats. 7 1. Program Context, Development Objectives, and Design 1. The single-tranche stand-alone DPL of US$216 million to the Brazilian State of Amazonas was designed to assist the State Government to improve the management of its financial resources, namely through tax administration and procurement procedures, and to strengthen service delivery, particularly citizen security and women’s access to justice and social services. 2. In the context of economic deceleration and growing fiscal constraints both nationally and locally, reforms were required to increase efficiency and improve service delivery. Since 2011, federal transfers and tax incentives to states and municipalities began to decline. The State Participation Fund (Fundo de Participação dos Estados e Distrito Federal), Amazonas’ second largest source of revenue, was reduced from an average of 15.3 percent of total revenues from 2009 to 2011 to 13.7 percent in 2012. Tax revenue also began to decrease. This is particularly true for taxes on industry, which alone was responsible for 40 percent of the state’s tax revenues. In this context, the DPL was fundamental to support the GoAM’s efforts to increase efficiency and improve service delivery. At the same time, and given the economic conditions, the GoAM also had to undertake important institutional rearrangements and tighten its budget. 3. The operation was a direct request from both the Federal Government and the GoAM and had to be prepared in a particularly short period of time. The World Bank was able to engage rapidly to fill a financing gap that was essential to the success of the reform program initiated by the State of Amazonas, while opening an important space for policy dialogue. The operation complemented one of the IADB, which was unable to provide all the required financing. The World Bank agreed to complement IADB’s operation, expediting approval processes to enable the two operations to move forward simultaneously. The operation was prepared in a short period of time (six months from concept review to Board approval) to meet the fiscal deadlines of the GoAM, as well as to continue the reform efforts that required additional funding. 4. The World Bank’s operation was designed to complement the IADB’s operation (Program of Consolidation of Fiscal Equilibrium for Social and Economic Development in the State of Amazonas, PROCONFIS AM). The IADB focused primarily on legal and regulatory aspects of the reform program and the World Bank on the implementation of policies. Diagnostics to identify specific needs and issues were conducted in partnership with the IADB. While the World Bank supported priorities for improving citizen security, it also agreed with the Government to add a critical component to improve women’s access to justice and social services to the reform program. The latter was instrumental in advocating for the advancement of important reforms that were elevated to political priorities. 1.1 Context at Appraisal Brazil’s Macroeconomic Context 5. At the time of the operation’s appraisal (March 2014), Brazil had achieved notable progress with regard to economic management, poverty reduction, and social indicators. The country’s gross domestic product (GDP) grew at an average of 4.0 percent from 2003 to 2010, slowing to 2.8 percent from 2011 to 2013. During this time, marked by large-scale job creation and implementation of social assistance programs, the share of Brazilians living below the national 8 poverty line of R$70 a month decreased from 10.8 percent in 2001 to 4.36 percent in 2012.1 While inequality in Brazil remained among the highest in the region, the Gini coefficient declined from 0.60 to 0.53 in the same period. 6. However, Brazil’s economic downturn revealed structural shortcomings, putting the country’s achievements at risk. With uncertainty in advanced economies, and slowing growth in major emerging markets with the end of the commodity super-cycle in 2011, Brazil’s revenues fell, leading the country to tighten its monetary and fiscal policies. Inflation reached 5.9 percent in 2013, increasing to 6.2 percent in March 2014. External instability also provoked an exchange rate depreciation and increased volatility. In sum, future growth in Brazil appeared to be limited by structural bottlenecks, making the spending commitments and expansion of the social security system increasingly unsustainable in the context of a declining growth potential.2 State of Amazonas’ Economic and Social Context 7. The incidence of poverty in Amazonas continues to be a pervasive problem. The state is sparsely populated with only 3.6 million inhabitants and had the fourth highest poverty and inequality rates in Brazil, and the highest in the northern region, at the time of appraisal. Of the total population, 17.8 percent were living below the poverty line, with 7.4 percent in extreme poverty. Amazonas also reduced poverty at a lower rate than the national average during Brazil’s economic boom (with rates declining from 22.6 percent in 2004 to 9.4 percent in 2012 nationally versus a 32.1 percent to 17.8 percent decline in Amazonas). Similarly, with regard to inequality, Amazonas attained a value of 13.1 percent for the Shared Prosperity Convergence Index in 2009, as compared to 18.3 percent nationally. 8. While the GDP of Amazonas grew faster than Brazil’s overall GDP since 2000, the state still contributes to a small portion of the overall national GDP (1.6 percent in 2010). GDP growth in Amazonas averaged 5.3 percent through 2010, compared to the national average of 3.6 percent. Growth helped boost current revenues by 7.2 percent annually in real terms between 2006 and 2012, totaling R$11.2 billion in 2012. Part of this growth was spurred by the advancement of the Manaus Special Economic Zone (Zona Franca de Manaus), established in 1960 as a free trade area with fiscal incentives supported by the Federal Government. The Manaus Special Economic Zone —industry sector—accounted for 42.3 percent of Amazonas’ GDP, on average, from 2007 to 2010. 9. The GoAM had made important progress in improving tax collection and expenditure management, but challenges remained.3 At the time of appraisal, state revenue from tax collection surpassed both national and regional rates, representing 56 percent of total revenue, compared to 40.6 percent at the regional level and 53 percent at the national level between 1 Data from Project Document - Pesquisa Nacional por Amostra de Domicílios (PNAD, national household survey) using the extreme poverty line of the Brasil sem Miséria program. 2 World Bank. 2016. Retaking the Path to Inclusion, Growth and Sustainability: Brazil Systematic Country Diagnostic. Washington DC: World Bank. http://documents.worldbank.org/curated/en/239741467991959045/pdf/106569-SCD-P151691-PUBLIC-non-board- version.pdf. p. xvi. 3 A debt sustainability analysis was carried out for the State of Amazonas, and it was also concluded that the GoAM expenditure program and fiscal arrangements with the Federal Government were adequate for the purpose of the operation. The GoAM was also compliant with the Fiscal Responsibility Law. 9 2006 and 2012. The Brazilian Value-Added Tax (Imposto sobre a Circulação de Mercadorias e Serviços, ICMS) generated the most, accounting for 52 percent of revenues in 2012. There were difficulties in enforcing tax collection in the transit of goods between states and in strengthening the effectiveness of and coordination between agencies dealing with tax crimes. Additionally, information on business transactions was not fully integrated, and paper receipts made it difficult for the Government to validate submissions, requiring costly tax audits. 10. Additional measures were also required to improve the efficiency of resource management and service delivery and to control recurrent expenditure. With a lack of standardized specifications for contracting outsourced services, compounded by weak supervision and execution of contracts, expenditures for the state’s outsourced services (for example, security, cleaning, and administrative services) increased by about 200 percent from R$598 million in 2004 to R$1.8 billion in 2012 (an average increase of 9 percent annually in real terms). Inefficiencies also arose from differentiated contracting criteria, producing divergent prices for identical or similar services. Inefficient processes also resulted in delays in payments, generating substantial liabilities to providers and high fines. In 2012, the average delay for payments was 45 days, which in the case of electricity services resulted in R$2.6 million in fines. 11. Homicide rates in northern Brazil have increased exponentially, more than doubling in Amazonas between 2000 and 2010, requiring significant additional investments by the state. Homicide rates increased from 19.8 per 100,000 inhabitants in 2000 to 30.6 in 2010. They remained high but relatively stable in the following years, at 31 per 100,000 inhabitants in 2011 and 30 in 2012. In the capital city of Manaus, violence has been particularly acute, reaching 56.2 homicides per 100,000 inhabitants in 2011, compared to the national average of 30. In 2011, the state also had the second highest robbery rate in the country (1,038 per 100,000 inhabitants). As a result, one of Amazonas’ primary budget expenditures at the time of the operation’s appraisal was citizen security. Between 2010 and 2012, expenditures in the sector increased by 31.8 percent in real terms. 12. The GoAM sought to introduce measures to address gender-based violence (GBV) and underlying issues of gender inequality in the state. According to the World Health Organization (2005), 29 percent of women ages 15 to 49 years in Brazil have reported experiencing physical or sexual violence at least once in their lifetimes, and Manaus had the third highest female homicide rate in Brazil, at 11.5 murders per 100,000 women in 2010. The GoAM established a technical committee to strengthen cross-sectoral cooperation in the implementation of the National Pact to Tackle Violence Against Women (Pacto Nacional pelo Enfrentamento à Violência contra a Mulher) and expanded the Network of Services for Women (Rede de Atendimento à Mulher). The GoAM also joined the federal program, ‘Women, Living Without Violence’ (Mulher, Viver sem Violência), which would involve the construction of centers called ‘Houses of the Brazilian Woman’ (Casas da Mulher Brasileira) in several state capitals throughout the country, providing a range of services for women and survivors of GBV. The federal program also provided two mobile unites that help provide services in areas with difficult to access such as those bordering with Colombia and Peru. 10 Rationale for World Bank Assistance 13. The single-tranche stand-alone DPL to the GoAM was designed to assist in improving the management of its financial resources (through tax administration and procurement procedures) and to strengthen service delivery (with a focus on citizen security and women’s access to justice and social services). The operation was a complement to and involved working closely with the IADB. With limited previous engagement with the State of Amazonas, the operation drew from World Bank experience from similar engagements in other Brazilian states, such as Rio de Janeiro, Pernambuco, and Bahia, especially in the areas of tax administration, citizen security, and GBV. Moreover, the operation was directly linked to support the objectives, priorities, and needs for implementing the Government’s Multiyear Plan (Plano Plurianual, PPA) (2012–2015). The operation was also consistent and closely aligned with the World Bank’s Country Partnership Strategy (2012–2015) for Brazil,4 which underscored the need to focus on second-generation development problems that require innovative solutions, with a focus on subnational governments. The gender component also aligned with broader efforts to scale up the World Bank’s portfolio for promoting gender equality in Brazil, while addressing pertinent gender issues affecting development outcomes in the state. 14. The World Bank was well positioned to support the GoAM’s priorities through the operation. This included experience supporting other Brazilian subnational governments in the implementation of reforms in tax and financial administration to strengthen fiscal capacity, spending efficiency, and public service delivery in a range of sectors. The focus on the citizen security sector was motivated both by the expressed priorities of the GoAM to improve service delivery in this area—as a fundamental contributor to citizen welfare and economic growth—and by the need to ensure security spending efficiency. While the World Bank assistance in the area of citizen security is relatively new compared to other sectors, relevant expertise and lessons were acquired through programs in other states (for example, Bahia and Pernambuco). The partnership of the World Bank with Amazonas also opened new opportunities and resulted in a subsequent DPL (P153203) with the Government of the Municipality of Manaus, which was interested in learning from and emulating the reforms of the State of Amazonas. 15. The operation was designed to contribute to the World Bank’s twin goals of ending extreme poverty and promoting shared prosperity. By improving the efficiency of revenue and expenditure management, the operation was designed to address spending pressures, which threatened the sustainability and improvement of the GoAM’s capacity for service delivery to its citizens. By fostering improvements in citizen security and women’s living conditions, the operation also sought to directly address key issues affecting development outcomes in the state. 1.2 Original Program Development Objectives (PDO) and Key Indicators (as approved) 16. As approved, the program development objectives (PDOs) of the operations were to assist the Government of Amazonas to (i) improve tax administration and procurement procedures and (ii) strengthen the delivery of citizen security services and women’s access to justice and social services. 4 World Bank. 2012. Country Partnership Strategy Brazil 2012–2015. Report Number 63731-BR. 11 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and Reasons/Justification 17. Not applicable 1.4 Original Policy Areas Supported by the Program (as Approved) Pillar I Policy Area I - Tax Administration 18. This policy area sought to increase compliance with the tax system through modernizing and simplifying tax administration systems and procedures. In particular, it aimed at increasing compliance of the ICMS—the primary tax revenue for the state—by improving methods for the collection of taxes in this area, as well as reducing bureaucracy for retailers and improved information on taxpayers. Actions addressed the low tax compliance arising, in particular, from the high cost, with regard to money and time, incurred by small businesses complying with the system, as well as the high costs of auditing through the gradual rollout of an Electronic Consumer Receipt (Nota Fiscal do Consumidor Electronica, NFC-e) across the state. These actions substituted the previous system of paper receipts, reducing the cost of compliance and decreasing the demand for in-depth audits through real-time and reliable information on collection and an improved taxpayer database. Furthermore, through a public awareness campaign, and a lottery through which those who request receipts can instantaneously be awarded cash and prizes, actions were also designed to boost voluntary compliance while improving the business environment.5 Policy Area II - Procurement 19. This policy area aimed to control expenditures and increase transparency and efficiency of government procurement. In particular, it sought to improve procurement planning, enhance contract management and savings from outsourced services, and reduce recurrent expenditures. Proposed actions supported the adoption of new policies, standards, and tools that would help plan and execute expenditures in a better way. These included a series of reforms, using a new procurement regulatory framework with a frameworks agreement modality, an online procurement platform, e-Compras.am, with publicly disclosed procurement plans to improve market information, standardizing technical specifications to establish reliable market reference prices for basic services, and a Contract Management System to register all contracts with outsourced services, linked to e-compras.am and the Integrated Financial Management System (Sistema de Administração Financeira Integrada). Pillar II Policy Area III - Citizen Security 20. Prior actions under this policy area aimed at improving efficiency, accountability, and transparency in the citizen security sector. As a response to the increasing crime trend in the state, 5 This is similar to the Nota Paulista in the State of Sao Paulo. For results see: Naritomi, Joana. 2013. “Consumers as Tax Auditors.” Harvard University. 12 the GoAM implemented the Ronda no Bairro (Neighborhood Watch) program, which consists of the implementation of joint actions by the civil and military police working in Integrated Citizen Security Areas (Áreas Integradas de Segurança Cidadã, AISCs). Actions aimed to improve coordination between the civil and military police and to mitigate duplication of efforts and misuse of resources. Efforts to improve coordination and security outcomes included an institutional reorganization and the introduction of a results-oriented approach and a georeferenced and integrated information system, as well as police force trainings. With respect to accountability and transparency, actions were taken to report indicators (security expenditures and crime rates) consistently and on time and make them available to the general public, together with broader civil society engagement. Policy Area IV - Women’s Access to Justice and Social Services 21. Actions under this policy area aimed at increasing women’s access to justice and social services, mainly through the coordination of support and protection services for victims of domestic violence, as well as training and increasing awareness in communities about gender issues and services for women. Activities included the creation of mobile units to reach women in distant communities and to strengthen the implementation of the Maria da Penha Law to combat violence against women. To operationalize these initiatives, the GoAM created SEPM, which is responsible for coordinating and implementing the state’s plan for gender-related policies. 2. Key Factors Affecting Implementation and Outcomes 2.1 Program Performance 22. The Board of Executive Directors approved the operation on May 23, 2014. All conditions in the Loan Agreement were met, and the operation became effective on October 1, 2014. The loan was quickly disbursed in a single tranche on October 3, 2014. Table 1. Tranche # Amount Expected Release Date Actual Release Date Release Tranche 1 $216 million 30-Sep-2014 03-Oct-2014 Regular Status as of Prior Actions Effectiveness Pillar I - Tax Administration and Procurement Prior Action #1: The Government has adopted the NFC-e to be applied on a (a) voluntary basis from Met July 1, 2013 until December 31, 2014 and (b) mandatory basis from January 1, 2015. Prior Action #2: The Government has improved its public procurement regulatory framework through the establishment of (a) the Integrated Management System of Procurement and Contracts (Sistema Integrado de Gestão de Compras e Contratos Met do Estado), which allows to manage public procurement procedures in an integrated manner, and (b) a price registration* modality aimed at recording in a systematic manner prices of goods and services procured by the state’s secretariats, agencies, and entities. Prior Action #3: The Government has established an electronic system (Sistema Eletrônico de Met Gestão de Contratos) for managing electronically the contracting of services 13 outsourced by the secretariats, agencies, and entities within the State Executive Branch (Poder Executivo Estadual). Prior Action #4: The Government has established a regulatory framework requiring specific budget allocations for essential recurrent expenditures (including, among others, Met electricity, telephone, water, and sanitation) incurred by the secretariats, agencies, and entities within the State Executive Branch (Poder Executivo Estadual) and restricting modifications of said budget allocations during a given budget period. Pillar II - Citizen Security Sector with a Focus on Gender-Based Violence Prior Action# 5: The Government has integrated the management of its civil and military police units through (a) the creation of AISCs in 12 of the state’s municipalities, (b) the establishment of integrated work routines aimed at coordinating the activities of said police units, (c) the adoption of operational procedures to integrate the Met activities of said police units and promote the exchange of information between them, and (d) the establishment of the Integrated Commission for the Systematization of Operational Procedures of the Civil and Military Police of Amazonas - Comissão Integrada para Sistematização dos Procedimentos Operacionais das Policias Civil e Militar do Amazonas (CISPO). Prior Action# 6: The Government has established a performance evaluation system (including, Met among others, indicators and objectives to be met) to monitor and measure the joint activities of its civil and military police units in the AISCs. Prior Action #7 The Government has regulated the public disclosure of expenditure and costs of its Met civil and military police units. Prior Action #8 The Government has established SEPM, which aims at adopting policies to improve women’s living conditions in the state’s territory through among others, Met (a) the coordination of support and protection services for victims of domestic violence and (b) the provision of training and the carrying out of awareness initiatives to communities regarding gender issues and services for women. Note: * Price registration is internationally known as framework agreements. 2.2 Major Factors Affecting Implementation Economic Recession and Political Crisis 23. The deterioration in both the external environment and domestic policies resulted in deep- protracted crisis starting in 2014. Brazil is going through the greatest recession in decades in 2015– 2016, while undergoing a major political crisis that has led to President Rousseff’s impeachment (August 2016). The recession in 2015–2016 was deepened by the impact of a difficult political environment and an unfavorable external environment. As a result of the crisis, there has been a considerable decrease in industrial production in the State of Amazonas, 17.2 percent in comparison to a decrease of 8.3 percent nationally in 2015. That same year, the State of Amazonas suffered a decrease of 12.1 percent in revenues, worse than the decrease in revenue during the 2008–2009 crisis (10.6 percent). As the rest of this report notes, the effects of the economic downturn had a negative impact on the implementation of certain policies: the GoAM had to make important organizational readjustments (for example, merge certain secretariats and close others) and readjust investments across all sectors, even in priority ones (for example, citizen security). The operation was also directly affected given that the Federal Government was not able to provide 14 the necessary funding for constructing the Casa da Mulher Brasileira.6 For a summary of the fiscal impact of the crisis on the State of Amazonas, see table 2. The political crisis did not have a direct impact on the operation. 24. As can be seen in table 2, current revenues fell by 10 percent in 2015 after reporting virtually no growth in the previous year. The slowdown in economic activity was the main reason behind this pattern, affecting both the national and subnational levels. ICMS collection, which accounts for 90 percent of tax revenues, dropped 14 percent in real terms. Meanwhile, transfers to the state contracted 8 percent. This impact was partly absorbed by cost reductions because of the reform program and cuts in investments. Payroll expenditures dropped 3 percent in real terms in 2015. Investments suffered a much larger cut, dropping 64 percent. The overall result was that although the GoAM was restructured, essential services were maintained and, most importantly, the state was able to meet all of its obligations, unlike some other states in Brazil (including Rio de Janeiro and Rio Grande do Sul), which have been unable to meet all wage payments as a result of the fiscal crisis. Moreover, the state was able to deliver a primary surplus in 2015, despite the weak performance of revenues. Table 2 R$ as of 2015 2011 2012 2013 2014 2015 Total Revenues 13,998.5 15,813.9 16,706.6 16,772.2 13,808.1 Current Revenues 13,408.1 14,595.7 14,883.5 14,918.7 13,361.8 Tax Revenues 8,455.8 8,953.3 9,518.7 9,461.9 8,195.0 ICMS 7,804.3 8,255.0 8,744.9 8,598.0 7,375.6 Transfers 3,635.0 3,798.9 4,285.6 4,322.7 3,989.6 Capital Revenues 590.4 1,218.2 1,823.1 1,853.5 446.3 Total Expenditure 13,166.4 14,278.6 16,171.2 16,527.4 14,200.7 Current Expenditure 11,126.5 12,402.8 13,313.7 14,029.7 13,001.5 Payroll 5,731.3 6,454.3 6,206.2 6,632.0 6,420.7 Interests 246.4 226.0 229.0 284.6 326.6 Capital Expenditure 2,039.9 1,875.7 2,857.5 2,497.7 1,199.2 Investments 1,602.0 1,413.5 2,323.9 2,113.8 758.1 Amortizations 328.5 425.2 435.9 366.4 389.3 Primary Revenues 13,567.1 14,878.3 15,716.5 15,648.6 13,767.1 Primary Expenditures 13,440.1 14,514.3 16,492.8 15,876.4 13,484.8 Primary Result 127.0 364.0 -776.3 -227.7 282.4 Source: GoAM. Government Commitment 25. The overall success of the development policy operation in triggering and supporting reforms was rooted in the strong commitment and ownership of the GoAM. In spite of the challenges that resulted from the economic crisis, the GoAM took all the necessary actions to achieve the PDO. Amazonas’ Secretariat of Finance (Secretaria da Fazenda do Amazonas, SEFAZ), responsible for the operation, provided strong leadership and a technically competent team. Their work included coordination and continuous supervision and engagement with the 6 However, judicial and social services for women continued to be offered through the existing state institutions— over 23,000 women received services during this operation. 15 sectoral government entities implementing the DPL-supported policies, namely Amazonas’ Secretariat of Planning (Secretaria de Planejamento), Secretariat of Citizen Security (Secretaria de Segurança Pública, SSP), and Secretariat for Gender Policies, within the State Secretariat for Justice, Human Rights, and Citizenship (Secretaria de Estado de Justiça, Direitos Humanos e Cidadania, SEJUSC). In addition, the continuity of government commitment was facilitated by a relatively smooth political transition following the 2014 gubernatorial election. As the former vice- governor, the current governor had participated in the negotiation process for the operation, maintaining a strong buy-in and general understanding of the reforms’ technical basis. The reforms also had a strong grounding in the state’s PPA. Institutional Rearrangements 26. The GoAM undertook major institutional rearrangements across its structure as part of a broad process of fiscal adjustment. These rearrangements had consequences for Pillar II: Citizen Security Sector with a Focus on Gender-Based Violence. In 2015, the Ronda no Bairro Secretariat was dismantled and replaced by a smaller unit—the Integrated Citizen Security Management Advisory Unit (Gabinete de Assessoramento da Gestão Integrada de Seguridad Cidadã, GAGIS). GAGIS, however, did maintain the essential functions of the Ronda no Bairro Secretariat and, according to the GoAM, the adjustment led to savings of 52 percent on personnel costs. The gender component was also affected. SEPM was integrated into the structure of SEJUSC. SEPM had been reporting directly to the governor, and the move involved three consecutive changes in leadership, which, compounded by the reduction in personnel, affected institutional memory. SEPM, however, maintained its status of Executive Secretariat, and the shift also came with additional responsibilities as well as financial resources. SEPM’s budget increased fivefold and it is now also responsible for additional gender-related services (Serviços de Atenção a Mulher) such as overseeing the major centers that provide services for GBV victims. Soundness of Background Analysis 27. The policy reforms undertaken by the GoAM with support from the operation were informed largely by existing analytical work and lessons from other Brazilian states and operations. They also relied on some original analysis specific to the operation. The Poverty and Social Impact Assessment (PSIA) was developed, providing information on the broader context and needs. Two technical notes were also produced on procurement and tax administration by the GoAM. Additionally, the World Bank commissioned a risk assessment on the security sector, informed by the prior analysis of the IADB and also supported by field consultants who had been accompanying the state reforms. For the gender component, analysis was informed by previous reports that were countrywide in scope: Towards a More Comprehensive Domestic Violence Policy in Brazil (World Bank 2013) and Brazil Gender and Social Inclusion Non-Lending Technical Assistance (P132325). In addition, the gender analysis was included in the PSIA. Relevance of Risks Identified at Appraisal and Effectiveness of Mitigation Measures Political and Governance Issues 28. The reelection of the governing coalition in 2014 contributed favorably to the continuation of the Government’s commitment to the program. However, changes in leadership within sectors 16 (as mentioned in the previous section) did create challenges during implementation. In this respect, follow-up by the World Bank and constant dialogue with the GoAM proved helpful in maintaining buy-in and institutional memory. Another risk involved the electoral cycle and the complexities of local political relations. For example, the rounds of the mobile units providing services to women were suspended during municipal election campaign periods to avoid their misuse for political purposes. Similarly, the Ronda no Bairro program systematized administrative processes for public participation, through Community Security Councils (Conselhos Interativos Comunitários Segurança Pública - CONSEGs). However, the frequency of the meetings had to be reduced in an attempt to quell their potential politicization in the run-up to local elections. Sector Policies and Stakeholders 29. For the citizen security component, a risk assessment of the Ronda no Bairro program was conducted building on an additional assessment conducted by the IADB. Informed by the assessment, the appraisal deemed the operation to be compliant with the World Bank’s legal note for engagement in the criminal justice sector. During implementation, some sections of the police did resist the reforms, particularly in establishing and linking criteria for monitoring and evaluation (M&E) to performance. There was reported pressure from within the police corporations to review the System of Performance Evaluation, which, they argued, had become too complex and that some indicators were deemed irrelevant for their work (for example, number of CONSEG meetings and minutes). This potential resistance had been previously identified and was mitigated through the enactment of laws and decrees (for example, Executive Decrees No. 23/2014 and No. 24/2014) institutionalizing the reforms. At the time this report was developed, the SSP had established a committee to discuss this issue and streamline procedures. Operational Design, Implementation, and Sustainability 30. Close supervision and continuous communication with the GoAM helped address the identified risks in policy coordination with institutions involved in the operation. The risk materialized visibly in the gender component, for which the incorporation of an indicator linked to a federally funded program was established (Casa da Mulher Brasileira) for service provision to survivors of GBV. As the fiscal situation in Brazil worsened, it was clear that the funding to enable this intervention would not move forward, independent of any action that the State Government could pursue to advance it. Steadfast in its commitment to address GBV, the GoAM continued providing services to women through its existing networks and facilities. Macroeconomic Risks at the Country Level 31. Brazil’s macroeconomic environment continued to deteriorate during the operation, as weaker-than-expected growth and inflationary pressures were compounded by the context of political crisis. In this context, the country’s budget revenues fell from 35.6 percent of the GDP in 2013 to 34 percent in 2014,7 and in 2015, the Federal Government’s revenue fell by 5.6 percent;8 7 World Bank. 2016. Brazil Systematic Country Diagnostic. Report Number 101431-BR. 8 Source: Secretaria da Receita Federal, Análise da Arrecadação 2015 . http://idg.receita.fazenda.gov.br/dados/receitadata/arrecadacao/relatorios-do-resultado-da-arrecadacao/arrecadacao- 2015/dezembro2015/apresentacao-arrecadacao-dez-2015.pdf 17 all of which has led to a fall in transfers to states. Pillar I has therefore become increasingly relevant in this context by enabling more responsible fiscal management. 2.3 Monitoring and Evaluation (M&E) Design, Implementation, and Utilization M&E Rating: Substantial 32. The progress of the operation was monitored principally through results indicators linked to each of the prior actions. Out of fourteen indicators that tracked the desired progress toward PDOs, two indicators lacked clarity9 and the achievement of another one was not fully within the control of the GoAM.10 Throughout the implementation period, SEFAZ was responsible for the data collection and consolidation of information, and for keeping the World Bank informed. It did so consistently. 33. The technical support incorporated into the content of the operation strengthened the state’s M&E systems and yielded notable legacies. The World Bank carried out regular implementation support missions to supervise progress, while also providing consistent remote support. This was particularly extensive for Pillar I, for which the World Bank went well beyond the Results Framework (see Assessment of Outcomes section for details). The World Bank also supported the GoAM to establish monitoring dashboards through a comprehensive procurement data analysis and to provide a tracking mechanism for a range of indicators integrated with the state’s broader procurement systems. The procurement monitoring dashboards have been integrated in institutional processes to inform decision making, and it is fair to say that few Governments have such sophisticated tools at their disposal. 34. In addition to Pillar I, the security component also incorporated data into its decision- making processes. The implementation of a performance-based indicators system to measure police performance, as well as the establishment of a georeferenced and integrated information system, has also been relevant for decision making. The Implementation Completion and Results Report (ICR) team interviewed officers that explained how they make tactical decisions based on data from the information system (for example, deploy teams to areas prone to crime and violence). Furthermore, the GoAM also established the Center for Studies and Applied Research in Public Safety (Núcleo de Estudos e Pesquisas Aplicadas em Segurança, NEPAS) in 2015. NEPAS is part of the push toward improving the analytical capacity and use of evidence throughout the work of the SSP. For the gender component, SEPM also consistently documents the number of women receiving services as well as the locations where they do so. 2.4 Expected Next Phase/Follow-up Operation 35. Despite being a single-tranche operation, there have been a number of actions that will provide continuity to the operation. The World Bank is conducting an impact evaluation on a 9 The indicator on ‘Percentage of procurement processes completed in less than 45 days’ could have benefited from further segmentation. The indicator on ‘Number of implementers of the Maria da Penha Law receiving training ’ should have clearly defined the term implementers. 10 With respect to the indicator ‘Number of women victims of violence receiving judicial and social services at the Casa da Mulher Brasileira,’ the construction of the Casa da Mulher Brasileira was dependent on Federal Government funding. 18 procurement reform, which will be introduced in 2017, and preliminary results will be available by mid-2017, for potential course corrections. 11 Furthermore, this operation opened new opportunities, namely a DPL with the Government of the Municipality of Manaus (P153203), which has led to an unprecedented partnership between the GoAM and the city of Manaus in launching the Nota Facil program that will allow for service coverage across more than 10,000 businesses.12 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design, and Implementation Relevance of Objectives: High Relevance of Design and Implementation: Substantial 36. The objectives of the operation continue to be highly relevant. Improving service delivery and strengthening state capacity continue to be integral parts of the strategic priorities of the GoAM in its new PPA 2016–2019. Improving planning and budgeting as well as expanding tax collection and increasing compliance with the tax system are particularly critical given the current slowdown of economic activity in Amazonas and across the country. Establishing the basis of an efficient fiscal system and procurement processes is fundamental for expanding the GoAM capacity in public service delivery. Moreover, as the next section will highlight, the establishment of the NFC-e reduced the financial burden that small businesses faced to comply with the requirements for issuing consumers’ receipts. 37. With respect to Pillar II, evidence also shows that crime and violence have a substantial negative impact on development. In Brazil, direct and indirect social costs, measured by health expenditures due to external causes, have been estimated at approximately 5 percent of the country’s GDP (Lima et al 2016). At the regional level, other studies have also shown that crime and violence have significant direct and indirect social costs through, for example, property loss; costs of public and private security; expenditures of the judicial system and health systems; opportunity costs of perpetrators; and loss of years of productive life of victims, to name a few (Jaitman 2015). Finally, the World Bank’s Systematic Country Diagnostic 2016 also highlights the importance of fiscal management and procurement as fundamental areas for improvement, as well as citizen security and GBV as critical obstacles for the country’s development. 11 The impact evaluation design is available at: http://operationsportal2.worldbank.org/wb/opsportal/ttw/about?projId=P156357. The GoAM will introduce an e- Documents platform to lower transaction costs for participating firms, which will facilitate the submission of a wide range of documents required for each bidding process. The evaluation focuses on the effects of transaction costs on the number and types of firms that participate in public procurement, the types of firms that win contracts, and the cost-effectiveness and quality of government procurement. 12 Nota Facil, a partnership between SEFAZ and the Municipal Finance Secretary, will enable the expansion of the NFC-e in the service sector in Manaus. 19 38. The design of the operation was appropriate, and the World Bank’s implementation was adequate in support of the GoAM objectives. The World Bank’s capacity for rapid engagement filled a financing gap that was essential to advance the intended impacts of the reforms, while opening an important space for policy dialogue around topics of both state and federal concern. The rapid response also permitted the positive momentum built by IADB’s engagement to continue. When the IADB was unable to finance the operation in its entirety, and at the request of the Federal and State Governments, the World Bank agreed to complement the operation, expediting approval processes to enable the two operations to move forward simultaneously. While the IADB’s contributions focused primarily on legal and regulatory aspects of fiscal reforms, the World Bank focused more heavily on the implementation of policies. In developing the operation’s design, the World Bank team was also instrumental in advocating for the inclusion of an additional complementary gender component, enabling the advancement of important reforms, which may not otherwise have been elevated to political priorities13. 3.2 Achievement of Program Development Objectives 39. As approved, the PDO was to assist the Government of Amazonas to (i) improve tax administration and procurement procedures and (ii) strengthen the delivery of citizen security services and women’s access to justice and social services. 40. See annex 4 for additional results by policy area. Pillar I Policy Area 1 - Tax Administration 41. The operation supported improvements in tax administration by increasing compliance with the tax system. The target indicator defined by the GoAM and the World Bank for this policy area was surpassed significantly, more than doubling the expected number of firms issuing the NFC-e.14 There were at least three main benefits that resulted from this policy area: (a) increased tax compliance in the retail sector; (b) reduced tax compliance costs for businesses; and (c) improved citizen engagement, fiscal transparency, and accountability. Accordingly, the efficacy of achievement of this policy area’s objective is rated High. 42. Increasing tax compliance in the retail sector. In spite of the current economic situation, which has produced a slowdown of economic activity, the policy actions implemented under this policy area were able to expand the tax base in this sector. According to SEFAZ, before the NFC- e, there were 2,744 businesses issuing receipts. As a result of the policy efforts and reforms, there are currently over 12,166 businesses issuing invoices. This represents an increase of 340 percent 13 The IADB’s operation has been assessed to be successful. See Programa De Consolidação Do Equilíbrio Fiscal Para Melhoria Da Prestação De Serviços Públicos Do Estado Do Amazonas – PROCONFIS/AM (BR-L1385 ; 3139/OC- BR) Relatório de Encerramento do Projeto. 14 It is worth highlighting that the effort undertaken by the GoAM to implement such a sophisticated system (NFC- e) is exceptional in Brazil, where only states with more capacity and resources such as Sao Paulo and Rio de Janeiro have been able to adopt such policies. 20 of regularized businesses. There was also a notable increase in the volume of invoices that resulted from the reform: in December 2015—with increased economic activity—the GoAM had a record of 18 million electronic invoices issued. As a result of the NFC-e, in 2015, the GoAM was able to collect R$905 million through the ICMS in spite of a 7.5 percent decrease in retail sales.15 43. Reducing tax compliance costs to businesses. The financial burden faced to comply with the requirements for issuing consumer receipts was reduced significantly. This was achieved in two ways. First, by important savings in eliminating time-consuming processes, which formerly included a number of bureaucratic steps that could lead to waiting times of up to 30 days for SEFAZ’s approval. This happened because businesses had to acquire equipment to generate the tax receipts from an accredited supplier of SEFAZ. With the NFC-e, the process became automatic and immediate, without the need for a specific equipment from an accredited supplier. Second, the equipment and the application required for businesses before the NFC-e represented an average total cost of R$3,000. With the reform, SEFAZ partnered with several commercial associations that produced free software for issuing the NFC-e and, therefore, any business can now issue invoices as long as they have a printer (which sell for R$300 onward). 44. Improving citizen engagement, fiscal transparency, and accountability. The GoAM launched a fiscal education campaign to incentivize consumers to ask for invoices to increase business accountability and fiscal transparency. The campaign was based on normative persuasion (that is, encouraging ‘fiscal citizenship’), to encourage consumers to participate in achieving the program’s objectives by requesting their receipts from retailers. After providing their National Identification Number in connection with their requested receipts, a consumer automatically participated in lotteries for immediate cash and prizes. Consumers could also select previously certified civil society organizations working in social areas that could also win prizes. The campaign became part of a law and had a budget of R$6 million for 2016. Following the campaign, which included radio, television, newspapers, social networks, and other forms of media, the volume of receipts submitted increased by 26 percent. In February 2016, there were 190,346 registered consumers, and over 16,000 prizes had been given away. The campaign was also an important vehicle to disseminate relevant information about tax obligations and processes among the population of Amazonas. 45. The new tax system also allows consumers to review online, in real-time, all their transactions. They can also issue complaints and denounce noncompliance by businesses or suspected illegal activities. Before the campaign, SEFAZ received an average of 240 complaints a year. After the NFC-e was put in place, they have registered over 5,000 complaints, out of which 3,260 have resulted in investigations (65.2 percent). In fact, as a result of the information collected through NFC-e, the Government was able to identify 3,549 fiscal infractions in 2015 worth R$206 million. With the new electronic system, real-time consumer data is securely saved in a new data center at SEFAZ, increasing capacity for effective controls and decreasing the need for costly audits through an improved triage. 15 The GoAM collected R$879 million in 2013, before the economic crisis. 21 Policy Area II - Procurement 46. The operation supported improvements in procurement procedures by controlling expenditures and increasing transparency and efficiency of government procurement. All target indicators were achieved, and there are at least three relevant outcomes that derive from this policy area: (a) efficient procurement processes, (b) government savings, and (c) increased transparency. Accordingly, the efficacy of achievement of this policy area’s objective is rated High. 47. Efficient procurement processes. The GoAM has been able to improve efficiency in procurement processes through a number of measures supported by the DPL. As the first indicator of Policy Area II shows, the percentage of procurement processes completed in less than 45 days has increased from 29.20 percent to 66 percent.16 Additionally, the GoAM introduced a process through which certain items need to be inspected through samples, with the aim of improving the quality of items purchased. To make processes more agile, SEFAZ also created a subcommittee for procurement processes in the health sector—the area with the highest volume of contracts— within the Procurement General Committee. While both SEFAZ and the World Bank acknowledge that the decrease in the volume of processes is partly due to the fiscal tightening in 2015, which led to significant improvements in the indicator due to reduced demands on the Procurement General Committee, it is important to note that a number of processes and regulations have been put in place that have made the system more efficient. Among them is the Contract Management System, which registers all contracts between outsourced service providers and the Government. This system, together with the Integrated Financial Management System, enables timely monitoring of contract execution, prices, and payments. The GoAM also worked closely with the World Bank to engage in an innovative survey with more than 150 suppliers, with the objective of identifying the main challenges faced by them when selling to the state and to improve processes. 48. Additionally, the GoAM was able to automate the process for estimating costs for procurement of goods by using an innovative database that collects pricing information in both the public and private sectors. This in itself is a major step in improving efficiency and unfortunately not widely used across Brazil. As a result, procurement of goods in Amazonas is faster when the entire cycle is measured, starting from the request to purchase up until an award is made, and the control of awarded prices is also more effective. The key improvement to the process is the utilization of the electronic invoice data (NFC-e) to come up with estimates—which also shows how SEFAZ built synergies between the DPL components. The use of big data from the NFC-e is quite innovative because it captures all commercial transactions in the state, thereby having prices even for transactions between private parties. 49. The GoAM also adopted measures, such as electronic signatures, to process purchases electronically and move away from in-person processes that significantly increase processing times. The GoAM was able to increase the value of purchases made through framework agreements, which also helps reduce the duration and inefficiency of procurement processes, while systematically keeping records about prices of goods procured. According to the GoAM, framework agreements allowed the Government to reduce the number of procurement transactions materially, lower prices, and improve quality. One framework agreement can replace hundreds of 16 This indicator could have benefited from segmentation that takes into account the size and complexity of purchased goods given that these also affect completion times. 22 individual procurement transactions, and smaller government agencies, for example, do not have to worry about procurement for a required item. As a result, resources that would be used to do procurement can be used elsewhere. Moreover, framework agreements also enabled the creation of annual purchasing plans. 50. Government savings. The Results Framework indicator focused on the Secretary of Education, as indicative of how the establishment of the electronic system for managing the contracting of services outsourced by the state’s executive branch could lead to savings. The results show that there was a 12.8 percent average reduction in unitary costs of selected services, namely school maintenance services, painting, and maintenance of air conditioning. According to the GoAM, savings are greater considering all secretariats and services to which the system was applied. The measures that resulted from the DPL, combined with the urgent need to reduce expenditures, have led to a reduction of approximately 20 percent in public expenditures in areas such as cleaning, security, and leasing of vehicles. In 2015, savings from measures under the DPL accounted for R$213 million, out of which R$91 million resulted from measures directly related to standardized services, and R$122 million were a result of less services and quantities being required as well as substitution of services. Also, through the establishment of a framework— Integrated Management System of Procurement and Contracts—the GoAM introduced mandatory procurement planning, as well as regulations for electronic purchasing. This initiative encouraged the use of framework agreements to enable the Government to take advantage of economies of scale through bulk purchasing of goods and services. The percentage of purchases made through this modality was 47.9 percent and is now 55.3 percent. According to the GoAM, the use of framework agreements resulted in average savings of 23 percent when compared to other purchasing mechanisms. 51. Increased transparency. A significant outcome of these efforts is the public access of information on government procurement. The GoAM established the e-compras.am procurement portal, which concentrates and discloses all procurement information to the public. All government entities are now required to use the portal, including to publish their procurement plans. As a result, citizens can access procurement information online, providing a platform for enhanced accountability. Businesses can also identify bidding opportunities, available framework agreements, and the results of bidding processes, as well as all the necessary manuals and legislation to inform them on required processes. Pillar II Policy Area III - Citizen Security 52. The operation supported the strengthening of citizen security service delivery by improving efficiency, accountability, and transparency in the sector. All relevant target indicators were achieved. Key outcomes that resulted include (a) improved efficiency and coordination, (b) increased use of evidence for decision making, and (c) improved transparency and community engagement. Accordingly, the efficacy of achievement of this policy area’s objective is rated Substantial. 53. Improved efficiency and coordination. This was achieved by integrating resources and actions of civil and military police and the implementation of results-based management to 23 improve performance. The expansion of the Ronda no Bairro program resulted in the integration of operations of military and civil police through the establishment of 30 Integrated Police Units (Delegacias Integradas de Polícia, DIPs). This is a particularly significant effort given the duplication of efforts, inefficient use of resources, and sometimes unclear roles and responsibilities of the two forces, as well as confusion within the population with respect to their mandates. The new integrated operations are spread throughout the 22 AISCs, of which 6 are in Manaus, 5 are in the Manaus metropolitan area, and 11 are in the interior. This distribution accounts for about 90 percent of the total population and areas where the vast majority of violent crimes occur (94 percent are registered in Manaus). The Government would like, however, to eventually reach all municipalities but was unable to do so given the economic downturn. 54. The training program in community policing methodology was particularly useful for helping police integration. A total of 2,017 officers were trained, which represents approximately 18 percent of the total workforce of both military and civil police. The GoAM claims that there was a significant cultural change because of this type of training and integrated work. The ICR team met with a number of police officers who explained that the training was an important arena where they met their counterparts from the different police forces, creating a space to exchange perspectives and improve understanding of roles, responsibilities, and tactics. Moreover, the training had a gender component, which was also important to sensitize participants to GBV and which created synergies with the services provided under Policy Area IV. 55. Increased use of evidence for decision making. The DPL also helped incorporate 31 performance indicators within the SSP through a System of Performance Evaluation, which are discussed on a monthly basis. This surpassed the original target of 15 indicators, and, according to GAGIS, the establishment of these indicators has had a fundamental impact in policy and operations.17 Furthermore, the GoAM invested in information and technology systems that provide relevant data to monitor operations and plan interventions, including the establishment of a georeferenced and integrated information system managed by the Commission on Geoinformation and Geographic Intelligence. This system provides real-time georeferenced data for all crime- related activities, enabling faster deployment of vehicles and officers and also allowing for better monitoring of all SSP data. This georeferenced data, according to police officers interviewed during a field visit, allows them to deploy teams to areas prone to crime and violence. It also enables them to conduct special interventions and track occurrences across the AISCs. The GoAM also established NEPAS in 2015 within the GAGIS and the Integrated Institute of Public Safety Education, part of the push toward improving the analytical capacity and use of evidence of the SSP. NEPAS has directly contributed to development of the conceptual and organizational basis for the implementation of a new pilot program that builds upon the successes of Ronda no Bairro— All for Life (Todos Pela Vida). 56. Improved transparency and community engagement. As a result of the DPL, the GoAM is now publishing seven indicators with data online, on crimes across the AISCs, police activity, as well as security costs on a monthly basis. This is certainly a step in the right direction and, at the time of the evaluation, the SSP was working on a communications strategy to encourage people to download the data, focusing on different target groups (for example, university students). It is worth highlighting that the engagement with the community improved through the promotion of a 17 See the section on Key Factors Affecting Implementation and Outcomes for details on pushback. 24 closer engagement of civil society, by following a strong community policing approach and through the establishment of periodic community meetings in all AISCs with the local police commanders through CONSEGs and by providing relevant and timely information to the general public through an open information system that publicizes periodic reports on all state crime- related data. The leadership of the local captain played a significant role in the effectiveness of the community meetings. In one DIP area, for example, the local commander established a WhatsApp group with the community by which residents could report on issues, as well as follow up on agreements made at the meetings. In other communities, the intensity of interactions may have been less significant. Also, as previously mentioned, the frequency of the meetings decreased considerably during the electoral cycle, in an effort by police offices to ensure the meetings were not politicized. 57. It is worth noting that throughout the operation’s implementation, new challenges emerged that led to setbacks in crime indicators and an increase in homicide rates since 2014. By the end of 2015, Amazonas had a rate of 31 homicides per 100,000 inhabitants. The Government has yet to establish specific causes for this gradual increase.18 It is certainly not possible to assess what would have happened without Ronda no Bairro; however, it is likely that the crime rate could have been worse without it. Some evidence does suggest that this could have been the case. In the 2000s, violent crime rates increased significantly in the northeastern and northern regions of Brazil, with rates climbing from 18 to 28.8 per 100,000 and from 22.50 to 32.97, respectively, between 2003 and 2008 (World Bank 2012). While crime rates started declining in Amazonas in 2012, after the implementation of Ronda no Bairro, the same trend was not found in some neighboring states, where homicide rates increased. At the same time, crime rates in the state of Pernambuco, which has a similar policy approach to Ronda no Bairro in Amazonas, declined by 9.7 percent over the same period (Forum Brasileiro de Segurança Pública 2014). Policy Area IV - Women’s Access to Justice and Social Services 58. The operation supported the strengthening of women’s access to justice and social services by seeking to increase access. The indicators for this component were partially achieved or not achieved. This is, however, mostly the result of events that the GoAM could not influence and that were directly linked to the component’s indicators (the transfer of federal resources to construct the Casa da Mulher Brasileira and SEPM’s inability to use its mobile units during the electoral cycle). Nevertheless, there were important improvements with regard to access to justice and social services that go beyond the Results Framework (over 23,000 women had access to services). Furthermore, the fact that the GoAM was able to bring gender issues to the forefront of its agenda is particularly relevant given the prevalence of GBV in the state. Accordingly, the efficacy of achievement of this policy area’s objective is rated Modest. 59. The attainment of the Casa da Mulher Brasileira indicator depended on investments/transfers from the Federal Government, and therefore, its achievement was outside the State of Amazonas’ control. The Casa da Mulher Brasileira was intended to be a state program for integrated service provision to women, implemented in partnership with the Federal and 18 Different sources commented that the GoAM’s new focus on incarcerating the leaders of drug -trafficking factions could be part of the cause of the increase in the number of homicides (due to the resulting splintering and power disputes within and among factions). 25 Municipal Governments. The Municipality of Manaus did concede the land required for the project but the Federal Government did not provide the funding. It was expected that 6,000 survivors of GBV would receive services in the new facility. In spite of not receiving the federal funds, the GoAM continued to provide services to this target population through existing channels; therefore, victims were not left unattended. Throughout the operation, according to the GoAM, these services reached over 23,000 women. 60. An important outcome was the improved coordination in the provision of services to victims of GBV in the DIPs, which created awareness of gender issues among officers through specialized trainings conducted under Policy Area III, and by providing survivors with immediate and targeted assistance, generally in situ. Additionally, the Ronda Maria da Penha project was established as a program that focuses on preventing recidivism of domestic and violence against women. The project, which is a new initiative and was not envisioned under the operation, follows a successful model implemented in Rio Grande do Sul, but adds some innovative elements, such as the ‘purple’ vehicle and the branded uniforms to distinguish the specialized police unit (which has allowed them to gain confidence in hotspot areas where women might otherwise feel intimidated to talk to the police) and the strong coordination with other key agencies and other relevant stakeholders (for example, social workers, representatives from the Secretary of Gender Policies, and youth). It also goes beyond the establishment of protective measures by providing a close interaction and presence of a specialized police unit in the areas where victims of domestic violence reside. Furthermore, since the program is led by security forces, it gained respect from the community and improved the enforcement of legal measures. The program has shown positive results in its first months of implementation, and the Government wants to expand it to other areas. However, given the financial constraints affecting Amazonas and the country, this has not yet been possible. 61. SEPM has been proactive in reaching out to civil society organizations through statewide conferences with the aim of creating awareness of gender issues. SEPM has also been able to influence other policy areas through direct engagements that has resulted in specific programs. For example, in addition to the abovementioned work in citizen security, it has been working with the Health Secretariat to secure specialized assistance to women that have been victims of sexual violence and with municipalities to directly confront human trafficking. SEPM was able to improve access to justice and social services in hard-to-reach areas of the state where these services are mostly unavailable. For example, in 2015, SEPM supported 4,486 women through its program, Itinerant Citizenship: Woman in the Field, Forest, and Waters program (Cidadania Itinerante a Mulher do Campo, da Floresta e das Aguas). Finally, as a result of its incorporation to SEJUSC, it will be able to build additional synergies and reduce costs through joint activities such as missions to remote locations. 3.3 Justification of Overall Outcome Rating Table 3. Summary of Relevance and Efficacy Ratings Relevance Efficacy Overall Design and Tax Citizen Objectives Procurement Gender Outcome Implementation Administration Security High Substantial High High Substantial Modest Satisfactory Substantial Substantial 26 62. The overall outcome rating for the DPL is Satisfactory. The relevance and efficacy of the DPL are rated Substantial. The operation supported important reforms and is well aligned with the Government of Brazil, the Government of Amazonas, and the World Bank priorities. Moreover, given the economic and political context in which the operation took place, the results achieved are particularly significant because they set the pillars for long-term improvements in the management of financial resources (tax administration and procurement procedures) and service delivery (citizen security and women’s access to justice and social services). The operation benefited from an active and technically sound implementing agency and a World Bank team that was fully engaged throughout the entire process, in spite of some minor shortcomings in a couple of the M&E indicators. 3.4 Overarching Themes, Other Outcomes, and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 63. The time frame of the DPL was too short for the operation to have identifiable impact on poverty and social development. Nevertheless, the operation’s PSIA showed that there can be positive distributional effects on poor and vulnerable groups through actions under both Pillar I and Pillar II. The results in savings and increased tax efficiency can eventually translate into more resources for public services that attend some of the basic needs of these groups. Moreover, while improved citizen security benefits the entire population, the PSIA argues that poor women and male Afro-descendant youth may have additional benefits because they are overrepresented among the victims of domestic violence and homicides, respectively. Additionally, it is likely that the actions under the DPL will foster social accountability by publicly disclosing criminality rates and police activities, as well as public procurement data. Finally, Policy Area IV can also become the stepping stone for improving services to women. (b) Institutional Change/Strengthening 64. Section 3.2 has referred to the institutional changes that emerge from the DPL, most notably in SEFAZ through the introduction of the NFC-e and the multiple frameworks and processes to improve procurement. The use of data and the focus on community policing to improve citizen security has already had an impact on the most relevant institutions in the sector, and SEPM is making use of its new functions and increased budget to promote gender issues across the state in partnership with multiple secretariats and municipalities. (c) Other Unintended Outcomes and Impacts 65. The Assessment of Outcomes Section already referred to the numerous actions and policies that the GoAM undertook that went beyond the DPL. There are, however, additional unintended outcomes. For example, it is important to recall that with the previous tax collection system, tax audits were costly, partly due to the difficulty of cross-referencing information between receipts and transactions. With the upcoming launch of the Malha Fiscal, a real-time information system that uses data from electronic receipts and cross-checks information to identify inconsistencies, resources will be deployed to undertake in-depth audits. Finally, the overall performance of the 27 World Bank in this operation opened new opportunities, namely it resulted in a demand for another DPL with the Government of the Municipality of Manaus (P153203). 4. Assessment of Risk to Development Outcome Rating: Moderate 66. Most of the policies and reforms implemented appear to be well established, in particular, with regard to tax and procurement reforms, which have gained increasing relevance in the current context of fiscal constraints. For the tax and procurement components, the reforms introducing NFC-e and procurement measures are well consolidated, demonstrating an impact on tax revenue and savings that would make their discontinuation unlikely. A minor but potential risk with this intervention would be maintaining the momentum built through the NFC-e public awareness campaign, identified as a major factor in increasing participation. Similarly, the efficiency gains demonstrated by the procurement reforms contribute to reduce their risk of being dismantled, supported by the fact that, as the Project Implementation Unit surmised, they would be ‘too complex for a political decision to reverse.’ Both the tax administration and procurement policy areas have also attracted attention for their success, enabling exchanges with other states and with the Municipality of Manaus, further decreasing the risk of reversal. 67. Budget constraints can affect the consolidation of the new integrated community policing. The Todos Pela Vida program, which will serve to strengthen some of the reforms achieved through the Ronda no Bairro program, faces some risks with regard to its capacity to integrate a wide range of sectors and institutions. The most pressing current risk, however, is that the Todos Pela Vida pilot will not move forward due to the current fiscal situation. Its launch has been postponed due to budget cuts. With respect to gender, SEPM’s transition to SEJUSC may ultimately be beneficial in the context of resource constraints—given that its own budget increased substantially—by enabling better harmonization of efforts and beneficiary targeting. However, outreach of service delivery to the state’s remote communities remains a challenge due to the high costs of mobilizing personnel across the vast and sometimes difficult-to-access territory. 68. At the time of writing this report, there is an ongoing judicial process against Governor Melo in the Electoral Court. The process charges him with illegitimate use of the military police and purchasing votes to support his 2014 campaign. If found guilty, he would lose his mandate and be ineligible for public office for eight years. While this process did not affect the operation, it is still premature to assess the extent to which the final sentence can affect the sustainability of some reforms. Nevertheless, as has been explained in this section, most reforms have been locked- in through the legal arrangements and legislation that resulted from the prior actions leading to this operation, as well as through the implementation of the reforms. Furthermore, the reforms are central priorities in the state’s PPA 2016–2019, which is supported by multiple stakeholders. 28 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Satisfactory 69. The World Bank was able to prepare a solid operation in a particularly short period of time with the goal of allowing the GoAM to build on the momentum of an ongoing IADB loan and continue the reform efforts that required additional funding. The World Bank was able to successfully coordinate efforts with the IADB, again, in a short period of time. The Bank team composition contributed to the quality of the operation given their extensive experience in Brazil, which allowed it to incorporate lessons from other operations in the country. The World Bank also worked closely with the GoAM during preparation to ensure strong alignment between the operation’s objectives and activities and was able to help include gender issues in the agenda. The team was able to lay out the risks to the operation clearly. This report, however, mentioned minor shortcomings, such as an indicator being dependent on a federally funded program, and lack of clarity in two others. Nevertheless, as explained in the Assessment of Outcomes section, these minor issues did not hinder the success of the operation. (b) Quality of Supervision Rating: Satisfactory 70. The operation benefited from continuous World Bank supervision with the steady deployment of technical teams for the policy areas supported. The World Bank accompanied the GoAM in the implementation of reforms satisfactorily and promptly responded to the borrower’s requests for assistance. The World Bank supervised progress, while also providing consistent remote support. As previous sections have discussed, supervision and support were particularly extensive for Pillar I, for which the World Bank supported the GoAM into pursuing activities that go beyond the achievement of the desired results indicators. The team leader and specialists are well regarded by the GoAM and command respect from their counterparts. This is also reflected in the request by the Government of the Municipality of Manaus to have a new DPL, which, by their request, includes the same team members (team leader and procurement specialist). Additionally, the World Bank is conducting an impact evaluation on a procurement reform, which will further contribute to improving outcomes. The operation ran on time with no changes, supervision reports were done on schedule, and the operation was closed as scheduled. (c) Justification of Rating for Overall Bank Performance Rating: Satisfactory 71. Based on the World Bank performance with regard to both Quality at Entry and Quality of Supervision, the overall rating for World Bank performance is Satisfactory. 29 5.2 Borrower Performance (a) Government Performance Rating: Satisfactory 72. The GoAM demonstrated ownership and commitment to the implementation of the reforms at the highest levels, as mentioned in the Key Factors Affecting Implementation and Outcomes and the Assessment of Outcomes sections. The design and implementation processes were smooth, partly given the strong technical capacities of the staff involved. The government counterparts were also well informed and open about challenges that needed to be overcome and reported and shared data promptly. It showed that the Government has the ability to set goals and coordinate across multiple agencies. The Government was able to deliver on its commitments, and its performance stands out, given the difficult economic and political situation that prevails. (b) Implementing Agency or Agencies Performance Rating: Satisfactory 73. The performance of SEFAZ, implementing agency of Pillar I—Policy Areas I and II—of the DPL, was Satisfactory. The rating reflects the strong performance and leadership, as well as capacity of SEFAZ that enabled positive results and institutional changes. SEFAZ participated actively in the preparation and implementation of the operation, coordinated actions with other implementing agencies, and reported on time to the World Bank on progress. SEFAZ’s commitment can also be illustrated by its active collaboration for the implementation of an impact evaluation of procurement reforms in the state, including the time-consuming gathering and sharing of large amounts of procurement and tax data with the World Bank. SEFAZ’s procurement team also worked closely with the World Bank to engage in an innovative survey with more than 150 suppliers, with the objective of identifying the main challenges faced by them when selling to the state. The performance of the SSP in charge of Policy Area III was Satisfactory given the efforts in trying to solve longstanding problems in the sector such as lack of coordination between police forces and lack of evidence to inform decision making, as well as the economic constraints. SEPM’s performance, for Policy Area IV, was Moderately Satisfactory given the shortcomings mentioned. However, it is important to notice that it suffered from important institutional rearrangements, staff turnover, and budget constraints throughout most of the operation that were beyond its control. (c) Justification of Rating for Overall Borrower Performance Rating: Satisfactory 74. Based on the above discussion, the rating for overall borrower performance is Satisfactory. 6. Lessons Learned 75. Donor collaboration. The operation is an example of successful collaboration between the IADB and the World Bank. The DPL was designed to complement and extend the IADB’s Program of Consolidation of Fiscal Equilibrium for Social and Economic Development in the State of Amazonas, PROCONFIS AM. The World Bank and the IADB established a successful partnership that reduced the burdens of the GoAM (for example, by using the same Project Implementation Unit, constantly sharing information, and linking prior actions) that also 30 guaranteed a continuation of a broader reform process. Moreover, before this DPL, the World Bank had limited engagement in the State of Amazonas. The partnership with the IADB not only enabled the World Bank to engage in the state, but also led to a subsequent DPL, with the capital city of Manaus (approved by the Board in December 2015). 76. Capacity for rapid engagement and team composition. The World Bank demonstrated a great ability to rapidly fill a financing gap that was fundamental to advance reforms. This allowed the GoAM to build on the momentum of an ongoing IADB loan and continue with its reform efforts at a time of fiscal constraint. Additionally, team composition was a key factor that contributed to the quality of entry and supervision because there were constant inputs from World Bank staff and sectoral experts in public sector management, procurement, and security, all of whom have extensive experience in Brazil. This was particularly relevant given that the World Bank had not engaged with the GoAM before. The World Bank technical teams have continued a productive dialogue with the GoAM through continuous technical advice for follow-up reforms, as illustrated, for example, with an ongoing impact evaluation in procurement. 77. Indicators. Two types of issues should be avoided in the future. First, teams should avoid including indicators that are not entirely under the control of the implementing agency (for example, the construction of the Casa da Mulher Brasileira, which depended on the transfer of funds from the Federal Government and land concession by the capital city). Second, teams should avoid results indicators that are difficult to measure and are affected by a multitude of factors. Moreover, although a DPL program document does not contain a specific section detailing how an indicator will be measured, teams should make an effort to clearly define, agree, and register them, together with the client counterparts. For example, one of the gender-related indicators, could have been better defined—policy multipliers. It is important to acknowledge, during the design of any operation, that the results indicators can influence implementation and are therefore relevant policy tools. 78. Citizen security knowledge. Interventions in this sector will become more frequent as security issues become a priority for more client countries. The World Bank has already supported a number of efforts in this sector and is therefore accumulating important knowledge that should be captured systematically to incorporate lessons into future operations. The risk assessment on the security sector was key in designing the operation and contributed to this effort. However, it seems that most of the World Bank’s existing knowledge in this sector is mostly tacit or is particularly fragmented. It is therefore important that the World Bank steps up its efforts to document its experiences in citizen security beyond what is reported in ICRs or risk assessments and in a systematic way. 79. Electoral cycles. This DPL surfaced different aspects of electoral cycles that need attention. During the municipal election campaign period, the rounds of the mobile units providing services to women were suspended to avoid their misappropriation. This had a direct impact on the number of women that received services. Similarly, the electoral cycle had an impact on the CONSEGs, where police forces and citizens engage in dialogue, because the frequency of the meetings was reduced, in an attempt to quell their potential politicization in the run-up to elections. Despite affecting service delivery for a period of time, the precautionary measures worked well given that the legitimacy of the programs (SEPM and Ronda no Bairro) would have been hurt significantly if citizens perceived that these programs were used for electoral purposes. In this 31 regard, future operations should carefully take into consideration the potential impact of election campaign periods and design appropriate mitigation measures. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/Implementing Agencies 80. The GoAM provided written comments on the ICR, which were received by the Bank on January 4th, 2017. These comments were sent as track changes within the ICR document itself and included minor corrections and some suggestions for improved phrasing in the report. The report has incorporated these comments and suggestions directly into the document. Additionally, on January 10th, 2017, the Secretary of Finance from the GoAM, Mr. Afonso Lobo Moraes, sent a letter –see Annex 2 - praising the partnership and work of the World Bank, and emphasizing some of the achievements referred to in this Report. (b) Cofinanciers N/A (c) Other Partners and Stakeholders N/A Annex 1. Bank Lending and Implementation Support/Supervision Processes (a) Task Team Members Responsibility/ Names Title Unit Specialty Lending Alexandre Borges de Oliveira Lead Procurement Specialist GGO04 Procurement Roland N. Clarke Program Leader LCC5C Program Leader Katherine Elizabeth Grau Consultant GGO16 Operations Ana Mie Horigoshi Reis Consultant GMF04 Operations Angela Nieves Marques Porto Consultant GGO16 Operations Miriam Muller Social Scientist GPV04 Gender Specialist Monica Moura Porcidonio Silva Program Assistant LCC5C Program Assistant Senior Public Sector Management Fanny Weiner GGO16 Specialist Laura De Castro Zoratto Senior Economist GGO16 Task Team Leader Supervision/ICR Alexandre Borges de Oliveira Lead Procurement Specialist GGO04 Procurement Flavia Fonseca Carbonari de Consultant GMF04 Citizen Security Almeida Stephanie Anne Kuttner Consultant GPV01 Gender Specialist Courtney Price Ivins Consultant GGO16 ICR Team member Angela Nieves Marques Porto Consultant GGO16 ICR Team member Daniel Ortega Nieto Operations Officer GGHVP ICR Team Leader Laura De Castro Zoratto Senior Economist GGO16 Task Team Leader (b) Staff Time and Cost 32 Staff Time and Cost (Bank Budget Only) Stage US$, Thousands (Including No. of Staff Weeks Travel and Consultant Costs) Lending FY14 36.48 166.93 FY15 14.95 71.78 Total: 51.43 238.71 Supervision/ICR FY15 0 12.98 FY16 26.03 106.21* FY17 3.83 17.51 Total: 29.86 136.7 Note: * This includes the amounts incorrectly charged to the lending code, as reflected in SAP. 33 Annex 2. Summary of Borrower’s ICR and/or Comments on Draft ICR The GoAM provided written comments on the ICR, which were received by the Bank on January 4th, 2017. These comments were sent as track changes within the ICR document itself and included minor corrections and some suggestions for improved phrasing in the report. The report has incorporated these comments and suggestions directly into the document. Additionally, on January 10th, 2017, the Secretary of Finance from the GoAM, Mr. Afonso Lobo Moraes, sent a letter –see below- praising the partnership and work of the World Bank, and emphasizing some of the achievements referred to in this Report. 34 35 36 37 Annex 3. List of Supporting Documents 1. Da Silva Soares. 2013. Nota Tecnica: Gestão Pública nas Áreas Integradas de Segurança Cidadã (AISC’s) em Manaus, Região Metropolitana de Manaus e Interior. Unpublished document. 2. Da Silva Soares. 2013. Nota Tecnica: Custos das AISC’s em Manaus, Região Metropolitana de Manaus e Interior. Unpublished document. 3. Inter-American Development Bank. 2016. Programa De Consolidação Do Equilíbrio Fiscal Para Melhoria Da Prestação De Serviços Públicos Do Estado Do Amazonas – PROCONFIS/AM (BR-L1385 ; 3139/OC- BR) Relatório de Encerramento do Projeto. 4. Perova, Reyndolds, and Muller. 2013. Towards a More Comprehensive Domestic Violence Policy in Brazil. Washington DC: World Bank. 5. World Bank. 2014. Program Document for Amazonas: Modernizing Public Sector Management, Citizen Security, and Gender Policies. Report No. 82539-BR. Washington DC: World Bank. 6. World Bank. 2014-2015. Implementation Status and Results Reports 1-3. Washington DC: World Bank. 7. World Bank. 2013. Poverty and Social Impact Analysis for Amazonas: Modernizing Public Sector Management, Citizen Security, and Gender Policies. Unpublished document. 8. World Bank. 2013. BR Gender and Social Inclusion NLTA (P132325). Washington DC: World Bank. 9. Jaitman. Laura. 2015. The Welfare Costs of Crime and Violence in Latin America and the Caribbean. Washington DC: IADB. 10. World Bank. 2016. Retaking the Path to Inclusion, Growth and Sustainability: Brazil Systematic Country Diagnostic. Washington DC: World Bank. http://documents.worldbank.org/curated/en/239741467991959045/pdf/106569-SCD- P151691-PUBLIC-non-board-version.pdf. 11. Lima, R. S., Ferreira, M. P., Costa, A. B.; Bueno, S. 2016. Relatório de Consultoria para o Banco Mundial. Unpublished working paper. 38 Annex 4. Achievements by Policy Area Results Indicators Baseline Target Actual Comments Objective 1: Increase sales tax compliance (Rating: High) 1. Number of firms issuing the 0 4,000 9,396 Achieved. NFC-e (electronic invoice) Target achieved and surpassed significantly 2. Increasing tax compliance 2,744 12,166 This represents an increase of in the retail sector businesses businesses 340% of businesses issuing issuing issuing invoices. receipts electronic (paper) receipts 3. Improved tax collection R$873 R$905 In spite of a 7.5% decrease in million million overall retail sales in the state, the (2013) (2015) GoAM increased revenues by 3%. 4. Reducing tax compliance Average Cost Cost and time to comply with costs to businesses cost of reduced to requirements for issuing consumer R$3,000 as low as receipt were reduced considerably. and up to R$300, and 30 days’ adopted an wait. automatic process 5. Citizen engagement and tax 190,346 Fiscal Education Campaign led to obligations registered consumers registering with consumers SEFAZ. The campaign was also an important vehicle to disseminate relevant information about tax obligations and processes among the population of Amazonas. 6. Increased accountability 240 5,000 The new tax system also allows complaints complaints consumers to review online, in a year a year real-time, all their transactions. They can also issue complaints and denounce noncompliance by businesses or suspected illegal activities. Before the campaign, SEFAZ received an average of 240 complaints a year. After the NFC-e was put in place, they have registered over 5,000, out of which 3,260 have resulted in investigations (65.2%). 7. Expand tax compliance in A partnership between the GoAM services and the city of Manaus in extending the GoAM’s program through the Nota Facil program that will allow to cover services across more than 10,000 businesses 39 Results Indicators Baseline Target Actual Comments 8. Identify fiscal infractions 3,549 As a result of the information infractions collected through NFC-e, the worth Government was able to identify R$206 fiscal infractions. million Objective 2: Controlling expenditures and increasing transparency and efficiency of government procurement (Rating: High) 1. Percentage of procurement 29.20% 40% 66.6% Achieved. processes completed in less Target achieved and surpassed than 45 days. significantly 2. Percentage of value of 47.9% 55% 55.3% Achieved. purchases made through Price Target achieved Registration 3. Percentage of procurement 0% 80% 100% Achieved. plans of GoAM published on Target achieved and surpassed e-Compras.am 4. Number of services for 9 19 21 Achieved. which standard specifications Target achieved and surpassed for contracting are obligatory 5. Percentage reduction in 0% 10% 12.8% Achieved. unitary cost of selected Target achieved services 6. Number of government 22 (36.6%) 60 (100%) 49 (100%) Achieved. institutions adopting the new Target achieved. Due to the budget execution practices economic downturn, there were institutional rearrangements. However, the percentage of institutions adopting new budget practices is 100%. 7. Tools for efficient The GoAM to establish monitoring procurement processes dashboards through a comprehensive procurement data analysis, and to provide a tracking mechanism for a range of indicators integrated with the state’s broader procurement systems 8. Agile procurement The GoAM adopted measures, processes such as electronic signatures, to process purchases electronically and move away from in-person processes that significantly increase processing times. 9. Improve quality of The GoAM introduced a process purchases through which certain items need to be inspected through samples, 40 Results Indicators Baseline Target Actual Comments with the aim of improving the quality of items purchased. Also, SEFAZ created a subcommittee for procurement processes in the health sector—the area with the highest volume of contracts— within the Procurement General Committee. 10. Processes and regulations The Contract Management System have been put in place that registers all contracts between have made procurement more outsourced service providers and efficient. the Government. This system, together with the Integrated Financial Management System, enables timely monitoring of contract execution, prices, and payments. 11. Efficient procurement The GoAM was able to automate processes the process for estimating costs for procurement of goods by using an innovative database that collects pricing information in both the public and private sector. As a result, procurement of goods in Amazonas is (a) faster when the entire cycle is measured, starting from the request to purchase up until an award is made; and (b) the control of awarded prices is also more effective 12. Government savings The measures that resulted from the DPL, combined with the urgent need to reduce expenditures, have led to a reduction of approximately 20% in public expenditures in areas such as cleaning, security, and leasing of vehicles. In 2015, savings from measures under the DPL accounted for R$213 million, out of which R$91 million resulted from measures directly related to standardized services, and R$122 million were a result of less services and quantities being required as well as substitution of services. 13. Economies of scale and Purchases Purchases According to the GoAM, the use of savings made made framework agreements resulted in through through average savings of 23% when framework framework compared to other purchasing mechanisms. 41 Results Indicators Baseline Target Actual Comments agreements agreements 47.9% 55.3% Objective 3: Improving efficiency, accountability, and transparency in the citizen security (Rating: Substantial) 1. Number of AISCs 6 18 22 Achieved. operating in the state Target achieved and surpassed 2. Number of performance 0 15 31 Achieved. indicators implemented Target achieved and surpassed significantly 3. Number of police trained 1,030 2,000 2,017 Achieved. Target achieved 4. Number of disclosures per 0 4 7 Achieved. year of indicators of Target achieved expenditures of AISCs in periodic reports 5. Evidence for decision The GoAM established a making and efficient use of georeferenced and integrated resources information system that informs tactical operations. Also, the GoAM also established NEPAS in 2015 within the Integrated Institute of Public Safety Education, part of the push toward improving the analytical capacity and use of evidence of the SSP. 6. Government savings in Average Average The costs are adjusted to 2016 AISCs cost of cost of prices. Initial investments were AISCs AISCs large but there was also a reduction R$77 R$22 in overall costs by both military million million and civil police. (2013) (2015) Objective 4: Increasing women’s access to justice and social services (Rating: Modest) 1. Number of communities 2 60 47 Partially Achieved. reached with mobile units 78% of the target was achieved. providing judicial and basic SEPM is continuously reaching out services to women to women across the state. In 2015, SEPM supported 4,486 women through its program - Itinerant Citizenship: Woman in the Field, Forest, and Waters program (Cidadania Itinerante a Mulher do Campo, da Floresta e das Aguas). 2. Number of women victims 0 6,000 0 Not Achieved. of violence receiving judicial The attainment of the indicator depended on funding from the 42 Results Indicators Baseline Target Actual Comments and social services at the Casa Federal Government to construct da Mulher Brasileira the Casa da Mulher Brasileira. However, judicial and social services for women continued to be offered through the existing state institutions—over 23,000 women received services during this operation. 3. Number of implementers of 320 2,750 1,699 Partially Achieved. the Maria da Penha Law 62% of the target was achieved. receiving training The definition of implementer changed over the course of the operation—due to changes both in the government team and World Bank staff—to include implementers and ‘multipliers.’ The definition of implementer expanded to include both implementers in the traditional sense (that is, service providers such as social workers, psychologists, policy, and judiciary) as well as ‘multipliers’ (including members of the general population, schools, and so on). This report only considered those that fall under the traditional definition of implementers. 4. SEPM increases capacity to SEPM is now also responsible for provide services additional gender-related services (Serviços de Atenção a Mulher) such as overseeing the major centers that provide services for GBV victims. 5. Additional resources for Budget Budget R$3 SEPM’s budget increased SEPM R$0.5 million substantially. million (2016) (2013– 2015) 6. Collaboration among The Ronda Maria da Penha was secretariats and different established as a program that levels of Government focuses on preventing recidivism of violence against women. SEPM is working with the Health Secretariat to secure specialized assistance to women that have been victims of sexual violence and with municipalities to directly confront human trafficking. 43 Map: State of Amazonas - Brazil 44