59618 IEG Working Paper 2010/6 Poverty Reduction Support Credits: Armenia Country Study Norman Hicks 2010 The World Bank i Washington, D.C. © 2010 Independent Evaluation Group, The World Bank Group 1818 H St., NW Washington, DC 20433 IEG: Improving Development Results Through Excellence in Evaluation The Independent Evaluation Group is an independent unit within the World Bank Group; it reports directly to the Bank's Board of Executive Directors. IEG assesses what works, and what does not; how a borrower plans to run and maintain a project; and the lasting contribution of the Bank to a country's overall devel- opment. The goals of evaluation are to learn from experience, to provide an objective basis for assessing the results of the Bank's work, and to provide accountability in the achievement of its objectives. It also im- proves Bank work by identifying and disseminating the lessons learned from experience and by framing recommendations drawn from evaluation findings. IEG Working Papers are an informal series to disseminate the findings of work in progress to encourage the exchange of ideas about development effectiveness through evaluation. The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessari- ly reflect the views of the Board of Executive Directors of the World Bank or the governments they represent, or IEG management. The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply on the part of the World Bank any judgment of the legal status of any territory or the endorsement or acceptance of such boundaries. ISBN-10: 1-60244-149-9 ISBN-13: 978-1-60244-149-1 Contact: IEG Communication, Learning and Strategies (IEGCS) e-mail: ieg@worldbank.org Telephone: 202-458-4497 Facsimile: 202-522-3125 http:/www.worldbank.org/ieg ii Contents ACRONYMS ................................................................................................................................................................. V ACKNOWLEDGMENTS .............................................................................................................................................. VI EXECUTIVE SUMMARY ............................................................................................................................................ VII 1. BACKGROUND ..............................................................................................................................................1 Country Context .............................................................................................................................................................. 1 Situating the PRSC in the Country Assistance Program ............................................................................................ 3 2. PRSC DESIGN ................................................................................................................................................3 To what extent is there a discernable difference between the PRSC series and earlier adjustment lending/development policy lending? ........................................................................................................................... 3 3. PRSC PROCESS ............................................................................................................................................6 How effective have PRSCs been as a vehicle to help Armenia operationalize a country-driven poverty reduction strategy?......................................................................................................................................................... 6 How effectively have PRSCs contributed to donor harmonization around a country-owned, medium-term poverty reduction strategy?......................................................................................................................................... 10 How well have PRSCs been aligned with country policy formulation, budgeting, and planning processes, and how effectively have they contributed to predictable resource flows? ........................................................... 11 4. PRSC RESULTS ...........................................................................................................................................13 How effectively have the PRSCs helped Armenia strengthen its public financial management systems?......... 13 Outcomes ..................................................................................................................................................................................... 15 How relevant and effective a vehicle have PRSCs been in helping Armenia set conditions for poverty-reducing growth? .......................................................................................................................................................................... 18 How effectively has the PRSC helped advance the dialogue and achieve results in sectors delivering services to the poor? ................................................................................................................................................................... 24 PRSC Contributions to Sector Development: Health ................................................................................................................... 24 PRSC Contributions to Sector Development: Education ............................................................................................................. 26 PRSC Contributions to Sector Development: Social Protection .................................................................................................. 28 Assessment of Overall PRSC Outcomes ..................................................................................................................................... 29 PRSC Contributions to Poverty Outcomes................................................................................................................................... 29 5. BANK PERFORMANCE ...............................................................................................................................29 6. CONCLUSIONS AND LESSONS LEARNED ...............................................................................................30 iii REFERENCES ............................................................................................................................................................33 Tables Table 1. Summary of Adjustment Lending to Armenia, FY95-FY08 ..............................................................................3 Table 2. Financial Role of PRSCs in Armenia..............................................................................................................11 Table 3. PRSCs, ODA, and GBS in Armenia, 2001-2007 ............................................................................................12 Table 4. Evolution of Aid Volatility ................................................................................................................................13 Table 5. Macro Targets and Accomplishments ............................................................................................................19 Table 6. Outcomes on Growth and Poverty, 2004-2007 ..............................................................................................20 Table 7. Selected Public Sector Indicators (% of GDP) ...............................................................................................22 Table 8. Monitoring Indicators for Infrastructure and Private Sector Development ......................................................23 Table 9. Results in the Education Sector .....................................................................................................................26 Figure Figure 1. Armenia Quarterly IDA Disbursements for DPLs 1995-2008 ..........................................................................6 Appendix Appendix Table 1. Share of PRSC and Adjustment Lending in Total Disbursement to PRSC Countries ....................35 Appendix Table 2. PRSC Disbursements as Share of GBS, Government Expenditures .............................................36 Appendix Table 3. Predictability of PRSC and Other GBS Resource Flows ................................................................37 Appendix Table 4. Alignment of PRSC Process with Domestic Planning and Budgeting Processes, for 2004 ...........38 Appendix Table 5. Alignment of PRSC Process with Domestic Planning and Budgeting Processes, for 2006 ...........40 Appendix Table 6. Alignment of PRSC Process with Domestic Planning and Budgeting Processes, for 2007 ...........42 Appendix Table 7. PEFA Scores for Armenia (2006) ...................................................................................................44 Appendix Table 8. Relationship of the PRSC Program to Other Bank Lending Projects ............................................. 45 Appendix Table 9. Results in the Health Sector ...........................................................................................................46 Appendix Table 10. Selected Indicators of Non-Income Poverty in Armenia ...............................................................47 Appendix Table 11. Relationship of PRSC Program to Bank AAA Work .....................................................................48 iv Acronyms AAA Analytical and advisory activities MDG Millennium Development Goal APR Annual Progress Report MDTF Multi-donor trust fund BEEPS Business Environment and MoFE Ministry of Finance and Economy Enterprise Performance Survey MTEF Medium-Term Expenditure CAS Country Assistance Strategy Framework CPS Country Partnership Strategy NCD Non-communicable disease CFAA Country Financial Accountability NGO Nongovernmental organization Assessment ODA Official development assistance CIS Commonwealth of Independent PAF Performance assessment States framework CPAR Country Procurement Assessment PCR Project Completion Report Report PD Program Document DfID Department for International PEFA Public Expenditure and Financial Development Accountability DPL Development policy loan PER Public Expenditure Review DTI Direct Trade Input PFM Public financial management EC European Commission PHC Primary health care EU Euro (or European Union) PPAR Project Performance Assessment FSU Former Soviet Union Report FY Fiscal year PRGF Poverty Reduction and Growth GBS General budget support Facility (IMF initiative) GDP Gross domestic product PRS Poverty reduction strategy GFS Government Finance Statistics PRSC Poverty reduction support credit GNP Gross national product PRSP Poverty Reduction Strategy Paper ICA Investment Climate Assessment RC Rehabilitation credit ICR Implementation Completion and SAC Structural Adjustment Credit Results Report SHA State Health Agency IDA International Development SIDA Swedish International Association Development Corporation IEG Independent Evaluation Group SNOC State non-commercial IFAD International Fund for organization Agricultural Development TA Technical assistance IFRP Integrated Financial TF Trust fund Rehabilitation Plan UN United Nations IGR Institutional and Governance USAID United States Agency for Review International Development IMF International Monetary Fund USD U.S. dollars IPSAS International Public Sector VAT Value-added tax Accounting Standards WDI World Development Indicators M&E Monitoring and evaluation v Acknowledgments This paper is one of a series of seven background country case studies, prepared for the IEG evaluation of the World Bank's support for Poverty Reduction Support Credits (PRSCs), coordi- nated by Anjali Kumar. The paper was prepared by Norman Hicks, supported by the World Bank's country department for Armenia, which helped to organize the field visit. The program of in-country work benefitted especially from support by Mushegh Tumazyan, and support in Washington was provided by Andrew Waxman, Research Analyst. Valuable comments are ac- knowledged from Armenia Country Team members, including Senior Country Economist Robert R. Taliercio and Saumya Mitra. The paper has been edited by Anjali Kumar and Jesse Torrence. vi Executive Summary Armenia gained its independence from the Soviet Union in 1991 and inherited a traditional So- viet-style command economy. After a period of hyperinflation and economic contractions, progress on basic reforms led to greater macroeconomic stability. The reform effort streng- thened after 2000, aided by return flows of investment from the diaspora. Recognized for its prudent fiscal management and improved business environment, Armenia received a Rehabili- tation Credit (RC) from the World Bank in 1995 and a series of five Structural Adjustment Cre- dits (SACs) from 1996-2003 to maintain stability, accelerate growth, and sustain reform. A World Bank evaluation--undertaken by its Operations Evaluation Department (OED)--of the first three Structural Adjustment Credits found that they placed insufficient emphasis on pri- vate sector development. This led to a stronger focus on second-generation private sector and institution-building reforms in subsequent operations. This emphasis was carried over into the Poverty Reduction Support Credit (PRSC) program, which consisted of four operations from 2004 to 2008. However, the PRSC represented a major departure from the prior adjustment lending model in that it employed a single-tranche disbursement mechanism, a more flexible approach to loan conditionalities, and strong linkages to Armenia's Poverty Reduction Strategy Paper (PRSP). With the Armenian government strongly committed to implementation of a Po- verty Reduction Strategy, the robust links between the PRSC and PRSP have helped deepen Government buy-in for PRSC supported reforms. The PRSC program in Armenia centered on four main themes that supported Armenia's pover- ty-reduction objectives: i) consolidating macroeconomic discipline and strengthening gover- nance; ii) sharpening competition and entrenching property rights; iii) mitigating social and en- vironmental risks; and iv) modernizing the rural economy. The PRSC also identified good governance and reduction of corruption as essential to achieving pro-poor growth. Further- more, reforms implemented under the PRSC were intended to make the social sectors more effi- cient, raise the quality of services, and increase access for the poor. Although many of the reforms supported by the PRSC program were institutional in nature and, therefore, will require some time to fully realize their full impact, thus far, the results in these four main areas are mixed. Notable advances were made in the area of public financial management (PFM), especially with regards to eliminating arrears, rationalizing extra- budgetary funds, and increasing budget transparency and accountability. Despite significant efforts and some progress, scope for improvement remains in persistently weak areas such as internal and external audit functions, transparency, anticorruption, and procurement. In health, overall access has improved but access by the poor has not, and health indicators have not im- proved. In education, overall enrollment rates have increased, but not in primary education. Per-student educational spending has increased, but there are relatively few indications of im- proved learning outcomes. For social protection, the PRSC receives high marks for reforms that have tightened targeting mechanisms and helped increase access for the poor. The family bene- fit program, for example, has had a major impact on reducing poverty. All in all, poverty de- vii clines have been dramatic during the PRSC period: poverty rates fell by 29 percent between 2004 and 2006, and extreme poverty rates by 36 percent. Although progress in reduction of non- monetary poverty is less pronounced, income distribution seems to have improved, making the entire experience a pro-poor growth episode. The lack of a true results focus was probably the weakest aspect of the PRSC program in Arme- nia. Despite commendable efforts to link each PRSC to a table of expected results indicators and actions, and to employ the government's own PRSP monitoring systems, there was little actual monitoring of results on the basis of these indicators. Rather, each new PRSC produced a new table with slightly different indicators. While each new PRSC reviewed general progress on past policy actions, there was no review of policy indicators and progress made in meeting goals. In regards to donor coordination, PRSC was not used to harmonize donor support; only one other donor provided general budget support, and this donor already linked its assistance to the PRSC. Donors worked well together on policy issues like tax reform and local aid coordination. The Armenian reform program has taken a gradualist approach and moved forward cautiously. Overall Bank performance on the PRSC was very good and supported the PRSC process well. The PRSC clearly built on, and benefited from, prior experience with policy-based lending. The Government also learned from and adapted its approach based on previous Structural Adjust- ment Credit experience. All in all, its engagement under the PRSCs was successful and some- what smoother than under the adjustment credits, with the PRSP providing a useful framework for determining joint priorities. In effect, the PRSCs were an instrument that complemented project lending in many key sectors, and Bank staff were seen by counterparts as valuable colla- borators in the process of strategy formulation and implementation. While PRSC disbursement may no longer be critical for Armenia, the Government continues to place value on being in compliance with the World Bank program. In general, the PRSC has proven its usefulness in Armenia as a vehicle for developing policy di- alogue on key issues and providing support for a government-led poverty reduction strategy. From the Bank side, it would not have been possible to implement these reforms using project finance or other instruments. The PRSCs provided a vehicle that supported and complemented project lending, since it allowed larger policy issues to be raised at the cabinet level. For the Government, the PRSC was a valuable instrument allowing it to implement a reform program while drawing on Bank-funded technical advice; it also permitted a valuable two-way relation- ship between the Ministry of Finance and operational ministries. Several important lessons emerge from Armenia's experience, including: Corruption was problematic for the PRSCs' strategy of promoting poverty reducing growth. Not enough attention was paid to key public financial management issues-- quality of audits, procurement, and management of state non-commercial organizations. PRSCs do not automatically lead to more regular disbursements of budget support. In the case of Armenia, this lack of regular budget support did not severely hamper the government's finances since it occurred during a relatively more stable fiscal period. viii The results orientation of the PRSCs was one of its weaker aspects. Progress reporting against a stable, agreed-upon set of objectives based on available data from the Govern- ment needs to be adopted more completely. Monitoring of past actions is critical to ensure that back-tracking does not occur but was insufficient in the case of the PRSCs in Armenia. In Armenia, poverty declined and income distribution improved, but unemployment rates and other social indicators stagnated or improved only slowly. Given that one of the principal goals of the PRSCs was to not only promote growth, but also reduce pover- ty and improve social conditions, poverty monitoring should track non-monetary, as well as income-based, measures of poverty. This is consistent with the PRSC's multi- dimensional nature and with the aid context within which the PRSC was introduced. Much can be accomplished with a PRSC even when it constitutes a small amount of support vis-à-vis the overall budget. Its value is not only its financing, but also its ability to give shape to a reform program and serve as a nexus for organizing assistance. The PRSCs work best in conjunction with project lending. Working in tandem, they can influence a policy framework, with projects providing funding for the necessary tech- nical assistance and capacity building. In essence, the PRSCs and project lending can -- and should -- serve as mutually reinforcing elements. ix x 1. Background Country Context 1.1 Economic performance. Armenia gained its independence from the Soviet Union in 1991 and inherited a traditional Soviet-style command economy, based on heavy intervention by the state and large investments in infrastructure. After a period of hyperinflation and severe eco- nomic contractions from 1991 to 1993, macroeconomic stability was achieved, inflation was re- duced, structural reforms were successfully carried out, and recovery began. The process of moving to a more open, market economy has taken some time, but has seen steady progress. Over the past three decades, the country has managed to maintain political stability within a democratic multi-party framework, and economic reforms have successfully dismantled most of the Soviet-style system. Most small and medium-sized enterprises have been privatized, and the Government's intervention in the economy has been considerably reduced. 1.2 Nonetheless, the reform effort has not been entirely smooth. Rapid progress on basic re- forms was made during 1991-1992, but this stalled during the armed conflict of 1992-1994 with Azerbaijan, which led to the closure of transport routes and had a negative effect on trade and donor support. Reforms accelerated again from 1994 to 1996, but slowed in the late 1990s. The reform effort strengthened once again after 2000, with a shift toward: improving the business environment and public utilities management; strengthening governance, competition, and fi- nancial intermediation; and promoting structural changes in public service delivery. The coun- try's fiscal stance has been prudent, with an overall fiscal deficit of 2.3 percent of Gross Domes- tic Product (GDP) during 2001-2006. Tax revenues, while increasing, remain relatively low at 14 percent of GDP. Social expenditures have been rationalized to eliminate excess capacity and overstaffing, and have increased from 7.7 percent of GDP in 2003 to 9.7 percent in 2006. 1.3 Growth averaged 5.5 percent per year during the period 1994-2000, one of the highest rates among the former Soviet Union countries, although, by 2000, per capita GDP was still only 66 percent of the 1990 level. A growth rate of 12 percent was achieved over the 2001-2006 pe- riod, fueled by export growth, expansion in construction, and service sector growth. A major source of financing for the construction boom has been return flows of investment in real estate from the diaspora of Armenians overseas. Growth in 2007 continued to be high at an estimated 14 percent (though this dropped to 7 percent in 2008) while inflation remained low and within the target of 3 percent. Poverty rates have fallen from 55 percent in 1999 to 27 percent by 2006, with extreme poverty falling from 21 to 4 percent.1 Income distribution improved over the same period, with the Gini coefficient falling from 0.59 in 1999 to 0.40 in 2005. Yet growth has had lit- tle impact on employment: nearly one-third of the labor force remains unemployed. 1 The poverty figures used here come from Table 2 of the PRSC IV President's Report and differ from the poverty numbers in the draft PRSP-II. Gini numbers come from the PRSP-I and draft PRSP-II. 1 1.4 Prior to the beginning of its Poverty Reduction Support Credit (PRSC) program, Arme- nia received a Rehabilitation Credit (RC) in 1995 and a series of five Structural Adjustment Cre- dits (SACs) from 1996 to 2003. The SAC program was designed to establish macroeconomic sta- bility, improve financial discipline in the private and public sectors, develop the private sector, and ensure the social sustainability of reforms. These credits facilitated: the removal of nearly all price controls and trade barriers; privatization of nearly all state-owned enterprises; streng- thening of the financial sector; improvement of payment discipline and establishment of a strong regulatory regime in the energy sector; significant strengthening of irrigation and drink- ing water firms; establishment of a legal framework suitable for a market economy; important education reforms; and improved benefits targeting. Subsequently, Armenia received four PRSCs between 2004 and 2007 (PRSC IV was approved in November 2007, disbursed in early 2008, and closed June 30, 2008).2 1.5 A 2002 OED3 Project Performance Assessment Report (PPAR) rated the Rehabilitation Credit highly satisfactory, the first structural adjustment credit satisfactory, the second adjustment credit moderately unsatisfactory, and the third adjustment credit moderately satisfactory, although they found that all three placed insufficient emphasis on critical aspects of private sector devel- opment (OED, 2002). Responding to these findings, later structural adjustment credits, especial- ly the fourth and fifth, focused less on macro stability and core liberalization efforts and more on "second generation" reforms related to providing an environment for private sector devel- opment and building institutions. This emphasis was carried over into the PRSC program. Im- plementation Completion Reports (ICRs)4 by the Bank's operational teams for the four PRSC operations rated them all as satisfactory. 2 The Poverty Reduction Support Credit (PRSC) was first introduced by the World Bank in early 2001 in the context of global changes in aid architecture that recognized the importance of country ownership, government reform commitment, and multi-dimensional poverty reduction, PRSCs were intended to aid country-owned Poverty Reduction Strategies, support comprehensive growth, improve social conditions, and reduce poverty. Compared to previous adjustment lending, PRSCs aimed to ease conditionality, provide more predictable annual support and strengthen budget processes, all in a results-based frame- work. Many of its principles were reflected in the Paris Declaration of Aid Effectiveness. Within four years of their introduction, PRSCs came to account for almost 60 percent of IDA policy based lending and a quarter of total Bank policy based lending. From FY01-FY08, the Bank approved 87 PRSC operations amounting to US$ 6.6 billion. By end-September 2009, PRSC approvals increased to 99 operations, with another 20 in the pipeline. 3 The Operations Evaluation Department (OED) of the World Bank was renamed the Independent Evalu- ation Group (IEG) in 2005. The two terms are used interchangeably in this report depending on the time period in question (pre-2005 or beyond). 4 Bank operational staff prepare a self-evaluation (known as an Implementation Completion Report or ICR) for every completed project. IEG staff then review every ICR, validate the self-rating, and identify projects that offer good potential for further learning (because of particularly good or bad performance) as candidates for a project performance assessment. One in four completed projects (or about 70 per year) is subject to a Project Performance Assessment Report (PPAR), which takes about six staff weeks to pro- duce and normally includes a field mission. Project Performance Assessment Reports rate projects in terms of their outcome (taking into account relevance, efficacy, and efficiency), sustainability of results, and institutional development impact. PPAs carried out after Bank funds have been fully disbursed to a project, are similar to the completion evaluations carried out by many development agencies, and are the main project-level evaluations conducted by IEG. http://www.worldbank.org/oed 2 Situating the PRSC in the Country Assistance Program 1.6 The PRSC program was preceded by a Rehabilitation Credit (RC) and five Structural Adjustment Credits to Armenia during FY96-03, which totaled USD 342.4 million and averaged 43 percent of the total lending program of the Bank during that period. The four PRSCs from FY05 to FY08 (summarized in Table 1) amounted to a total of USD 88.6 million, or about 31 per- cent of the total lending program to Armenia (see Appendix Table 1).The PRSC program was seen as a continuation of the Structural Adjustment Credit program and as a vehicle for ad- dressing policy issues that could not be adequately addressed in project lending. This included issues not covered by project loans and those that helped complement project loans. Project staff reported useful complementarities between the PRSC and their projects, and even project staff who were initially skeptical of the PRSC approach became "believers." The PRSC program was never designed to replace project lending, but maintained a level of policy based lending of 25- 30 percent of total lending established during the structural adjustment period. Bank staff and government counterparts confirmed that the Government was comfortable with this level of policy-based lending, as it gave the Ministry of Finance and Economy (MoFE) leverage over other ministries and control over the policy process. Table 1. Summary of Adjustment Lending to Armenia, FY95-FY08 Operation Board Approval Date Amount No. of Tranches (USD millions, equiv) Rehabilitation Credit (RC) Feb. 1995 69.0 2 SAC-I Feb. 1996 60.0 2 SAC-II Aug. 1997 60.0 2 SAC-III Dec. 1998 63.4 4* SAC-IV May 2001 50.0 3 SAC-V May 2003 40.0 2 PRSC I Nov. 2004 21.2 1 PRSC II Jan. 2006 20.3 1 PRSC III March 2007 28.6 1 PRSC IV Nov. 2007 18.5 1 Source: Business Warehouse, World Bank * The third tranche of SAC-III was split into two 2. PRSC Design To what extent is there a discernable difference between the PRSC series and earlier adjustment lending/development policy lending? 2.1 Design and content. There were no major changes in focus between the Structural Ad- justment Credit program and the PRSCs. In fact, the PRSC program was mostly a direct contin- uation of the structural adjustment program in terms of sector content and often continued or completed reforms initiated under the adjustment credits. Although the adjustment credits had more conditions on macro policies, public enterprise restructuring, privatization, and the power 3 sector compared to the PRSCs, they both seem to have paid equal attention to public financial management, health and education, and social protection. There is no overwhelming evidence that the PRSCs were more poverty focused, either in terms of the number of conditions or their composition. Specifically the PRSCs focused on: Public administration reform, including tax and customs reforms; Improved public management, including budget, audit, and procurement reforms; Corporate governance reforms; Liberalization in key services, mainly civil aviation and telecommunications; Regulatory reform in state owned utilities and improved property rights for secured transactions; Improvements in banking and insurance regulation; Improvements in social protection and social risk management, including pensions; Raising the quality of education services, including early childhood education; Improving health services, including rationalization of health facilities and staff; Improving forestry management, including controls on illegal logging; and Raising agricultural productivity, with a focus on agricultural extension, seeds, and food safety. 2.2 The main difference between the PRSC and the Structural Adjustment Credit, according to every government official familiar with the two programs, was how the programs were de- veloped. In the case of the adjustment credits, the conditions and focus of the program were largely determined by the Bank; in the case of the PRSCs, the Government determined the over- all strategy in its Poverty Reduction Strategy Paper, and the PRSCs then supported this strate- gy.5 2.3 The Government took the poverty reduction strategy process seriously and was commit- ted at the highest level, and the first strategy was based on a wide range of consultations with civil society and nongovernmental organizations. This strategy was, in turn, used to shape the Medium-Term Economic Framework (MTEF), and this framework was used to shape annual budgets. All counterparts interviewed agreed that there was very close harmonization between the Poverty Reduction Strategy Paper, Medium-Term Economic Framework, and the budget. 2.4 Approach to conditionalities. Another difference between the two programs was in the number of conditions and tranches. While all Structural Adjustment Credits had two or more tranches, the PRSCs were single-tranche operations, which helped reduce the tensions over tranche release conditions. PRSCs tended to have fewer conditions. The adjustment credits had extensive and detailed conditionality, with a peak of 66 conditions in the fourth Structural Ad- justment Credit. PRSCs I-III had about 33 conditions or prior actions each; of these, about 10-12 were core conditions.6 Most, but not all, PRSCs and Structural Adjustment Credits identified le- 5 However, some ex-government officials now in the private sector suggested that both the SAC and PRSC were developed in collaborative framework and that the degree of difference between the two was not that great. 6 Since the PRSCs are single-tranche operations, legally binding conditions are, in reality, all prior actions. Hence, they are referred to as simple "prior actions". Triggers, on the other hand, are non-legally binding tranche release conditions (also called "indicative prior actions") for moving to the next operation in a PRSC series. 4 gally binding prior actions (sometimes referred to as "core conditions"). PRSC IV was an excep- tion to this, as it included only seven such conditions with no identified conditions for a future operation. The reduction in the number of non-binding and binding conditions in PRSCs over time reflected pressure from within the Bank to simplify Development Policy Loans and to have fewer conditions. 2.5 PRSC flexibility compared to earlier lending instruments. Overall, the PRSC program ap- pears to have been as flexible as the structural adjustment program. With the Rehabilitation Credit, disbursements went forward despite lack of progress on electricity tariffs, and, during the structural adjustment period, the Bank demonstrated an unusual level of flexibility in inter- preting tranche release conditions. Under the first Structural Adjustment Credit, the Bank re- leased a second tranche despite the Government falling short on its ambitious privatization program. Under the second structural adjustment credit, a second tranche was released even though compliance with some key tranche release conditions was considered less than satisfac- tory (OED, 2002). In the third Structural Adjustment Credit, progress on meeting third tranche release conditions was good, with the exception of the sale of four electricity corporations. A waiver was eventually granted on the basis that the Government had made a good-faith effort to privatize these corporations, but no reasonable bids had been received, and the sales were in- corporated into a floating tranche as part of the fourth adjustment credit. 2.6 As single-tranche operations, PRSCs had no difficulties with tranche release conditions. However, these are programmatic credits, in which triggers for additional lending are laid out in advance. Hence, PRSC I contained an indicative program for PRSC II and III, and, more im- portantly, the specific triggers7 for proceeding with PRSC II. PRSC II, in turn, provided an in- dicative program for III and IV, and so forth. Of 13 prior actions or "core" (that is, legally- binding) conditions listed in PRSC I for PRSC II, 11 were completed without modification, and the other two were partially implemented. Of the 10 triggers listed in PRSC II for PRSC III, one related to controls on illegal logging was reduced in scope due to slower-than-expected progress. Another, related to pension reform, was expanded in scope to encompass the entire pension system rather than just the private system. All seven triggers identified in PRSC III for PRSC IV were met. 2.7 Government opinion on PRSC vs. SAC differences. The better alignment of PRSC actions with Government capacity and intentions reflects a number of factors, including the maturation of the relationship between the Bank and the Government, as well as the presence of an agreed upon overall strategy in the form of the Poverty Reduction Strategy Paper. A higher level of economic prosperity and a general reduction in opposition to structural reforms probably also helped. Government counterparts indicated that, in the structural adjustment period, the Gov- ernment was somewhat naïve in its dealings with the Bank and tended to accept Bank ideas and conditions without debate, resulting in problems down the road. Under the PRSC, however, not only did the Government possess a viable framework in the form of the Poverty Reduction Strategy Paper, but it had also become more confident in its ability to relate conditions to Ar- menian realities and make the appropriate adjustments. 7 Triggers are non-binding conditions (also called "expected/indicative prior actions") that are laid out in advance as conditions for moving forward with the next operation in a programmatic series. Because they are not legally binding, they provide the flexibility to accommodate the unpredictability and uncer- tainty of complex policy reforms. 5 2.8 Regularity of disbursements. There is no evidence that PRSC disbursements were more regular than those under Structural Adjustment Credits (see Figure 1). Government finance of- ficials commented, however, that the PRSCs occurred during a period when the budget crisis had diminished. The Government has been able to manage its fiscal position with short-term borrowings and use of cash reserves and does not depend on the timing of the PRSCs. The ad- justment credits, on the other hand, occurred during a time of extremely tight fiscal resources in which few alternatives were available. The first tranche of the first adjustment credit amounted to about 10 percent of government spending. Hence, delays in tranche releases meant that the Government could not make key disbursements. Figure 1. Armenia Quarterly IDA Disbursements for DPLs 1995-2008 Fiscal Year: Jan-Dec $70 $60 $50 USD Millions $40 Q4 PRSC 3 $30 Q3 PRSC 1 PRSC 2 PRSC 4 $20 Q2 Q1 $10 $0 Source: Business Warehouse, World Bank 3. PRSC Process How effective have PRSCs been as a vehicle to help Armenia operationalize a country-driven poverty reduction strategy? 3.1 Alignment of the PRSC with Armenia's national development strategy and sector strategies. On the whole, Armenia's PRSCs have been strongly aligned with the country' overall development strategy (or poverty reduction strategy). The PRSCs tend to follow the first strategy paper's fo- cus on public sector reforms, including expenditure management and tax policies, as well as reform programs in health, education, and social protection. However, several differences sep- arate the PRSCs from that national development plan. For example, Armenia's first Poverty Re- 6 duction Strategy Paper8 placed heavy emphasis on public sector reforms and human develop- ment and raised issues that did not appear in the PRSCs such as employment, roads, and cul- tural development. The PRSCs place much more emphasis on private sector development and competition. Also, in many cases, the PRSC goes well beyond the treatment in the strategy pa- per and provides a more extensive reform program. For example, the first Poverty Reduction Strategy Paper noted the need to enhance private sector development and improve financial in- termediation, including improvements in the secured transactions network and development of better infrastructure. The PRSC goes beyond that framework and includes specific additional measures to improve civil aviation, enact a telecommunications regulatory framework, create an electricity market and improve the regulatory framework for the banking and insurance sec- tors.9 3.2 The PRSC's Project Appraisal Document (PAD) indicates that the PRSC is fully aligned with the Country Assistance Strategy (CAS) objectives, but in reality this alignment is less than exact. For example, the Country Assistance Strategy objectives are actually more poverty- oriented than the PRSC and focus on: 1) promoting private sector-led growth; 2) making growth more pro-poor; and 3) reducing non-income poverty. The first Poverty Reduction Strategy Pa- per also endorsed natural resource protection in general, and improved forestry management in particular, but did not focus on controls on illegal logging as does the PRSC. In the strategy pa- per, targets for education and health were quite detailed and were more directly reflected in the PRSC, although the latter places relatively greater emphasis on early childhood education. The PRSC IV contains actions for strengthening corporate governance which were not in the original PRSC program and did not appear in the first Poverty Reduction Strategy Paper. Hence, the alignment with the strategy paper is, in many cases, more on the basis of principles rather than on well-defined actions lifted from the strategy and put into the PRSC. 3.3 Ownership. The reform program under the Poverty Reduction Strategy Paper/PRSC program has been, according to most country counterparts, very country-led, rather than Bank/donor-led. The Bank is clearly the biggest and most influential budget support donor, and draft strategy papers are discussed with the Bank; thus, there is a certain amount of mutual determination of the poverty strategy. However, the Government is clearly in charge. No coun- terpart pointed to any actions in the PRSC matrices that were not consistent with the strategy paper or part of the Government's program. Officials in all ministries are aware of the Poverty Reduction Strategy Paper, though awareness of the PRSC program is somewhat lower. In cases 8 In October 2008, the government approved the second Poverty Reduction Strategy Paper (PRSP-2), called the Sustainable Development Program (SDP). It has 3 key goals covering the period 2009-2021 (with specific medium term targets indicated for 2009-2015): reducing poverty, including the elimination of extreme poverty; ensuring human development; and deepening economic growth and accelerating de- velopment of lagging regions. 9 The PRSC's choice of sectors reflects a number of factors, including: the content of policy work in the preceding Structural Adjustment Credit series; the PRSP's focus on poverty reduction and health, educa- tion, and social protection sectors (issues identified via economic and sector work and subsequent policy dialogue); and the need to support objectives of the Country Assistance Strategy. 7 where the International Development Association (IDA)10 has project operations, ministry offi- cials often confuse what was done for the PRSC with what was being done under the project. 3.4 Participation. Although participation by the wider Armenian society in the implementa- tion and monitoring of the Poverty Reduction Strategy Paper was very limited, participation in its preparation was broad and inclusive at both the national and regional levels. The process in- volved consultation with the executive branch of the Government and members of the National Assembly, as well as nongovernmental organizations, academia, the private sector, trade un- ions, local communities, farmers, and members of the Armenian diaspora. It also included pub- lic opinion surveys, workshops, and wide dissemination of information in the local media. The process was led by a steering committee headed by the Minister of Finance and Economy and coordinated by a working group led by the Deputy Minister of Finance. The committee re- ported to the Prime Minister and involved line ministries, committees of the National Assem- bly, political parties, and nongovernmental and donor organizations. This consultative process was revived for the production of a draft for a second Poverty Reduction Strategy Paper. 3.5 The Bank's role in supporting country-led PRSC formulation. Bank influence has been par- ticularly useful in operationalizing the strategy in cases where overall goals have been agreed upon but modalities need to be worked out. Bank staff (particularly the resident mission staff) are seen by many counterparts as valuable collaborators in the process of strategy formulation and implementation. Government counterparts value especially their ability to access interna- tional experience through Bank technical assistance in various forms, including Bank staff and consultants financed via trust funds or projects. Several noted that implementation of a given reform would have been virtually impossible without World Bank support.11 3.6 Economic and sector work (ESW)12 is also frequently cited as having played an impor- tant role. Several counterparts pointed to the Bank's book "Caucasian Tiger" as having been in- fluential, particularly as it relates to the need for a strong growth orientation with an emphasis on the private sector environment and the need to improve transparency and efficiency in gov- ernment services. Other reports mentioned include the 2003 Public Expenditure Review, the poverty assessments, and a study on social benefits. Economic and sector work on transporta- tion revealed the huge premiums that Armenia had been paying for air services due to the mo- nopoly position of the local carrier, and the inefficiencies and high costs of the railway system. The Bank's 2005 Report on Observance and Standards of Codes on Corporate Governance di- rectly led to the inclusion of steps on improving corporate governance in PRSC III. 10 The International Development Association (IDA) is the part of the World Bank that helps the world's poorest countries. Established in 1960, IDA aims to reduce poverty by providing grants and interest-free credits for programs that boost economic growth, reduce inequalities and improve people's living condi- tions. IDA complements the World Bank's other lending arm­the International Bank for Reconstruction and Development (IBRD)­which serves middle-income countries with capital investment and advisory services (www.worldbank.org). 11 For instance, the Bank introduced the idea of self-assessment in the tax system and the use of the tax system to collect pension contributions. 12 Economic and sector work represents the core of the Bank's knowledge base and is the foundation on which policy dialogue with external clients is carried out. As the largest category of analytic and advi- sory activities (AAA) carried out by the Bank, ESW activities are: (i) meant to influence the policies and/or programs of an external client; (ii) involve original analytic effort and (iii) "owned" by a specific Bank unit, representing the views of the Bank (not those of individual staff or consultants). 8 3.7 Effect on governmental policy dialogue. The real value of the PRSCs to the Government is no longer the amount of budget support it provides, but the access it gives to expertise and the relationship it creates between the Ministry of Finance and the operational ministries. The PRSC has served as a useful vehicle by which operational ministries can bring issues to the attention of the Prime Minister/Ministry of Finance and ensure a place for their individual programs in the Medium-Term Economic Framework. Likewise, the Ministry of Finance sees the PRSC as a way of pushing sectoral ministries to complete their reforms. Although there is less evidence that the PRSC significantly improved dialogue among ministries on Armenia's overall develop- ment strategy, it does seem to have improved coordination between ministries on specific, key reforms. 3.8 Results focus and monitoring and evaluation system. Commendably, each PRSC is attached to a table of expected results indicators, which are linked to actions, and these indicators at- tempt to use (and, thereby, strengthen) the government's own PRSP monitoring system. How- ever, there is little monitoring of results on the basis of these indicators. Rather, each new PRSC produces a new table with slightly different indicators, reflecting the focus of the new PRSC (except PRSC III and IV, which have almost identical tables). While each new PRSC reviews past progress on policy actions in a general sense, there is no review of the specific policy indi- cators and the progress that has been made in meeting these goals. For instance, the target for tax revenue to GDP in PRSC I was set to rise from 14.1 percent in 2003, to 16.2 percent by 2006, and to 17 percent by 2007. However, when actual revenue for 2006 fell short at only 14.4 per- cent, the Project Appraisal Document refers to this as a "modest rise" and makes no mention of the shortfall in the target. Instead, it merely sets a new target of 15.2 percent for 2007. This pat- tern is repeated in other sectors as well. In addition, there is no general discussion of previous targets or the lack of government tracking of PRSC indicators. Several important indicators, in- cluding all those on public procurement included in PRSC I, were dropped from PRSC III and IV. 3.9 There were similar deficiencies in the monitoring system of the first Poverty Reduction Strategy Paper. While the Government, in theory, established a poverty reduction strategy mon- itoring and evaluation framework in 2004, it has never functioned effectively. The framework identified 177 indicators, of which 36 were targets and 141 were intermediate indicators. How- ever, no information was available on 44 of these indicators, and many of the others were in- complete or of substandard quality. Ministries lacked the incentives or capacity to improve and report indicators. The Poverty Reduction Strategy Paper Progress Report in 2006 listed only 20 indicators, on which it found generally satisfactory progress. These indicators, which were heavily weighted with expenditure data, included no measures on social conditions such as health, education, or unemployment. 3.10 The second Poverty Reduction Strategy Paper (called the Sustainable Development Pro- gram) again commits to improving the monitoring system and measures progress over Poverty Reduction Strategy Paper-1 on the basis of a somewhat different list of about 22 indicators, in- cluding GDP, exchange rate, two social indicators, and indicators on social expenditures. How- ever, there is no evidence that the Government uses these indicators in a meaningful way to as- sess progress or modify policies. Progress in reducing poverty is measured solely in terms of income poverty, with little reference to other measures such as social indicators. 9 3.11 The Bank has made available to the Government trust fund resources, which will streng- then the capacity of the National Statistical Services and units within ministries. It has also supported the clarification of responsibilities through a series of decrees. The Government's re- sults focus will also be improved through the gradual adoption of program budgeting. Program budgets were adopted on a trial basis in four ministries in 2006 (Health, Education, Social Af- fairs/Labor, and Culture) and extended to an additional ten agencies in 2008. These fourteen agencies and ministries account for some 70 percent of all expenditures, excluding defense, po- lice, and national security. Program budgets seem to help the Government make a clearer case to the National Assembly for funds through better explanation of program content and objec- tives. The National Assembly also finds the program approach valuable and has asked for it from agencies not yet participating in the program. Thus, the 2010 budget will contain an annex with all ministry/agency budgets presented in program classification in addition to the tradi- tional economic/functional/administrative and mixed classifications. Compared with 2009, this represents important progress in the sense that ministries/agencies such as defense, police, and national security will be joining the process for the first time. Although the program budgeting format will still not be the primary format for the Parliamentary approval of the budget, it will again generate discussions and provide information for Parliament to better understand the purpose of intended budget allocations. Meanwhile, the Ministry of Finance is preparing amendments to the Law on Budget System to that will make the programmatic format the pri- mary basis for budget approval and require the government to submit the annual budget to the parliament in this format. How effectively have PRSCs contributed to donor harmonization around a country- owned, medium-term poverty reduction strategy? 3.12 Relationship to other general budget support programs. PRSC disbursements are a relatively small share of total Government expenditures. In 2007, total expenditures were 560 billion dram, and PRSC II disbursements comprised 10 billion dram, or about 1.8 percent of this (see Table 2). Outside of IDA, the Netherlands is the only donor to provide general budget support to Armenia, and this is a very small amount--EU 5 million per year in 2006 and 2007 (see Ap- pendix Table 2).13 The Dutch condition their aid on agreement between the Bank and the Gov- ernment on a PRSC program (parallel financing) and do not participate in setting a common policy framework or attempt to influence the PRSC's policy framework. 13 The EC provides budget support, though not general support, as part of a food security package with funds provided to import agricultural goods, which are assigned to specific ministries and programs. 10 Table 2. Financial Role of PRSCs in Armenia 2004 2005 2006 2007 Expected amount, USD million 20.0 0.0 20.0 28.0 Actual disbursement, , USD million 21.2 0.0 20.3 28.6 PRSC as a share of total foreign credit 24.1 0.0 24.4 24.1 disbursements, % PRSC disbursement as a share of GBS, % 100.0 0.0 64.6 91.6 PRSC disbursement as a share of Govt. 3.1 0.0 1.9 1.8 expenditures, % Source: PRSC Project Documents and Ministry of Finance and Economy Note: Data for 2007 for Government expenditures and total disbursements are budget programs. 3.13 Collaboration on underlying analytical and diagnostic work. Because other donors provide little or no general budget support, there is no need for coordination on a common framework, and the donors and Government accept the Bank's lead role in policy dialogue. Coordination with the IMF on macroeconomic issues has been strong, although the Fund tends to take the lead in this area. In general, the Bank has taken a lead role in the areas of governance and public sector management, with two key bilaterals (U.K., U.S.) and the EU playing an important role. The Bank shares responsibility on financial sector issues with the IMF and DfID, and USAID has also had a prominent role. While the Bank is also the lead partner in social sector reforms, in certain aspects the United Nations agencies have played a role, as has DfID. In the environment and rural economy, the Bank has worked with such agencies as the Swedish International De- velopment Agency (SIDA), USAID, the International Fund for Agricultural Development (IFAD), the EU, and various United Nations agencies. Some analytical work has been done jointly with external partners, as well as the Government. A programmatic Public Expenditure Review (PER) in 2005, as well as a Country Financial Accountability Assessment (CFAA) in 2004, were prepared by the Bank with Government participation. The IMF, DfID and USAID al- so participated in various parts of this work. External partners, including the U.S. Treasury, the EU, DfID, and the IMF, have jointly supported the Public Expenditure and Financial Accounta- bility (PEFA) assessment (World Bank, 2003c). 3.14 In 2005, the Government agreed with the IMF on a USD 34 million Poverty Reduction and Growth Facility, in effect for three years. Performance under the Poverty Reduction and Growth Facility has been strong, with almost all performance criteria met on time, with a few minor exceptions. A Neighborhood Agreement with the EU was signed in 2006 to provide EU 100 million in budget support over five years; as of January 2010, USD 25 million has been dis- bursed under the 1st Vocational Education and Training program and another USD 8 million under the Justice project. This aid, conditioned on actions in the education sector, is not coordi- nated with the PRSC. How well have PRSCs been aligned with country policy formulation, budgeting, and planning processes, and how effectively have they contributed to predictable resource flows? 3.15 Alignment with domestic processes. The PRSC seems to have been well aligned with the country's Poverty Reductions Strategy Paper /national development strategy, and Government 11 officials expressed satisfaction with the process. PRSC disbursements have occurred on a fairly regular basis, normally in the first or second quarter of the Armenian fiscal year, which follows the calendar year (see Appendix Table 4, Appendix Table 5, and Appendix Table 6). PRSC I was disbursed in December 2004, in time for incorporation into the 2005 budget. PRSCs II and III were disbursed in April of 2006 and 2007, respectively, having been approved by the Board some months earlier. PRSC IV was approved by the Bank's Board in November 2007, and was disbursed in March 2008 in time for the 2008 budget. The timing of PRSC IV was advanced somewhat to ensure that approval occurred before the Presidential elections in February 2008. However, the triggers for PRSC IV had been fully met. The tracking of the Poverty Reduction Strategy Paper implementation is done by the Government, which is separate from the tracking of PRSC implementation, done largely by the Bank. All PRSCs fell under the 2003 poverty re- duction strategy, and were generally well aligned with it in terms of overall objectives. 3.16 Predictability of disbursements. Disbursements seem to be broadly in line with government expectations, although there is no data on the actual month of disbursement by the Government to line agencies. In general, PRSC formulation has occurred during the year preceding dis- bursement and at the time of discussion of the new budget, so that any budget implications re- lating to policy measures can be taken into account. 3.17 Government officials indicated that, since the amounts involved are relatively small, de- lays have not been an issue, as the Government has other resources and can cover temporary shortfalls, knowing the disbursements will eventually come in the year expected. This is in con- trast to the situation during the Structural Adjustment Credit period when disbursements were bigger and more apt to be delayed over tranche-release issues. 14 At that time, the Government lacked other viable financial management options, such as short-term borrowing or the use of cash reserves. The PRSC has not leveraged other general budget support flows, and PRSC dis- bursements appear as predictable as the one other general budget support provider (Nether- lands). The PRSC does not appear to have increased the predictability of total official develop- ment assistance flows. The general budget support share in total assistance has not increased over the PRSC period (see Table 3). Board approval by the Bank has not been a factor affecting predictability. The overall volatility of assistance flows has not improved during the PRSC pe- riod (see Table 4). Official development assistance volatility was high during the 1990s and rela- tively low during the 1999-2003 structural adjustment period. It appears to have risen slightly during the PRSC period (2004-2006). Table 3. PRSCs, ODA, and GBS in Armenia, 2001-2007 2001 2002 2003 2004 2005 2006 2007 General Budget Support/ODA, % 28.2 1.5 19.2 8.3 3.3 18.3 NA PRSC/General Budget Support, % 0.0 0.0 0.0 100.0 0.0 64.6 91.6 General Budget Support/Government Expenditures, % 12.9 1.0 8.7 3.1 0.7 3.0 2.0 Source: OECD online database; Government Revenue and Expenditure data for 2007 are program data from Budget Law 14 For example, the first tranche of first Structural Adjustment Credit of USD 30 million was equal to al- most 10 percent of budget expenditures in 1996. 12 Table 4. Evolution of Aid Volatility 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 ODA, net flows 2720 22460 10971 19127 21759 29216 16613 19420 20933 21589 19839 (USD million) 3 yr. average 12050 17519 17286 23367 22529 21750 18989 20647 20787 23591 standard dev 9914 5911 5625 5233 6337 6617 2192 1112 884 1237 coeff. of variation 0.823 0.337 0.325 0.224 0.281 0.304 0.115 0.054 0.043 0.052 2002 2003 2004 2005 2006 ODA, net flows 29346 24875 25379 19327 18960 3 yr average 24687 26533 23194 21222 standard dev 4756 2449 3358 3605 coeff. of variation 0.193 0.092 0.145 0.170 Sources: ODA 1991-2005 from WDI data bank. Flows are net of repayments, ODA 2006 estimated from Appendix Table 1 Note: Years with PRSC disbursements marked in bold 3.18 Exit strategy. There was no planned exit strategy for the PRSC in Armenia. At one point, the Government considered dropping the program on the grounds that it no longer needed budget support and would then be able to direct IDA resources to project lending. However, the Government concluded that the program's ability to interact with operational ministries and help implement reforms was of value, so the size of the last PRSC was slightly reduced and the savings redirected to project lending. PRSC IV was the last PRSC in the current series, fulfilling actions laid out in PRSC I. A new Country Assistance Strategy (now called the Country Partner- ship Strategy) was endorsed by the Bank's Board in June 2009 and includes a new development policy loan series to be disbursed in annual single tranches during FY 2010-2012. 4. PRSC Results How effectively have the PRSCs helped Armenia strengthen its public financial management systems? 4.1 Diagnostic work. The need to strengthen public financial management had been identi- fied before the PRSC program and was part of the Structural Adjustment Credit series. Under the fourth and fifth Structural Adjustment Credits, steps were taken to introduce the Medium- Term Economic Framework, reform the budgetary process, reduce unclassified expenditures, and curtail arrears. The Government had made progress in budget reform prior to 2004 with the introduction of a new Budget Law and the first Public Investment Program (1994), as well as re- visions to budget presentation and execution. Deficiencies nevertheless existed, as noted by the Bank's Institutional and Governance Review (IGR) in 2000. 13 4.2 The first full Public Expenditure Review, completed in 2003, expanded on the Institu- tional and Governance Review's findings related to public financial management and served as a backdrop to reforms introduced under the PRSC program. Further diagnostic work was de- veloped during the PRSC period, including a Country Procurement Assessment Report (2004), special Public Expenditure Reviews on health (2006), infrastructure (2007), and education (2008), and a budget reform study (2007). The main recommendations of the 2003 Public Ex- penditure Review were the following: Advance fiscal adjustment and budget consolidation Improve tax collection Improve budget transparency and accountability Advance expenditure rationalization and reallocation Strengthen strategic approach to budgeting Streamline budget execution Advance sectoral reforms to ensure more effective use of funds at the sectoral level. 4.3 Almost all of these items were subsequently picked up, at least to some degree, by the Structural Adjustment Credits or the PRSC program. 4.4 Public financial management content and design. In general, reforms in public financial management were directed at achieving greater budget transparency and accountability, im- proving budget planning and expenditures and strengthening the public procurement system. The specific conditions in the PRSCs included four major items: Approval of a new budget classification system according to government finance statis- tics (GFS) 2001 principles, introduced in 2008, and development of an integrated chart of accounts, accounting, and reporting standards; Development of a strategy for introduction of a new system of staff performance ap- praisal; Adoption and implementation of a strategy for developing internal and external audit functions, including establishment of a new Department for Internal Audit Assessment and Financial Supervision, as well as implementation of International Public Sector Ac- counting Standards (IPSAS) cash basis standards for accounting; Implementation of financial reporting by state non-commercial organizations (SNOCs), as well as sanctions for non-compliance; and Preparation of an action plan and cost estimates for introduction of an electronic pro- curement system by 2007, as well as publication of technical specifications for goods and services of common use. 4.5 With the assistance of DfID and other donors, the Government also began to introduce program budgeting in pilot ministries. The PRSC included steps to collect performance indica- tors and program descriptions in the pilot social sectors. Although the Bank has supported the introduction of program budgeting, this was not an action in the PRSC policy matrix, and the lead agency on this reform was DfID. Program budgets were prepared for four state budget agencies in 2007, which was broadened to 14 in 2008; the aim was to have the entire budget on a program basis by 2010. The experience of the four pilot agencies varies with respect to their ability to identify and use performance indicators and capacity to integrate the performance ap- 14 proach into agency management (Taliercio, 2007). One problem is that aid for implementation of program budgeting from DfID is being eliminated. 4.6 In general, reforms introduced covered the whole government rather than individual agencies. In program budgeting, however, the reforms have been introduced in a limited num- ber of agencies as a trial before extending them to the rest of government. The PRSC did not fo- cus on capacity building in the sector, which was instead handled by on-going projects from Df- ID, USAID and the EU, as well as the Bank's own Public Sector Modernization Project. The reforms did not focus on revenue mobilization, although the PRSC did address reforms in tax and customs. OUTCOMES 4.7 Overall achievements. The introduction of better budget classification system and the move toward program budgeting have led to greater budget transparency and accountability. Counterparts report that Parliament is particularly happy with the program budgeting ap- proach, because it gives them a greater understanding of government programs. However, defi- ciencies still exist in terms of internal audits and reporting by state non-commercial organiza- tions. The original PRSC targets in these areas have not been met. In the area of budget planning, the introduction of the Medium-Term Economic Framework, linked to the poverty reduction strategy, has improved budget planning and expenditure control. Budget execution follows budget allocations. Most counterparts confirmed the link between the poverty reduction strategy, the medium term framework, and the budget, and admitted that this has been a useful innovation. As for the public procurement system, while some limited progress has been made, the system is still weak. Government officials still find ways to circumvent the system. The in- troduction of e-procurement has been delayed, and the target set out in PRSC I has not been met. 4.8 Budget classification system. In 2008, the Government introduced its new budget classifi- cation system with the new budget which will also assist in the introduction of program bud- geting. It is also making progress introducing the International Public Sector Accounting Stan- dards (IPSAS) in the public sector. A new chart of accounts based on Government Finance Statistics 2001 standards has also been introduced. Steps were also taken to move from a cash basis accounting system to an accrual basis. Budget function has also been improved by the in- corporation in 2008 of the State Social Insurance Fund, a major extra-budgetary fund previously not in the budget. 4.9 State non-commercial organizations. Decentralization reforms in the social sector resulted in the formation of a large number of autonomous institutions (state non-commercial organiza- tions) in health, education, and culture sectors. The result was a loss of control of the Treasury, both in terms of control on expenditures and disbursements as well as in financial reporting and auditing. The PRSC supported reforms designed to reduce fiscal risk by strengthening financial management and reporting. While progress has been made in the introduction of financial and accounting standards, reporting by these organizations varies greatly in quality, and the reform is still in the process of being implemented. The goal outlined in PRSC I was to achieve com- pliance of 80 percent of state non-commercial organizations, accounting for not less than 75 per- cent of aggregate income of these organizations by 2007. This indicator, which has not been 15 carefully tracked, does not appear to have been met.15 Government counterparts appreciated the assistance of the Bank in the formulation of the law regulating state non-commercial organiza- tions, which also put a limit on the establishment of new organizations and helped define con- ditions for reorganizing state bodies into non-commercial entities. In their opinion, this reform could not have been done successfully without World Bank support. 4.10 Auditing. As part of the PRSC, the Government agreed to take steps to strengthen insti- tutional arrangements for internal audit, implement an external audit action plan, and move toward IPSAS cash basis standards for accounting. The overall goal was to have internal and ex- ternal audit systems fully functional in all ministries by 2007 (World Bank, 2004). But counter- parts report that internal audits are still weak and of variable quality. While substantial efforts have been made for the last several years by the Government and donor community to streng- then internal auditing, real progress has been slight. There are unresolved jurisdictional prob- lems between the Ministry of Finance and Economy, Chamber of Control, and control agencies within each ministry, and this adversely impacts the effectiveness of the process. Currently, in- ternal auditors lack independence and the ability to perform audit services within agencies. 4.11 The Ministry of Finance and Economy has made efforts to strengthen internal audit functions in public agencies, including capacity-building exercises. Even so, in many agencies auditors become part of financial department staff. Lack of staff who are trained in audit func- tions continues to hamper the system. New legislation has been drafted to speed up reforms in the internal auditing system, reduce overlap with different control agencies, and increase inter- nal auditor effectiveness. More resources will also be put into audit staff training. Clearly, the goal set in PRSC I has not yet been met. 4.12 Likewise, the lack of a functioning external audit system remains a serious deficiency. The Chamber of Control still lacks the ability to function as an independent external auditor; it cannot yet produce audits of government accounts, thus leaving the credibility of government accounts at the end of the year in question. Thus, the PRSC I matrix goal for 2007 of a "function- al external audit system" remains unmet. 4.13 Procurement. A Country Procurement Assessment Report (CPAR) in 2004 led to a re- vised procurement law in 2005. The report's action plan is being implemented, and the Gov- ernment reports that most of the action plan's recommendations have been accomplished. A new Country Procurement Assessment Report is planned for 2009 to review the situation. A strategy for introduction of e-procurement has been adopted by the Government, and a World Bank Institutional Development Fund (IDF) grant and the Bank's Public Sector Modernization Project are providing technical assistance; an e-procurement system would significantly reduce corruption and increase transparency. The first module of this--an e-tendering system--is not likely to be introduced completely before 2010, significantly behind the original PRSC I target date of 2007. 4.14 Single source procurement remains a challenge. According to the Country Procurement Assessment Report, single source procurement was as high as 80 percent in 2003. Targets in PRSC I and II called for a reduction in single-source procurement to 50 percent in 2005 and 20 percent in 2006. Implementation of several recommendations contained in the 2004 Country 15 The goal is in PRSC policy matrix, but the indicator was not included in the monitoring table. 16 Procurement Assessment Report action plan has led to significant improvements; although sin- gle source procurement still appears to be higher in Armenia than generally accepted interna- tional norms (10 percent in quantity, 5 percent in value).16 PRSC I also called for a target of 75 percent for procurement documents adhering to agreed standards. This target cannot be met, as the standards were never adopted (they were expected in late 2008), and it is still not clear (as of early 2010) when they will be. Global Integrity gives Armenia a score of 63 out of 100 on the ef- fectiveness of the public procurement system, and 42 in terms of citizen access. Specifically, the deficiencies include: Lack of regulation on the handling of complaints, along with a complaint review system that makes no use of the nongovernment sector or the media in its process; A high level of corruption risk in the non-transparent selection of tender participants; Lack of transparent selection criteria, which in turn are not reflected in the rationale for tender awards; and Evidence that agencies reclassify expenditures to avoid procurement regulations. 4.15 As for pro-poor tracking, there is currently no effort made to track expenditures in terms of what percentage reaches the poor. This was never covered by the PRSC program. 4.16 An assessment of public expenditure management was undertaken by the region in 2006, and the standard Public Expenditure and Financial Accountability (PEFA) scores were calculated. Unfortunately, this is the only available source of PEFA scoring for Armenia (see Appendix Table 7). Armenia receives high marks in some areas, such as the comparison of budget and revenue outcomes relative to expected values, classification of the budget, availabil- ity of information, and quality of the budget process. Shortcomings identified include deficien- cies in the collection of tax payments, lack of transparency in taxpayer obligations and liabilities, and internal audits. Other deficiencies are found in the lack of payroll controls, and in the scope and nature of external audits. It is too soon to judge whether these Public Expenditure and Fi- nancial Accountability indicators have improved in recent years as a result of the PRSC, or represent an improvement over the status of 2004. 4.17 Alignment of public expenditures with policy priorities. In general, most counterparts agree that the Poverty Reduction Strategy Paper, the Medium-Term Economic Framework, and the budget are well aligned with each other. This alignment seems to have happened because the Government takes the poverty reduction strategy process seriously, not because of reforms un- der the PRSC. Alignment of overall programs with objectives started with the introduction of the economic framework prior to the strategy paper in 2005. Comparing expenditure allocations before the poverty reduction strategy period (that is, pre-2003) with the Poverty Reduction Strategy Paper period, we can see a discernable shift toward education and health, as per pover- ty reduction strategy priorities. This shift is not so much a reflection of public expenditure man- agement reforms as it is of reforms supported by the PRSC in these sectors, as discussed below regarding the health and education sectors. 16 Single source procurement continues under the pretext of lack of competition and from those holding (registry system) certificates issued by the Anti-Monopoly Commission. No clear distinction is made be- tween "emergency" and "urgent" needs, resulting in the use of single source procurement. Late dis- bursement of budgetary funds is also problematic, as it does not allow use of competitive procurement methods since procuring entities must use their funds before the end of the fiscal year, or lose them. 17 4.18 Other shifts include increased allocations to defense, as compared to 2002, and lower al- locations for energy and agriculture. While reforms in program budgeting have helped intro- duce a results orientation to the budget process and more of a performance-based management culture in the line ministries, these reforms have only just begun to be implemented. 4.19 Expenditure predictability. Expenditure predictability appears to be high in Armenia (see Public Expenditure and Financial Accountability indicators). Actual expenditures closely paral- lel the allocations foreseen in the PRSP, but there are some deviations. The level of deviations, usually in terms of under-spending, seems to have been reduced in recent years, but the trend is not entirely clear. In general, capital spending shows consistent under-spending, and this has not changed during the PRSC period. In terms of sectors, allocations for health, education, and social security/social insurance seem to be slightly higher now than in the pre-PRSC period. How relevant and effective a vehicle have PRSCs been in helping Armenia set conditions for poverty reducing growth? 4.20 Extent of growth focus. Setting conditions for poverty-reducing growth was never the di- rect objective of the PRSC program in Armenia. Instead, the strategy was centered on four main themes, which appeared consistently in PRSCs I-IV and which supported, directly or indirectly, the objectives of the Poverty Reduction Strategy Paper. These four themes were: Consolidating macroeconomic discipline and strengthening governance Sharpening competition and entrenching property rights Mitigating social and environmental risks Modernizing the rural economy 4.21 The first two objectives were designed to strengthen the economy, and, inter alia, the growth process, by improving the functions of the public sector and the environment for private investment. However, there was no specific pro-poor aspect of the growth objective; in fact, the growth objective was subsumed under the broader objective of macroeconomic stability and linked to private sector development. In doing so, the region was responding to the findings of the Project Performance Assessment Report (OED, 2002), which had suggested that the Struc- tural Adjustment Credit program had placed insufficient emphasis on critical aspects of private sector development, including building government capacity to support a market economy. The Bank's focus clearly shifted toward governance and private sector development during this pe- riod, which were the main areas examined in its flagship volume "Caucasian Tiger," a publica- tion that did not focus on "pro-poor" growth (Mitra and others, 2007). Poverty concerns in the PRSC were not entirely ignored but addressed more explicitly in the last two themes, which fo- cused on improving health, education, social protection services, and agricultural productivi- ty.17 4.22 Macroeconomic management. A general objective of all PRSCs was maintaining macroeco- nomic stability as a pre-condition for growth. Objectives included limiting fiscal deficits to 2-3 17 It should be noted that the CAS objectives for the period were: 1) Promote Private Sector Led Growth; 2) Make Growth More Pro-Poor; and 3) Reduce Non-Income Poverty. The PRSCs are seen as supporting CAS objectives. 18 percent of GDP, keeping inflation under 3 percent and maintaining a market-oriented exchange rate. A projected growth rate of 6 percent was foreseen in PRSC I. These targets were consistent with the IMF's program and were generally met. 4.23 In fact, in some cases, actual accomplishments far exceeded expectations. GDP growth averaged about 13 percent in 2004-2006, and inflation averaged 2.3 percent, while the fiscal defi- cit averaged 1.7 percent (see Table 5). The economy benefitted from the continuation of large re- turn flows from the diaspora, much of which was invested in real estate. In addition, there was a positive impact from trade liberalization and improvements in infrastructure services on the economy. Fiscal prudence and low resulting inflation also helped maintain a growth rate re- markably high by world standards. Table 5. Macro Targets and Accomplishments Measure 2003 2004 2005 2006 GDP growth Target 13.9 7.0 6.0 6.0 Actual 14.0 10.5 13.9 13.2 Inflation (eop) Target 8.6 3.0 3.0 3.0 Actual 8.6 2.0 -.2 5.2 Fiscal Deficit Target 1.5 2.7 3.0 3.0 Actual 1.3 1.7 1.9 1.5 Source: Targets from PAD, PRSC I. Actuals from IMF 4.24 Relevance of growth-enhancing reforms in light of poverty reduction and capacity constraints. The impact on poverty was equally impressive during the PRSC period. Per capita GDP in- creased by 30 percent, and the poverty rate fell from 35 to 27 percent. By the same token, the structural adjustment period had seen equally strong performance in terms of growth and po- verty reduction (see Table 6). On income distribution, the results were nothing short of remark- able. Gini coefficients are normally fairly stable and not prone to major changes in the medium term. In the case of Armenia, however, the Gini coefficient dropped (that is, improved) by 33 percent between 1999 and 2005. Thus, growth was clearly "pro-poor," to the extent that poorer groups received more than their share of the benefits. Some part of the improvement in income distribution can be attributed to modifications of the family benefit18, which helped raise in- comes among the poor, both reducing poverty and improving income distribution. 18 The Poverty Family Benefit System, introduced in 1999, consists of cash benefits paid to eligible fami- lies, based on a proxy-means test and family characteristics. 19 Table 6. Outcomes on Growth and Poverty, 2004-2007 Indicator 1999 2004 % change 2006 % change `99-`04 `04-`06 GDP growth rate 3.3 10.5 10.6** 13.2 13.5** GDP per capita (USD as of 2000) 581.0 987.0 70.0 1283.0 30 Poverty Rate (%) 56.1 34.6 -38.0 26.5* -29 Extreme Poverty (%) 21.0 6.4 -70.0 4.1* -36 Infant Mortality (%) 15.4 11.6 -25.0 13.8 19 (2001) (2001-04) % of Rural population with access to 64.7 74.0 14.3 78.0 5 safe water Unemployment Rate (%) 35 31.0 -11.0 29.0 -6 (2002) (2002-04) Gini Coefficient 59.3 39.7 -33 (2005) (1999-2005) Sources: World Bank, Government of Armenia 4.25 Nonetheless, unemployment remains a serious problem. Despite rapid economic growth, 29 percent of the labor force identified itself as unemployed, only a slight drop from 35 percent in 2002. Many of these "unemployed" however, may be informal sector workers who desire jobs in the formal economy and do not consider themselves gainfully employed in the in- formal sector. High unemployment and rapid poverty reduction seems somewhat contradicto- ry. However, many Armenian workers have migrated abroad and send substantial remittances home. In 2007, inward remittances amounted to USD 1.3 billion, and it is believed that this un- derstates the true amount. These inflows have doubled since 2003 (Ratha and others, 2007). This amounts to about USD 600 per capita, compared to per capita GNI of about USD 2,200.19 4.26 The PRSC identifies good governance and the reduction of corruption as essential to achieving pro-poor growth, a major Country Assistance Strategy objective. Unofficial payments to government officials -- bribes or facilitation payments for services -- occur relatively fre- quently in Armenia. In 2005, 18 percent of firms reported such payments in relation to taxes and tax collection, compared to only 5 percent in 2002 (World Bank, 2007b). Other areas in which such payments occurred frequently were contracts, customs/imports, and the courts. A 2005 Business Environment and Enterprise Performance Survey found that 30 percent of Armenian firms saw corruption as a significant constraint to doing business in the country. 4.27 The PRSCs also placed heavy emphasis on reform of the tax and customs systems, with the objective of improving governance via reduced discretion and increased transparency (which is also a way of raising overall tax collections rates) Reforms in tax administration have included the establishment of a large taxpayers unit dedicated to fighting cases of tax evasion, introduction of tax identification numbers, development of tax manuals, and development of an informational Web site. Reforms have provided the option of filing tax returns by mail or sub- mitting returns electronically. The audit and taxpayer service functions of the State Tax Service have been augmented, and tax arrears have been reduced. 19 Estimated for 2007, current prices based on Bank Atlas method GNI per capita of 1930 in 2006. 20 4.28 Steps were also intended to have been taken to reduce outstanding VAT refund claims. However, exporters continue to report substantial delays in getting VAT refunds processed.20 Exporters report this as one of their biggest problems. This impacts small exporters, in particu- lar, who lack the time and money to pursue cases in court. Delays seem to stem from overzeal- ous officials who try to meet revenue targets by delaying VAT refunds, even when entirely legi- timate.21 The 2008 government budget appropriated USD 20 million to cover these refunds, which may help reduce the problem. 4.29 In customs, clearance procedures had been archaic and inefficient, and a high degree of discretion is reported. Reforms supported by the PRSC included adoption of a self-declaration system at all customs houses and the extension of Direct Trade Input (DTI) facilities to impor- ters and custom brokers. A program was introduced to reduce the role of reference prices, cus- toms control was made more selective based on risk management, and Post Clearance Audit capacities were enhanced. However, private sector counterparts report that arbitrary valuation of reference prices by customs officials is still common. 4.30 Other pathways to growth identified in the PRSC were improving property rights, im- proving the functioning of credit and insurance markets, liberalizing key public services, intro- ducing more competition and improving the public regulation of infrastructure, including rail- ways, aviation, telecommunications, and public utilities. Relevant key reforms included in the PRSCs were the following: Strengthening of the utilities regulatory framework for state-owned utilities, including improved governance, creation of an electricity market, and development of rules for service provision and procedures for monitoring quality in energy and drinking water; Adoption of a new telecommunications law to allow for the progressive licensing of ad- ditional telecommunications service providers, along with a revised sector regulatory policy; Revision of the corporate registration law that provides for public access to company records; revision of a securities law to require public disclosure of beneficial owners of traded companies; Amendment of air service agreements to allow for increased competition in air services; Adoption of a plan for the restructuring of railways, including a vertically integrated concession for railways services and improvements in regulation that would allow more competition; Reforms to improve secured transactions, including strengthened creditor rights, im- proved civil procedures for debt recovery, and simplification of procedures for certifica- tion for property sales and registration of secured credit; Approved amendments to the law on banks and banking, including regulations to im- prove corporate governance for banks; and Adoption of prudential regulations and reporting guidelines for insurance companies. 4.31 A final PRSC supported area, rural development, was more clearly aligned with the idea of "pro-poor" growth, since the objective was to raise incomes of some of the poorer groups in 20 Only one exporter this year received a refund--about USD 1 million on amounts due from 2002, and only after a lengthy process of litigation. 21 It also should be noted that fraudulent applications for VAT refunds are also a problem. 21 Armenian society. Here the main focus was to increase productivity and farm incomes, develop the agro-processing industry, and encourage development of commercial activities in rural areas. Major actions supported by the PRSC included: Improvement of the framework for agricultural research, extension, and education; Putting in place a new Seeds Law covering seed certification procedures; Harmonization of the development of food quality standards; Development and launching of a plan for reforming rural infrastructure (roads, water, energy, and infrastructure); and Implementation of an effective system for the control of illegal logging, including estab- lishment of an autonomous State Forest Monitoring System. 4.32 Of these actions, the last is probably the most important, as it promises to curtail the abuses of large-scale illegal logging and help develop community forest resources to benefit the poor. The Bank provided technical support for establishing the forest monitoring system, and the Bank's influence was key in preventing the Government -- driven by large logging inter- ests -- from forcing its closure. The PRSC had a major influence in this debate and helped bring the issue of illegal logging to the attention of the Ministry of Finance. Reforms in other areas were useful and reinforced actions supported by other projects, such as the Rural Enterprise Project, the Agriculture Reform Support Project, and the Irrigation Development Project. 4.33 In the area of tax administration, despite certain improvements, there has been no dis- cernable increase in the overall ratio of taxes to GDP (see Table 7), which remains at about 14.5 percent of GDP. This is low in comparison to similar countries and falls short of expected progress. The PRSC I target was to reach 17 percent by 2007, but current estimates for the year place it at no more than 15-16 percent. Fiscal management has been good, with the overall defi- cit remaining less than 2 percent of GDP. There has been no accumulation of payment arrears and no arrears outstanding. Table 7. Selected Public Sector Indicators (% of GDP) 2002 2004 2005 2006 2007 Total Revenue, excl. grants 15.6 15.1 16.2 16.0 17.9 Tax revenue 14.6 14.0 14.3 14.5 16.1 Overall budget balance -2.6 -1.7 -1.9 -1.5 -1.5 Corruption Perception Index 3.0 (2003) 3.1 2.9 2.9 3.0 (Transparency International) Sources: Government of Armenia, Transparency International. Poverty numbers from PRSC IV, PAD ,p.11; GNP growth and per capita GDP from WDI; infant mortality from PRSP-1 for 1999, and AM for Health Modernization Project 10/07 for 2004 and 2006; Gini coefficients from PRSP-1 and draft PRSP-2; unemployment from PRSP-2 4.34 On corruption, Transparency International gave Armenia a score of 3.0 out of 10 (10 is the best score) on its Corruption Perception Index. Armenia ranks 99th out of 179 countries, suggesting a high level of perceived corruption among business and other leaders. There has been no improvement in the overall score since 2003, which implies little impact from Govern- ment and PRSC reform efforts. (see Table 7) 22 4.35 In the private sector, PRSC reforms produced a number of significant improvements in the environment for investment and growth. Some of the more important include: A 27 percent increase in air traffic between the first half of 2006 and the first half of 2007, and progress toward an "open skies" agreement with the U.S.; Establishment by the Public Service Regulatory Commission of a direct contracting me- chanism between the power distribution company and generator and service providers, and abolition of the state electricity wholesale company; Adoption of the Integrated Financial Rehabilitation Plan (IFRP) for public utilities (elec- tricity, drinking water, irrigation, and urban electric transport); Reduction in the monopoly rights of the fixed line operator, permitting introduction of competition for Internet and cellular services; Establishment of a new facility for registering pledges of moveable property in the cen- tral cadastre registry, along with legislative amendments, which together provide an adequate legal framework for most secured credit transactions; and Regulatory reforms in the banking industry, including abolishing the security and in- surance inspectorates and moving their responsibilities to the central bank. 4.36 The overall impact of these reforms has helped in the development of the private sector. As detailed in Table 8, the private sector has grown from 78 percent of GDP in 2003 to 85 per- cent of GDP by 2007/2008, and bank loans and deposits have also risen. In the case of infra- structure, there were no monitoring indicators in the PRSC and hence no targets for perfor- mance. Information from the Public Service Regulatory Commission shows a mixed picture. Electric power generation system is somewhat more efficient in terms of reduced power losses in transmission, but the average frequency of power outages has actually worsened since 2005. In drinking water, the losses remain high, at over 80 percent, and, if anything, have worsened over the last three years. Irrigation water losses, however, have declined (see Table 8). Table 8. Monitoring Indicators for Infrastructure and Private Sector Development Indicators 2003 2006 2006 2007 (actual) (target in PRSC I) (actual) (target) Private Sector (% of GDP) Share of Private Sector 78 84 85 85 Commercial Bank loans to private sector 6 9 9 10 Bank deposits (resident physical and legal entities) 9.0 9.6 10.2 11.0 Infrastructure 2004 2005 2006 2007 Electricity generation (mill. KWH) 6033 6316 5941 5898 Total losses, elect. Distribution (%) 17.3 16.3 15.1 14.4 Average freq. of cessation of users supply 5.4 5.7 5.9 Losses of Drinking Water in system (%) 80.7 83.6 84.9 Losses of Irrigation Water (%) 27.6 14.6 11.6 Source: Monitoring Indicator Tables, PRSC I and IV, and Public Service Regulatory Commission files. 4.37 As for the rural economy, it has done well in recent years. Between 2003 and 2006, value added grew on average by more than 7 percent per annum, although with significant variation from year to year, mostly weather related. Bank project assistance in extension, seeds, and irri- 23 gation probably helped make this growth possible,22 with the positive effect reinforced by the PRSC. However, the direct contribution to this growth performance by the PRSC seems fairly minimal. The PRSC's major contribution in this area was to help establish an effective system to control illegal logging, which will produce most of its benefits sometime in the future. Indica- tors are not available for some of the PRSC's other objectives, such as the development of agro- processing and commercial activities in rural areas. How effectively has the PRSC helped advance the dialogue and achieve results in sectors delivering services to the poor?23 4.38 Reforms in the social sectors had two broad goals: to improve overall efficiency and im- prove accessibility, especially by the poor. In education and health, Armenia had inherited a large establishment of facilities that were in excess of what was needed and were overstaffed. In social insurance, the family benefit program took up a substantial share of social expenditures but was not well targeted to the poor. The Bank worked with the Government through the PRSCs, project lending, and analytical and advisory services to implement reforms that were agreed upon as part of the poverty strategy. While projects provided technical help in imple- mentation of reforms, the PRSC opened a dialogue with the Ministry of Finance and helped en- sure that reforms were on track and that the necessary decrees and legislation were adopted. PRSC CONTRIBUTIONS TO SECTOR DEVELOPMENT: HEALTH 4.39 The post-transition drop in social spending had an adverse impact on health care, and health facilities were underutilized. Health care availability discriminated against rural areas and the poor. In 2004, public spending on health care was only 1.5 percent of GDP, significantly less than the average of 2.5 percent for lower income countries. Thus, there was a need for in- creased spending and better targeting of health services. 4.40 Objectives and policy content. The PRSC established the following objectives for health: Enhance the accessibility and quality of health services, especially for the poor Increase spending in real terms on a sustainable basis, with intra-sectoral reallocation of resources toward more cost-effective care modalities (such as primary care, especially in rural areas) Optimize the hospital network and enact efficiency improvements set for the hospital system by 2009. 4.41 One of the key ingredients for expanding services to the poor was a rationalization of the hospital structure. Payments to hospitals were shifted to a payment on the basis of patients treated and type of treatment, not the cost of running the facility. This resulted in the closing of redundant facilities and reduction in redundant staff. The savings were shifted toward the ex- 22 Progress has been made relative to the 2003 level on certified seeds and irrigation cost recovery. How- ever, actual performance was about 50 percent or more below the level expected at the time of PRSC I. 23 This section focuses on health, education, and social benefits. The only treatment of water supply in the PRSC program came indirectly in the establishment of water utilities performance indicators as part of general improvements in public utility management. Technically, social benefits are not "service deli- very," but these constitute a major element of the "pro-poor" emphasis of the PRSCs. 24 pansion of the polyclinic system (out-patient services), which provides free and/or highly sub- sidized services to poor families. Fees for primary health care have been abolished, further in- creasing access for the poor. The overall public health budget, which in 2003 was allocated 73 percent for hospital costs and 27 percent for outpatient services, was shifted so that, by 2007, 49 percent was going to hospital costs and 51 percent for outpatient care. Salaries in polyclinics have been raised and the quality of care has increased. Measures were adopted for the secured financing and management of the primary health care (PHC) departments of polyclinics, and a program was prepared for the prevention and control of non-communicable diseases (NCDs). This led to a program of preventive services for the control of such illnesses (tobacco control, mammography, high blood pressure, diabetes). Recently, the Government has given more em- phasis to maternal and child health care; all pregnant women receive free care. 4.42 Relevance and design. The PRSC program focused on a number of key steps, which in- cluded: Setting targets for the hospital rationalization program in terms of beds and physicians per person Establishing measures for secured financing and management of primary health care departments of polyclinics Adopting policies for defining scope of publicly funded healthcare services Preparing draft options for private health insurance and defining a regulatory frame- work for voluntary health insurance Preparing for, and beginning implementation of, a program for the control and preven- tion of non-communicable diseases 4.43 Implementation. Bank involvement with the sector was linked to two health moderniza- tion projects. The PRSC provided strong support for the policy dialogue that was necessary for project success, while the projects provided technical support for the reforms. 4.44 Results. Results are somewhat mixed. In terms of enhancing accessibility and quality, in- dicators show (see Appendix Table 9) that overall utilization by the population of clinics and hospitals has increased. However, there has been no increase in utilization of health facilities by the lower two quintiles. Utilization of PHC facilities by the bottom two quintiles is substantially below the target set in PRSC IV (3.5 percent vs. 10 percent), and about the same as it was in 2003. While percentage of ill people in the bottom two quintiles using primary health care facili- ties has risen in 2006, it seems unlikely that the target set forth in PRSC III will be met. Vaccina- tion coverage of children has declined. In terms of quality, an increase in the quality of services should result in better health indicators. However, little progress seems to have been made so far in improving health status (see Appendix Table 9). Indicators such as infant and maternal mortality have actually risen, technically speaking, since 2003, although much of this is most likely the result of small, random fluctuations24 or of combining data with different definitions. 4.45 In terms of increasing spending and improving intra-sectoral allocation, the results are better. Overall expenditures on health rose from 1.2 percent of GDP in 2002 to 1.5 percent of 24 Armenia has a very small population; and, since mortality figures are reported per 1000 people, small changes in the number of deaths can cause wide fluctuations in mortality rates. 25 GDP in 2006, but are below the target set in PRSC II of 1.9 percent (for 2007).25 The share for primary health care has also risen from 33 percent to 36 percent, but is somewhat below the ex- pected target of 39 percent for 2007. 4.46 Progress has been made in optimizing the hospital network. Occupancy rates (for APL hospitals) have risen from 40 percent to 67 percent between 2003 and 2006. The number of hos- pitals under contract to the State Health Agency (SHA) has fallen from 532 in 2004 to 406 in 2006, as a result of hospital rationalization, which includes merger and closure (see Appendix Table 9). PRSC CONTRIBUTIONS TO SECTOR DEVELOPMENT: EDUCATION 4.47 Objectives and policy content. Armenia, with a large educational establishment, has achieved a 93 percent primary enrollment rate (see Table 9) and appears capable of reaching its Millennium Development Goal Target of 100 percent by 2015. The country inherited a large, costly, inefficient education establishment from the Soviet Union in 1999, one ill-suited to its development needs. Working with the Bank, it instituted a number of structural reforms. Even so, by 2003, Armenia's education establishment was too big, with relatively small class sizes and woefully underpaid teachers. Average teacher salaries in 2003 were about USD 35 per month, inducing teachers to seek side payments for tutoring and encouraging various forms of corrup- tion. Table 9. Results in the Education Sector 2007 1999 2003 2004 2005 2006 (target) Primary enrollment, gross 81.7 93 90.0 91.8 93.1 99 Total enrollment, primary, general, secondary and vocational 56.6 69.0 70.9 72.0 Public Education Spending, % of GDP 2.3 2.1 2.5 2.7 2.7 Public spending on general education, per student (index 2003=100) 100 226 244 Proportion of children participating in pre-school services 16 16 18 Proportion of children from lowest quintile participating in pre-school services 19 19 20 Source: Appendix Table 10 and PAD, PRSC IV, monitoring indictors table. 4.48 There is no clear statement in the PRSCs of objectives in education, other than "Raising the Quality of Education."26 Key intermediate objectives included increasing budget allocations to education, increasing teacher salaries, and increasing school enrollment and completion rates, including pre-school enrollments. 4.49 Relevance and design. The Bank has worked with the Government on education reform, both under the PRSCs as well as in the process of implementing two projects on the quality of education implemented during the same period. Under the PRSCs, the Government agreed to do the following: 25 One reason health expenditures did not reach their expected target could be that GDP rose so rapidly during this period, growing between 12 to 14 percent per annum between 2002 and 2007. 26 Stated as a goal in the policy matrix for PRSC I-IV. 26 Undertake a school rationalization program, along with significant increases in teacher salaries; Make improvements in system efficiency through consolidating high schools, reviewing per capita financing formula, and reaching a target ratio of 16:1 pupils per teacher Develop a new law on Early Childhood Education, design standards for early childhood education institutional services, and begin implementation of a pilot project; Begin implementation of higher education reforms, including sustainable financing and piloting and expansion of the course credit system. 4.50 Implementation. A school rationalization program was undertaken that eliminated over 20,000 teaching posts and raised the average student to teacher ratio from 7:1 in 2003 to 14:1 in 2007. Teacher salaries were increased in three steps between 2004 and 2007, and now average about USD 200 per month. This salary should reduce rent-seeking activities by teachers, which tend to discriminate against the poor who cannot afford to pay. 4.51 To improve the quality of education, more resources have been allocated to the sector. Education's share of total public expenditures rose from about 10.2 percent in 2003 to 14 percent by 2006, while the sector's expenditures relative to GDP have risen from 2.1 percent of GDP in 2003 to 2.7 percent in 2007. Spending per student more than doubled since 2003, but a large part of this increase is due to higher teacher salaries. The share relative to GDP is still low compared to the 4.1 percent average for Europe and Central Asia and Former Soviet Union countries. About 70 percent of public expenditures go for primary and general secondary education. Un- der the present structure only 11 years of education is required; however, the fact that many students do not complete secondary education, particularly among the poor, is due more to challenges of physical and financial access than to the lack of a 12 year compulsory education law Expansion of secondary education is a priority, and the Government is already in the process of moving to a mandatory 12-year education cycle27and has approved a number of oth- er important pieces of education-related legislation.28 4.52 Results. According to the most 2008 PER for the Education sector in Armenia, over 95 percent of 7-15 year-olds are enrolled in school (based on 2005 data), but enrollment rates are much lower for other age groups (for example, preschool enrollment has declined in the last 10 years to 22 percent in 2005.29 Total school enrollment has expanded, rising from about 56 per- 27 In 2006-07, half of all 6-year olds (those aged 6.0-6.4 years) started school under the new 12-year curri- culum and will leave school in 2018. At the same time, the other half of these 6-year olds (those aged 6.5- 6.9 years) started primary school under the 11-year curriculum. This was meant to smooth the transition from the 10/11 to 12-year system. Also, in 2009, the new High School system with a three-year curricu- lum was introduced, so that, by 2012, the first group of students will completed high school under the 12- year system, despite having only completed 11 years of study. 28 For example, in November 2005, the government passed a new Law on Preschool Education, which stipulates the roles for the central and local governments in the context of the overall state education framework, specifics the different types of preschool education institutions, their responsibilities, and the organization of the learning process and financing mechanisms. In early 2008, it approved the Strategic Program for the 2008-15 Reforms in Preschool Education as well as a related Pilot Program. 29 The 2008 PER in Education attributes this largely to "the transfer of management responsibility in 1996 from central to local governments under the Local Self-Governance Law. With limited local own budgets, 27 cent in 1999 to 72 percent in 2006 (due mostly due to increases in upper secondary enrollment-- at 86% in 2006--since lower secondary levels are already quite high at over 99 percent) Primary enrollment is over 97 percent.; however, 30data on primary school completion show that com- pletion rates have stagnated at about 90 percent, but the data do not go beyond 2005. There ap- pears to be no data on secondary school completion.. As there are no results indicators for achievement scores, it is difficult to judge if the quality of education has risen. However, Arme- nia did perform well in the 2007 Trends in International Mathematics and Science Study (TIMSS) test, being the only ECA country to have improved its scores in all four categories compared to the 2003 TIMSS. Monitoring seems to have focused on inputs and efficiency goals (budget, student/teacher ratios), not on results, and the objectives in terms of results are not de- fined or monitored. PRSC CONTRIBUTIONS TO SECTOR DEVELOPMENT: SOCIAL PROTECTION 4.53 Objectives and policy content. Armenia has a substantial social protection system based on the Poverty Family Benefit System, introduced in 1999. This consists of cash benefits paid to eli- gible families, based on a proxy-means test and family characteristics. It has had a major impact on reducing poverty. According to the 2005 household survey, the family benefit reduced po- verty in the country by 1.8 percentage points, and extreme poverty by 2.5 percentage points. About 17 percent of all families in Armenia now receive the family benefit, a reduction from about 27 percent in 2000. This decline reflects both better targeting and more rapid income growth. The family benefit is equal to about 38 percent of the food poverty line. However, only about 25 percent of poor families, and 51 percent of the extreme poor, were beneficiaries in 2005. About 40 percent of beneficiaries were families not classified as poor. Still, about 80 per- cent of the benefits paid went to the lower 50 percent in the income strata. One important prob- lem was that, in 2005, only about half of poor families were even registered to receive the family benefit. 4.54 Relevance and design. Under PRSC supported reforms, the family poverty benefit was sharpened through: Real increases in benefits Further differentiation of the benefit amount according to family social parameters Regular annual review of the targeting formula Continued monitoring and evaluation of the program's poverty impact Further improvements in benefit administration, including staff training 4.55 Results. Preliminary data indicate that approximately 55 percent (estimated from 2009 Q2 household survey data) of poor families are currently recipients of the family benefit pro- gram, though it is likely that this figure will decline once the full year's data are available. This figure is expected to stabilize at 50% percent by 2011. local governments have been unable to provide the same level of service as before. Private provision of preschool education is also quite limited, covering only 1,351 children in 17 institutions in 2006." 30 Because many Armenian children enroll in primary school underaged, net enrollment rates for prima- ry may actually underestimate enrollment. Another factor that can negatively affect the enrollment rate is the emigration of families, many of which fail to check-out from the community registers when they move away. 28 ASSESSMENT OF OVERALL PRSC OUTCOMES 4.56 Reforms implemented under the PRSC were intended to make the social sectors more efficient, raise the quality of services, and increase access for the poor. They have been sup- ported by project operations in these sectors which have helped build institutional capacities and finance needed civil works and technical assistance. The PRSC helped raise policy issues at a higher level than what would have been possible with project lending alone. It also improved dialogue with the Government about new directions in the strategy (for example, early child- hood education). 4.57 However, the results so far appear mixed. In health, indicators show that overall access has improved, but access by the poor has not improved. Despite increase in access and more spending on health, health indicators also have not improved. If the quality of services had im- proved, then health indicators should reflect this. In education, overall enrollment rates have increased, but not those in primary education. Spending on education has increased dramatical- ly (per student), but there are few indications of the impact of this spending in terms of im- proved learning outcomes. In social protection, benefits going to the poor have increased, and the efficiency of administration has improved. PRSC CONTRIBUTIONS TO POVERTY OUTCOMES 4.58 Poverty has declined dramatically. As discussed earlier, poverty rates fell by 29 percent and extreme poverty rates by 36 percent between 2004 and 2006. At the same time, income dis- tribution seems to have improved, making the entire experience a very pro-poor growth epi- sode. However, progress in areas that measure the non-monetary aspect of poverty are more subdued. Unemployment rates, infant and maternal mortality, vaccination rates, utilization of health services, and access to water have either improved slowly or been stagnant.31 5. Bank Performance 5.1 Assessment of instrument adequacy. Overall, Bank performance was very good. The PRSC clearly followed and built on, and profited from, experience with policy-based lending. The Government adapted its approach based on previous structural adjustment experience. All in all, engagement under the PRSCs was successful and somewhat smoother than under the Struc- tural Adjustment Credits, with the Poverty Reduction Strategy Paper providing a useful framework for determining joint priorities. The PRSC clearly demonstrated their utility in rais- ing policy issues at a high level, beyond what could have been achieved with an exclusively project-lending approach. In effect, as an instrument, the PRSCs complemented project lending in key sectors. 31 For data see tables in this Chapter. For water, see Appendix Table 11. 29 5.2 Capacity assessment. The Government clearly had the willingness, if not always the ca- pacity, to implement reforms; from the Bank side, technical assistance was provided to build capacity and promote understanding of reform. This came in a variety of forms, including Poli- cy and Human Resources Development Fund grants, economic and sector work, resident mis- sion informal contacts, and expertise supplied through projects. There were often synergies be- tween the PRSCs, which promoted policy changes, and projects to provide the Government with technical skills and capacity to carry out reforms. Many country counterparts commented that their implementation of reforms would not have been possible without the technical help of the Bank. 5.3 Adequacy of Bank inputs. No Bank staff interviewed complained of inadequate resources. Bank staff were able to use a variety of funding sources to provide the necessary technical in- puts, including Policy and Human Resources Development Fund grants, Bank budget, project loans, and the work of other donors. Government staff appreciated the skills of Bank staff and technical expertise supplied by the Bank, and in many cases felt reforms would not have been possible without such assistance. There was generally high praise for the work of the resident mission, and the Government appreciated having Bank staff readily available for consultation. 5.4 Effectiveness in donor harmonization. Harmonization and cooperation with other donors has been good. Although there has been some overlap with the work of the IMF and USAID in tax reform, no serious conflict or duplication appears to have arisen. Since the Bank is, in effect, the only supplier of general budget support, there has been no need to harmonize the PRSC ma- trix with other donors. The Bank maintains contacts with other donors via a local donor group, which it co-chairs. Coordination with European Community budget support could become an issue in the future if the PRSC program is continued. 6. Conclusions and Lessons Learned 6.1 In general, the PRSC has proven its usefulness in Armenia as a vehicle for developing policy dialogue on key issues and providing support for a Government-led poverty reduction strategy. From the Bank side, it would not have been possible to implement these reforms using project finance or other instruments. The PRSCs provided a vehicle that supported and com- plemented project lending, since it allowed larger policy issues to be raised at the cabinet level. However, the presence of projects often provided the necessary technical expertise and capacity building that was critical for the success of the PRSC. 6.2 For the Government, the PRSC was a valuable instrument allowing it to implement a reform program while drawing on Bank-funded technical advice. The PRSC also permitted a valuable two-way relationship between the Ministry of Finance and operational ministries. The operational ministries used the PRSC vehicle to raise issues to the level of the Prime Minis- ter/Ministry of Finance, and to ensure a place for their programs in the Medium-Term Expendi- ture Framework. Likewise, the Ministry of Finance saw the PRSC as a way to emphasize high- 30 priority policy objectives and push sectoral ministries to complete their reforms. Government counterparts attest that the amount and quality of reform would not have been as great had the PRSC not existed. 6.3 Major lessons for future PRSCs include: The Bank needs to focus more frontally on corruption, which is endemic and a potential constraint to growth. More attention should be paid to key public financial management issues that remain unresolved -- for example, quality of audits, procurement, and man- agement of state non-commercial organizations. There needs to be a better results orientation to PRSCs, with progress reporting against an agreed-upon set of objectives that does not change from year to year, based on readi- ly available information that can be reported by the Government. There should be better monitoring of past actions to ensure that back-tracking has not occurred (for example, progress on VAT refunds and Customs valuation seem to be overstated). Poverty monitoring should consider non-monetary measures of poverty in addition to traditional income-based measures. Much can be accomplished with a PRSC, even when it constitutes a small amount of support vis-à-vis the overall budget. In essence, the PRSC's value is not exclusively in what it finances, but also in its ability to give shape to a reform program and serve as a nexus for organizing technical assistance. 6.4 The PRSCs work best in conjunction with project lending. Working in tandem, they can influence a policy framework, with projects providing funding for the necessary technical assis- tance and capacity building. In essence, the PRSCs and project lending can--and should--serve as mutually reinforcing elements. 31 32 References Government of Armenia. 2005. Armenia Demographic and Health Survey. National Statistical Service of the Republic of Armenia. Yerevan. Government of Armenia. 2007a. Government Revenue and Expenditure data from Armenian Budget Law. Ministry of Finance, Republic of Armenia. Yerevan. Government of Armenia. 2007b. Demographic Handbook of Armenia. National Statistical Service of the Re- public of Armenia. Yerevan. Government of Armenia. 2007c. Food Security and Poverty. National Statistical Service of the Republic of Armenia. Yerevan. Government of Armenia. 2007d. Social Snapshot and Poverty in Armenia. National Statistical Service of the Republic of Armenia. Yerevan. Mitra, Saumya, Douglas Andrew, Gohar Gyulumyan, Paul Holden, Bart Kaminski, Yevgeny Kuznetsov, and Ekaterine Vashakmadze. 2007. Caucasian Tiger: Sustaining Economic Growth in Armenia. Washington, D.C.. The World Bank. OECD. 2009. Aid Aggregates Statistics Database. http://stats.oecd.org/qwids/ OECD and World Bank. 2003. Budget Practices and Procedures Database. http://ocde.dyndns.org/ OED. 2002. "Project Performance Assessment Report on a Rehabilitation Credit, Structural Adjustment Credits I, II, and III and a Structural Adjustment Technical Assistance Credit to Armenia". Re- port No. 24459. Washington, D.C. OED. 2004. "Implementation Completion Report Review for the Fourth Structural Adjustment Credit to Armenia". Washington, D.C. IEG (Independent Evaluation Group). 2010a. "Implementation Completion Report Review for the First Poverty Reduction Support Credit to Armenia. World Bank. Washington, D.C. IEG (Independent Evaluation Group). 2010b. "Implementation Completion Report Review for the Second Poverty Reduction Support Credit to Armenia". World Bank. Washington, D.C. IEG (Independent Evaluation Group). 2010c. "Implementation Completion Report Review for the Third Poverty Reduction Support Credit to Armenia". World Bank. Washington, D.C. IEG (Independent Evaluation Group). 2010d. "Implementation Completion Report Review for the Fourth Poverty Reduction Support Credit to Armenia". World Bank. Washington, D.C. Ratha, Dilip and Zhimei Xu. 2007. "Migration and Remittances Factbook". Migration and Remittances Team, Development Prospects Group, Washington, D.C.. The World Bank. Accessible online at: www.worldbank.org/prospects/migrationandremittances. Taliercio, Robert. 2007. "Aide Memoire, Programmatic Public Expenditure Review." World Bank. Wash- ington, D.C World Bank. 1999. "Implementation Completion Report on Structural Adjustment Credits I and II to Ar- menia". Washington, D.C. World Bank. 2001. "Implementation Completion Report on the Third Structural Adjustment Credit to Armenia". Washington, D.C. World Bank. 2003a. "Implementation Completion Report on the Fourth Structural Adjustment Credit to Armenia". Washington, D.C. 33 World Bank. 2003b. "Implementation Completion Report on the Fourth Structural Adjustment Credit to Armenia". Washington, D.C. World Bank. 2003c. "Public Expenditure Review of Armenia". Report No. 26973. Washington, D.C. World Bank. 2004. "Program Document for the First Poverty Reduction Support Credit to Armenia". Washington, D.C World Bank. 2005. "Program Document for the Second Poverty Reduction Support Credit to Armenia". Washington, D.C World Bank. 2006. Budget Reform Study Internal Draft, Annex 20. ECA Regional Vice Presidency. World Bank. 2007a. "Program Document for the Third Poverty Reduction Support Credit to Armenia". Washington, D.C World Bank. 2007b. "Program Document for the Fourth Poverty Reduction Support Credit to Armenia". Washington, D.C World Bank. 2008. World Development Indicators database. http://data.worldbank.org/data-catalog World Bank. 2009. "Implementation Completion Report on Poverty Reduction Support Credits I, II, III, and IV to Armenia". Washington, D.C. 34 Appendix A Appendix Table 1. Share of PRSC and Adjustment Lending in Total Disbursement to PRSC Countries Adjustment Lending/Total Country Disbursement Pre- Pre- Pre- Pre- Pre- PRSCFY5 PRSCFY4 PRSCFY3 PRSCFY2 PRSCFY1 PRSC PRSC PRSC PRSC Country (%) % % % % FY1 % FY2 % FY3 % FY4 % Albania 56.2 54.4 38.5 36.8 28.0 29.9 16.0 Armenia 53.3 37.4 28.4 55.5 34.5 28.2 36.9 33.6 24.0 Azerbaijan 54.8 49.8 32.9 Benin 23.2 34.8 45.0 55.2 50.4 Burkina Faso 29.2 36.9 60.9 50.0 39.7 50.5 Cape Verde 57.1 41.4 45.9 48.6 0** 42.1 Ethiopia 33.1 35.2 29.2 36.8 Georgia 43.3 33.7 37.4 26.1 28.9 Ghana 12.1 23.2 40.8 50.2 54.2 56.0 0** 39.7* 61.2 Guyana 9.3 Honduras 0.4 31.0 29.3 Lao PDR 17.0 26.2 0.2 26.7 15.5 Madagascar 59.1 27.5 24.2 38.6 37.2 21.9 Mali 23.8 61.7 34.6 38.2 Moldova 29.1 35.7 24.0 Mozambique 72.6 38.5 26.9 39.0 26.5 Nepal 74.3 Nicaragua 32.1 24.6 16.8 25.4 0** 39.1 46.0 Pakistan 53.2 58.7 56.8 63.1 30.8 12.4 29.7 Rwanda 69.5 63.7 46.2 50.1 53.1 48.1 Senegal 66.2 43.3 36.1 14.4 0** 20.3 Sri Lanka 63.1 Tanzania 31.0 16.6 26.1 51.6 54.5 39.3 32.7 44.1* 49.7 Uganda 29.8 26.7 28.5 25.5 45.7 57.6 47.9 52.5 49.2 Vietnam 24.2 14.2 30.4 35.0 25.0 25.4 Source: Business Warehouse, World Bank Note: * = There was more than one PRSC disbursement in a given year for country; ** = no PRSC in given year for country 35 Appendix Table 2. PRSC Disbursements as share of GBS, Government Expenditures 2001 2002 2003 2004 2005 2006 2007 Pre-PRSC Pre-PRSC Pre-PRSC PRSC1 YR PRSC2 YR PRSC3 YR YR3 YR2 YR1 Credits/Loans (State Budget 29.2 0.0 23.8 10.3 0.0 9.2 10.2 Support), bln AMD PRSC 0.0 0.0 0.0 10.3 0.0 9.2 10.2 SAC 29.2 0.0 23.8 0.0 0.0 0.0 0.0 Grants (State Budget 2.2 2.6 3.3 0.0 2.9 5.0 0.9 Support), bln AMD Netherlands 2.2 2.6 3.3 0.0 2.9 5.0 0.9 Total General (State) 31.4 2.6 27.1 10.3 2.9 14.2 11.1 Budget Support, bln AMD Total ODA*, bln AMD 111.5 171.6 140.8 123.3 86.8 77.5 NA PRSC/General Budget 0.0 0.0 0.0 100.0 0.0 64.6 91.6 Support, % General Budget 28.2 1.5 19.2 8.3 3.3 18.3 NA Support/ODA, % Government Revenues and Expenditures** Total Expenditures, bln 244.4 263.9 312.7 334.0 417.5 481.2 558.7 AMD Total recurrent 188.6 219.4 222.3 262.8 329.1 356.6 401.4 expenditures Total capital expenditures 55.8 44.5 90.4 71.2 88.4 124.6 157.2 (including net lending) Total Domestic 184.3 211.3 241.6 288.5 363.1 426.3 475.1 Revenues, bln AMD ODA/Government 45.6 65.0 45.0 36.9 20.8 16.1 NA Expenditures, % General Budget Support/Government 12.9 1.0 8.7 3.1 0.7 3.0 2.0 Expenditures, % PRSC/Government 0.0 0.0 0.0 3.1 0.0 1.9 1.8 Expenditures, % Domestic Revenues/Government 75.4 80.1 77.2 86.4 87.0 88.6 85.0 Expenditures, % Source: OECD online database. Government revenue and expenditure data for 2007 are program data from budget law 36 Appendix Table 3. Predictability of PRSC and Other GBS Resource Flows Commitment made prior Month Country Amount Amount to budget disbursement Month of FY Expected, Disbursed, Disbursed/Expected GBS Donor being expected per actual quarter in % mill AMD mill AMD submitted to MOF budget disbursement which Parliament plan disbursed (yes/no) FY 2004 PRS-I 11,786 10,290 87.3 Yes NA Dec Q4 (USD 20 mill) Netherland's 884 - 0.0 Yes Dec - - Government (EU 1.5 mill) FY 2005 PRSC - - - - - - - Netherland's 1,097 2,909 265.2 Yes June March Q2 Government (EU 1.65 mill) FY 2006 PRSC2 9,364 9,168 97.9 No na April Q2 (USD 20 mill) Netherland's 2,520 5,028 199.5 Yes March May, Dec Q2, Q4 Government (EU 5 mill) FY 2007 PRSC3 9,996 10,199 102.0 No NA April Q2 (USD 28 mill) Netherland's 2,190 932 42.6 Yes Dec Dec Q4 Government (EU 4.7 mill) Sources: Government of Armenia, World Bank 37 Appendix Table 4. Alignment of PRSC Process with Domestic Planning and Budgeting Processes, for 2004 Medium Term Expenditures National Budget/Planning PRS review and update Month Framework of PRSC Process Process process Government (MTEF) Process January 1 Start of Financial Year Government adopted the 15 2004-2006 Action Plan of PRSP-1 Implementation (in 31 January 22, 2004) February 1 15 28 March 1 15 31 April 1 15 30 May 1 Submission of the 2003 State Budget Implementation Report to the National Assembly 15 Examination, Discussions and Adoption of the 2003 State Budget Implementation Report by National Assembly 31 June 1 15 30 July 1 Government discus and Start of the 2005 Annual adopt MTEF for the next Budget Process (budget 3 year by the July 10 formulation) 15 Government publish MTEF by the July 20 31 August 1 15 31 September 1 15 30 October 1 38 Appendix Table 4. Alignment of PRSC Process with Domestic Planning and Budgeting Processes, for 2004 15 31 November 1 2005 State Budget was Government has adopted PRSC 1 has been adopted in December 24, the PRSP Monitoring approved (November 2004 System (November 11, 18, 2004) 2004) 15 30 Start of the formulation of the draft MTEF (1st stage of Budget process) December 1 Actual Disbursement of the PRSC1 (December 30, 2004) 15 31 End of Financial Year Sources: Government of Armenia, World Bank 39 Appendix Table 5. Alignment of PRSC Process with Domestic Planning and Budgeting Processes, for 2006 Medium Term Expenditures National Budget/Planning PRS review and update Month Framework of PRSC Process Process process Government (MTEF) Process January 1 Start of Financial Year (2006) PRSC 2 has been 15 approved (January 19, 2006) 31 February 1 15 28 March 1 15 31 April 1 Actual Disbursement 15 of the PRSC 30 May 1 Submission of 2005 State Budget Implementation Report to the National Assembly 15 31 June 1 PRSP 2nd Progress Report (for 2004-05) 15 Examination, Discussions and Adoption of the 2005 State Budget Implementation Report by National Assembly 30 July 1 Start of 2007 Annual Budget Process (budget formulation) 15 31 August 1 WB/IMF Joint Staff WB/IMF Joint Staff Advisory Advisory Note- PRSP Note- PRSP Progress Report Progress Report (August (August 26, 2006) 26, 2006) 15 31 September 1 Start of regional Discussions of PRSP 2 15 30 Submission of the 2007 State 40 Appendix Table 5. Alignment of PRSC Process with Domestic Planning and Budgeting Processes, for 2006 Budget Draft Low to the National Assembly October 1 15 31 November 1 2007 State Budget was adopted in November 29, 2006 15 30 Start of the formulation of the draft MTEF (1st stage of Budget process) December 1 15 31 End of Financial Year Sources: Government of Armenia, World Bank 41 Appendix Table 6. Alignment of PRSC Process with Domestic Planning and Budgeting Processes, for 2007 Medium Term Expenditures National PRS review and update Month Framework of Budget/Planning PRSC Process process Government (MTEF) Process Process January 1 Start of Financial Year 15 31 February 1 Start of drafting the PRSP 15 2 28 March 1 15 PRSC III has been approved (March 8, 2007) 31 April 1 15 Actual Disbursement of the PRSC III (April 28, 30 2007) May 1 Submission of the 2006 15 State Budget Implementation Report to 31 the National Assembly June 1 Examination, Discussions 15 and Adoption of the 2006 State Budget 30 Implementation Report by National Assembly July 1 Government discus and adopt MTEF for the next 3 year by the July 10 15 Government publish MTEF by the July 20 31 August 1 15 31 September 1 15 30 October 1 15 31 November 1 Regular (monthly or 15 bimonthly) Supervisory Missions; Preparation of 42 Appendix Table 6. Alignment of PRSC Process with Domestic Planning and Budgeting Processes, for 2007 30 Start of the formulation of the Program Document (Sept draft MTEF (1st stage of 24, 2007) Budget process) December 1 15 31 Sources: Government of Armenia, World Bank 43 Appendix Table 7. PEFA Scores for Armenia (2006) Source: Europe and Central Asia Region, "Budget Reform Study Internal Draft" September, 2006, Annex 20 44 Appendix Table 8. Relationship of the PRSC Program to Other Bank Lending Projects Sources: PAD, PRSC IV 45 Appendix Table 9. Results in the Health Sector 1999 2003 2004 2005 2006 2007 Target (PRSC #) Health Resources and Access Health Exp. /GDP 1.4 1.2 1.3 1.4 1.5 (est.) 1.9 (II) PHC Exp./ Total Budget 33.2 32.3 38.9 36.4 (est.) 38.5 (II) Rationalization Number of SHA hospital 532.0 418.0 406.0 contracts (from APL-2) Occupancy Rate for APL 40.0 50.0 60.0 67.0 project hospitals Visits to ambulatory 2.1 2.1 2.4 2.5 outpatient clinics, per person, annual average Utilization/Coverage Hospital admissions per 100 6.9 7.3 7.9 8.4 population Utilization of PHC by bottom 3.5 4.8 3.7 3.5 10.0 (IV) two quintiles (%) % of ill people in bottom two 27.1 21.1 20.1 28.5 33 (III) quintiles using PHC facilities Vaccination coverage, 92.4 93.4 89.6 88.3 Children (%) Health Results Indicators Infant Mortality (per 1000 15.4 12.0 11.6 12.3 13.9 births) Maternal Mortality (per 32.9 18.0 26.7 18.7 26.6 100,000 births) Sources: AM from Health Modernization APL, Oct, 2007, PRSP-2 draft (data on clinic and hospital usage), AP:-1 PAD, and published sources. Targets taken from various PRSC monitoring tables, and APL-1. 46 Appendix Table 10. Selected Indicators of Non-Income Poverty in Armenia 1999 2004 2005 2006 Source 1 Health Care Aspect 1.1 Life expectancy at birth (years) 72.7 73.4 73.5 73.3 a 1.2 Under-five mortality rate, per 1000 live births 19.3 13.0 13.7 15.8 a 1.3 Infant mortality (under 1), per 1000 live births 15.4 11.6 12.3 13.9 a 1.4 Maternal mortality, per 100 000 live-births 32.9 26.7 18.7 26.6 a Vaccination coverage for children by types of diseases, % of children to be 1.5 92.4 93.4 89.6 88.3 b vaccinated Poliomyelitis 96.7 93.0 89.4 87.4 b Diphtheria and tetanus 91.1 93.5 85.6 86.8 b Whooping cough 89.6 93.5 84.4 85.0 b Measles 91.1 91.5 94.4 91.9 b Tuberculosis 93.6 95.8 94.8 90.4 b 1.6 Prevalence of underweight children under five years of age, in % 2.6 na 4.0 na c (2000) 1.7 Prevalence of stunted children under five years of age, in % 13 na 13.0 na c (2000) 1.8 Prevalence of wasted children under five years of age, in % 2 na 5.0 na c (2000) 1.9 Expenditures of State and Community Budgets in Health Care, % of GDP 1.4 1.3 1.4 1.5 b 2 Educational Aspect Enrolment of children in primary schools/classes as a % of children of the 2.1 81.7 90.0 91.8 93.1 b corresponding age Enrolment of children in the general education system, as a % of the number 2.2 77.4 87.4 89.2 88.9 b of children of the corresponding age Enrolment of pupils and students (from state and non-state primary, secondary 2.3 and higher vocational educational institutions), % of the number of population 56.6 69.0 70.9 72.0 b of the corresponding age 2.4 Expenditures of State and Community Budgets in Education, % of GDP 2.3 2.5 2.7 2.7 b 3 Living Conditions 3.1 Share of households with access to safe drinking water, % 83.7 88.9 89.4 91.3 d Urban areas 96.2 96.5 97.9 98.1 d Rural areas 64.7 74.0 72.8 78.0 d 3.2 Share of households with access to centralized sewerage system, % 57.9 61.6 65.1 83.1 d Capital city (Yerevan) 90.6 93.4 94.9 96.7 d Urban areas (excluding Yerevan) 81.0 79.0 83.2 83.1 d Rural areas 15.4 12.8 17.7 19.1 d 3.3 Share of households who use natural gas for heating, % 2.3 15.9 24.1 39.2 d Capital city (Yerevan) 0.2 11.0 13.6 31.4 d Urban areas (excluding Yerevan) 3.7 24.6 39.5 51.2 d Rural areas 2.6 12.5 20.5 34.7 d Sources: a - Demographic Handbook of Armenia, NSS of RA. b - Food Security and Poverty, NSS of RA. c - Armenia Demographic and Health Survey, NSS of RA. d - Social Snapshot and Poverty in Armenia, NSS of RA. 47 Appendix Table 11. Relationship of PRSC Program to Bank AAA Work Sources: Government of Armenia, World Bank 48 49