Document of The World Bank Report No: ICR00003766 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-H9060 AND IDA-D0060) ON TWO GRANTS IN THE AMOUNTS OF SDR 3.4 MILLION AND SDR 2.1 MILLION (US$5.2 MILLION EQUIVALENT AND US$3.0 MILLION EQUIVALENT) TO THE REPUBLIC OF KIRIBATI FOR A FIRST ECONOMIC REFORM DEVELOPMENT POLICY OPERATION AND A SECOND ECONOMIC REFORM DEVELOPMENT POLICY OPERATION June 29, 2016 Macroeconomics and Fiscal Management Global Practice Timor-Leste, Papua New Guinea, and Pacific Islands Country Management Unit East Asia and Pacific Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Republic of Kiribati GOVERNMENT FISCAL YEAR January 1 – December 31 CURRENCY EQUIVALENTS (Exchange Rate Effective as of June 1, 2016) Currency Unit US$1.00 = AUD1.38 ABBREVIATIONS AND ACRONYMS ADB Asian Development Bank AUD Australian Dollar CSO Community Service Obligation DPO Development Policy Operation ERP Economic Reform Plan ERT Economic Reform Taskforce FY Fiscal Year GDP Gross Domestic Product GoK Government of Kiribati ICR Implementation Completion and Results Report IDA International Development Association ISR Implementation Status and Results Report MFED Ministry of Finance and Economic Development MFMRD Ministry of Fisheries and Marine Resource Development M&E Monitoring and Evaluation PAYE Pay As You Earn PDO Program Development Objective RERF Revenue Equalization and Reserve SDR Special Drawing Rights SOE State-Owned Enterprise US$ United States Dollar VDS Vessel Day Scheme Vice President: Victoria Kwakwa Acting Country Director: Mona Sur Senior Practice Director: Carlos Felipe Jaramillo Practice Manager: Mathew Verghis Program Leader: Robert Utz Task Team Leader: Tobias Haque ICR Team Leader: Virginia Horscroft REPUBLIC OF KIRIBATI First Economic Reform Operation Second Economic Reform Operation 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 Program Performance in ISRs H. Restructuring 1. Program Context, Development Objectives and Design 1 2. Key Factors Affecting Implementation and Outcomes 5 3. Assessment of Outcomes 10 4. Assessment of Risk to Development Outcome 17 5. Assessment of Bank and Borrower Performance 18 6. Lessons Learned 21 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners 22 List of Tables Table 1: Selected Macroeconomic Indicators as at Board Approval, October 2014 2 Table 2: Revisions to Results Indicators 4 Table 3: Changes to Indicative Triggers in the Second Operation 6 List of Annexes Annex 1: Bank Lending and Implementation Support/Supervision Processes 24 Annex 2: Comments of Borrower on Draft ICR 25 Annex 3: Comments of Partners on Draft ICR 26 Annex 4: List of Supporting Documents 27 MAP A. Basic Information Program 1 Kiribati Economic Country Kiribati Program Name Reform Operation Program ID P144602 L/C/TF Number(s) IDA-H9060 ICR Date 06/29/2016 ICR Type Core ICR MINISTRY OF FINANCE & Lending Instrument DPL Borrower ECONOMIC DEVELOPME Original Total XDR 3.40M Disbursed Amount XDR 3.40M Commitment Implementing Agencies Ministry of Finance and Economic Development Cofinanciers and Other External Partners Ministry of Foreign Affairs and Trade Australian Department of Foreign Affairs and Trade Asian Development Bank Program 2 Second Economic Country Kiribati Program Name Reform Development Policy Operation Program ID P149888 L/C/TF Number(s) IDA-D0060,IDA-H9060 ICR Date 06/29/2016 ICR Type Core ICR GOVERNMENT OF Lending Instrument DPL Borrower KIRIBATI Original Total XDR 2.10M Disbursed Amount XDR 2.10M Commitment Implementing Agencies Ministry of Finance and Economic Development Cofinanciers and Other External Partners Ministry of Foreign Affairs and Trade Australian Department of Foreign Affairs and Trade Asian Development Bank B. Key Dates Kiribati Economic Reform Operation - P144602 Revised / Actual Process Date Process Original Date Date(s) Concept Review: 04/11/2013 Effectiveness: 01/31/2014 02/04/2014 Appraisal: 10/21/2013 Restructuring(s): Approval: 12/11/2013 Mid-term Review: Closing: 06/30/2014 06/30/2014 Second Economic Reform Development Policy Operation - P149888 Revised / Actual Process Date Process Original Date Date(s) Concept Review: 05/14/2014 Effectiveness: 02/16/2015 02/16/2015 Appraisal: 09/22/2014 Restructuring(s): Approval: 11/14/2014 Mid-term Review: Closing: 06/30/2015 06/30/2015 C. Ratings Summary C.1 Performance Rating by ICR Overall Program Rating Outcomes Satisfactory Risk to Development Outcome Substantial Bank Performance Satisfactory Borrower Performance Highly Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Overall Program Rating Bank Ratings Borrower Ratings Quality at Entry Satisfactory Government: Not Applicable Implementing Quality of Supervision: Satisfactory Not Applicable Agency/Agencies: Overall Bank Overall Borrower Satisfactory Highly Satisfactory Performance Performance C.3 Quality at Entry and Implementation Performance Indicators Kiribati Economic Reform Operation - P144602 Implementation QAG Assessments (if Indicators Rating: Performance any) Potential Problem Quality at Entry Program at any time No None (QEA) (Yes/No): Problem Program at any Quality of Supervision No None time (Yes/No): (QSA) DO rating before Satisfactory Closing/Inactive status Second Economic Reform Development Policy Operation - P149888 Implementation QAG Assessments (if Indicators Rating: Performance any) Potential Problem Quality at Entry Program at any time No None (QEA) (Yes/No): Problem Program at any Quality of Supervision No None time (Yes/No): (QSA) DO rating before N/A Closing/Inactive status D. Sector and Theme Codes Kiribati Economic Reform Operation - P144602 Original Actual Sector Code (as % of total Bank financing) Animal production 25 25 General public administration sector 50 50 Telecommunications 25 25 Theme Code (as % of total Bank financing) Debt management and fiscal sustainability 25 25 Public expenditure, financial management and procurement 13 13 Regulation and competition policy 12 12 Rural non-farm income generation 25 25 State-owned enterprise restructuring and privatization 25 25 Second Economic Reform Development Policy Operation - P149888 Original Actual Sector Code (as % of total Bank financing) Animal production 25 25 General public administration sector 50 50 Telecommunications 25 25 Theme Code (as % of total Bank financing) Debt management and fiscal sustainability 25 25 Rural non-farm income generation 25 25 State-owned enterprise restructuring and privatization 50 50 E. Bank Staff Kiribati Economic Reform Operation - P144602 Positions At ICR At Approval Vice President: Victoria Kwakwa Axel van Trotsenburg Country Director: Mona Sur Franz R. Drees-Gross Sr. Practice Director: Carlos Felipe Jaramillo Practice Manager: Mathew A. Verghis Vivek Suri Task Team Leader: Tobias Haque Tobias Haque ICR Team Leader: Virginia Horscroft ICR Primary Author: Virginia Horscroft Second Economic Reform Development Policy Operation - P149888 Positions At ICR At Approval Vice President: Victoria Kwakwa Axel van Trotsenburg Country Director: Mona Sur Franz R. Drees-Gross Sr. Practice Director: Carlos Felipe Jaramillo Practice Manager: Mathew A. Verghis Mathew A. Verghis Task Team Leader: Tobias Haque Tobias Haque ICR Team Leader: Virginia Horscroft ICR Primary Author: Virginia Horscroft F. Results Framework Analysis Program Development Objectives The Program Development Objective is to support the Government of Kiribati in strengthening public services, while improving fiscal sustainability. Revised Program Development Objectives The Program Development Objectives are: (i) improving the management of fisheries revenues; (ii) improving the management of public assets and liabilities; and (iii) expanding private sector opportunities Indicator(s) Kiribati Economic Reform Operation - P144602 Original Target Actual Value Formally Baseline Values (from Achieved at Indicator Revised Value approval Completion or Target Values documents) Target Years Second Economic Reform Development Policy Operation - P149888 Original Target Actual Value Formally Baseline Values (from Achieved at Indicator Revised Value approval Completion or Target Values documents) Target Years Improved predictability of fishing license fee revenues as indicated by the Indicator 1 : variance between budgeted and actual fishing license revenues Value less than +/- 50 (quantitative or +/- 188 percent + 164 percent percent Qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 The first results indicator for the programmatic series has to be assessed as not Comments achieved, which is very unfortunate given the extremely positive results of the (incl. % reform program under this pillar implemented by the GoK with support from the achievement) Bank. SOEs become less reliant on ad hoc government subsidies as indicated by ad hoc Indicator 2 : subsidies to SOEs Value less than AUD 0.8 (quantitative or AUD 2.5 million AUD 0.0 million million Qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Comments (incl. % The second results indicator is assessed as being achieved. achievement) Inefficient expenditure through the copra price subsidy is reduced as indicated by Indicator 3 : actual copra price subsidy expenditure Value (quantitative or AUD 7.3 million AUD 6.8 million AUD 7.65 million Qualitative) Date achieved 12/31/2013 12/31/2015 12/31/2015 Over the course of the programmatic series, Kiribati succeeded in reducing the Comments administrative costs of the copra subsidy scheme, but it did not reduce total (incl. % expenditure on the scheme, so the third results indicator has to be assessed as not achievement) achieved. No increase in non-concessional debt as indicated by central government non- Indicator 4 : concessional debt balances Value less than AUD 1 (quantitative or AUD 14 million AUD 0 million million Qualitative) Date achieved 06/30/2012 12/31/2015 12/31/2015 Comments (incl. % The fourth results indicator is assessed as being exceeded. achievement) Closer alignment between RERF portfolio performance and relevant market Indicator 5 : benchmarks as indicated by the performance of the RERF portfolio relative to benchmark Value better than -2 performance within (quantitative or -8.29 percent percent 0.3 percent of the Qualitative) underperformance benchmark Date achieved 03/31/2014 12/31/2015 12/31/2015 Comments The fifth results indicator is assessed as being exceeded. (incl. % achievement) Reduced prices in the telecommunications sector as indicated by the cost of 3- Indicator 6 : minute peak-hour mobile-to-mobile telephone call within Tarawa AUD 0.69 (for existing 3G customers of the Value previous operator); (quantitative or AUD 0.90 AUD 0.70 AUD 0.78 including Qualitative) VAT (for customers with sim cards issued by the new operator) Date achieved 06/30/2014 12/31/2015 12/31/2015 Comments (incl. % The sixth results indicator is assessed as being partially achieved. achievement) G. Ratings of Program Performance in ISRs Kiribati Economic Reform Operation - P144602 Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 06/04/2014 Satisfactory Satisfactory 5.28 H. Restructuring N/A 1. Program Context, Development Objectives and Design 1. This Implementation Completion and Results Report (ICR) assesses the achievement of the expected results of the programmatic series of Economic Reform Development Policy Operations for Kiribati. The programmatic series intended to support the Government of Kiribati (GoK) to improve the management of fisheries revenues, improve the management of public assets and liabilities, and expand private sector opportunities. The first operation in the series for SDR 3.4 million (US$ 5.2 million equivalent) was approved by the World Bank’s Board of Directors on November 4, 2013 and disbursed upon loan effectiveness on March 18, 2014. The second operation in the series for SDR 2.1 million (US$ 3.0 million equivalent) was approved by the Board on October 22, 2014 and disbursed upon loan effectiveness on March 18, 2015. The second and last operation in the series closed on June 30, 2015. 1.1 Context at Appraisal 2. In the decade preceding the operations, economic growth in Kiribati was low and volatile – averaging less than 1 per cent per annum and actually declining in per capita terms. Growth opportunities in Kiribati are highly circumscribed by structural factors. Very small size, extreme remoteness from major markets, and vast internal dispersion raise production costs, and mean that growth opportunities are concentrated in areas where economic rents can be generated, such as fisheries and – to a limited extent – niche tourism. The livelihoods of people in the outer islands are largely dependent on semi-subsistence agriculture, while those in the capital depend primarily on the public sector (some 80 percent of formal sector jobs in Kiribati are in the public sector, and government expenditure is roughly equivalent to GDP). The level of economic growth is determined mainly by fishing license fees and development assistance (which together fund the bulk of public expenditure), remittances from seafarers, global food and fuel prices, and natural disasters. Alongside providing a supportive environment for economic growth, the major concern of macroeconomic management is cushioning the impact of volatility and shocks – with fiscal policy the only available tool because Kiribati uses the Australian Dollar as its currency. 3. The fiscal situation worsened in the decade preceding the operations, due to weak growth, volatile fishing license fees and declining tax revenue (see Table 1). Kiribati’s Revenue Equalization and Reserve Fund (RERF) provided something of a fiscal buffer, but a significant decline in real expenditure occurred nonetheless. Of most significance to the poor, education and health spending declined in real terms by 7 and 13 percent, respectively, over the decade preceding the first operation. Over the same period, the real per capita value of the RERF declined by more than a third, reducing the extent to which it could serve as a fiscal buffer in the future. There was, therefore, a pressing need to begin the process of restoring fiscal sustainability – including strengthening revenue performance, improving expenditure prioritization, and ceasing unsustainable drawdowns from the RERF – as well as to take the available opportunities to strengthen economic growth prospects. 4. The weakness and unpredictability of Kiribati’s fisheries revenues, and poor management of its state-owned enterprises (SOEs), were two key drivers of the fiscal deficits that led to unsustainable drawdowns from Kiribati’s RERF in the decade preceding the operations. The causes of the weak and unpredictable revenues that Kiribati 1 was earning from its extremely valuable fishery included weaknesses in the management of the fishery itself, weaknesses with joint venture arrangements, weak enforcement, and weaknesses in information sharing on fisheries revenues within government. With respect to Kiribati’s SOEs, the liquidity and broader financial problems of its poorly managed SOEs led them to depend on intermittent injections of funds from the government, injections that were unbudgeted and unpredictable, causing significant problems for fiscal management. Table 1: Selected Macroeconomic Indicators as at Board Approval, October 2014 Source: Bank staff estimates 5. In the lead-up to the operations, the GoK sought assistance from the Bank to formulate an Economic Reform Plan (ERP) in 2012. This marked the beginning of the Bank’s economic policy engagement with the GoK. Thus, that engagement was still relatively new when the Bank prepared its first ever DPO in Kiribati in 2013. The Bank began the engagement by working with the GoK on a public expenditure analysis, identifying the key problems with revenue, expenditure and asset management that needed to be addressed. The Bank then worked with the GoK and other development partners to prioritize these economic and fiscal challenges and prepare a coherent set of reforms to address them, with the longer-term objective of achieving fiscal sustainability and improving living 2 standards.1 The Bank supported the implementation of a selection of these reforms through a set of technical assistance engagements and through a programmatic series of two DPOs, helping maintain the momentum behind the ERP in its critical initial phase. Similarly, Australia, New Zealand and the Asian Development Bank (ADB) provided technical assistance and/or budget support linked to the implementation of selected reforms in the ERP. The extremely thin capacity of Kiribati’s public sector and the broader fragility of the state had significant implications for the Bank’s operations, which cannot be overemphasized. 1.2 Original Program Development Objectives (PDO) and Key Indicators 6. The original PDO was to support the government in strengthening public services, while improving fiscal sustainability. The PDO was to be achieved through actions under three pillars, which corresponded with three of the objectives in Kiribati’s ERP. These were: (i) increasing revenues; (ii) improving the management of public assets and liabilities; and (iii) expanding private sector opportunities.2 The key outcome indicators were: (i) increasing revenue from exploitation of fisheries resources; (ii) SOEs become less reliant on government subsidies; (iii) no increase in non-concessional debt; (iv) RERF asset allocations and investment strategy better reflect government investment objectives; and (v) increased competition in the telecommunications sector.3 1.3 Revised PDO and Key Indicators, and Reasons/Justification 7. In the second operation in the series, the PDO was revised to be more specific and more closely aligned with the pillars of the operation. The PDO of the programmatic series became: (i) improving the management of fisheries revenues; (ii) improving the management of public assets and liabilities; and (iii) expanding private sector opportunities. Effectively, then, the PDO was re-specified at the level of the three pillars being supported by the program, so the revision did not represent a discontinuity or change in direction in the program itself. In the process of that re-specification, the first pillar was also made more specific (from revenue to fisheries revenue), to more accurately reflect the content of the policy matrix. The PDO that had been specified in the first operation – of supporting improved public services and fiscal sustainability – was retained as a broader aspiration in the second operation. The reason for the revision was to ensure that the PDO consisted of specific objectives that related directly to the supported actions, so could be expected to be achieved if the actions were undertaken successfully (this explanation was not provided explicitly in the program document, however). 8. There were revisions to the results indicators in the second operation, reflecting revisions to the indicative triggers and attempts to make the results indicators more attributable to the policy actions themselves. These revisions are summarized in Table 2. Table 2: Revisions to Results Indicators Results Indicators (1) Results Indicators (2) Reasons for Revisions                                                              1 This engagement was aligned with the first pillar of the Bank’s Country Assistance Strategy for Kiribati, FY2011-FY2014, approved by the Board on January 31, 2011, on building resilience against external shocks. 2 There were minor differences in the wording of the pillars in the program summary and in the policy and results matrix, but these wording differences are not indicative of any difference in substance. 3 Again, there were minor differences in the wording between the outcome indicators in the program summary and in the policy and results matrix, which are not indicative of any difference in substance, except that the results indicator relating to the RERF was inadvertently omitted from the program summary. 3 Improving the management of fisheries revenues Result: Increasing revenue Result: Improved predictability Revised in an attempt to from exploitation of of fishing license fee revenues make it more attributable fisheries resources Indicator: Variance between to the policy actions Indicator: License revenue budgeted and actual fishing being supported per ton of catch license revenues Improving the management of public assets and liabilities Result: SOEs become less Result: SOEs become less Revised to make it more reliant on Government reliant on ad hoc Government attributable to the policy subsidies subsidies actions being supported Indicator: Subsidies to Indicator: Ad hoc subsidies to SOEs as a proportion of SOEs GDP Result: Inefficient expenditure New indicator, reflecting through the copra price an additional prior action subsidy is reduced in the second operation Indicator: Actual copra price not included initially, subsidy expenditure because reform progress had not expected to be made so quickly Result: No increase in non- Result: No increase in non- No change concessional debt concessional debt Indicator: Central Indicator: Central government government non- non-concessional debt concessional debt balances balances Result: RERF asset Result: Closer alignment Revised to reflect allocations and investment between RERF portfolio changes in the related strategy better reflect performance and relevant indicative trigger Government investment market benchmarks objectives Indicator: Performance of the Indicator: Proportion of RERF portfolio relative to RERF assets under passive benchmark management in fixed income assets and equities Expanding private sector opportunities Result: Increased Result: Reduced prices in the Revised to reflect competition in the telecommunications sector changes in the related telecommunications sector Indicator: Cost of 3-minute indicative trigger Indicator: Number of peak-hour mobile-to-mobile providers of internet and telephone call within Tarawa mobile phone services 1.4 Original Policy Areas Supported by the Program 9. The program supported policy actions under three pillars, articulated in the first operation as: (i) increasing revenues; (ii) improving the management of public assets and liabilities; and (iii) expanding private sector opportunities. A set of prior actions and 4 triggers over the two operations in the program were mapped to these pillars, each of which was aligned to a key objective in the GoK’s ERP. 10. The pillar on increasing revenues was focused on fisheries revenues, with the reform plan for this pillar underpinned by the adoption of a new national fisheries policy. The fisheries policy was aimed at strengthening the sustainable management of fisheries resources and maximizing fisheries license fee revenue. The latter involved the implementation of a number of actions over time, including revising Kiribati’s joint venture arrangements. The program supported by the DPO series was drawn from the GoK’s ERP. 11. The pillar on improving the management of public assets and liabilities covered strengthening the management of SOEs, strengthening debt management, and strengthening the management of Kiribati’s RERF. The reform plan for SOEs was underpinned by the adoption of new legislation aimed at achieving sound corporate governance of SOEs. The reform plan for debt management centered on the adoption of a debt policy that included criteria for borrowing consistent with sustainable macroeconomic management. The reform plan for the RERF was focused on its management, in particular the need to align RERF management with government investment objectives. The programs supported by the DPO series in each of these areas were drawn from the GoK’s ERP. 12. The pillar on expanding private sector opportunities was focused on the telecommunications sector, with the reform plan for this pillar underpinned by the introduction of legislation to liberalize the telecommunications sector. The legislation provided a basis for commercializing the incumbent SOE provider and introducing competition into the sector. The program supported by the DPO series under this pillar was aligned with the GoK’s ERP. 1.5 Revised Policy Areas 13. In the second operation, the first pillar was revised to align it more closely with the actions being supported. It became ‘improving the management of fisheries revenues’, thus making it more specific to the supported actions, which focused exclusively on fisheries revenues. 1.6 Other Significant Changes 14. N/A 2. Key Factors Affecting Implementation and Outcomes 2.1 Program Performance 15. The two operations in the series provided Kiribati with a total of SDR 5.5 million (US$ 8.2 million equivalent) in budget support. The operations were conducted in line with a schedule agreed with the GoK, providing it with a predictable flow of resources over the two years of the program. The first operation in the series for SDR 3.4 million (US$ 5.2 million equivalent), was approved by the Board on November 4, 2013 and disbursed on March 18, 2014. The second operation in the series for SDR 2.1 million (US$ 3.0 million equivalent) was approved on October 22, 2014 and disbursed on March 18, 2015. All prior actions for both operations were completed prior to Board approval. There were no subsequent reversals of these prior actions. 5 16. There were a number of changes to the indicative triggers for the second operation, foreshadowed in the program document for the first operation as risks posed by the very thin capacity environment, the fragile state context and the tackling of key development challenges where strong vested interests were at play. At the time of the second operation, the GoK had made solid progress in implementing the reform agenda in each policy area. Indicative triggers were adjusted to account for slower-than-expected or greater-than-expected progress with some steps in those reform processes, and the evolution of the Bank’s and GoK’s thinking on the most appropriate sequence of reforms in other reform processes. These revisions to the indicative triggers are shown in Table 3. Table 3: Changes to Indicative Triggers in the Second Operation Prior Action (1) Indicative Trigger Prior Action (2) Improving the management of fisheries revenues The Recipient has a) The Ministry of The Recipient has produced a joint report approved a Fisheries and Ministry of its Ministry of Fisheries and Marine national fisheries of Finance will jointly Resource Development and Ministry of policy to review the consistency Finance and Economic Development on strengthen the of fishing access the sources of fisheries revenue containing sustainable arrangements with disaggregated data regarding the sources of management of government policy and all fishing revenues, and has published the fisheries resources regional management report on the website of the Ministry of and maximize agreements; and b) Finance and Economic Development. license revenue Cabinet will approve income. revisions to access arrangements based on the recommendations of this review. Improving the management of public assets and liabilities The Recipient has: Key provisions of the The Recipient has adopted the following i) approved a bill new SOE Act will be measures consistent with the State Owned to establish a legal implemented, Enterprises Act: (a) ensured full framework for including: i) full compliance of all SOE board appointments improved compliance with with the provisions of the SOE Act; (b) governance, legislative provisions ensured that the five largest SOEs have strengthened for all board submitted to it full financial statements financial reporting, appointments; ii) the within three months of the end of the and commercial five largest SOEs will Recipient’s 2013 fiscal year; and (c) management of submit full financial ensured the inclusion of SOE community state owned statements to service obligation subsidies in the enterprises and has Government within Recipient’s 2014 annual budget. submitted said bill three months of the end to Parliament for of the 2013 fiscal year; The Recipient, through its Cabinet, has approval; and ii) and iii) a concession approved a roadmap for reform and brought to the contract for the rationalization of its copra sector, including point of management of the the merger of the Kiribati Copra transaction a Betio Shipyards will be Cooperative Society and the Kiribati Copra concession brought to the point of Mill Limited. contract for private transaction. 6 operation of the Otintaai Hotel. Cabinet approves key The Recipient, through its Cabinet, has The Recipient has RERF management approved key reforms in the management approved a debt reforms, including: i) of its Revenue Equalization Reserve Fund policy establishing reallocation of RERF to achieve consistency with clearly-stated policy criteria for assets to achieve investment objectives, including: (a) concessional and consistency with reallocation of RERF’s assets to new asset non-concessional clearly-stated managers; and (b) reorientation of RERF’s public borrowing investment objectives; portfolio towards lower-risk instruments. consistent with ii) application of new sustainable concentration and macroeconomic deviation limits; and management. iii) application of more appropriate benchmarks to improve monitoring of asset manager performance. Expanding private sector opportunities The Recipient has New ICT licenses are The Recipient has invited pre-qualified approved a bill to offered permitting at firms to bid for purchase of liberalize the least a second national Telecommunication Services Kiribati telecommunication mobile operator to Limited following completion of pre- sector in the supply wireless mobile qualification processes by the Recipient’s Recipient’s services to the public Ministry of Communications. territory, including using its own commercialization infrastructure. of the incumbent state-owned operator and the introduction of competition. 17. The revision to the indicative trigger under the first pillar of the program was informed by a deepened understanding of the weaknesses in the management of the fisheries sector which emerged as the implementation of reforms progressed. The revision of the indicative trigger recognized that the information sharing and evidence base was not yet in place for it to be possible to review existing fisheries access arrangements, and that – particularly without the necessary data – supporters of reform within the GoK were unlikely to be strong enough to overcome vested interests opposed to revising those access arrangements at that juncture. An incremental step – which would provide the necessary evidence base and potentially strengthen the reform momentum – was supported instead, with the Ministry of Fisheries and Marine Resource Development (MFMRD) and Ministry of Finance and Economic Development (MFED) jointly producing and publishing a report providing disaggregated information on the sources of all fisheries revenues, a marked break with the past on information sharing by MFMRD and transparency by the GoK as a whole. 18. The changes to the indicative triggers under the second pillar of the program on the management of public assets and liabilities largely reflected reform progress 7 occurring at a different pace than originally expected. Overall, progress on implementing the SOE reform agenda was strong, with changes made to the indicative triggers to reflect one area (the concession contract for the Otintaai Hotel) proceeding more slowly than initially anticipated, while two others (explicit budgeting for community service obligations (CSOs) and reform of the SOEs operating the copra subsidy scheme) proceeded more quickly than initially anticipated. On the RERF reform agenda, the Bank’s technical assistance engagement deepened, market conditions changed, and the government’s understanding of the situation and commitment to reform strengthened. These factors all contributed to the strengthening of the trigger for the second operation. 19. The revision to the indicative trigger under the third pillar of the program was informed by the need to re-sequence the reforms in the telecommunications sector. In the lead up to the second operation, the Bank’s telecommunications team advised that the most important next step in the reform process for the sector was not in fact to bring in competition (as had been envisaged originally) but to restructure and privatize the assets of the incumbent SOE, so the trigger was revised accordingly. 2.2 Major Factors Affecting Implementation 20. The commitment of the GoK to implementing the reform program supported by the operations was very strong. The reform program was drawn from Kiribati’s ERP, which the GoK had formulated with assistance from the Bank. In the year preceding the preparation of the first operation, the Bank had established an influential macro-fiscal policy engagement with the GoK, rapidly building up expertise on the key challenges facing Kiribati through a joint public expenditure analysis. This approach enabled the Bank to provide the GoK with policy advice when it really needed it, correctly identifying the key challenges Kiribati faced and setting the GoK in the right direction for tackling these. But the relative newness of the engagement meant the Bank did not have a detailed knowledge of all the factors that would affect subsequent steps in the reform processes being initiated. In this context, an iterative approach was taken to the reforms, with subsequent steps in reform processes adjusted in accordance with the knowledge gained from parallel technical assistance engagements and on the basis of what was learned during the course of the initial reform steps. This contributed to the flux evident in the policy matrix, with some revisions to the indicative triggers reflecting a deepened understanding of the most appropriate sequencing of reforms. 21. The key risks to the operations were correctly identified at the time of appraisal, particularly the extremely limited capacity of the GoK, the fragile nature of the state, and the significant interests opposed to some of the reforms that tackled critical issues for Kiribati’s development. The public service in Kiribati is very small, with large policy responsibilities – like budget formulation or monitoring of SOEs – typically handled by only a handful of people. This makes reform very challenging. The people who are formulating and implementing reforms are typically already fully employed operating the existing systems, leaving little time for reform efforts. And because the reforms depend on so few people, if they are pulled away from their work for any reason (training, leave, other responsibilities, etc.), the reform process often comes to a halt. Predicting and sustaining the pace of reforms is thus very difficult. In addition, Kiribati is one of only a handful of countries formally considered a fragile state by the Bank given its CPIA of 3.0. The fragility of the state is reflected in weak internal coordination, which also makes predicting and sustaining the pace of reforms very difficult. Finally, in tackling some critical issues for 8 Kiribati’s development (like fisheries management and SOE reform) the operations necessarily encountered vested interests opposed to reform, intensifying the challenges of weak internal state coordination. These risks were clearly recognized in the program documents for the two operations. The mitigation measures of focusing on a small set of key reforms where strong support existed within the GoK for change, and coordinating closely with other donors both to provide technical assistance in support of the reforms and to harmonize donor policy advice, were appropriate. But as the program documents made clear, the risks could only be partially mitigated and it remained likely that the reform program would need to be adjusted over time where particular elements could not be implemented as successfully or as quickly as initially planned. The flux evident in the policy matrix for this program was in part a reflection of this. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization 22. The M&E mechanism for the program was designed to operate through the Economic Reform Taskforce (ERT). The ERT brings together representatives from all relevant government agencies and development partners involved in the implementation of the ERP, and is responsible for overseeing the implementation of those reforms. Appropriately, particularly given the extreme capacity constraints in Kiribati, the M&E for the program was designed to work through an existing government system, thus avoiding imposing an additional administrative burden on the government. 23. M&E implementation was effective, with the ERT assessing the progress of each of the reform processes supported by the program and addressing emerging challenges with the reforms. The ERT worked effectively to adjust reform steps and sequences in light of implementation progress, and to maintain momentum for the reforms. Appropriately, the use of the ERT for M&E for the operations helped to reinforce its operation and importance. 24. The utilization of the M&E system to track the results indicators for the operations was more limited than for tracking the implementation of the reforms. Understandably, the focus of the GoK was on the successful implementation of the reforms and – following that – the implementation of subsequent steps in the reform processes. While there was less emphasis on tracking the results indicators chosen for the Bank’s operations, including with changes in personnel participating in the ERT on the Bank side, the ERT did still work effectively to gather the requisite information at the time of the ICR. 2.4 Expected Next Phase/Follow-up Operation 25. A new programmatic series of DPOs is being prepared in Kiribati, with the first operation in the series expected to go to the Board in FY17. At the close of the first programmatic series, it was anticipated that a new programmatic series would be prepared for approval in FY16, to maintain continuity in the dialogue and momentum for the reforms. However, delays caused on the Bank side by the transition between TTLs and then on the government side by elections and the need to determine whether the new government shared the previous government’s commitment to the ERP, put the program back to FY17. 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation Relevance of Objectives: High 9 26. The objective of the first pillar of the program was highly relevant to the overarching development problem Kiribati faced with fiscal deficits and unsustainable drawdowns from the RERF. The weakness and unpredictability of Kiribati’s fisheries revenues were a key driver of the fiscal deficits that led to unsustainable drawdowns from Kiribati’s RERF. Through the joint public expenditure analysis, the Bank had worked with the GoK to ascertain the significance of fisheries revenues to this overarching problem. 27. The objective of the second pillar of the program was also highly relevant to the overarching development problem Kiribati faced with fiscal deficits and unsustainable drawdowns from the RERF, as well as with the management of the RERF directly. Poor management of SOEs was a key driver of the fiscal deficits that led to unsustainable drawdowns from the RERF, with the liquidity and broader financial problems of SOEs leading them to depend on unbudgeted injections of funds from the government. These same problems also underpinned the failure of several SOEs to pay corporate taxes and the practice of some SOEs to withhold the PAYE taxes of their employees, contributing to shortfalls in government revenue. Building on existing work by the ADB, the Bank and other development partners worked with the GoK to build a shared understanding of the extent and causes of these problems, and a shared commitment to reform. A further key driver of the fiscal deficits that led to unsustainable drawdowns from the RERF were weaknesses in debt management, with unexpected fiscal deficits partially financed by unplanned borrowing via a bank overdraft, incurring very high – and unbudgeted – interest charges which in turn exacerbated the deficits. The Bank worked with the GoK and other development partners to build awareness of the weaknesses of Kiribati’s approach to debt management, particularly with respect to the overdraft. Poor management of the RERF directly undermined its sustainability, with problems with the existing asset managers’ adequacy of reporting to the GoK, engagement with the GoK, adherence to their investment mandate, and problems with the ability of the GoK to monitor the asset managers. 28. The objective of the third pillar of the program was relevant to the development problem Kiribati faced with low and volatile economic growth. The high cost and low quality of telecommunications services in Kiribati was one of the weak points of the business environment in Kiribati, in turn constraining private sector development. It was not the major weak point of the business environment – the major weak point was Kiribati’s economic geography, with its extreme remoteness, very small size and vast internal dispersion significantly raising the cost of doing business. But it was arguably the biggest weak point that the GoK could actually do something about. More broadly, Kiribati’s people would enjoy a better standard of living with better quality and lower cost services. Relevance of Design: High 29. The design of the reform program under the first pillar of the operations was highly relevant to addressing the weakness and unpredictability of fisheries revenues. The Bank worked with the GoK and other development partners to identify the causes of the poor management of fisheries revenues, and address those as part of the ERP. These included weaknesses in the management of the fishery itself, most notably not fully implementing the Vessel Day Scheme (VDS) agreed with the other Pacific Island Countries with significant tuna fisheries, to manage access within the overall catch limits established by the scheme (as part of DPO preparations, the Bank provided specific technical assistance to the GoK on 10 fisheries revenues). They also included weaknesses with joint venture arrangements, weak enforcement, and the refusal of the MFMRD to provide any information to the MFED on the sources of revenues, the agreements they are generated from, or the performance of joint ventures. The reform program was designed to address each of these causes of the poor management of fisheries revenues over time. 30. The design of the reform program under the second pillar of the operations was highly relevant to strengthening the management of public assets and liabilities. On SOEs, the reform agenda began with establishing a sound legislative framework for corporate governance and financial reporting, with a view to putting SOEs on a commercial footing. It then focused on implementing key sections of the new legislation and making progress with the restructuring of particular SOEs. In one aspect of the design of the SOE reform program, greater attention could have been paid to balancing reform ambition with capacity. From a fiscal management perspective, compliance with the legislative provisions on board appointments was fairly immaterial beyond the five largest SOEs, so it may have been better to focus element (a) of the trigger on board appointment on those five SOEs, as was done in element (b) on financial statements, rather than using up scarce capacity tackling all board appointments in one year. On debt management, the reform agenda centered on the formulation and adoption of a new debt policy with clear criteria for concessional and non- concessional borrowing. On the RERF, the Bank worked with the GoK to design a reform program that would address the identified causes of the mismanagement of the RERF, and provided parallel technical assistance to help the GoK implement the reforms. 31. The design of the reform program under the third pillar of the operations was relevant to improving the quality and reducing the cost of telecommunication services, thereby improving the business environment in Kiribati. The Bank’s telecommunications team worked with the GoK to design a reform program for the sector and incumbent SOE, with particular milestones from the reform agenda – including a new legislative framework and the introduction of competition – included as policy actions in the programmatic series. Relevance of Implementation: Substantial 32. The implementation of the reform program under the first pillar of the operations was relevant to addressing the weakness and unpredictability of fisheries revenues. The GoK made good progress with implementing the reform agenda, from the solid foundation of the adoption of the new national fisheries policy. As the Bank’s fisheries engagement with Kiribati continued (this being a fairly new area for the Bank in the Pacific at the time), it gained a fuller understanding of the extent of the weaknesses in the sector, and the reform agenda also encountered more resistance. Both considerations led to the recognition that an incremental step needed to be added to the reform process, before it would be possible or feasible for the planned review of existing access arrangements to proceed. This adjustment of the timing and content of detailed aspects of the reform program was appropriate in the circumstances, and was in keeping with the risks that had been identified in the program documents for the operations. 33. The implementation of the reform program under the second pillar of the operations was relevant to strengthening the management of public assets and liabilities. On SOEs, the reform momentum was strong. There were adjustments to the timing of detailed reform steps as some areas of reform moved more slowly than initially planned, and 11 others progressed more quickly. On debt management, the reform program was implemented as initially envisaged. On the RERF, the deepening of the Bank’s technical assistance engagement and strengthening of the GoK’s understanding of the situation and commitment to reform enabled the level of ambition of the reform program to be increased over time, to significantly exceed initial expectations. 34. The implementation of the reform program under the third pillar of the operations was relevant to improving the quality and reducing the cost of telecommunication services. The new legislative framework was reformed on schedule. As the reform agenda then progressed, the Bank’s telecommunications team recommended re- sequencing the reform agenda, to place the restructuring and privatizing of the assets of the incumbent SOE before the introduction of competition. In the estimation of the Bank’s team, the pre-sale restructuring of the SOE and improvement in its management would result in significant service quality improvements and cost reductions. 3.2 Achievement of Program Development Objectives Overall Rating: Substantial 35. The PDO of the programmatic series was evaluated using six results indicators. Out of the six indicators, two were exceeded (no increase in non-concessional debt and closer alignment between RERF portfolio performance and relevant market benchmarks), one was achieved (SOEs become less reliant on ad hoc government subsidies), one was partially achieved (reduced prices in the telecommunications sector) and two were not achieved (improved predictability of fishing license fee revenues and reduced expenditure through the copra price subsidy). The results indicators and their current status are discussed below. Objective 1 – Improving the management of fisheries revenues: Substantial 36. In improving the management of fisheries revenues, the programmatic series achieved substantial development. The reforms were broadly implemented as envisaged, and although the chosen results indicator itself was not achieved, other indicators – including a more appropriate results indicator that had been dropped – suggest that the policy actions were successful. Outcome Indicator 1 – Improved predictability of fishing license fee revenues as indicated by a reduction in the variance between budgeted and actual fishing license revenues from 188 percent in 2013 to less than 50 percent in 2015: Not Achieved 37. The results indicator chosen for this pillar of the program was an improvement in Kiribati’s prediction of its fisheries revenues, which was not achieved. In the view of the ICR team, this was not an appropriate indicator to have chosen, because it underestimated the complexity and risks of forecasting fisheries revenues in the period covered by the programmatic series, and did not represent the core of what the fisheries reform program was aimed at achieving. Having the MFMRD share detailed information with MFED on the sources of revenues (as supported by the program) is vital to revenue forecasting, but two key drivers of the revenues remained very difficult to forecast – first, the migration of fish stocks and second, the impact of the relatively new regional VDS on access prices. An indicator based on the predictability of revenue per tonne of catch may have better controlled for the former, but as it turned out the latter was the key driver in the period covered by the 12 programmatic series because the series coincided with the transitional period to the regional VDS which had a significant positive price effect.4 In addition, fiscal risks of overestimating revenues are substantial relative to the risks of underestimating revenues, a possible cause of forecast bias that may not have been fully accounted for. 38. A better overall indicator for this pillar would have been improving the returns on Kiribati’s fisheries assets (as measured by revenue per tonne of catch), which was the results indicator chosen in the first operation but then dropped. The reform agenda, underpinned by the new national fisheries policy, was geared at increasing the returns to Kiribati on its fisheries resources. It is true that a key driver of revenue per tonne of catch was the impact of the regional agreement initiating the VDS itself, thus not all of any increase in revenue per tonne of catch would have been attributable to the policies supported by the operations. But the two were obviously closely related, given that the policies supported by the operations included – as part of its national fisheries policy – Kiribati’s commitment to implement that regional agreement. In 2015 the fisheries revenue of the GoK was about AUD 682 per tonne of catch, compared with AUD 103 per tonne in 2012 (the baseline year for the original indicator), a 560 percent increase that significantly over-achieved the original target of a 20 percent increase.5 39. The Bank’s dialogue helped to raise the causes of the weaknesses in the management of fisheries resources and revenues to the highest level in the GoK, enabling significant progress on the reform agenda against entrenched interests opposed to the reforms. The new national fisheries policy, reflecting a new commitment on the part of the GoK to implement the VDS, was a critical achievement. Kiribati’s previous practice of granting access to its fisheries outside the VDS was a key reason why it was failing to maximize fisheries revenues. The dialogue also resulted in the MFMRD sharing detailed information regarding the sources of fisheries revenues with the MFED for the first time, a marked break with the past on information sharing by MFMRD and transparency by the GoK as a whole. On the basis of the dropped indicator, showing the extremely significant improvement in returns on Kiribati’s fisheries resources, and the other achievements of this reform agenda, the outcome of this pillar is assessed as substantial. Objective 2 – Improving the management of public assets and liabilities: High 36. In improving the management of public assets and liabilities, the programmatic series achieved high development. The reforms were broadly implemented as envisaged, three of the four results indicators were exceeded or achieved, and the results indicator that was not achieved was not well aligned with the objective of the related reform program. Outcome Indicator 2 – SOEs become less reliant on ad hoc government subsidies as indicated by a reduction in ad hoc subsidies to SOEs from a baseline of AUD 2.5 million in 2013 to less than AUD 0.8 million in 2015: Achieved                                                              4 Going forward, the GoK’s ability to forecast fisheries revenues should improve, because the transition period to the regional VDS, with its significant price effect, appears to have now been completed and access prices have become more stable at a new, much higher level. 5 There appears to have been an error in the calculation of the baseline for 2012 in the program document, so this ICR works from the percentage increase targeted rather than the absolute numbers. 13 37. Over the course of the programmatic series, Kiribati successfully eliminated ad hoc subsidies to SOEs. As part of the SOE reform agenda, the GoK sought to confine subsidies to SOEs to CSOs funded through an adequate budget appropriation, and reduce the extent of ad hoc (defined as unbudgeted) subsidies to SOEs. The targeted 68 percent reduction in ad hoc subsidies was exceeded, with ad hoc subsidies to SOEs totally eliminated by 2015. Two important factors were behind that achievement. First, through the GoK’s commitment to restructuring SOE boards and putting competent managers in charge of SOEs, the management of SOEs improved, resulting in less need for ad hoc subsidies. Secondly, the Bank’s dialogue with the GoK on the fiscal costs and risks of poor SOE management significantly strengthened fiscal discipline in this regard, so the SOEs that did apply for support through the designated process to access budgeted CSO funds had their requests rejected wherever they exhibited weaknesses in performance that ought to have been addressed prior to making applications for CSOs. One weak point in the implementation of the reform agenda to date relates to the capacity to calculate CSOs for individual SOEs as a basis for CSO budget allocations to individual SOEs. The GoK has found that it lacks the capacity to make these calculations. Instead, the budget allocation provides a pool of funds that can be accessed for the purpose of funding CSOs on the basis of an assessment of the application in accordance with a preset process. This deals with the fiscal problems caused by the unpredictability of (unbudgeted) ad hoc subsidies and – with the management in place to date – has been operated with very careful scrutiny. To ensure SOEs are adequately and reliably compensated for their CSOs going forward and to mitigate the risk of the general CSO allocation becoming a poor for general subsidies (a risk that is assessed to be small, given that this would be a clear breach of the legislative provision that subsidies be paid only for specified community goods), progress is still needed on SOE-specific CSO calculations. Outcome Indicator 3 – Inefficient expenditure through the copra price subsidy is reduced as indicated by a reduction in actual copra price subsidy expenditure from a baseline of AUD 7.3 million in 2013 to AUD 6.8 million in 2015: Not Achieved 38. Over the course of the programmatic series, Kiribati succeeded in reducing the administrative costs of the copra subsidy scheme, but it did not reduce total expenditure on the scheme. The Bank’s technical assistance to the GoK on the copra price subsidy scheme recommended, in the first instance, an improvement in the management of the scheme so it was not so inefficient (the Bank’s analysis suggested potential savings of up to 40 percent of the cost of the scheme, through improved administrative efficiency). This was to be achieved through the implementation of a roadmap for reforming and rationalizing the copra sector, including a merger of the two SOEs involved in the operation of the scheme and stronger financial oversight from MFED. Reforms in this area moved faster than anticipated at the time of the first operation, enabling the approval of the roadmap to be included as a policy action in the second operation. A results indicator for this area of reforms was added in the second operation, but unfortunately it was specified in terms of total expenditure on the copra price subsidy scheme. That may have been appropriate for the implementation of the roadmap over the medium-to-long term, but was not appropriate for the initial phase of the roadmap where an indicator of reduced operational costs of the copra SOEs would have aligned better with the supported program. The chosen results indicator of total expenditure would not be met if production levels increased or producer subsidies per unit of production were increased (as occurred on September 1, 2015). For 2015, total expenditure on the subsidy scheme is estimated at AUD 7.95 million, higher than the target. However, the 14 operational costs of the copra subsidy scheme fell in 2015, and it was the savings from these reduced operational costs that funded the increase in the actual subsidy paid from AUD 0.80 per kg to AUD 1 per kg in September. Thus, though the chosen results indicator was not achieved, the actual objective of the first phase of reforms under the roadmap – lower administrative costs – was in fact achieved. Outcome Indicator 4 – No increase in non-concessional debt as indicated by a reduction in central government non-concessional debt balances from a baseline of AUD 14 million in mid-2012 to no more than AUD 1 million in 2015: Exceeded 39. Over the course of the programmatic series, Kiribati ceased borrowing on non- concessional terms altogether. The debt policy supported by the first operation was implemented effectively. The GoK ceased borrowing on non-concessional terms, and cleared its commercial overdraft, so by 2015 it no longer had any non-concessional debt, exceeding the target set for the results indicator. Outcome Indicator 5 – Closer alignment between RERF portfolio performance and relevant market benchmarks as indicated by an improvement in the performance of the RERF portfolio relative to the benchmark from a baseline of –8.29 percent in the year to March 2014 to better than –2 percent underperformance in 2015: Exceeded 40. Over the course of the programmatic series, the management of the RERF improved dramatically. As discussed above, an effective technical assistance engagement from the Bank underpinned the strengthening of the trigger on the RERF for the second operation, with the GoK agreeing to reallocate its RERF assets to new asset managers (itself a remarkable achievement of the Bank’s engagement) and reorient the portfolio towards lower- risk instruments. As part of the broader reform agenda for the RERF supported by the Bank, these reforms have been very successful in improving the management of the RERF. In 2015, the performance of the RERF portfolio was within 0.3 percent of the relevant benchmark, easily exceeding the results indicator of a better than –2 percent underperformance relative to the benchmark.6 This excellent result was achieved despite the fact that 2015 was a year of considerable transition for the RERF, with the assets transferred from the previous asset managers to money market accounts, and a transition manager hired to execute the new passive investment strategies and to achieve the least cost transition out of the old holdings into the new holdings. With the transition to passive mandates now complete, the RERF is not at risk of underperforming the benchmark in future because the investments will replicate the benchmark portfolio very closely, an extremely good outcome for Kiribati. Objective 3 – Expanding private sector opportunities: Modest 41. In expanding private sector opportunities through an improvement in the quality and reduction in the cost of telecommunication services, the programmatic series achieved modest development. The reforms were broadly implemented as envisaged, and the chosen results indicator of reduced mobile call costs was partially achieved. Outcome Indicator 6 – Reduced prices in the telecommunications sector as indicated by a reduction in the cost of a 3-minute peak-hour mobile-to-mobile telephone call within Tarawa                                                              6 Note also that at the end of 2015, the value of the RERF was 11.4 percent higher than at the end of 2014. 15 from a baseline of AUD 0.90 in mid-2014 to AUD 0.70 by the end of 2015: Partially Achieved 42. Over the course of the programmatic series, the quality of telecommunications services in Kiribati improved significantly and their costs declined, as a result of the reform program in the telecommunications sector supported by the Bank. Following on from the successful implementation of the new legislative framework under the first operation, in the second operation the GoK restructured and privatized the assets of the incumbent SOE. This was an impressive achievement of the reform process, with the buyer bringing considerable new investment to Kiribati and significantly improving the quality of services. Subscriber numbers have risen rapidly since the new operator began, with 36,000 subscribers at present, and the mobile network is now within reach of 85 percent of the population. The cost of mobile phone calls has also declined substantially. Against the target of AUD 0.70, mobile call costs have fallen to AUD 0.69 for a three minute call for existing 3G customers of the previous operator, but for customers with sim cards issued by the new operator the call costs include VAT, taking the rate to AUD 0.78. On this basis, the results indicator is only partially achieved. Broader progress on the reform agenda has continued since the reforms supported by programmatic series were implemented, with a second operator licensed in December 2015, and set to begin operations in September 2016. 3.3 Justification of Overall Outcome Rating Rating: Satisfactory 43. This overall assessment is based on the high degree of relevance of the objectives and design, the substantial degree of relevance of the implementation, and the substantial achievement of the PDOs. The relevance of the objectives, design and the implementation of the operations was a clear strength of the programmatic series. The achievement of the PDOs was also strong, with an overall rating of substantial. Within this, there was substantial achievement of the first PDO (containing one reform area), high achievement of the second PDO (containing four reform areas), and modest achievement of the third PDO (containing one reform area). 3.4 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 44. The impact of the programmatic series on poor households and vulnerable groups, including women, was indirect. In the design of the programmatic series, it was recognized that poor households and vulnerable groups were disproportionately dependent on public services for their wellbeing. In addition, consultations with women’s groups suggested that women were particularly concerned about access to basic services, including education, health and sanitation, and bore a disproportionate burden of the costs of dealing with disruptions to and unavailability of these services. By improving fiscal sustainability (both by improving the management of fisheries revenues and by improving the management of public assets and liabilities), the operations would increase fiscal space in Kiribati over time, thereby enabling the GoK to maintain and improve the public services that are of such significance to poor households and vulnerable groups, including women. In addition, by making the fiscal situation more sustainable, the GoK would be able maintain employment opportunities in the public sector, by far the largest source of formal employment for women in Kiribati. 16 (b) Institutional Change/Strengthening 45. The joint government-donor process through which Kiribati prepared its ERP assisted in strengthening the prioritization and sequencing of the GoK’s reform agenda. This occurred through the identification of necessary reforms, and the targeting of reform effort on those reforms likely to have the greatest impact. In addition, donor coordination was substantially strengthened through the ERP process, helping to harmonize the approaches of development partners, coordinate technical assistance requirements, and reinforce the GoK’s focus on key reform priorities. There was also joint monitoring of progress against the ERP, and a cooperative effort to address challenges encountered and keep reforms on track. (c) Other Unintended Outcomes and Impacts 46. N/A 3.5 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops 47. N/A 4. Assessment of Risk to Development Outcome Rating: Substantial 48. The risk to the development outcome of the programmatic series is assessed to be substantial, due to political and governance risks, macroeconomic risks and capacity risks. With respect to political and governance risks, the government in office during the preparation and implementation of the DPOs strongly supported the program of policy actions. However, as identified in the program documents for the DPOs, a number of the components of the program – particularly SOE reforms but also fisheries reforms – had been the subject of political contestation in the past, and did not necessarily enjoy cross-party support. Among these, the copra subsidy scheme has been particularly politically sensitive. The government has now changed hands, raising the possibility of changes in direction on the policies supported by the program, though to date the new government has expressed support for continuing with the reforms in key components of the program. The new government has acted to raise the level of copra subsidies paid to producers, but is conscious of the costs of administering the copra subsidy scheme. Given the strong international interest in the fisheries sector in Kiribati, and the losses to some fishing nations with political influence in Kiribati from the implementation of the national fisheries policy, there are also risks to the sustainability of the outcomes achieved under the first pillar of the operation. 49. The development outcome of the series is also subject to macroeconomic risks and capacity risks. These risks arise because the sustainability of the reforms implemented depends in part on stable macroeconomic conditions, whereas Kiribati is highly vulnerable to economic shocks, particularly from natural disasters and from global capital markets (through the RERF). Where shocks undermine macroeconomic stability, policy effort may be diverted away from the areas supported by the programmatic series and the reform progress achieved may not be sustained, in turn threatening the outcomes achieved. Similar considerations arise from the very thin capacity of the public sector in Kiribati, with the capacity needed to sustain the implemented reforms easily lost through staff turnover and diversion to other 17 policy areas. More broadly, the fragility of the state poses significant risks for the sustainability of program outcomes. 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Satisfactory 50. At entry, the operations focused on two of the most critical challenges affecting Kiribati’s development, as well as a further important challenge for private sector development. The first critical challenge affecting Kiribati’s development that the operations focused on was the management of fisheries revenues, with weak and unpredictable fisheries revenues being a key driver of the fiscal deficits that had led to unsustainable drawdowns from Kiribati’s RERF in the decade preceding the operations. The second critical challenge was the management of public assets and liabilities, with subsidies to SOEs and weaknesses in debt management being further key drivers of the fiscal deficits that had led to unsustainable drawdowns from the RERF, and with poor management of the RERF itself directly undermining its sustainability. The operations also tackled the high cost and low quality of telecommunications services, which posed an important challenge to private sector development. From the point of view of addressing Kiribati’s critical development challenges, the operations were therefore very well designed. 51. Through the dialogue and technical assistance engagements underpinning the operations, the Bank helped the GoK to formulate solid reform plans for each of the policy areas covered by the operations. The components of these reform plans were well selected and the sequencing envisaged for them was appropriate. Over time, and through the related technical assistance engagements, both the Bank and the GoK deepened their knowledge of the policy areas covered by the program. In some areas it was necessary to amend the timeframes for the implementation of the reform plans or add in incremental steps. Since some of the components affected had been selected as triggers for the second operation at the time the first operation was prepared, this involved reformulating the prior actions for the second operation so that they would continue to represent the most relevant and appropriate next steps in these reform processes. This represented an appropriate use of the flexibility built in to programmatic series, in a context where the Bank’s economic policy dialogue with the GoK was at a fairly early stage of development when the operations were designed, and there was room on both sides to deepen knowledge of the policy areas being addressed. It was also an appropriate use of the flexibility built in to programmatic series, in the context of designing and implementing operations focused on critical development challenges in a fragile state. The risks to the timely and successful implementation of reforms in this very thin capacity context were flagged upfront in the program documents. 52. The design of the results framework for the programmatic series had a number of strengths, but also had weaknesses. Among the strengths were the approach of focusing on indicators that could be measured with readily available data, to avoid adding to the capacity constraints of the GoK by requiring additional data to be generated for the DPO series itself. Most of the chosen indicators were also well-specified. In two cases, however, 18 the results indicators could have been better specified, by being aligned more closely with the central aspects of the policy actions being supported. With respect to the fisheries indicator, in the view of the ICR team it would have been better to retain a results indicator linked to the broader objective of increasing revenue per tonne of catch (as was the case under the first operation), than to have included a results indicator focused on the narrower objective of improving the accuracy of revenue predictions at a time when the region was in a transitional phase to the new VDS, with its very significant (and unpredictable) price effect. With respect to the copra subsidy indicator, in the view of the ICR team it would have been better to specify an indicator of the administrative costs of the subsidy scheme, rather than its overall cost. These two weaknesses in the design of the results indicators make the outcomes achieved by the operations look less successful than they in fact are. Weighed up against the strengths of the rest of the results framework, together with the extremely strong design of the reform program targeting critical development challenges and coordinating technical assistance from multiple donors in support of these reforms in a fragile state, the performance of the Bank in ensuring quality at entry is rated as Satisfactory. (b) Quality of Supervision (including M&E arrangements) Rating: Satisfactory 53. The Bank team’s supervision of the reform program supported by the operations was Satisfactory. The team liaised very closely with the GoK throughout the implementation of the reforms and undertook frequent missions to help support reform momentum, identify obstacles to progress, and help the GoK find solutions to the problems encountered. Following on from its initial public expenditure analysis, the Bank also provided technical assistance to the GoK in a number of related areas to support the program (including on the VDS, the copra subsidy scheme, and alternative social protection options). The Bank team’s assistance was not narrowly focused only on the prior actions and triggers, but supported the broader reform processes that these prior actions and triggers were drawn from, in order to increase the prospects that the GoK would be successful at implementing its reform program. The Bank team worked very closely with other development partners, to increase the extent and frequency of the support being provided to the GoK through the coordinated approaches of these other development partners, and to ensure that Kiribati received technical assistance that was comprehensive of all of the areas of the reform program. This was an important factor in enabling the Bank to achieve a greater degree of supervisory support for the program than would have been possible with the small, remotely located Bank team alone. (c) Justification of Rating for Overall Bank Performance Rating: Satisfactory 54. This overall assessment is based on the satisfactory performance of the Bank in ensuring quality at entry and the satisfactory performance of the Bank in ensuring quality of supervision. At entry, the operations focused on two of the most critical challenges affecting Kiribati’s development, and a further important challenge for private sector development. Through dialogue and technical assistance, the Bank helped the GoK to formulate and implement solid reform plans for each of the policy areas covered by the operations. Alongside its strengths, the design of the results framework did have weaknesses, with two of the results indicators not well aligned with the core objectives of the related 19 reform areas. But given the remaining strengths of the results framework, the extremely strong design of the reform program targeting critical development challenges and coordinating technical assistance from multiple donors in support of these reforms in a fragile state, and the satisfactory supervision of the reform program supported by the operations, the overall performance of the Bank is rated as satisfactory. 5.2 Borrower Performance Check here if the Government and the Implementation Agency is the same or X indistinguishable. Rating: Highly Satisfactory 55. For the purpose of this review, Government performance and Implementing Agency performance are considered the same. Therefore, the above rating should be viewed as an overall rating for the Borrower. 56. The GoK’s ownership of the process of designing its reform program and the solid progress it made with the implementation of that reform program are impressive, particularly given the significance of the development challenges it was tackling, the fragility of the state, and the extreme capacity constraints of the public sector. From the outset, the GoK recognized the seriousness of the fiscal and broader development challenges it faced, and also recognized that it needed technical assistance to help it identify the main causes of these problems and to formulate and implement a reform plan to tackle these. The GoK embraced a reform program that confronted its most critical development challenges, many of which were complex, contentious and involved strong vested interests opposed to reform. It implemented the components of these reform plans relatively successfully, though in some instances the timeframes needed to be adjusted or incremental steps added in light of the GoK’s capacity constraints or weaknesses of internal coordination that limited its ability to tackle vested interests opposed to reform. In instances where parts of the government took actions that were contrary to the spirit of the reform plans – for instance, the opening of a separate offshore bank account by the MFMRD – the GoK successfully addressed these weaknesses with support from the Bank, and was thus able to remain on track with the implementation of its reform program. For a fragile state with extremely limited capacity, tackling significant development challenges, Kiribati’s performance was really impressive. 6. Lessons Learned 57. Budget support to small states with thin capacity can be a key element of development assistance, providing benefits far beyond the financial resources transferred. These benefits can include an evidence-based policy dialogue, an enhanced capacity to identify key development constraints and formulate focused reform plans, the provision of related pieces of analytical work, the provision of technical assistance engagements and capacity-building activities to support reform implementation, and close donor coordination. This last benefit is particularly important for such thin capacity environments, because joint budget support operations can bring donors together around a tightly-focused reform plan addressing key government priorities, with reform momentum and implementation capacity supported through a policy dialogue with coherent and consistent advice, and through related analytical work and technical assistance. 20 58. Project-level engagements can assist in identifying key priorities to be supported through policy-based operations, and vice versa. This programmatic series successfully leveraged the Bank’s project-level engagement in the telecommunications sector, to provide reform support at the policy level for the telecommunications sector. At the same time, the Bank’s broader economic policy dialogue on the reform program for the programmatic series provided the basis for the Bank to develop a deep engagement with the GoK on asset management, which has had a significant positive impact on Kiribati’s critical RERF assets and which has since been extended to other Pacific Islands with trust fund resources. 59. It is important that task teams consider results indicators at the same time as they are working with counterparts on policy actions when preparing DPOs. Fairly understandably, the focus during the preparation of DPOs is primarily on the policy actions being supported, with results indicators often formulated at a very late stage in the process. The problem with a near-exclusive focus on the policy actions is that it reduces opportunities for the results indicators to be reconsidered and refined over time, including through consultation with counterparts. This, in turn, increases the chances of including results indicators that are not as well specified as they could be, risking making operations look less successful than they in fact are – an unfortunate situation that should of course be avoided. 60. The flexibility built into the Bank’s approach to programmatic series of DPOs is vital in fragile states and very thin capacity environments. In fragile states and very thin capacity contexts, it is likely that changes to the timeframes for reforms and/or to the ambition of different steps in reform processes is going to be necessary, because the internal coherence of the state is relatively weak and the capacity to implement reforms is so easily undermined by staff turnover and/or diversion to other policy areas. These weaknesses can be mitigated to a limited extent, but by no means resolved. Where clients make progress in good faith with the implementation of agreed reform plans, but this progress is undermined by internal state weakness or thin capacity, it is important that the Bank uses the flexibility it has to adjust triggers in subsequent operations to accommodate these changes. By doing so, the Bank retains the incentive for the client to make progress on the other reform areas according to the original schedule, gives due acknowledgement to reform endeavors made in good faith, and supports macroeconomic stability through the predictable flow of resources that have been included in national budgets. These risks should of course be clearly stated upfront in program documents, as they were in this programmatic series in Kiribati. 61. The flexibility built into the Bank’s approach to programmatic series of DPOs is also vital when the Bank’s economic policy dialogue with the client is in its early stages of development. In this case, the Bank was able to begin a programmatic series of DPOs within a year of establishing an economic policy dialogue with the client. The Bank and the GoK had enough knowledge of the macro-fiscal context to be able to identify critical development constraints and formulate reform plans to address key policy areas. Not all the relevant analytical work had been completed when the series was initiated, nor had all the relevant technical assistance been provided – those proceeded in parallel with the commencement of the series. Through those engagements, the Bank and the GoK deepened their knowledge of the reform areas that were the focus of the operations, which led both parties to recognize that the resequencing of reform plans and/or the addition of some incremental steps was needed in particular reform processes. Appropriately, the Bank used the flexibility it had to adjust triggers in the second operation to accommodate these changes. 21 Again, this meant that the incentive for the client to make progress with its reforms was retained, reform endeavors undertaken in good faith were acknowledged, and macroeconomic stability was supported through a predictable flow of budget support resources. 62. The final lesson learned from this programmatic series is that tackling more contentious policy areas – particularly in fragile states – involves taking greater risks, and that the Bank needs to support this. During the course of the policy dialogue for these operations, it appeared to the Bank team that there was an opportunity for traction on tackling some contentious policy areas (like SOE reform in general, and the copra subsidy scheme in particular) and some areas with very strong vested interests (like fisheries). The team took the risk of engaging in these areas with a fragile client, making the programmatic series more challenging than it would otherwise have been (in terms of the related technical assistance needed, the frequency of missions, the extent of support required for the dialogue, and the difficulties of negotiating changes in triggers through the Bank’s internal processes). To support the taking of this kind of risk in these kind of contexts, the Bank needs to ensure that it can maintain the Bank’s policy engagement with the client at a similar level of intensity over an extended period, to give adequate time for reforms to be implemented and bedded down, and to reduce the risks of policy reversions in future. This kind of continuity is impeded by significant decreases in the resources available to sustain policy dialogues and by transitions in TTLs that do not occur in a timely manner. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/Implementing Agencies 63. The comments provided by the GoK indicate the ongoing importance to the GoK of the ERP that the Bank has supported in Kiribati, including through the programmatic series of DPOs reviewed in this ICR. The new programmatic series of DPOs now under preparation by the Bank is designed to support key elements of the new ERP, which has just been approved by Kiribati’s Cabinet. (b) Cofinanciers 64. N/A (c) Other partners and stakeholders 65. The ICR team concurs with Australia’s Department of Foreign Affairs and External Trade, that the results indicators as specified do not adequately reflect the positive outcomes of the programmatic series. Had the fisheries and copra subsidies indicators been better specified, it is indeed likely that it would have been possible to rate the achievement of outcomes of the operations as High Satisfactory. 22 Annex 1: Bank Lending and Implementation Support/Supervision Processes (a) Task Team Members P144602 – Kiribati: First Economic Reform Development Policy Operation Name Title Unit Responsibility/Specialty Lending Tobias Haque Economist EASPN Team Leader Stephen Hartung Financial Management Specialist EAPFM Financial Management Robert Gilfoyle Senior Financial Management Specialist EAPFM Financial Management Marjorie Mpundu Senior Counsel LEGES Country Lawyer Patricia Hoyes Senior Finance Officer CTRLN Loan Officer Samantha Evans Team Assistant EACNF Team Member Supervision Tobias Haque Economist EASPN Team Leader P149888 – Kiribati: Second Economic Reform Development Policy Operation Name Title Unit Responsibility/Specialty Lending Tobias Haque Economist GMFDR Team Leader Stephen Hartung Financial Management Specialist GGODR Financial Management Marjorie Mpundu Senior Counsel LEGES Country Lawyer Lead Financial Officer/ Ekaterina Gratcheva FABRP Asset Management Reserve Advisory Management Dmitry Pevzner Senior Investment Officer FABRP Asset Management Natasha Beschorner Senior ICT Policy Specialist GTIDR Telecommunications Samantha Evans Team Assistant EACNF Team Member Supervision Tobias Haque Economist EASPN Team Leader (b) Staff Time and Cost Staff Time and Cost Cost (US$) Stage Staff Time including travel (No. of Staff Weeks) and consultant costs Lending First Economic Reform Operation 27 156,318 Second Economic Reform Operation 17 271,804(a) Total: 44 428,122 Supervision First Economic Reform Operation - - Second Economic Reform Operation - - Total: - - Note: (a) While the investment of staff time was greater for the first operation, the related technical assistance provided to Kiribati depended on a greater degree of consultancy support for the second operation, accounting for the higher total cost. 23 Annex 2: Comments of Borrower on Draft ICR Firstly, the Government and the people of Kiribati would like to express their sincere and profound appreciation for the Bank’s continued support and assistance in developing, implementing, and refining the Government’s first and second Economic Reform Plans. The overall positive performance results or ratings reported in this ICR on those plans would not have been possible without the Bank’s enduring technical and financial supports. Second, the Government is very pleased to confirm that its Third Economic Reform Plan for 2016 to early 2017 which was under negotiation has now been approved by Cabinet and as such we now look forward to our continued partnership with the Bank in carrying out this Economic Reform Plan. The Government would like to reiterate its commitment to the Bank to continue and support the country’s ongoing and new ERPs. 24 Annex 3: Comments of Partners on Draft ICR Australia’s Department of Foreign Affairs and External Trade: The review provides a good summary of the context, process and outcomes of the DPO. Currently the assessment is based on the extent to which the results indicators have been met, with the conclusion that the DPO was satisfactory. It is arguable that the results indicators do not adequately reflect the positive outcomes achieved by the DPO in a new and challenging context and therefore the rating of satisfactory undervalues the impact of the program. This is particularly the case with Pillar One, where the broader benefit was increases in fisheries revenue, even though predictability of revenues was not achieved. However, on this point it would also be useful to assess the extent to which such increases in revenue are structural versus attributable to reforms flowing from the DPO. 25 Annex 4: List of Supporting Documents 1. Program Document, November 2013 2. Letter of Development Policy, October 2013 3. Program Document, October 2014 4. Letter of Development Policy, September 2014 26 MAP