The World Bank (P149555) Document of The World Bank Report No: ICR00004518 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IBRD-83660, IBRD-86740, IDA-54390, IDA- 59280 and IDA-59290) ON LOANS AND CREDITS IN THE AMOUNT OF USD 75 MILLION TO THE REPUBLIC OF MOLDOVA FOR A SERIES OF DEVELOPMENT POLICY OPERATIONS October 30, 2018 Macroeconomics, Trade and Investment Global Practice Belarus, Moldova and Ukraine Country Unit Europe and Central Asia Region The World Bank (P149555) CURRENCY EQUIVALENTS (Exchange Rate Effective October 26, 2018) Currency Unit = Moldovan Lei (MDL) 16.88 = US$1 US$1 = SDR 0.7221 FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS BEM Banca de Economii CAR Capital Adequacy Ratio CPS Country Partnership Strategy CSD Central Securities Depository DPO Development Policy Operation ECF Extended Credit Facility EFF Extended Fund Facility EGDO Economic Governance Development Policy Operation EU European Union FSAP Financial Sector Assessment Program GDP Gross Domestic Product GMI Guaranteed Minimum Income GOM Government of Moldova HVA High-Value Agriculture IMF International Monetary Fund ISR Implementation Status and Results Report LE Large Exposure LR Liquidity Ratio M&E Monitoring and Evaluation MOF Ministry of Finance NBM National Bank of Moldova NCFM National Commission of Financial Markets NDS National Development Strategy PA Prior Action PAD Project Appraisal Document PDO Project Development Objective PIM Public Investment Management PSIA Poverty and Social Impact Assessment RBI Risk-Based Inspections RI Results Indicator RIA Regulatory Impact Analysis ROC Regional Operations Committee The World Bank (P149555) SMEs Small and Medium Enterprises SORT Systematic Operations Risk-rating Tool TA Technical Assistance TNC Total Normative Capital The World Bank (P149555) TABLE OF CONTENTS Data Sheet A. Basic Information B. Key Dates C. Ratings Summary D. Sector and Theme Codes E. Bank Staff F. Results Framework Analysis G. Ratings of Project Performance in ISRs 1. Project Context, Development Objectives and Design ............................................................................. 8 2. Key factors Affecting Implementation and Outcomes............................................................................ 11 3. Assessment of Outcomes ........................................................................................................................ 24 4. Assessment of Risk to Development Outcome ....................................................................................... 30 5. Assessment of Bank and Borrower Performance ................................................................................... 30 6. Lessons Learned ...................................................................................................................................... 33 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners ........................................ 34 Annex 1: Bank Lending and Implementation Support/Supervision Processes ........................................... 35 Annex 2: Borrower's Comments and Comments of Development Partners on Draft ICR ......................... 36 Annex 3: List of Supporting Documents...................................................................................................... 37 Annex 4: Map of Moldova……………………………………………………………………………………………………………………. 38 A. BASIC INFORMATION Program 1 First Development Policy Country Moldova Program Name: Operation Program ID: P143283 L/C/TF Number(s) IBRD-83660,IDA-54390 ICR Date: 05/13/2018 ICR Type: 05/13/2018 Financing Instrument: DPL Borrower REPUBLIC OF MOLDOVA Original Total USD 30.00M Disbursed Amount USD 30.23M Commitment Implementing Agencies: State Chancellery, Ministry of Finance, Ministry of Agriculture, Regional Development and Environment, National Bank of Moldova, Ministry of Health, Labor and Social Protection, Competition Council Program 2 Country Moldova Program Name: Moldova - DPO2 IBRD-83660,IBRD- Program ID: P149555 L/C/TF Number(s) 86740,IDA-54390,IDA- 59280,IDA-59290 ICR Date: 05/13/2018 ICR Type: 05/13/2018 Financing Instrument: DPL Borrower REPUBLIC OF MOLDOVA Original Total USD 45.00M Disbursed Amount USD 44.62M Commitment Implementing Agencies: State Chancellery, Ministry of Finance, Ministry of Agriculture, Regional Development and Environment, National Bank of Moldova, Ministry of Health, Labor and Social Protection, Competition Council B. KEY DATES First Development Policy Operation P143283 Process Date Process Original Date Revised / Actual Date(s) Concept Review: 10/28/2013 Effectiveness: 04/29/2014 Appraisal: 02/03/2014 Restructuring(s): Approval: 03/28/2014 Mid-term Review: Closing: 10/31/2014 10/31/2014 Moldova - DPO2 P149555 Process Date Process Original Date Revised/Actual Date(s) Concept Review: 06/05/2014 Effectiveness: 12/23/2016 Appraisal: 11/04/2016 Restructuring(s): Approval: 12/16/2016 Mid-term Review: Closing: 09/30/2017 09/30/2017 Page 1 of 38 C. RATINGS SUMMARY C.1 Performance Rating by ICR Overall Program Rating Outcomes Moderately Satisfactory Risk to Development Outcome High Bank Performance Moderately Satisfactory Borrower Performance Moderately Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Overall Program Rating Bank Ratings Borrower Ratings Quality at Entry Moderately Satisfactory Government: Moderately Unsatisfactory Implementing Quality of Supervision: Moderately Unsatisfactory Moderately Satisfactory Agency/Agencies: Overall Borrower Overall Bank Performance Moderately Satisfactory Moderately Satisfactory Performance C.3 Quality at Entry and Implementation Performance Indicators First Development Policy Operation P143283 Implementation Performance Indicators QAG Assessments (if any) Rating Potential Problem Program at any No Quality at Entry (QEA) None time (Yes/No): Problem Program at any time No Quality of Supervision (QSA) None (Yes/No): Moldova - DPO2 P149555 Implementation Performance Indicators QAG Assessments (if any) Rating Potential Problem Program at any No Quality at Entry (QEA) None time (Yes/No): Problem Program at any time No Quality of Supervision (QSA) None (Yes/No): D. SECTOR AND THEME CODES First Development Policy Operation P143283 Original Actual Major Sector Agriculture, Fishing and Forestry Other Agriculture, Fishing and Forestry 11 11 Public Administration Central Government (Central Agencies) 22 22 Financial Sector Capital Markets 11 11 Banking Institutions 22 22 Industry, Trade and Services Other Industry, Trade and Services 34 34 Major Theme/Theme/Sub Theme Economic Policy 6 6 Page 2 of 38 Fiscal Policy 6 6 Tax policy 6 6 Finance 6 6 Finance for Development 6 6 Agriculture Finance 6 6 Private Sector Development 122 122 Business Enabling Environment 22 22 Regulation and Competition Policy 22 22 Jobs 100 100 Public Sector Management 62 62 Public Administration 11 11 Transparency, Accountability and Good Governance 11 11 Public Finance Management 17 17 Domestic Revenue Administration 6 6 Public Expenditure Management 11 11 Rule of Law 34 34 Legal Institutions for a Market Economy 34 34 Urban and Rural Development 6 6 Rural Development 6 6 Rural Markets 6 6 Moldova - DPO2 P149555 Original Actual Major Sector Agriculture, Fishing and Forestry Agricultural Extension, Research, and Other Support 9 9 Activities Public Administration Central Government (Central Agencies) 9 9 Financial Sector Banking Institutions 55 55 Social Protection Social Protection 9 9 Industry, Trade and Services Other Industry, Trade and Services 18 18 Major Theme/Theme/Sub Theme Environment and Natural Resource Management 9 9 Climate change 9 9 Adaptation 9 9 Finance 6 6 Finance for Development 6 6 Agriculture Finance 6 6 Financial Stability 55 55 Financial Sector oversight and policy/banking 55 55 regulation&restructuring Private Sector Development 122 122 Business Enabling Environment 22 22 Investment and Business Climate 9 9 Regulation and Competition Policy 22 22 Jobs 100 100 Page 3 of 38 Public Sector Management 62 62 Public Administration 11 11 Public Assets and Investment Management 9 9 Social Development and Protection 9 9 Social Protection 9 9 Social Safety Nets 9 9 Urban and Rural Development 6 6 Disaster Risk Management 9 9 Disaster Response and Recovery 9 9 E. BANK STAFF First Development Policy Operation P143283 Positions At ICR At Approval Vice President: Cyril Muller Laura Tuck Country Director: Satu Kahkonen Qimiao Fan Practice Gallina Andronova Vincelette Carolina Sanchez-Paramo Manager/Manager: Task Team Leader: Ruslan Piontkivsky/Mame Fatou Digne ICR Team Leader: Jeffrey Allen Chelsky ICR Co-Team Leader Hilda Shijaku ICR Primary Author: Jeffrey Allen Chelsky/Hilda Shijaku Moldova - DPO2 P149555 Positions At ICR At Approval Vice President: Cyril Muller Cyril Muller Country Director: Satu Kahkonen Satu Kahkonen Practice Gallina Andronova Vincelette Gallina Andronova Vincelette Manager/Manager: Task Team Leader: Ruslan Piontkivsky ICR Team Leader: Jeffrey Allen Chelsky ICR Co-Team Leader Hilda Shijaku ICR Primary Author:1 Jeffrey Allen Chelsky/Hilda Shijaku F. RESULTS FRAMEWORK ANALYSIS Program Development Objectives The series had three PDOs, each with three subcomponents. No revisions were made to the PDOs throughout the series. 1. Strengthen the regulatory framework to improve predictability of the business environment, facilitate competition and reduce regulatory compliance costs 2. Strengthen financial sector stability, promote transparency of shareholding, and ease conditions for access to finance 3. Improve the public investment management framework, make investment subsidies in agriculture more efficient and equitable, and improve the coverage of well-targeted social assistance programs. 1 The team is grateful for the assistance and advice provided by Marcel Chistruga, Economist, GMET3. Page 4 of 38 Results Indicators First Development Policy Operation (P143283) and Second Development Policy Operation (P149555) Original Target Value Achieved at Baseline Formally Revised Indicator (from approval Completion or Value Target documents) Target Years Pillar A: Strengthen the regulatory framework to improve predictability of the business environment, facilitate competition and reduce regulatory compliance costs 1. Inspection coverage (except 100% (2013) 80% (2015) 50% (2017) 30% (2017)2 Uncertain -- With documentation tax, customs and financial) on calculations lost, the definition of this indicator is uncertain. It may relate to the number of firms visited by inspectors. 2. Share of proposed customs- None (2012) 100% (2015) 80% (2017) (2017) Not achieved (achieved for and tax-related legislation and 60% (tax and customs, not achieved for tax, not regulations that affect customs) achieved overall) companies which has been subjected to a Regulatory 0% (tax) Impact Analysis (RIA) by the 100% (customs) RIA Secretariat before adoption 3. Number of state aid providers None (2013) 183 (2015) 100% (2017) 100% of state aid Partially achieved. According to that are connected and report providers are the GoM, some entities did not to the state aid automated connected of report using the system due to an information system which 88% are insufficient level of computer reporting on the skills. These entities reported in a system (2017) non-electronic format. All state aid providers were connected by 2018. 4. Number of new registered 145 (2013) 160 (2015) 210 (2017) 198 (2017) Almost achieved. The GOM varieties of seeds and ) excluded from Moldova’s seedlings Catalogue of Plant Varieties, those varieties in the EU catalogue that, according to testing by the State Commission, did not show better results in Moldova. Nevertheless, they did meet the less ambitious original 2015 target and were only about 5% below the 2017 target. Pillar B: Strengthen financial sector stability, promote transparency of shareholding, and ease conditions for access to finance 5. Share of public interest None (2013) 100% (2015) 100% (March 2018) 0% (August 2018) Not achieved. With the roadmap entities whose registers of for the establishment of the CSD securities’ owners were approved by the NBM in March transferred to a Central 2018, the CSD was only able to Securities Depository (CSD) begin it activities in July 2018. A proposed amendment to CSD Law would postpone the deadline for banks and insurance companies to transfer securities records to the CSD to Spring 2019. 6. Share of loans in the system 33% (June 36.3% (2015) 37% (June 2017) 29.0% (end 2017) Not achieved. Share declined secured by movable collateral 2013) 27.5% (June 2018) instead of increasing. Related indicators show a decline in the share of movable collateral in total collateral. 2The authorities have reported a value of 30 percent. However, the underlying definition and calculation is unclear and cannot be replicated. Page 5 of 38 7. (a) BEM operates in CAR: 26.03%; CAR->=16%; Dropped in DPO-2 accordance with prudential TNC MDL555 TNC >=MDL200 norms as measured by Capital mn, LR mn, LR>=20%, Adequacy Ratio (CAR), Total 66.06%, LE LE <15% of TNC Normative capital (TNC), 16.63% to (2015) Liquidity Ratio (LR), Principle II 18.89% of LR and Large Exposure (LE) TNC (end established by NBM. 2013) 7. (b) For BEM, Banca Sociala, none 3 Added in DPO-2 3 (June 2018) Achieved with delay and Unibank: non-insider deposits transferred; payment functions transferred; liquidators appointed (number of banks) 8. Time-bound action plans for none 3 Added in DPO-2 2 (2018) Partially achieved. Action plans Victoriabank, Moldincombank, for Victoriabank and Moldova and Moldova Agroindbank Agroinbank were developed and prepared and under implemented by February 2017 implementation based on the results of diagnostic studies undertaken as a PA to DPO-2. Moldincombank cancelled the contract with the company for a diagnostic study/special audit and no action plan was elaborated based on its findings. Instead, using results of on-site inspections, the bank developed its own, less robust, action plan to address shortcomings identified during inspections. This plan is still under implementation. 9. Share of tenders for banking 0% (2013) 100% (2015) 100% (2017) No new tenders Achieved. Law on Public Finances services by State Treasury and issued therefore and Budgetary-Fiscal fully or majority state-owned none contravene Accountability no.181 (July 25, companies that are price- DPO-2 PA 2014) requires that procedure to based and that include criteria select the commercial banks for of the regulatory minimum cash budget execution be done required capital adequacy every three years. The last ratio (CAR) and liquidity ratios agreement was signed September 18, 2015. Therefore no tenders have been issued for this purpose since DPO-2. A new tender will be issued that conforms to Order No. 162 at on 12/28/17 to supplement previous order with Section 1.6 to exclude privileged access of banks to public sector “means”. Pillar C: Improve the public investment management framework, make investment subsidies in agriculture more efficient and equitable, and improve the coverage of well-targeted social assistance programs. 10. Share of multi-year 33% (2006- 20% (2015) 20% (2017) 11.1% (2017) Achieved investment projects with non- 2011) continuous funding 11. (a) Project proposals prepared 0% (2013) 100% of all new Dropped in DPO-2 in compliance with the domestically- requirement of the regulation funded projects on public capital investment costing over for project identification and MDL5 mn that preliminary assessment are included in the 2015 budget submission. 11. (b) Share of new eligible 0% (2013) 100% (2017) Added in DPO-2 100% (2017) Achieved. The new PIM projects included in the regulations include a lower budget submission in threshold for project size for Page 6 of 38 compliance with the new PIM inclusion in the budget. The regulation authorities report that all new projects meeting this and other relevant criteria and included in the 2017 central government budget followed the new PIM regulation. While the baseline reported is 2013, this indicator was only introduced in DPO-2. A 2015 baseline would have therefore been warranted. 12. Share of subsidies supporting 38.6% (2013) 42.5% (2015) 58% (2017) 56.5 % (2017) Almost achieved -- For 2017, high-value agriculture (HVA) Government Decision no. 455 sectors (21.06.2017) allocated 56.5% of the subsidy fund to HVA sectors. 13. Share of subsidies allocated to 24.4% (2013) 27% (2015) 30% (2017) 29.1% Almost achieved—to 2,142 individual producers individual producers 14. Monthly number of Ajutor 23,339 30,000 126,000, o/w at 125, 189 o/w Almost Achieved--While the Social recipients. households households least 74,000 females 71,286 female number of benefiting households (Oct 2013)/ (2015) (October 2017) (October 2017)3 increased significantly between 68,150 2015 and 2017, the structure of persons o/w households changed to include 42,824 more small households. female Therefore, the target for the revised RI (# of individuals and females) fell short. 15. Monthly number of Heating 30,249 75,000 248,800 persons, 268,884 o/w Achieved Allowance recipients households households o/w at least 163,000 169,285 female (Nov (2015) females (November (November 2017) 2013)/78,524 2017) persons o/w 50,696 female G. RATINGS OF PROJECT PERFORMANCE IN ISRs No ISR completed or waiver requested. 3By November 2017, the number of females receiving Ajutor Social benefits had risen to 76,025 (more than 2,000 above the target). In number of households, October 2017 was 50,869, well above the original target for 2015. Page 7 of 38 The World Bank (P149555) 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal 1. Moldova experienced rapid growth and poverty reduction in the early 2000s. Nevertheless, Moldova remains the poorest country in Europe. In 2013, it initiated the signing of the Association Agreement and Deep and Comprehensive Free Trade Area with European Union (EU) which anchored the political and economic transformation agenda. At the same time, it revealed weaknesses in the financial sector and with respect to governance that required prompt and sustained policy actions. However, a deterioration in external conditions beginning in 2014 resulted in a slowdown in remittances and trade, and serious governance challenges, particularly in the financial sector, slowed growth. At the same time, a series of summer droughts and the banking fraud which came to light in the fall of 2014, reducing financial intermediation and halting external assistance, led to a contraction of the economy by 0.4 in 2015, a sharp contraction in public capital spending, with public debt and guaranties increasing to 46.5 percent of gross domestic product (GDP) from 38.2 percent in 2014. In 2016, due to better climate conditions, the economy recovered, growing 4.5 percent. Private investments remained subdued due to lack of confidence in the banking system and tighter monetary conditions. While the fiscal discipline was maintained, fiscal space was constrained by the suspension of external assistance. 1.2 Original Project Development Objectives (PDO) and Key Indicators (as approved) Program Development Objectives PDO 1/Pillar A: Strengthen the regulatory framework to improve predictability of the business environment, facilitate competition, and reduce regulatory compliance costs. Results Indicators 1. Inspection coverage (except tax, customs and financial) 2. Share of proposed customs- and tax-related legislation and regulations that affect companies which has been subjected to a Regulatory Impact Analysis (RIA) by the RIA Secretariat before adoption 3. Number of state aid providers that are connected and report to the state aid automated information system 4. Number of new registered varieties of seeds and seedlings PDO 2/Pillar B: Strengthen financial sector stability, promote transparency of shareholding, and ease conditions for access to finance. Results Indicators 5. Share of public interest entities whose registers of securities’ owners were transferred to a Central Securities Depository (CSD) 6. Share of loans in the system secured by movable collateral Page 8 of 38 The World Bank (P149555) 7. (a) Banca de Economii (BEM) operates in accordance with prudential norms as measured by Capital Adequacy Ratio (CAR), Total Normative Capital (TNC), Liquidity Ratio (LR), Principle II LR and Large Exposure (LE) established by NBM. 7. (b) For BEM, Banca Sociala, and Unibank: non-insider deposits transferred; payment functions transferred; liquidators appointed (number of banks) 8. Time-bound action plans for Victoriabank, Moldincombank, and Moldova Agroindbank prepared and under implementation 9. Share of tenders for banking services by State Treasury and fully or majority state-owned companies that are price-based and that include criteria of the regulatory minimum required CAR and liquidity ratios PDO 3/Pillar C: Improve public investment management framework, make investment subsidies in agriculture more efficient and equitable, and improve coverage of well‐targeted social assistance programs. Results Indicators 10. Share of multi-year investment projects with non-continuous funding 11. (a) Project proposals prepared in compliance with the requirement of the regulation on public capital investment for project identification and preliminary assessment 11. (b) Share of new eligible projects included in the budget submission in compliance with the new PIM regulation 12. Share of subsidies supporting high-value agriculture (HVA) sectors 13. Share of subsidies allocated to individual producers 14. Monthly number of Ajutor Social recipients, of which females. 15. Monthly number of Heating Allowance recipients, of which females. 1.3 Revisions to PDOs and Key Indicators 2. There were no revisions to the PDOs. However, several results indicators and targets were adjusted between Development Policy Operation (DPO)-1 and 2. Specifically: • Given the delay in approving DPO-2, targets that had been set for DPO-2 were extended from 2015 to 2017. • RI1: The target on the inspection coverage was reduced from 80 to 50 percent to better reflect implementation capacity and coverage ratio relative to more advanced economies. Page 9 of 38 The World Bank (P149555) • RI2: The target on the share of proposed customs and tax related legislation and regulation accompanied by a regulation impact assessment (RIA) was reduced from 100 percent (by 2015) to 80 percent (by 2017). This reflected input from the relevant Global Practice that 100 percent was infeasible for a transition economy over such a short period of time (no RIA’s were done prior to 2012). In fact, advanced economies rarely achieve 100 percent. • RI3: DPO I targeted 183 state aid providers that were connected and reported to the state aid automated information system in 2015 (from none in 2013). In DPO-2 RI3 indicator was formulated as a share total the state aid providers and set at 100 percent for 2017. This reflected two factors: (i) the Government of Moldova (GOM) created an inventory of mandated state aid providers and restricted the provision of state aid and (ii) fewer than 183 state entities were mandated to provide state aid that had been authorized by the GOM. • RI14 and 15: Results indicators in DPO-1 targeted the monthly number of households which were beneficiaries to existing social programs of Ajutor Social, and Heating allowance. The formulation of these indicators was changed in DPO2 to target the number of persons which benefitted from these programs as well as the number of female recipients. The change was done to include a gender dimension to the associated prior action (PA). • Two new RIs were introduced in relation to PAs 5 and 6: (i) RI6--For BEM, Banca Sociala, and Unibank, the number of banks for which non-insider deposits were transferred, payment functions transferred, and liquidators appointed, with a target for the number of banks of three in 2017 (from 0 in 2013); and (ii) RI7-- Time-bound action plans for Victoriabank, Moldincombank, and Moldova Agroindbank prepared and under implementation, with a target of three by 2017. 1.4 Original Policy Areas Supported by the Program (as approved) 3. The three PDOs were also considered to be the policy pillars. They were unchanged throughout the series. 4. In Pillar A, “Strengthen the regulatory framework to improve predictability of the business environment, facilitate competition, and reduce regulatory compliance costs”, the supported policy areas included: • Facilitating competition through reducing anticompetitive practices including reporting and reducing advantages granted by the Government in the form of state aid. • Improving the predictability of the business environment through reducing regulatory compliance costs for businesses and reducing burdensome, overlapping, and often unclear regulatory requirements. 5. In Pillar B “Strengthen financial sector stability, promote transparency of shareholding, and ease conditions for access to finance” the supported policy areas included: • Improving shareholder transparency, enforcing property rights, Page 10 of 38 The World Bank (P149555) • Strengthening governance and enforcement powers of financial sector regulatory agencies by granting staff adequate protection to exercise their duties • Enhancing access to finance, which continues to constrain private sector development, especially for small and medium enterprises (SMEs) and entrepreneurs • Protecting government resources from impact of possible future bank failures. 6. In Pillar C, “Improve the public investment management framework, make investment subsidies in agriculture more efficient and equitable, and improve the coverage of well‐targeted social assistance programs”, the supported policy areas included: • Ensuring that resources are used more efficiently to implement good projects in a timely manner. • Allocating agricultural investment support to foster modernization and innovative practices, increase market competitiveness, reduce vulnerability and improve equity. • Adapting existing social support programs to reduce poverty in a more cost-effective way. 1.5 Other Significant Changes 7. Given delays in implementing DPO2, and to prevent the series from lapsing, a waiver was requested from the Europe and Central Asia Vice President in September 2016 (and granted retroactively), to extend the deadline for Board consideration of the second operation beyond 24 months from the approval of the first operation in March 2014. DPO-2 was approved in December 2016. 2. Key factors Affecting Implementation and Outcomes 2.1 Program Performance Table 1. First Operation in a Programmatic Series Prior Actions Status PA1: Government adopts decision establishing the Completed –Decision No. 694 on approving the framework Risk-Based Inspections (RBI) Methodology General Framework of the Methodology on Planning of State Control Base on Risk Criteria (Sept 5, 2013) PA2: Enact amendments to legislation to ensure that Completed – Law 324 on amending and customs- and tax-related legislation and regulations Supplementing Some Legislative Acts (December 23, that affect companies be accompanied by a RIA 2013) PA3: Enact secondary legislation required to implement Completed – For the Law on Competition, Issuance of the Law on Competition and Law on state aid, including enabling regulations for Borrower’s Law No. 183 (July a Parliamentary decision on the structure of the 12, 2012); for Law on State Aid issuance of enabling Competition Council regulations for Law 139; (June 15, 2012); adoption of Parliamentary Decision PA4: Government adopts a decision reducing the Completed – Government Decision No 964 on testing and registration period for seeds and seedlings Supplementing Para 29 of Regulation on Testing and to: (1) one year for all annual crops listed in the EU acceptance of Varieties in the Catalogue of Plan Varieties (December 4, 2013) Page 11 of 38 The World Bank (P149555) Prior Actions Status catalogue; and (2) one harvest year for multiannual (fruit trees) crops listed in the EU catalogue. PA5: Government adopts and submits to Parliament Completed—Draft Law on Amendments to Law No. amendments to the Law on Capital Markets to 171 pursuant to Government Decision No. 1028 consolidate the share registry function of corporate (December 19, 2013) submitted to Parliament securities and merge it into the depository function for (December 24, 2013) corporate securities. PA6: Enact the amendments to the Laws on NBM, on Completed—Law No. 343 on Amending and Administrative Court, and Code of Civil Procedure to Supplementing Some Legislative Actions (December clarify the legal frameworks for contesting the acts of 24, 2013) the NBM. PA7: NBM instructs and independent diagnostic audit Completed—NBM has instructed an independent of BEM and an auditor is selected competitively to diagnostics audit of BEM pursuant to Resolution No. carry out said audit in compliance with terms of 16-003/154-3602 (December 9, 2013); auditor reference agreed with NBM. competitively selected. PA8: Government adopts regulation on public capital Completed—Regulation on Public Capital Investments investment to set up transparent and efficient pursuant to Government Decision No. 1029 mechanisms for planning, implementing, and (December 19, 2013) monitoring public investment projects. PA9: Enact amendments to the Budget Law for 2013 Completed—Government Decision No. 793 on and to Government regulation to adjust parameters of Amending and Supplementing Government Decision Ajutor Social (increase GMI to MDL680) and Heating No. 1167 (October 8, 2013) and Law No. 277 on allowance (increase benefit size to MDL250) targeted Amending Law No. 249 on State budget for 2013 benefit programs. (November 15, 2013) Table 2. Second Operation in Programmatic Series PAs from Legal Agreement/ Program Document Status PA1: Enact amendments to Law on state Control of Completed—Amendments to Law on State Control of entrepreneurial Activity to reduce compliance costs of Entrepreneurial Activity (Law No. 131) through Law inspections 230: Amendment and Completion of Some Legislative Acts (September 23, 2016) PA2: Government approves a decree setting up an Completed—Decision No. 378 on Approving the institutional mechanism for information collection on all Concept of Automated Information system: State Aid state aid under the Law on State Aid and state aid Register (May 27, 2014) secondary legislation. PA3: Enact the amendments to the Law on the National Completed—Enactment of Law No. 62 on Amending Bank of Moldova to consolidate the corporate securities and supplementing Certain Acts (April 8, 2016) registration function into a single central securities depository. PA4: Enact legislative amendments to strengthen the Completed—Enactment of Law No. 108 on Approving supervisors’ independence through (i) strengthening of amendments and Completion of some Legislative the framework for contesting NBM acts related to acts (June 19, 2014) and Law No. 62 on Amending and banking regulation and supervision; (ii) providing legal supplementing Certain Acts (April 8, 2016) protection to all NBM and NCFM employees in the event of lawsuits for actions taken in good faith; and (iii) removing the Ministry of Justice’s power to amend the content of NBM and NCFM regulations. PA5: NBM withdraws the licenses from BEM, Banca Completed—Decision No. 62 of the NBM on Approving Sociala, and Unibank and initiates liquidation of all three the Withdrawal of the License for Conducting Financial banks. Page 12 of 38 The World Bank (P149555) PAs from Legal Agreement/ Program Document Status Activities and Initiation of the Liquidation Process of JS Banca Sociala (October 23, 2015) PA6: NBM issues an order on launching special audits Completed—NBM Decisions Nos. 156-159, on (diagnostics studies) for the three largest banks Establishment of Special surveillance at (Victoriabank, Moldincombank, and Moldova Moldincombank, Victoriabank and Moldova Greenbank) Agroinbank, respectively (June 11, 2015) PA7: Enact amendments to the Pledge Law to facilitate Completed—Enactment of amendments to the Law on use of movable assets as collateral. Pledge (Law No. 173 on Approving of Amendments and Completion of some Legislative Acts, July 25, 2014) PA8: Government approves a policy requiring the Completed—Government Decision No. 387 on Ministry of Finance (State Treasury) and fully or majority approving the Regulation on the Procurement of state-owned companies to conduct price-based public financial Services by State and Municipal Enterprises tenders for banking services which incorporates at least and Trading Companies with Integral or Majority the regulatory minimum CAR and LRs. Public Capital (May 28, 2014) and Government Decision No. 428 on Approving the Selection of Banking Services of Financial Institutions for Public Institutions and Authorities (June 9, 2014) PA9: Strengthening the regulatory framework for PIM Completed—Law No. 181 on Approving Public Finance by: (i) enacting provisions of the Law on Public Finance and Fiscal Budgetary Responsibility (July 25, 2014), and Fiscal Budgetary Responsibility on the principles of Government Decision No. 1181 on Approving the PIM system and public investment financing; (ii) Amendments to Appendix 3 to the Government amending the regulation on public investment to Decision No. 2019 (October 26, 2016) and Order No. expand its coverage; and (iii) adopting (by the Ministry 185 from the Ministry of finance on approving the of Finance) guidelines for the project preparation cycle. Instructions on the Management of Capital Investment Projects (November 3, 2015) PA10: The Government adopts the annual regulation on Completed—Adopted regulation on the agricultural the agricultural support fund for 2016 to: (i) narrow the support fund for agricultural producers for 2016 with scope of the largest investment subsidy program and (ii) Government Decision No. 910 (July 25, 2016) cap (or reduce the caps of) subsidy sizes allowed per one beneficiary. PA11: In the amendments to the 2015 Budget Law, Completed—Law no. 200 on amendments to the State increase the eligibility threshold of the Ajutor Social Budget for 2015 (November 20, 2015) program to MDL900. Table 3. Program Performance Tranche Expected Release Amount Actual Release Date Release # Date DPO-1 US$9 mn (IBRD) and US$21.0 mn (IDA) April 15, 2014 April 29, 2014 Regular DPO-2 US$27.2 mn (IBRD) and US$17.8 mn (IDA) December 15, 2016 December 23, 2016 Regular Prior Actions/Triggers 8. Table 4 DPO-2 describes the adaptation of triggers for DPO-2 into DPO-2 PAs. Pillar B reforms, which dealt with financial sector issues, were made more specific and augmented in the wake of major fraud revealed after the approval of the first operation in some domestic banks and alleged fraud in others. These developments helped build domestic support to address the weaknesses that had been identified in the governance of the banking sector. The resulting PAs for DPO-2 were significant given the major events that transpired in the more than 24 months between the two operations and the pressing need to respond to the impact of the bank fraud which came to light in the fall of 2014. Page 13 of 38 The World Bank (P149555) 9. PA3 was strengthened to require consolidation of the securities registration function into a single central securities depository. At the time of DPI-1, Trigger 3 sought consolidation but, reflecting a lack of consensus on the appropriate model, the function was to be divided between the NBM (for bank securities) and the NCFM (non-bank securities). With respect to the legal framework governing the NBM, and in agreement with the International Monetary Fund (IMF), trigger 4 was considerably strengthened and made more precise with PA4 now requiring legal protection for NBM staff in carrying out their duties and removing the Ministry of Justice’s power to interfere in the decision of the NBM. Triggers 5 and 6 evolved to close BEM in response to the findings of wrong doing in BEM (partly from the independent diagnostic audit of BEM that was a PA in DPO-1) and it was extended to two other banks for which evidence had become available. An additional PA (PA 6) was introduced to require an audit of three other large banks under suspicion. 10. Triger 9 on the implementation of the regulation on public investment by local governments was met and strengthened by introducing specific PIM provisions in the Law on Public Finance and Fiscal Budgetary Responsibility on the principles of the PIM system and public investment financing; (ii) amending the regulation on public investment to expand its coverage and include donor funded projects; and (iii) adopting (by the Ministry of Finance) guidelines for the project preparation cycle. 11. Trigger 10 on agriculture subsidy programs was substantively met in 2014 and was extended in DPO-2 (PA 10) to 2016, allowing this measure to now impact three rather than one fiscal year. 12. The eligibility threshold of Ajutor Social in Trigger 11 was raised from MDL700 to MDL900 in DPO- 2 PA11 to reflect the higher cost of living relative to 2014. Table 4. Adaptation of DPO 2 Triggers into PAs DPO 2 Triggers DPO 2 PAs Explanation Pillar A: Strengthen the regulatory framework to improve predictability of the business environment, facilitate competition, and reduce regulatory compliance costs Trigger 1: Government adopts PA1: Enact amendments to Law on Trigger was met and PA1 introduced to decisions establishing sector RBI state Control of Entrepreneurial reduce the number of bodies entitled to methodologies based on the Activity to reduce compliance costs carry out inspections to 13 (from 33 in the framework RBI Methodology and of inspections. original law). The exhaustive nature of the creates an on-line State Registry list curtailed the exercise of inspection of Inspections which services as functions by 25 bodies that had no and inspection management inspection mandate, but which were system. exercising inspection functions outside the scope of the law. Pillar B: Strengthen financial sector stability, promote transparency of shareholding, and ease conditions for access to finance Trigger 3: Enact the amendments PA3: Enact the amendments to the Initial plan envisaged share registry for to the Law on Capital Markets to Law on the National Bank of banks to reside with the NBM and the consolidate the share registry Moldova to consolidate the share registry for non-banks to reside function of corporate securities corporate securities registration with NCFM. By DPO, it was agreed to and merge into depository function into a single central consolidate into a single registry in a more function for corporate securities. securities depository. independent NBM. This reflected an initial lack of consensus on the appropriate model and represented a significant strengthening of the original reform. Page 14 of 38 The World Bank (P149555) DPO 2 Triggers DPO 2 PAs Explanation Trigger 4: Enact measures to PA4: Enact legislative amendments In light of the experience of the bank further enhance the framework to strengthen the supervisors’ fraud, the triggers were made for contesting NBM acts, in independence through (i) considerably more precise and concrete. particularly related to banking strengthening the framework for This represented a significant regulation and supervision. contesting NBM acts related to strengthening of the proposed reform. banking regulation and supervision; (ii) providing legal protection to all NBM and NCFM employees in the event of lawsuits for actions taken in good faith; and (iii) removing the Ministry of Justice’s power to amend the content of NBM and NCFM regulations. Trigger 5: Complete an external PA5: NBM withdraws the licenses As the details of the bank fraud came to independent audit of BEM; and from BEM, Banca Sociala, and light, this trigger was adapted to respond Unibank and initiates liquidation of to the findings of the audit on BEM, acting Trigger 6: BEM adopts a time all three banks. to withdraw the license and liquidate the bound medium-term BEM instead of seeking its restructuring. It restructuring strategy designed was also extended to two other to restore long-term viability implicated banks. approved by BEM Supervisory Board and acceptable to the NBM. In addition, the following new PA was introduced: PA6: NBM issues and order on launching special audits (diagnostics This PA was introduced given evidence of studies) for the three largest banks (Victoriabank, Moldincombank, and potential wrong doing in three other large Moldova Agroindbank) banks. It represented a significant strengthening of the reach of the DPO in terms of addressing governance shortcomings in the banking sector. Pillar C: Improve the public investment management framework, make investment subsidies in agriculture more efficient and equitable, and improve the coverage of well‐targeted social assistance programs Trigger 9: Government adopts PA9: Strengthening the regulatory The change reflected a significant guidelines for implementation of framework for PIM by: (i) enacting strengthening of the initial trigger. First, the regulation on public provisions of the Law on Public an analysis conducted after DPO1 investment by local governments Finance and Fiscal Budgetary revealed that the share of investment Responsibility on the principles of financed or co-financed by local the PIM system and public governments was small, which suggested investment financing; (ii) amending that the trigger was not relevant. On the the regulation on public other hand, the overall public investment investment to expand its coverage; and (iii) adopting (by the Ministry framework - which included public of Finance) guidelines for the investments of local governments - project preparation cycle. needed a further upgrading of the legal and regulatory frameworks and implementation. The scope of the DPO-2 PA was expanded to take this into account. Page 15 of 38 The World Bank (P149555) DPO 2 Triggers DPO 2 PAs Explanation Trigger 10: Government PA10: The Government adopts the Implementation of measures was introduces the following changes annual regulation on the extended into 2016 (after having been to the annual regulation on the agricultural support fund for 2016 substantively met in 2014). This deepened agricultural support fund for to: (i) narrow the scope of the the impact of the underlying effort to 2014: (i) narrowing the scope of largest investment subsidy promote equity. the largest investment subsidy program and (ii) cap (or reduce the programs; and (ii) capping (or caps of) subsidy sizes allowed per reducing the caps of) subsidy one beneficiary sizes allowed per one beneficiary Trigger 11: In the Budget Law for PA11: In the amendments to the Implementation of measures was 2014, adjust parameters of the 2015 Budget Law, increase the extended into 2016 (after having been Ajutor Social program (increase eligibility threshold of the Ajutor substantively met in 2014). This deepened GMI to MDL700). Social program to MDL900. the impact of the underlying effort. 2.2 Major Factors Affecting Implementation Unanticipated Developments 13. The Country Partnership Strategy (CPS, FY14-17) which was discussed at the Board in August 2013 envisaged one DPO per year. However, only two were approved, with a gap of more than 24 months between them. 14. Clear evidence of the large bank fraud, which had been percolating since at least 2011, came to light in the fall of 2014, only a few months after the concept note review for DPO-2 (June 5, 2014). This reduced the relevance of banking-sector related DPO-2 triggers 5 and 6 and, given the high cost of the banking sector bail out (more than 10 percent of GDP), required a reassessment of the adequacy of the underlying macroeconomic framework. At the same time, the capacity of the authorities to commit to concrete and sufficiently ambitious reforms that were to be supported by DPO-2 was undermined by a period of profound political uncertainty following parliamentary elections in late 2014 (with seven substantive or acting governments between November 2014 and January 2016). 15. The bank fraud and other crimes related to poor banking-sector governance required a fundamental re-think of the design of DPO-2.4 In February 2015, once the extent of the fraud became clear, the Bank, in collaboration with the IMF, identified a set of measures deemed critical for restoring confidence and stability in the banking sector. These were communicated to the Prime Minster of Moldova in a letter dated February 26, 2015 in which it was made clear that risks to the financial sector were such that corrective action was necessary for further budget support to be justified from a fiscal standpoint. The financial sector measures set out at that time became PAs for DPO-2 and were implemented with support from the Bank. Relevance of Risks Identified and Adequacy of Mitigating Measures 16. The Bank team and management were aware of key risks to the success of the operation at the time that DPO-1 was approved (particularly those related to the banking sector). This was clearly reflected 4Revelations of the fraud also triggered a stop to external budget support from development partners. This constrained public consumption and investment and heavily eroded investors’ perception of the Moldovan business and investment climate. Page 16 of 38 The World Bank (P149555) in the Macroeconomic Outlook and Debt Sustainability section of the PAD5 and figured prominently in discussions at the ROC Review in January 2014 although it received only passing mention in the Summary of Risks and Mitigation, with little discussion of mitigating measures6. The text highlighted the political and governance risks that could potentially undermine the independence and effectiveness of public institution and agencies which implicitly included constraints on the conduct of banking supervision (which allowed the bank fraud to propagate). On macroeconomic risks, there was a reference to “credit quality, liquidity and capital adequacy of select banks” in the list of “main risks”. These were considered to have been “partly mitigated” by the measures in Pillar B of the operation, Moldova’s “prudent and flexible monetary policies” and by “continuing consultations with the IMF” despite the absence of agreement on a new IMF-supported arrangement. There was also an expectation that the findings of the FSAP that was underway at the time would also inform the reform agenda for future World Bank support. 17. Much of the discussion of the problems in the banking sector that would eventually derail the series was in the section on “Recent Economic Developments” where the text clearly sets out the problems related to the Constitutional Court’s ruling on the NBM’s ability to carry out its mandate. But this concern is downplayed by the assertion that “the government moved swiftly to put in place mitigating legislative measures and committed to further remedial actions”. This contrasts somewhat with the discussion in Box 1 which speaks of the “serious ramifications” of the Constitutional Court ruling. While the GOM may have moved swiftly with mitigating measures to address risks to the conduct of monetary and foreign exchange policy, internal documents show an awareness of “remaining concerns in the areas of banking regulation and supervision”, with the IMF conveying the view that even with the amendments introduced by the GOM, risks in the area of banking regulation and supervision remained a “significant concern and pose[d] a risk to the macroeconomic situation”. 18. With the benefit of Systematic Operations Risk-rating Tool (SORT, which was introduced in October 2014), risks were more clearly articulated in DPO-2 which was appropriately assigned an overall risk rating of substantial, based on similar ratings for “political and governance”, “macroeconomic”, “institutional capacity for implementation and sustainability” and “fiduciary” categories. “Sector strategies and policies” which included the financial sector, was assigned a risk rating of moderate. While experience with DPO-1 might have suggested the need for a rating of high for “political and governance” and at least substantial for “sector strategies and policies”, the extensive and detailed PAs in DPO-2 requiring enactment of key legislation on banking sector governance was a credible effort to reduce associated risks. That said, the events of 2014 and 2015, ongoing resistance from vested interests in both the judicial and financial sectors, and the complexity of the ongoing fraud investigation, might have warranted higher risk ratings in this area. And while the substantial rating for risks associated with “institutional capacity for implementation and sustainability” was justified, the explanation in the text was generic, with little reference to specific reforms or agencies of concern. For example, technical and administrative factors were cited by the authorities to explain significant delays in transferring the registrars of securities owners to a Central Securities Depository which had been targeted for March 2018, but which is not expected to be completed before Spring 2019. Moreover, the justification for the 5 The text explicitly described risks to the macroeconomic outlook and banking sector as being “high”. While this was largely attributed to external developments (e.g., weakening of trade partners), concern was voiced about the effectiveness of banking regulators’ enforcement power. There was also concern about the about the lack of shareholder transparency masking deeper underlying problems in the banking sector. 6 The PAD argued that consultations with Parliamentarians on proposed measures requiring the approval of Parliament “partly mitigated” risks as did the ongoing engagement of donors and the Bank in the policy dialogue, technical assistance (TA), and awareness campaigns. Page 17 of 38 The World Bank (P149555) “institutional capacity for implementation and sustainability” rating focused on political dimensions of implementation rather than capacity which should have instead informed the rating on “political and governance” risk. Adequacy of Background Analysis/Analytical Underpinnings 19. The analytical underpinnings of the operation are relatively comprehensive and current, particularly for Pillar A, which benefited from timely work (“Policy Priorities for Private Sector Development in Moldova (June 2013) and the 2014 and 2015 “Cost of Doing Business Survey”). PAs related to social supported were informed by the ongoing project on “Strengthening the Effectiveness of the Social Safety Net”. While Moldova-specific background analysis for DPO-1 PA4 on the testing and registration for seeds and seedlings might appear dated (with analysis from 2007), this measure was largely an attempt to promote competition in a sector dominated by clear vested interests; the analytical foundations were well established and were informed by good practice in the EU. 20. The analytical underpinnings for banking sector reform in DPO-1 were contained in internal memoranda from the Bank’s Legal Department and the views of IMF staff. The analytical underpinnings of related PAs in DPO-2 were informed by FSAP work completed in June 2014 and the emerging results of the audit of BEM, which facilitated a strengthening and refinement of PAs. 21. A notable analytical gap was with respect to the PA related to PA10 in DPO-2 on efforts to use the agricultural support fund to make investment subsidies more efficient and equitable. Much of the impact of the PA (at least in terms of its aspiration to promote innovation) appears to hinge on an assumption about the types of machinery that support more innovative farming practices (with machinery not deemed to support innovative practices removed from the list of capital investment eligible for subsidies). This does not appear to have been based on a rigorous systematic assessment of the relationship between different types of machines and innovative farming but rather the undocumented opinion of a consultant about which machinery should and should not be subsidized. Incorporation of Lessons Learned 22. The series explicitly considered the need for PAs/triggers to involve both government and parliamentary adoption, as well secondary legislation (where relevant). 23. However, the PAD states that, among the lessons learned from previous operation was the importance that the operation’s design is selective and adjusted to the country’s administrative capacity, with fewer policy areas. This programmatic series (generically named “Development Policy Operation”), however, has three policy areas/pillars/PDOs, each with three sub-objectives, many of which are only marginally connected. Pillar A7 is the most integrated of the pillars, focusing on the lowering the costs of doing business. While the first two sub-components of Pillar B8 are closely connected, the addition of “access to finance” is only peripherally related, and supported only by DPO-2, PA7 on amendments to the Pledge Law. This pillar also contains PA8 that has little to do with strengthening financial sector stability but is instead motivated by a concern with the fiscal consequences for the GOM of putting its money in 7 Pillar A: Strengthen the regulatory framework to improve predictability of the business environment, facilitate competition, and reduce regulatory compliance costs. 8 Pillar B: Strengthen financial sector stability, promote transparency of shareholding, and ease conditions for access to finance Page 18 of 38 The World Bank (P149555) financial institutions that could subsequently fail. As such, it represents an additional objective/policy area that does not support any of the PDOs. Pillar C9 contains three largely unrelated policy reform areas. Overall, this implies six distinct policy areas, drawing into question the extent to which the operations’ design was selective, with fewer policy areas. Coordination with Development Partners 24. The key development partners for this series were the IMF, the EU, and the Government of Romania, with the IMF largely concerned with financial sector governance and stability, the EU focusing on public investment management, and Romania providing additional budget support. 25. Coordination with the EU was productive and contributed meaningfully to PAs in both operations on PIM. 26. The relationship with the IMF was more complex. The Fund welcomed proposed PAs to improve targeting of social assistance and triggers to improve the selection of banking institutions for the banking services of state and state-controlled institutions to protect public resources. But there were differences of opinion on how to respond to the governance shortcomings in the banking sector. Both the Bank and Fund were concerned with the adequacy of the amendments adopted in December 2013 to address the impact of the Constitutional Court ruling. However, while the Bank was prepared to proceed given that the amendments were a “step in the right direction”, the Fund concluded that, without strengthening, the amendments were insufficient for an IMF program. The Bank proceeded with the DPO based on a commitment from the GOM to (i) clarify ambiguities that were reducing the law’s effectiveness and (ii) begin an independent audit of BEM as a PA. This gave the Bank an opportunity to remain closely engaged and directly influence the policies to strengthen financial sector governance after the fraud came to light. Given the willingness of the Bank to proceed before the law’s shortcomings were fully addressed (a precondition for a resumption of Fund financial support), the Fund expressed concern that this would delay or prevent the needed steps on the Constitutional Court ruling from being taken. 27. In hindsight, both the Fund’s concern and the World Bank’s chosen strategy were both justified (see Assessment of Overall Design below) and the Bank’s calculated risk paid off. As a result, once the bank fraud was uncovered in the fall of 2014, and with the benefit of a (joint) FSAP completed in June 2014 and findings of the independent audit of BEM, the Bank and Fund were able to quickly to agree on a set of remedial and preventive measures that would be required for a resumption of both Bank and Fund lending. From this point onward (and leading up to the approval of DPO-2), the Bank and Fund worked in lock step, with a unified and carefully coordinated position on financial-sector reform priorities. Assessment of Overall Design 28. The design of the operation covered many of the main challenges faced by Moldova in the areas of World Bank expertise. About one half of the PAs came directly from the 2011 Country Economic Memorandum which set out a broad policy agenda for fostering competitiveness and achieving shared growth.10 These included PAs on strengthening competition, access to seeds and seedlings, RBI, PIM, and 9 Pillar C: Improve the public investment management framework, make investment subsidies in agriculture more efficient and equitable, and improve the coverage of well‐targeted social assistance programs 10 “Moldova After the Global Crisis: Promoting Competitiveness and Shared Growth”, World Bank, June 2011, Report No. 55195-MD, Policy Recommendations pages vi to ix Page 19 of 38 The World Bank (P149555) banking sector supervision. However, the PDOs lacked the focus and selectivity that experience in Moldova would have suggested and which had been identified as “lessons learned”. That said, there is little evidence that excessive breadth was the main reason for failure to make progress on key reforms. Bank staff explained the breath of areas covered by noting that the DPO was being used to support policy objectives already being pursued in several sectors and other operations where the Bank was active in Moldova (e.g., strengthening social safety nets, improving the effectiveness of public spending, lowering costs of production and promoting competition in the agricultural sector). However, not all these efforts were equally successful, with vested interests continuing to generate significant headwinds in some areas, and partial measures being pursued in others. 29. DPO-1 PAs were generally well designed given feasible (and often limited) prospects of achieving consensus within government at the time. This was particularly the case for PAs on the legal framework underpinning NBM independence which sought to « clarify » the existing framework for contesting the acts of the NBM. Particularly impactful in hindsight was PA7 which required the NBM to initiate an independent audit of BEM in compliance with terms agreed with the NBM. This proved instrumental in helping to bring to light the extent of the banking fraud which, given its fiscal and economic impact, eventually motivated the authorities to substantively address governance shortcomings. In other areas, and with a delay in approving DPO-2 of more than 24 months, the authorities managed to implement some of the DPO-2 triggers early on, allowing the Bank to tighten and make more ambitious several DPO- 2 PAs, particularly around banking-sector governance and inspection reform. 30. Two DPO-2 PAs could have been more clearly articulated. PA7, to amend the Pledge Law to facilitate use of movable assets as collateral, did not appear to address fundamental obstacles to easing conditions for access to finance. There was limited evidence that expanding the types of movable capital that could be used as collateral would meaningfully address the major reasons that commercial banks in Moldova were reluctant to accept movable assets as collateral nor was the associated results indicator well designed to capture progress toward the overarching objective of increasing lending.11 Specifically, there was little clear evidence presented that expanding the types of movable collateral that could be pledged was the binding constraint to making banks less risk averse in expanding lending to SMEs in Moldova. More meaningful parts of the reform to the Pledge Law were the creation of an out-of-court settlement mechanism to facilitate rapid resolution in case of default and the establishment of a notification registry to ensure transparency and clarify priority of lenders but little evidence was presented to suggest that either of these measures would make movable assets any more attractive. Moreover, the analytical underpinnings of this measure identify the creation of a modern online pledge registration regime as a key determinant of expanded lending, not expanding the range of eligible capital. 31. PA10, which narrowed the scope of the largest investment subsidy program and capped (or reduced caps on) subsidies allowed per one beneficiary, was intended to make investment subsidies in agriculture more efficient and equitable. While the measure clearly contributed to the second objective (by reducing subsidies to a few large entities), its link to the first objective (efficiency) was less clear. « Efficiency » in this context was defined in the PAD as a “more targeted approach to stimulating investments in innovation and high value sectors” which was being pursued by directly subsidies to innovative and HVA-related machinery. However, little evidence had been collected to show that smaller 11 Moreover, paragraph 71 acknowledged that “the availability of movable collateral and anticipated response of lending institutions are unknown, so that the expected impact of this measure (DPO-2, PA7) on access to finance cannot be quantified a priori”. Page 20 of 38 The World Bank (P149555) farms in Moldova were more innovative or efficient than large farms. As such, pursuit of the equity objective could potentially undermine achievement of the efficiency objective. The decision to remove several items from the list of farming machinery that could be subsidized was also not clearly linked to the promotion of innovation or HVA or based on documented analysis. Further blurring the underlying theory of change (including the link between the type of machinery and agricultural output) was the absence of a documented definition of HVA sectors and a lack of clarity on whether this RI applied to price or profit margin. 32. A key design issue was the timing and content of the first operation in light of increasingly suspicious activity in the banking sector and the reluctance or inability of the GOM to take decisive action to address governance shortcomings. At the ROC Decision meeting for the series in January 2014, there was considerable debate about the adequacy of measures to strengthen banking-sector governance. As noted above, the IMF did not consider the framework in place to be adequate to proceed with a new arrangement, particularly given accumulating evidence of suspicion transactions. Fund staff were of the view that a budget support operation before shortcomings were addressed would reduce pressure on the authorities to adequately respond to the actions of the Constitutional Court that had undermined the independence of the NBM, particularly with respect to the conduct of effective banking supervision. Bank staff acknowledged shortcomings in the framework and of the GOM’s remedial actions at the time of the Decision Meeting on DPO-1 but described the GOM’s actions as a “step in the right direction” and as “reducing risks” for banking regulation and supervision.12 The PA intended to address concerns with the underlying legal framework (PA 6), while necessary, sought only clarification of the legal framework for contesting acts of the NBM. This was intended to strengthen the implementation of the existing (but inadequate law) by limiting the scope from some parties to use ambiguity in the existing law to prevent the NBM from carrying out its responsibilities in a timely manner. 33. However, the Bank’s decision to proceed in the Spring of 2014 effectively hinged on agreement with the GOM to launch an independent diagnostic audit of BEM (including selecting the auditor). While this action did not itself address weakness in the legal framework for banking supervision, it proved instrumental in uncovering evidence of fraud in BEM which helped persuade the authorities of the need to address the remaining shortcomings as well as to launch investigations into possible wrong doing in other banks. This was a pragmatic and politically astute strategy for revealing the extent of the problem with BEM, thereby strengthening the position of those within the GOM and NBM seeking to strengthen the governance framework in the banking sector. While this approach directly contributed to the more than two-year delay in approving DPO-2, it played a key role in bring about the concrete improvements in banking-sector governance embodied in PA3 and PA4 of DPO-2. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization 34. Results indicators covered the key pillars and policy areas, although there were some gaps and misalignment. A few indicators were not clearly defined, and documentation of definitions was insufficient13. Some indicators and targets were changed. While in some cases this improved the realism of the target or appropriateness of the indicator, there was little explanation of the reasons for changes. 12 The final decision on the timing of the operation was referred to Bank management. 13Documentation on RI definitions and calculations was lost when a Bank computer was stolen (there were no backups and the authorities did not retain documentation when the monitoring function was transferred from the State Chancellery to the Ministry of Finance). Page 21 of 38 The World Bank (P149555) After the series had closed, the authorities transferred responsibility for monitoring performance relative to the results indicators from the State Chancellery to the Ministry of Finance. However, capacity, institutional memory and familiarity with the framework were lost in the process. Given the length of time between the two operations, an ISR should have been prepared to assess performance but none was undertaken nor was a waiver requested. M&E Design 35. The first results indicator (RI) in DPO-1 is reduced “inspection coverage”. However, there is no common definition of “inspection coverage” and loss of documentation has made it difficult to assess progress. It is not clear if it refers to the share of firms that have had at least one inspection (relative to the universe of firms), the average number of inspections experienced by firms, or the number of firms visited by inspectors.14 While there is precedent in IFC to define it as the number of firms visited by inspectors, this RI is expressed as a share which implies an assumption about changes in the number of firms in existence but the basis on which such an assumption would be made is unclear. The PDO sub- component that this RI was associated with was to “strengthen the regulatory framework to improve predictability of the business environment and reduce regulatory compliance costs”. There were two elements to the associated PAs -- the use of a RBI methodology and the creation of an on-line State Registry of Inspections. The first would result in firms at low risk receiving fewer inspections, with inspection efforts concentrated on those that are deemed riskier. This would require a RI that capture the distribution of inspections among firms (not just the average number of inspections) which “inspection coverage” does not seem to do. The second element (the on-line registry) was intended to both improve the predictability of inspections and eliminate unannounced inspections but it is unclear how this RI captured this objective. 36. RI2 in DPO-1 referred to the share of proposed customs- and tax-related legislation and regulations that affect companies and that have been subjected to an RIA by the RIA Secretariat before adoption. An initial target was set at 100 percent by 2015 but changed to 80 percent for 2017 in DPO-2. There was no documentation to explain the reasons for the change in target. Moreover, no criteria were set to prioritize which legislation and regulations should include a RIA given the reduction in the target below 100 percent. 37. RI3 (State Aid providers connected and reporting to the State Aid automated information system) was initially defined as the number of State Aid providers reporting but was changed in DPO-2 to the share of State Aid providers. This change was appropriate given that absence of a clear inventory of providers of authorized State Aid at the time of DPO-1, which was rectified by the time of DPO-2 (although tax exemptions were underreported by the Ministry of Finance and were not included as part of State Aid). 38. DPO-1 contained a results indicator (RI-8) on the share of loans in the system secured by movable collateral, with a target of a 3.3 percentage point increase between 2013 and 2015 (changed to 4 percentage point increase by 2017 in DPO-2). This indicator--the share of loans secured by movable collateral relative to the stock of outstanding loans—was calculated as the share of the total value that was secured by movable collateral. However, the Moldova 2020 strategic priority on which this policy area was based was to reduce financing costs, which was expected to lead to a “significant increase in 14 As noted elsewhere, documentation of indicator definitions and calculations has been lost. Page 22 of 38 The World Bank (P149555) financing volumes”15 but the expected result of the associated PA was an increase in lending secured by movable collateral16. An RI linked to the volume of lending secured by movable collateral rather than share of loans secured by movable capital would have been better aligned. Alternatively, an RI linked to interest rates charged by financial institutions when lending against movable collateral could have been used to track the potential improvement in financing costs. Also, since this provision largely pertained to new lending, it might have been preferable to define the RI denominator as a share of new loans, with the baseline defined relative to the stock of existing loans to better isolate the impact of the reform. 39. As noted, RI-12 associated with DPO-1 on the share of subsidies supporting HVA sectors was problematic given a lack of documentation on the criteria used to categorize a sector as HVA as well as the sectors that were considered HVA. These were discussed and agreed with the authorities at the outset but not documented. 40. The unit of account for RI-13 and RI-13 on Ajutor Social beneficiaries and Heating Allowance recipients was changed from “number of households” in DPO-1 to “number of persons/women” in DPO- 2 to allow for a sub-indicator on the number of women recipients. While the introduction of a gender disaggregated indicator was well intentioned, eligibility for both benefits is defined by the GOM by “household”. Converting this to persons/women required assumptions that may have diluted the quality of the data. Since gender coverage was not identified as an explicit objective of the PA, the net value added of this change was unclear. 41. RI9 (share of tenders for banking series that include criteria related to regulatory standards), was more in the nature of a PA than an indicator of impact. The authorities achieved the 100 percent target for 2017 by issuing an order on the last working day of 2017 (No. 162 at on 12/28/17) that excludes privileged access to public-sector contracts for banking services (Section 1.6). As no tenders were required in 2017, none contravened the new requirement17 and the target of 100 percent for 2017 was “technically” met although it does draw into question the value added of a results indicator to support this PA. M&E Implementation 42. The State Chancellery was assigned responsibility for supervising the operation, monitoring progress in the policy areas supported by it, and for coordinating the monitoring of the results indicators.18 However, in 2017, this function was transferred to the Ministry of Finance although the extent to which the transfer included the necessary institutional capacity to conduct effective monitoring of results. As a result, there was some confusion as to what several of the RIs meant or how they were to be monitored (see para 42, for example). 15 Moldova 2020: National Development Strategy”, page 30 16 Program document for Moldova: Second Development Policy Operation, paragraph 39 17 The Law on Public Finances and Budgetary-Fiscal Accountability no.181 (July 25, 2014) requires that the procedure to select the commercial banks for cash budget execution be undertaken every three years. The last agreement was signed September 18, 2015. Therefore no tenders have been issued for this purpose since the PA was implemeented. A new tender should be issued that conforms to Order No. 162 at on 12/28/17 before the deadline in 2018. 18 Paragraph 67 in Moldova: Second Development Policy Operation. Page 23 of 38 The World Bank (P149555) M&E Utilization 43. With more than 24 months between the first and second operation, the team was required to prepare an Implementation Status and Results Report (ISR), but this was not done, and it does not appear that a waiver was requested (or granted). An ISR would have provided an important benchmark and input into the design of DPO-2, including with respect to the numerous adjustment of targets.19 For example, the target for the share of State Aid providers connected and reporting to the State Aid automated information system had been set at 100 percent (RI3). The target was not met because, according to the authorities, not all State Aid providers had the computer skills to use the new system. This capacity constraint would likely have been uncovered in an ISR, allowing technical assistance to be applied in time to meet the target. Also, with respect to the FY14-17 CPS, IEG raised a major concern that no ISRs were filed. Had this been done, it would have allowed a better assessment of progress and achievement of outcomes. 2.4 Expected Next Phase/Follow-up Operation 44. The next budget operation took about one and a half years to prepare and was a stand-alone operation (Economic Governance Development Policy Operation [EGDPO]). The EGDPO supported reforms to strengthen economic governance linked to the management of public resources to reduce fiscal risks and to level the playing field for private sector development, including in the core economic sectors of agriculture, banking, and energy. 45. Given the deep-seated structural governance problems that had been uncovered by the banking fraud, follow-up financial sector TA focused on addressing remaining risks in the system, including the design and execution of strategies to transform the ownership structures of distressed banks, allowing control to be transferred to reputable investors; identifying and declaring ultimate beneficial owners; improving the governance, operations, and risk management; drafting legislation to ensure the necessary legal powers to resolve and/or intervene in distressed banks appropriately and in a timely manner, and on bank resolution strategies. Current Bank TA in the financial sector is focused on the insurance sector. 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation Overall Relevance: Substantial Relevance of Objectives: Modest 46. Moldova’s National Development Strategy (NDS) Moldova 2020 set seven strategic priorities: (i) national education system aligned with labor market requirements to enhance labor productivity and increase employment; (ii) improving roads infrastructure to reduce transportation costs; (iii) reducing financing costs by increasing competition in the financial sector and developing risk management tools; (iv) improving the business environment, promoting competition and streamlining the regulatory framework; (v) increasing energy efficiency and encouraging reliance on renewable sources of energy; (vi) ensuring financial sustainability of the pension system; and (vii) increasing the quality and efficiency of 19 ISRs are to be prepared on a 12-month cycle and at least one ISR is prepared for each individual operation before it closes. Page 24 of 38 The World Bank (P149555) justice and fighting corruption. The CPS for FY14-1720 was organized around three pillars: (i) increasing competitiveness; (ii) enhancing human capital and minimizing social risks; and (iii) promoting a green, clean and resilient Moldova. 47. The objectives of this series21 were mostly supportive of the fourth and seventh strategic priorities of Moldova 2020, with modest support for the third priority. Pillar C contributed to increasing the efficiency of public expenditures, which, while not an explicit “strategic priority” was identified in the NDS as a determining factor in the achievement of national priorities. The series was mostly aligned with the first CPS pillar (“increasing competitiveness”) which sought to support “institutional reforms for a business enabling environment and governance, access to finance, transparency in the financial sector and targeted activities to help improve companies’ competitiveness”22 and somewhat aligned to aspects of the third pillar (“minimizing social risks”). In keeping with the Bank’s twin goals of reducing poverty and promoting shared prosperity, the series supported expanding coverage of well targeted social assistance programs and improving the equity of agricultural investment subsidy allocation. 48. The diffusion of PDOs implies varying degrees of relevance across the elements of the three pillars, with those pertaining to the financial sector increasing in relevance as the series was implemented. On balance, the policy areas in the objectives/pillars remained relevant to Moldova’s development agenda. However, the PDO for Pillar B was overly narrow, incorporating only financial-sector stability and transparency of shareholding rather than a broader banking-sector governance agenda which dominated risks to macroeconomic stability. As the series evolved, a shift in the formulation of this PDO toward a more explicit strengthening banking sector governance and supervision and promoting the rule of law in the banking sector would have strengthened the relevance of the series’ objectives (this shift occurred for the underlying PAs but not in the PDO itself). 49. For these reasons – diffusion of PDOs, several of which focused too narrowly, with no clear prioritization -- the relevance of the objectives is rated modest. Relevance of Design: Substantial 50. The PAs in the series supported (i.e., were relevant to) the PDOs to varying degrees, with some clearly linked to achieving the objectives via a coherent results chain, while a few were either not well linked or were not sufficient to achieve the associated objective. 51. PAs under DPO-1 (PAs 1 and 2) and the associated PA under DPO-2 related to RBI and RIA had a well-articulated impact on the predictability of the business environment and DPO-1 PA3 (and the associated PA under DPO-2) was clearly linked to the objective of facilitating competition (by reducing the scope for State Aid to be used to tilt the playing field toward particular firms and sectors). PA 4 on access to seeds and seedlings was intended to reduce the monopoly power of vested interests (including seed importers) and regulatory compliance costs, thereby lowering the costs of agricultural inputs and enhancing agricultural productivity. DPO-1 PA5 (and the associated PA in DPO-2) directly contributed to promoting transparency of shareholding which (along with DPO-2 PA 3) was intended to improve governance in the banking sector (thereby contributing to financial sector stability) which, as argued 20 August 2013 21 The PAD indicates that the PDOs for the series are also the policy “pillars”. 22 “Country Partnership Strategy for Republic of Moldova for the Period FY14 -17”, August 9, 2013, Report No. 79701-MD, page ii Page 25 of 38 The World Bank (P149555) above, was the overarching objective of much of Pillar B. DPO-1 PA6 and PA7 (and DPO-2 PAs 4, 5 and 6) were also motivated by a desire to address governance shortcomings in the banking-sector (and related concerns with financial sector stability)with DPO-1 PA6 and DPO-2 PA4 intended to strengthen NBM independence (particularly with respect to banking supervision) and DPO-1 PA7 and DPO-2 PAs 5 and 6 intended to help collect evidence on suspicious transactions on selected banks. DPO-1 PAs 8 and 9 (and the associated PAs under DPO-2) were clearly relevant to the achievement of the first and third elements of Pillar C. 52. On the other hand, DPO-2 PA8 (banking service tender) was somewhat removed from the financial stability objective and was largely a fiscal measure (for which there was not explicit PDO) intended to protect government deposits that might be placed with unstable banks for non-transparent reasons. Indeed, were the GOM seen to be removing its deposits from a bank that did not meet regulatory requirements, it could conceivably precipitate a bank run. Relevance of Implementation: High 53. The implementation assistance provided by the Bank was highly relevant to changing circumstances across all three pillars. Most notably, with the banking fraud coming to light in the fall of 2014 (partly aided by the information that became available due a PA in DPO-1), the team rightly responded with a set of comprehensive requirements that had to be met before budget support could resume. These were among the most significant adaptions of the original set of reforms supported by the series. The measures were agreed jointly with Fund staff, with negotiations on PDO-2 closely coordinated with IMF negotiations on a new Fund-supported program for Moldova. Both the revised set of PAs and the process by which they were pursued significantly increased leverage in support of a credible strengthening of banking-sector governance. 3.2 Achievement of Program Development Objectives PDO 1/Pillar A: Strengthen the regulatory framework to improve predictability of the business environment, facilitate competition, and reduce regulatory compliance costs. 54. Improvements in the predictability of the business environment and the costs of regulatory compliance were pursued with a more focused (risk based) approach to inspections of private-sector entities. This fostered greater awareness of the impact of laws and regulation on businesses through use of regulatory impact analysis (RIA), and reducing the time required to test new seeds and seedlings. Competition was increased through enhancements in transparency and accountability in the use of State Aid, the reduction in and better tracking of entities allowed to conduct inspections, and by reducing the dominance of vested interests in the supplying of seeds and seedlings to Moldovan farmers. The reforms supported by the series were successful in achieving these objectives although RIAs were undertaken for a smaller share of customs and tax legislation than planned (60 percent versus a target of 80 percent) and only 88 percent of State Aid providers were using the online system to report by the 2017 target date23. 55. Competition in the provision of seeds and seedlings available to farmers was increased by the target date, but with only 166 of the targeted 210 seed varieties registered by the 2017 deadline. This exceeded the original 2015 target but fell short of the revised target. Performance improved markedly in 23While all state aid providers were connected by the target date, only 88 percent used it to report due to “a lack of adequate computer skills” in some agencies. Page 26 of 38 The World Bank (P149555) the subsequent year (with 198 seed varieties registered), with the small shortfall attributed to the exclusion of those varieties for which testing by the State commission did not show better results for Moldova. PDO 2/Pillar B: Strengthen financial sector stability, promote transparency of shareholding, and ease conditions for access to finance 56. Financial-sector stability was being pursued by promoting transparency on bank shareholding (to improve clarity on ultimate beneficial owners), strengthening independence of the NBM, and investigating and, where appropriate, closing banks involved in illegal behavior. Conditions for access to finance were to be eased by removing impediments to the use of movable capital as collateral in borrowing. 57. The first objective was achieved, with the DPO series helping catalyze important changes in the legislative framework underpinning the independence of the NBM. The DPO series sought to track achievement of this objective by monitoring progress in liquidating the three problem banks, and by tracking the implementation of action plans for three other banks based on the findings of diagnostic studies (RI8). The latter action was partially achieved, with Moldincombank cancelling the contract with the audit company for the diagnostic study so that no action plan was formulated from it. The cancellation and decision to not base the action plan on the audit’s findings was inconsistent with the spirit of the policy measure. Instead, Moldincombank developed its own, less robust, action plan based on a series of inspections, with the plan still being implemented. More broadly, the strengthening of financial sector stability largely achieved via a large bailout of individual banks by the GOM, with investigations into wrong doing still ongoing. 58. Efforts to the promotion of transparency of shareholding by consolidating and centralizing the securities registry to facilitate the identification of beneficial owners of securities were partly successful. A lack of consensus on the preferred model to be pursued resulted in a longer than expected period of consultation which delayed the transfer of responsibilities and the establishment of a CSD. However, the CSD has not been set up as a legal entity and the transfer of government securities to the CSD was accomplished in late summer 2018. Agreement on a roadmap for the establishment of the CSD was only reached in March 2018, the original target date for transfer of all registers of securities owners, so that no records were transferred by the target date. A proposed amendment to the CSD Law would postpone the deadline for public interest entities (i.e., banks and insurance companies) to Spring 2019. 59. The RI used to track the share of loans secured by movable capital shows a decline over the period for which data was provided. The share of collateral that was not in the form of financial assets, official guarantees and immovable collateral declined from 33.6 percent at end December 2016 to 29.0 percent at end December 2017 and further to 27.5 percent by end June 2018. PDO 3/Pillar C: Improve public investment management (PIM) framework, make investment subsidies in agriculture more efficient and equitable, and improve coverage of well‐targeted social assistance programs 60. All parts of Pillar C were achieved or almost achieved. The objective of improving PIM was achieved through improvements in the associated regulatory environment. The necessary statutory changes were made but compliance still needs to be assessed more fully. While there were improvements Page 27 of 38 The World Bank (P149555) in the equity of agricultural investment subsidies through limits on the share of subsidies that could go to individual producers, the impact of efforts to use the allocation of subsidies to promote efficiency and innovation in the agricultural sector was not assessed nor was it clear how the desired result would follow from the prior action. The coverage of well -targeted social assistance programs, on the other hand, appears to have been achieved and sustained. 61. Table 5 shows Moldova’s scores on various dimensions of the World Economic Forum’s Global Competitiveness Report, comparing the results of the 2013-14 report with those of 2017-18 (the most recent) to assess improvements over the series in areas of relevance to the PDOs. The dimensions, to the extent possible, are linked to the relevant PDOs. The results show little or no improvement in most dimensions, including those related to the costs of regulatory compliance, competition, and public investment management. The notable exceptions are the substantial improvement in the ease of access to finance and a sharp deterioration in the soundness of banks (a proxy for financial-sector stability). Another indicator of relevance to assessing access to finance is the number of resident customers that are nonfinancial corporations (public and private) and households who obtained loans from commercial banks and other banks functioning as commercial banks. In 2013, Moldova had 46.7 per 1000 adults, falling to 40.4 in 2016 before jumping to 64.5 in 2017. It is unclear if the use of movable capital as collateral contributed significantly to improved access to credit as the change was instituted in mid-2014. Moreover, developments in the sector would have been impacted by the three bank failures that took place in the wake of the crisis. The increase in access also occurred alongside an increasing from NPL ratio, from 9.9 percent in 2015 to 18.4 percent in 2017. These developments, and the deterioration in the score for banking sector soundness, draw into question the sustainability of the increase in access to credit. Table 5: Change in Relevant Scores from the Global Competitiveness Index Global Competitiveness Indicator Relevant PDO 2013-2014 Score 2017-2018 Score Scale of 1 to 7 (worst to best) Burden of Government Regulation24 PDO1 2.8 2.8 Competition25 PDO1 4.2 4.3 Extent of Market Dominance26 PDO1 3.0 3.0 Ease of Access to Loans27 PDO2 2.4 3.2 Soundness of Banks PDO2 4.1 2.6 Efficiency of Government Spending28 PDO3 2.6 2.7 24 In your country, how burdensome is it for companies to comply with public administration’s requirements (e.g., permits, regulations, reporting)? [1 = extremely burdensome; 7 = not burdensome at all] 25 The Global Competitiveness Report indicators on competition are: intensity of local competition; extent of market dominance; and effectiveness of antimonopoly policy. 26 In your country, how do you characterize corporate activity? [1 = dominated by a few business groups; 7 = spread among many firms] 27 In your country, how easy is it for businesses to obtain a bank loan? [1 = extremely difficult; 7 = extremely easy] 28 In your country, how efficient is the government in spending public revenue? (1 = extremely inefficient; 7 = extremely efficient) Page 28 of 38 The World Bank (P149555) 3.3 Justification of Overall Outcome Rating Rating: Moderately Satisfactory 62. As noted, DPO-1 was approved in March 2014, despite concerns with banking sector governance and with growing concern with suspicious transactions by at least one bank. When the full extent of the fraud came to light later in 2014, the timetable for DPO-2 was appropriately delayed as development partners suspended aid and the budget experienced a large hit to bail out depositors. Achievement of some of the PDOs experienced setbacks as budgetary resources faced competing pressures and the subsequent political turbulence delayed key policy decisions. As a result, agreement on DPO-2 took more than 24 months from the approval of DPO-1, necessitating an extension of the target date for results indicators from 2015 to 2017 although two of the DPO-2 triggers (10 and 11) were achieved as planned in 2015,29 with the impact of the measures strengthened by extending the targets into 2017 and adjusting the eligibility thresholds to take into account increases in the costs of living between 2015 and 2017. 63. Performance of RIs relative to baselines and targets is summarized above (pages vii to x). Of the 15 targets, just over half were met or almost met. Two of the four targets for Pillar A results indicators were partially met, and there was a lack of clarity on the actual measure in the case of another indicator. Shortfalls were due to a range of factors, including capacity constraints, overly optimistic timeframes for implementation and resistance from vested interests. Two of the five Pillar B targets were met on time and one target was met with a delay. All Pillar C targets were met or almost met. 3.4 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 64. The PDOs supported by this operation were aimed at reducing poverty and promoting shared prosperity by (i) helping provide business growth opportunities over the medium to long run for the whole economy and (ii) introducing specific measures targeting the poor. In the second operation, a PA was introduced to rebalance subsidies in agriculture to increase participation of small farmers while increasing innovation. PAs were implemented to support well targeted social assistance programs (Ajutor Social and heating allowance programs) aimed at expanding the number of beneficiaries and reducing vulnerability of the poorest during the cold season. In addition, RI14 and RI15 under DPO2 were changed to disaggregate female beneficiaries to take into consideration gender although as discussed above, with implications for the quality of data.. 65. Increases in pensions, real wages and remittances drove the decline in poverty between 2012 and 2018. Measured at the international moderate poverty threshold of $5.5/day is estimated to have declined from 24.9 percent in 2012 to an estimated 13.6 percent in 201830. Labor market dynamics have also contributed to the reduction in poverty as employment rose in rural areas. Agriculture, which employs the poorest and most vulnerable Moldovan, grew much faster than the rest of the economy except for 2015. As agriculture expanded, real wages in this sector also rose faster and supported the poverty reduction. 29 See Table 2: Adaptation of DPO II Triggers into PAs. 30 The latest data available on the poverty headcount is of 2016 and is at 16.5 % Page 29 of 38 The World Bank (P149555) (b) Institutional Change/Strengthening 66. Since the end of the second operation, the governance of the banking system improved although major challenges remain. A new Banking Law was approved which significantly strengthened the capacity of the NBM to efficiently and promptly supervise the banking sector in Moldova. All major banks undertook external audits and the ultimate beneficial owners identified and scrutinized using a “fit and proper” mechanism. Prudential and supervision principles have been tightened and the Moldovan banking system is preparing to implement Basel III principles. The new Banking Law also introduced new corporate governance elements and risk management principles. The central depositary under the NBM has been created where shares in banks and public interest entities must be registered although implementation is pending. In addition, there have been improvements in capacity in public investment management, the burden of inspections on firms has been reduced, agricultural investment subsidies are being more equitably distributed, and competition has been enhanced. (c) Other Unintended Outcomes and Impacts 67. None 4. Assessment of Risk to Development Outcome Rating: High 68. The risks to development outcomes are high. Corruption remains a serious concern in Moldova31, including in the implementation of the current Economic Governance DPO. oligarchs retain significant influence including over the political process. Investigations into bank fraud are ongoing, with no major convictions to date. While the statutory changes affecting banking-sector governance are adequate in establishing the underlying legal framework, vested interests remain strong in Moldova, presenting challenges for implementation. The Prosecutor General continues to bring charges against employees of the NBM for actions taken in the conduct of their duties. While convictions have become rare, this suggests continuing efforts to see proper oversight of the banking sector undermined. Actions to ensure that targeted social program remain relevant and that agricultural subsidies are distributed equitably require regular reinforcement in the context of annual the budget process and cannot be guaranteed, particularly in a fiscally constrained environment, and in a political environment characterized by frequent changes in government. Improvements in public investment management have been promising but require ongoing political support and capacity building. 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Moderately Satisfactory 69. “Quality at entry” measures the extent to which the Bank identified, facilitated preparation of, and appraised the operation such that it was most likely to achieve planned development outcomes and was consistent with the Bank’s fiduciary role. The breadth of the series, the related abundance of 31Moldova continues to perform poorly on Transparency International’s Corruption Perception Index, ranking 122 out of 180 countries in 2017, with a score of 31, below its score in four of the last five years and an average between 2012 and 2016 of 34. Page 30 of 38 The World Bank (P149555) objectives and the diffusion of PAs, presented challenges in reinforcing reforms and in establishing clear reform priorities. In part, this was justified by the need to leverage work being pursued through other operations to technical support. However, some measures were too partial to have an impact (e.g., efforts to enhance SME access to finance and to promote innovation in the agriculture sector) or were left to the final operation in the series, providing no opportunity to adapt and reinforce based on experience in implementation. 70. Bank staff deftly used this series to catalyse meaningful change in the banking sector (Pillar B), moving forward when justified by the merits if taking a calculated risk, pausing when the policy making environment was not conducive to meaningful reform, and proceeding, in collaboration with development partners and in support of a complementary and coherent set of actions. DPO-1 took a calculated risk in the absence of an IMF program and with serious shortcomings in banking-sector governance still present. However, this risk was justified and at least partially offset by the inclusion of a PA requiring an audit of BEM intended to produce concrete evidence of alleged bank fraud. The risk paid off given the enhanced insight that the required audit provided into banking-sector activities, eventually helping to forge the political will to confront governance shortcomings and criminal behavior more directly. Moving forward with DPO-1 also enabled Bank staff to retain a “seat at the table” and help shape the substantive reforms that were put in place as PAs for DPO-2 where the Bank correctly took sufficient time to obtain agreement (including with the IMF) on Pillar B PAs which were detailed, concrete and monitorable. 71. Efforts to support targeted social safety nets were also well conceived and adapted as necessary as were efforts to improve competition and reduce regulatory compliance. This was less the case in other areas of the program, which were partial and minimally effective. For example, reforms to agricultural support were not well aligned with the expressed desire to promote efficiency and innovation and the measure to improve access to finance was too partial to have a meaningful impact. That said, these measures were largely peripheral to the series and their exclusion could have resulted in a more focussed program, perhaps opening the way for additional reforms that reinforced more central objectives. 72. The main weakness was with respect to the M&E framework. Several the results indicators could have been better, and more clearly, defined, with definitions included in program documentation. While explicit agreement to have the State Chancellery assume centralized responsibility for M&E of implementation and results over such a multi-sector operation was appropriate, the positive impact of this was lost when this function was transferred to the Ministry of Finance which lacked expertise and institutional memory. Because of a theft of a laptop, Bank staff did not independently retain records of indicator definitions and calculations. (b) Quality of Supervision (including of M&E arrangements) Rating: Moderately Unsatisfactory 73. “Quality of supervision” refers to the extent to which the Bank proactively identified and resolved threats to the achievement of relevant development outcomes and the Bank’s fiduciary role. The adaptation of the Pillar B triggers into a set of strengthened and concrete PAs was the clearest example of the Bank effectively addressing a major threat to the achievement of program objectives. However, M&E arrangements overall were sub-par, with a lack of clarity and documentation on definitions and calculations. The ISR that was required due to the time between operations was not done nor was a waiver requested. Page 31 of 38 The World Bank (P149555) (c) Justification of Rating for Overall Bank Performance Rating: Moderately Satisfactory 74. The series supported meaningful progress in several areas including reduced the administrative burden on businesses, more equitable distribution of agricultural investment subsidies, improved management of state aid, an expansion of targeted social assistance, enhanced competition, and better public investment management. But the most noteworthy and impactful achievement of this series was its contribution to far reaching reform of the banking sector and important improvement in PFM. Unfortunately, this was not well captured in the results indicators. The rating of moderately satisfactory reflects progress in the above areas, particularly the important impact on the banking sector which was approached by Bank staff in a nuanced and thoughtful way. Progress in supporting the extension of targeted social protection (which has been sustained) was also significant as was efforts to improve PFM and PIM. These achievements are substantive and central to the overall development of Moldova and overshadow shortcomings in other aspects of the series (e.g., support to innovation in the agriculture sector and improved access to finance) as well as a weak M&E framework. 5.2 Borrower Performance 75. This refers to the extent to which the borrower (including the government and implementing agencies) ensured quality of preparation and implementation, and complied with covenants and agreements, toward the achievement of development outcomes. (a) Government Performance Rating: Moderately Unsatisfactory 76. The implementation of the series was protracted partly due to a lack of consensus at the highest levels of government on the required reforms, with vested interest in the banking and agriculture sector complicating implementation. However, much of this—particularly with respect to the banking sector-- was overcome in the second operation even though the reform agenda remains incomplete in important respects. The loss of continuity and expertise in monitoring the impact of the operation through the results indicators when the function was transferred out of the State Chancellery was regrettable. (b) Implementing Agency Performance Rating: Moderately satisfactory 77. The main implementing agencies were the Ministry of Finance, Ministry of Agriculture, Regional Development and Environment, National Bank of Moldova, Ministry of Health, Labor and Social Protection, and Competition Council. Assessment of overall performance is complicated by the breadth of reforms supported by the DPO series. Important steps were taken to reduce the costs of regulatory compliance although the full benefit has yet to emerge and the coverage of well-targeted social assistance programs has been expended and benefits enhanced. There have been some improvements in transparency and accountability in the use of State Aid, but compliance with measures to track State Aid and to open the supplying of seeds and seedings to greater competition was incomplete, suggesting room to further enhance the benefits of the reforms undertaken. The allocation of agricultural investment subsidies has become more equitable but their contribution to a more efficient and innovative sector is less obvious. Public investment management framework has improved with public investment becoming more predictable and transparent. However, under-execution remains a problem due, in part, to heavy Page 32 of 38 The World Bank (P149555) reliance on external financing, the timing of which can be uncertain given delays in project appraisal, selection and implementation. Delays in road infrastructure spending, particularly in 2017, were also related to bottlenecks in implementation by contractor companies. 78. A significant number of program measures fell to the NBM to implement and these were largely successful. The securities depository function was centralized in the NBM and while the transfer of records is incomplete, this largely reflects delays by the GOM in settling on the appropriate model to follow and technical issues related to CSD software rather than a lack of NBM commitment to the reform. The legal framework for governance of the banking sector has been improved and regulations are being brought up to international standards. At the same time, employees of the NBM still subject to periodic legal challenges to actions in the conduct of their duties. The independent audit of BEM supported by DPO-1 contributed to the broader investigation of the nature and modalities of the fraud in the three closed banks. There have been numerous prosecutions, with efforts to recover misappropriated finances are ongoing. (c) Overall Borrower Performance Rating: Moderately Satisfactory 79. Despite delays in implementation, the ambition of the program and commitment to implementation increased over time. Reforms undertaken have been largely sustained; challenging, important and credible progress in strengthening banking-sector governance has been made. The relative importance of these positive developments more than offsets the shortcomings in more peripheral policy areas and problems in maintaining monitoring capacity. 6. Lessons Learned 80. PDOs should be clear, focused and not overly complicated; compound objectives should be used sparingly, if at all. In this operation, the three-pillar PDO structure embodied as least six distinct sub- objectives, some of which were only peripherally related to each other. The PD states that the PDOs were the same as the “pillars”. However, PDOs are not the same as “pillars” (or “policy areas”) which are the channels through which PDOs are to be achieved, with reforms/PAs serving as the mechanisms used to achieve PDOs. 81. While multi-sector DPOs can be effective, casting the net too widely with insufficient complementarity or synergy between reforms can result in partial and largely ineffective interventions. DPOs should settle on a limited number of sectors, supported with complementary reforms that generate synergies, are sufficient to achieve meaningful results, and that can be explained through clearly articulated results chains. Excessively diverse PAs risk diverting limited implementation capacity and political capital. They also complicate meaningful monitoring of impact and cloud messaging around prioritization and the primary objective of underlying reform measures. In the case of this DPO, the PA on amending the Pledge Law was not well supported by other actions and was too narrow to credibly confront underlying constraints. The PA on agricultural investment subsidies was adequate to meet the equity dimension of the PDO sub-component but did not meaningfully support the “efficiency” sub component. 82. Project design that includes in the first operation objectives or PAs in particular sectors that only apply in the final DPO in the series are risky as they provide no opportunity to adapt triggers or Page 33 of 38 The World Bank (P149555) targets to ensure adequate implementation and the achievement of results. Unless actions in these sectors provide clear reinforcement to other objectives in the DPO or are sufficiently comprehensive to achieve results they should be avoided. 83. Results indicators should be clearly and unambiguously defined as should the method for calculating them. This was the case with most but not all results indicators in this operation. To avoid a loss in indicator definitions and calculations, and promote transparency and accountability, this information could helpfully be summarized in program documentation or, at least contained in a single document. 84. Results indicators should have clear value added and should not only replicate prior actions. A few of the RIs were more in the nature of prior actions (i.e., the number of banks with implemented actions plans). Results indicators should, where possible, measure impact/outcome rather than mirror the activity required by the PA. 85. It is good practice that, when an operation goes off track, the Bank (in collaboration with relevant development partners) should clearly articulate and communicate to the authorities the actions required to resume budget support. This was done well in this DPO, and in the wake of the bank fraud, with required actions agreed with the main development partner (i.e., IMF) and communicated unambiguously in a letter to the Prime Minister. This minimizes the potential for the client to arbitrage between development partners and help focus the attention of decision makers at the highest level. 86. For programmatic DPOs, it is critical that ISRs be completed when required to increase opportunities to systematically reflect on progress and challenges in implementation and adapt operations as appropriate. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/Implementing agencies No comments or objections. (b) Other partners and stakeholders No substantive disagreements; factual corrections reflected as appropriate. Page 34 of 38 The World Bank (P149555) Annex 1: Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Responsibility/ Names Title Unit Specialty Lending Ruslan Piontkivsky Sr. Economist GMTLC TTL Lily Begiashvili Sr. Private Sector Specialist GFCE2 Private Sector Arcadi Capcelea Sr. Environmental Specialist GEN03 Environment Marcel Chistruga Economist GMTE3 Overall coordination Brett E. Coleman Program Manager GFCFT Financial Sector Elena Corman Procurement Specialist GGOPC Procurement Mame Fatou Diagne Sr. Economist GMTA2 Overall coordination Oxana Druta Fiscal Management Specialist GGOEE Financial Management Anatol Gobjila Sr. Agriculture Economist GFA03 Agriculture Irina Guban Consultant GHN03 Health Caterina Laderchi Ruggeri Sr. Economist GPV03 Economic framework Ala Muntean Executive Assistant ECCMD Administrative Support Cesar Niculescu Sr. Environmental Specialist GEN 03 Environment Carolina Odobescu Sr. Country Officer ECCMD Overall coordination Private sector, competition Georgiana Pop Sr. Economist GMTCI policy Felicia Pricop Private Sector Specialist GFCE1 Private Sector Melissa Metz Sr. Private Sector Specialist GFCEW Private Sector Iryna Shcherbyna Public Sector Specialist GGOEE Public Sector Yuliya Smolyar Sr. Social Protection Specialist GSP03 Social Protection Viorica Strah Program Assistant ECCMD Administrative Support (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including travel No. of Staff Weeks and consultant costs) Lending Total: 67.5 269 Supervision/ICR Total: 5.2 38 Page 35 of 38 The World Bank (P149555) Annex 2: Borrower's Comments and Comments of Development Partners on Draft ICR No comments or objections received from Borrower. From Development Partners, minor factual corrections were received and reflected in the draft. No major disagreement with the assessment was expressed Page 36 of 38 The World Bank (P149555) Annex 3: List of Supporting Documents International Monetary Fund, 2014 Article IV Consultation and First Post-Program Monitoring Discussions, IMF Country Report No. 14/190, July 2014 International Monetary Fund, 2015 Article IV Consultation and Third Post-Program Monitoring Discussions, IMF Country Report No. 16/19, January 2016 International Monetary Fund, 2017 Article IV Consultation and Second Reviews under the Extended Fund Facility and Extended Credit Facility, IMF Country Report No. 17/398, December 2017 International Monetary Fund, Republic of Molodva: Financial System Stability Assessment, IMF Country Report No. 16/70, February 2016 Republic of Moldova, Moldova 2020: National Development Strategy – Seven Solutions for Economic Growth and Poverty Reduction, 2015 World Bank, Country Partnership Strategy for the Republic of Moldova for the Period FY14-17, Report No. 79701-MD, August 2013 World Bank, First Development Policy Operation: Republic of Moldova, Report No. 83942-MD, February 2014 World Bank, Moldova After the Global Crisis: Promoting Competitiveness and Shared Growth, Report No. 55195-MD, June 2011 World Bank, Moldova: Public Expenditure Review – Capital Expenditures: Making Public Investment Work for Competitiveness and Inclusive Growth in Moldova, June 2013 World Bank, Second Development Policy Operation: Republic of Moldova, Report No. 110070-MD, November 2016 Page 37 of 38 The World Bank (P149555) Page 38 of 38