Report No: ACS7927 Samoa Samoa Public Expenditure Review Notes March 2014 EASPN EAST ASIA AND PACIFIC Standard Disclaimer: This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not 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 any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Copyright Statement: The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Samoa Public Expenditure Review Notes Taking stock of expenditure trends from FY06-FY12 March 2014 THE WORLD BANK Abbreviations and Acronyms CAPEX Capital expenditure MoF Ministry of Finance CEO Chief Executive Officer MoH Ministry of Health CSO Community Service Obligation NCD Non-Communicable Diseases DPT Diphtheria, Pertussis, Tetanus NGO Non-Governmental Organisation DSA Debt Sustainability Analysis NHS National Health Service EAP East Asia and the Pacific OMT Overseas Medical Treatment ECE Early Childhood Education PER Public Expenditure Revuew FMIS Financial Management Information PSC Public Service Commission Systems FY Fiscal Year SABER Systems Approach for Better Education Results GDP Gross Domestic Product SDS Samoa Development Strategy GFS Government Financial Statistics SOEMD State-Owned Enterprise Monitoring Division GoS Government of Samoa SPG South Pacific Games HQ Headquarters STR Student-Teacher Ratio LMIC Lower Middle Income Countries TATTE Tui Atua Tupua Tamasese Efi Building LTA Land Transport Agency WHO World Health Organisation MoESC Ministry of Education, Sports and Culture i Contents ACKNOWLEDGEMENTS ..........................................................................................................................................iv INTRODUCTION .........................................................................................................................................................1 SUMMARY ..................................................................................................................................................................2 AGGREGATE FISCAL TRENDS ................................................................................................................................1 GOVERNMENT PERSONNEL COSTS .................................................................................................................... 19 HEALTH ..................................................................................................................................................................... 28 EDUCATION .............................................................................................................................................................. 41 Data Annex .................................................................................................................................................................. 52 ANNEX TABLES ....................................................................................................................................................... 54 Figures Figure 1: External debt (% GDP) ..................................................................................................................................3 Figure 2: Decomposition of expenditure growth by education sub-sector(domestic) ...................................................3 Figure 3: Domestic expenditure growth by health sector output (FY06 – FY12) .........................................................4 Figure 4: Composition of expenditure (FY12) ..............................................................................................................3 Figure 5: Budget shares (% of total) in FY06 & FY12 ..................................................................................................3 Figure 6: Expenditure & Revenue (% GDP) .................................................................................................................4 Figure 7: Decomposition of expenditure growth (FY06-FY12) ...................................................................................4 Figure 8: External debt (% GDP) ..................................................................................................................................4 Figure 9 : Tax revenue/ GDP in PICs (2011/2012) .......................................................................................................5 Figure 10: % change in dom. revenue (FY06-12) .........................................................................................................5 Figure 11: Transfers to public agencies (Tala millions - constant) ................................................................................7 Figure 12: Transfers by public agency (% total in FY10-12) ........................................................................................7 Figure 13: Aid by type (% of total) ...............................................................................................................................9 Figure 14: Loans by purpose (Tala millions) ................................................................................................................9 Figure 15: Expenditure by sector (% of budget) ...........................................................................................................9 Figure 16: % change in expenditure by sector (FY06 -FY12) .......................................................................................9 Figure 17: Revenue outturns (actual/ original budget) ............................................................................................... 10 Figure 18: Expenditure deviations (original budget vs. actual) .................................................................................. 11 Figure 19: Expenditure deviations by ministry (FY06 –FY12) .................................................................................. 11 Figure 20: Disaggregated expenditure deviations (FY06-12)...................................................................................... 12 Figure 21: Breakdown of deviations by budget category in normal years................................................................... 12 Figure 22: Size of revised Budget Adjustment (% of original budget) ........................................................................ 13 Figure 23: Comparison between revised and original budget deviations (FY06-12 average) .................................... 13 Figure 24: Aggregate fiscal projections FY13-FY33 .................................................................................................. 14 Figure 25: Projection of fiscal space for priority sectors ............................................................................................. 15 Figure 26: Total personnel costs, in real and nominal terms ....................................................................................... 20 Figure 27: Personnel costs as proportion of recurrent expenditure and GDP .............................................................. 20 Figure 28: Shares of personnel costs ........................................................................................................................... 21 Figure 29: Growth in personnel costs by function (FY06 – FY12) ............................................................................. 21 Figure 30: Growth in personnel costs by ministry (FY06 – FY12) ............................................................................. 21 Figure 31: Share of overall growth in personnel costs by ministry (FY06 – FY12) .................................................... 21 Figure 32: International comparison, wage bill as proportion of recurrent expenditure .............................................. 22 Figure 33: International comparison, wage bill as proportion of GDP ........................................................................ 22 Figure 34: Government employment ........................................................................................................................... 24 Figure 35: Growth in payroll components over FY06-FY12 ...................................................................................... 24 Figure 36: Average annual salary (constant prices) ..................................................................................................... 25 Figure 37: Salary to GDP per capita ratio .................................................................................................................... 25 Figure 38: Personnel costs by type of remuneration .................................................................................................... 25 Figure 39: Salary scale range ....................................................................................................................................... 25 Figure 40: Government employee Lorenz curves, FY06 and FY12 ............................................................................ 26 ii Figure 41: Health outcomes in Samoa ......................................................................................................................... 29 Figure 42: Reproductive health outcomes ................................................................................................................... 30 Figure 43: Access to basic infrastructure services ....................................................................................................... 30 Figure 44: Healthcare expenditure ............................................................................................................................... 31 Figure 45: Cumulative growth in healthcare expenditure ............................................................................................ 31 Figure 46: Health expenditure by function .................................................................................................................. 32 Figure 47: Donor project expenditure, by function ..................................................................................................... 32 Figure 48: Domestic expenditure by economic category ............................................................................................. 32 Figure 49: Domestic expenditure on personnel ........................................................................................................... 32 Figure 50: Public healthcare expenditure as a percent of GDP and total expenditure ................................................. 33 Figure 51: Public health sector expenditure, regional comparisons ............................................................................ 33 Figure 52: Domestic expenditure growth by output (FY06 – FY12)........................................................................... 35 Figure 53: Domestic nursing & community services expenditure ............................................................................... 35 Figure 54: Administrative costs breakdown ................................................................................................................ 36 Figure 55: Preventative health care expenditure .......................................................................................................... 37 Figure 56: Vaccination expenditure (inc. donor programs) ......................................................................................... 37 Figure 57: Vaccination rate, measles ........................................................................................................................... 38 Figure 58: Vaccination rate, diphtheria-pertussis-tetanus (DPT) ................................................................................ 38 Figure 59: Domestic tertiary expenditure .................................................................................................................... 39 Figure 60: Total tertiary expenditure ........................................................................................................................... 39 Figure 61: Composition of MOESC spending – FY06 to FY12 average (share of total) ............................................ 42 Figure 62: Selected education indicators compared to PICs ....................................................................................... 43 Figure 63: Education spend as a share of total budget compared to PIC average ....................................................... 43 Figure 64: Education budget (2006 prices) .................................................................................................................. 44 Figure 65: Education expenditure per person (2006 prices) ........................................................................................ 44 Figure 66: Decomposition of expenditure growth by sub-sector(domestic) ................................................................ 45 Figure 67: Decomposition of expenditure growth by economic category (domestic) ................................................. 45 Figure 68: % change in education compared to total budget ....................................................................................... 46 Figure 69: Education budget growth compared to total budget ................................................................................... 46 Figure 70: Net student enrollment ratios in government schools ................................................................................ 47 Figure 71: Student-teacher ratios ................................................................................................................................. 47 Figure 72: Student-teacher ratios by district (primary) 2013 ....................................................................................... 48 Figure 73: Student-teacher ratios by district (secondary) ............................................................................................ 48 Figure 74: National year 8 (primary) exam results ...................................................................................................... 48 Figure 75: National year 12 (secondary) exam results .............................................................................................. 48 Figure 76: Primary and secondary progression rates (percent) ................................................................................... 49 Figure 77: Primary and secondary drop-out rates ........................................................................................................ 49 Figure 78: Potential cost pressures for MOESC ......................................................................................................... 51 iii ACKNOWLEDGEMENTS These public expenditure review notes were prepared by World Bank staff to strengthen the analytical basis for the Government of Samoa in the management of public expenditure. The lead authors are Shireen Mahdi and David Knight. The team is grateful to the government for their cooperation in providing data for the notes and for sharing feedback on the content. The team is also grateful to a number of World Bank colleagues who reviewed and provided guidance for this work. These include Robert Utz, Vivek Suri, Rosa Alonso I Terme, Maeva Betham Vaai, Virginia Horscroft, Eileen Sullivan and Stephen Close. iv INTRODUCTION Samoa’s fiscal position and the structure of its budget have evolved markedly in recent years. Like other Pacific Island countries, Samoa has an economy that is highly vulnerable to external shocks. Samoa had built up sufficient fiscal space in the early to mid-2000s to be able to respond to a major exogenous shock. Deficits were low, averaging less than half a percent of GDP in the three years to FY08, and debt had been reduced to just over 30 percent of GDP. Thereafter, the government has had to increase its expenditure significantly, in order to respond to the global economic crisis and successive major natural disasters (a tsunami in 2009 and cyclone in 2012). It has also invested in large infrastructure projects. These factors have driven a period of rapid fiscal expansion and external financing of the budget. At the same time, the structure of the budget has been evolving both as the number of, and share of the budget executed by, public agencies expanded. The objective of the PER notes is to assist the Government of Samoa (GoS) in taking stock of the evolving expenditure trends in recent years and to strengthen the analytical basis for the management of public expenditure. This series is composed of a concise set of PER notes analyzing forward projections for fiscal policy, fiscal trends, the wage bill and the two largest expenditure sectors: health and education. This first note reviews trends in public expenditure, revenues, grants, financing and debt accumulation to provide insight into which aspects of real resource allocation have changed the most over the period. The second note will analyze the trends in the wage bill to inform the GoS on the size and growth of its personnel budget. The third and fourth notes consist of an analysis focusing on the two of the largest priority sectors: education and health. These notes will assess expenditure trends in major sector programs and frame the budget trends against sector policies and outcomes to the extent possible. The notes are based on detailed analysis of disaggregated data covering the seven year period between FY06 and FY12. These notes are the first time Financial Management Information System (FMIS) data has been used to produce a consistent public expenditure analysis, and as such are expected to be a valuable source of information for policy-makers. The FMIS data was consistently classified and supplemented with project grant and loan related data from Ministry of Finance accounts to compile a full database of domestically and externally funded expenditure. The analysis of the public wage bill combines FMIS data with data from the payroll system. It also incorporates payroll data from the accounts of the largest public agencies to present approximate estimates of payroll trends for the whole government for the first time. Similarly, the health analysis represents the first time that unified data for the public healthcare sector has been presented since the creation of the National Health Service as an autonomous agency. 1 SUMMARY Rapid fiscal expansion and borrowing over the past seven years have left Samoa with a large stock of debt and an elevated cost base. With heightened fiscal needs arising from the December 2012 cyclone, concerted efforts will now be needed to bring the public finances back onto a sustainable footing. Samoa’s fiscal trends between FY06 and FY12 are characterized by increasing expenditure, modest revenue growth and an expanding deficit largely financed by aid inflows. As a result, the budget deficit steadily grew, peaking in FY10 at 7.5 percent but remaining high in subsequent years. Deficits were financed with external borrowing, and external debt grew to reach 55 percent of GDP at end FY12. Substantial new fiscal commitments arose from the need to recover from Cyclone Evan, placing increased pressure on the budget, and budget deficits are likely to need to be lowered and kept around 1 percent of GDP in order to reduce debt to a comfortable level in the foreseeable future. Samoa has generally run a well-managed budget in the face of large externals shocks such as spikes in commodity prices, and the global economic crisis. Nevertheless, recent borrowing has highlighted a need for more prudent debt management and in particular, for higher quality investments. The composition of expenditure since FY06 has been supportive of the government’s policy priorities, with two thirds of the budget funding recurrent expenditure and the remaining third being allocated to capital spending in the priority sectors. Expenditure growth was driven by recurrent and capital spending in similar proportions as the government increased resources to a broad range of areas including frontline service delivery personnel, post-disaster reconstruction and other construction projects. As the budget expanded, the sectoral composition of total spending remained balanced between social, economic and administrative sectors. Budget execution has also been generally strong at the aggregate level even during the peak of the post-tsunami period. Although grant funding buoyed much of Samoa’s growing budget, loan financing has also been growing rapidly and has been accompanied by a rising debt-to-GDP ratio. Tsunami and economic recovery related activities contributed to this trend. However, 57 percent of loan financed expenditure between FY09 and FY12 went towards large-scale non-economic infrastructure projects such as a number of large government buildings. 2 The wage bill was a major driver of the growing Figure 1: External debt (% GDP) recurrent budget, especially in the health and Fiscal strategy education sectors, and continued rapid growth in target these sectors would risk undermining the government’s overall fiscal consolidation efforts. Over the period, while payroll in the education sector grew at slightly above the average for the rest of government at 35 percent, payroll in the health sector grew much more rapidly, at 80 percent. In most of government increased personnel costs have financed an expansion of public sector employment. Average real salaries have only just kept up with prices, with rapidly growing average salaries in the health sector being the notable exception. Despite an upward trend in the wage bill, the proportion of resources that are devoted to wages in Samoa is not particularly high compared to regional peers or the average for lower-middle income countries. As the largest ministry and public sector employer, robust growth in employment drove expansion in the education sector and the overall recurrent budget. The Ministry of Education, Sports and Culture budget is both the single largest expenditure item in the domestic Figure 2: Decomposition of expenditure growth by budget and the ministry that experienced the education sub-sector(domestic) largest increase in expenditure since FY06. Most of the increase in the education budget went towards the primary and secondary teaching wage bill and the majority of this spending was on the recruitment of new teachers. Although teacher salaries increased, they did not keep up with prices, and have hence declined on average in real terms. Funding for sports and culture was the second most important driver of growth in the education budget. In contrast, total government funding for the tertiary education sector did not increase over the period. 3 From a lower base, the health budget grew much more rapidly and was largely driven by growing payroll costs, with a significant expansion in primary and healthcare Figure 3: Domestic expenditure growth by health services. Health is the second largest sector output (FY06 – FY12) expenditure area and one of the fastest growing areas in the budget. The sector saw signficant increases in resources channeled to primary and treatment services. Personnel expenditure, especially on senior nurses was the main area of growth for the health budget as part of the program to expand integrated community healthcare. In addition, the new referral hospital is now completed and tertiary treatment for non-communicable diseases which comes at a very high cost per patient has continued to rise. While primary healthcare has expanded rapidly, as per the sector plan, health prevention has not seen increases in resources, and has fallen as a share of total health spending. Whilst employment in the health sector grew rapidly, it also bucked the trend in the rest of government by realizing a large increase in the average real salary. These trends raise questions about how effectively the additional resources are being spent, and whether the expansion can be brought under control. While the exact reason for the rise in the average salary is not clear from the data, it seems likely that there has been an increased use of overtime payments over the period. It is notable that the stark difference in personnel policy outcomes for the health sector have occurred over the same period that the NHS has become independent of the PSC. Given the greater independence that the corporate body now enjoys, it will be all the more pressing for central government to make full use of the accountability and control mechanisms that remain in place to ensure that health sector policy is not inconsistent with overall fiscal and personnel policy, and that fragmentation of responsibility does not undermine the strong control and coherency of government fiscal policy. Although the expansions in the health and education sectors mark the government’s focus on strengethening key basic services, it is not clear that the current expenditure strategies are yielding the desired results. The education sector invested in increasing the number of teachers even though student-teacher ratios were close to national target levels. Hence, the average ratios for both primary and secondary schools have decreased to well below the target levels by 2012. Yet, it is not clear that the additional investments, especially the reduction of class sizes, have led to improving outcomes at the primary level as indicated by the lower progression rates from primary to secondary and weaker scores in some key subjects. For health, 4 while good, trained staff are undoubtedly crucial, it is not clear that the buildup in staffing levels of senior nurses has addressed binding staff constraints in a balanced manner. Some evidence has suggested that recruitment has been concentrated amongst senior nursing staff whilst gaps persist at the more junior levels. The growing share of the budget executed by autonomous public agencies has also been a key trend in recent years. Transfers to non-central government agencies expanded markedly as a larger share of government functions have been assigned to a series of new public agencies and as transfers to help fund public utility operations. Over the period between FY06 and FY12, the share of the budget executed by public agencies increased from 15 to 40 percent of recurrent expenditure. In the health sector especially, the establishment of a variety of autonomous public agencies tasked with delivering health services has entailed a significant increase in non-service delivery (administrative) costs. Overall, Samoa benefits from strong fiscal systems, a clear commitment to strengthening service delivery and pursuing important reforms. This is evident from its expenditure policy. Nevertheless, Samoa is and will continue to be highly vulnerable to both natural disasters and exogenous economic shocks, all of which have been painfully evident over the last decade. Recent fiscal trends and Cyclone Evan have set the budget on a tight balance looking ahead. Hence, achieving the twin goals of fiscal sustainability and improved services will require the government to exercise vigilance in the planning and implementation of expenditure policy, to subject debt-financed proposals to extra levels of scrutiny and to ensure value-for-money in government programs. Such efforts, taken together with the continued support of development partners, especially in times of crisis, will support the government’s policy and reform agenda. In this context, key recommendations from the PER are summarized in the following matrix: 5 Potential Lead timeline Agency(s) Aggregate fiscal control Debt Continue to tighten debt controls and prevent contracting of non-concessional loans On-going MoF, GoS management Where possible secure grant financing for project, and where not possible, refocus borrowing policy to On-going MoF prioritize high quality projects with a clear economic rationale funded by high concessional loans Expenditure Introduce a focus on value for money in expenditure across public sector budgets and performance FY15 budget MoF, line management frameworks, with an emphasis on the fastest growing expenditure programs ministries Increase oversight over public agency payroll by: Short-term PSC/ MoF  Establishing personnel reporting requirements for public agencies to MoF and PSC to facilitate monitoring of the whole of government wage bill.  Instituting a requirement that MoF and PSC clear large recruitment and remuneration decisions proposed by public agencies. Expenditure controls Public Limit the creation of new public agencies to limit the fragmentation of budget and personnel controls Immediate GoS agencies and reduce the extent of duplication in administrative costs Seek savings in transfers to agencies that do not affect front-line service delivery FY15 budget MoF, line ministries Strengthen MoF’s technical capacity to assess medium term expenditure plans and efficiency of public On-going MoF agencies Personnel Carry out a functional review of human resources across government, to seek opportunities where Short- medium PSC certain functions could be amalgamated or centralized to save costs term Health Seek to bring payroll expansion (both recruitment and remuneration levels) in the health sector under FY15 budget MoF, MoH, central control by setting and enforcing fiscal targets for personnel expenditure PSC Begin rationalizing Ministry of Health and National Health Service administration budgets to reduce FY15 budget MoH, MoF duplication of functions where possible Education Seek to bring employment expansion in the education sector under control by setting fiscal targets for FY15 budget MoF, MESC, personnel expenditure PSC Review human resource requirements in education and establish medium-term and annual recruitment FY15 budget MESC, MoF targets in line with defined goals and fiscal resources Review potential to bring coordinate all primary and secondary school development projects under Short- medium MESC central oversight of MoE, to facili term 1 AGGREGATE FISCAL TRENDS Summary Samoa ran a relatively well managed budget from FY06 to FY12: Expenditure growth was driven by recurrent and capital spending in similar proportions, and as the government invested in a broad range of areas including frontline service delivery personnel, post- disaster reconstruction and other construction projects. As the budget expanded, the sectoral composition of total spending remained balanced between social, economic and administrative sectors. Nevertheless, a rapid expansion of spending over this period and increases in debt levels placed the sustainability of the budget at risk: The wage bill was a major driver of the growing recurrent budget, especially in the health and education sectors. Transfers to non-central government agencies also expanded markedly as a larger share of government functions have been assigned to a series of new public agencies and as transfers to help fund public utility operations. Overall, the share of the budget executed by public agencies increased from 15 to 40 percent of recurrent expenditure since FY06. Although grant funding buoyed much of Samoa’s grow ing budget, loan financing has also been growing rapidly, causing a rising debt-to-GDP ratio. Tsunami and economic recovery related activities contributed to this trend. However, 57 percent of loan financed expenditure between FY09 and FY12 went towards large-scale non-economic infrastructure projects such as a number of large government buildings. Cyclone Evan exacerbated a fragile fiscal situation. The challenge looking ahead will be to consolidate government finances and reduce fiscal deficits in the long term to bring debt back to sustainable levels: The medium term fiscal framework targets a deficit of 3.5 percent of GDP after grants. An even lower deficit will be needed in the long-term to bring Samoa back on track vis-à-vis its target debt ratio. Achieving this requires that budget policy is underpinned by expenditure restraint and an emphasis on savings and value for money in spending. Particular scrutiny of public agencies is needed to ensure that their budgets are managed with prudence and efficiency. Tightening debt management should also be a priority looking ahead, including better prioritization of loan funded projects. 1.1 INTRODUCTION Samoa’s economy has been faced with slow growth and a series of economic and natural shocks in recent years. The country faced two major natural disasters in recent years with the damage caused by a tsunami in 2009 and the effects of Cyclone Evan in December 2012. Samoa is also vulnerable to external economic shocks stemming from the high level of pass though of global prices and its heavy reliance on imports. It was therefore significantly affected by the fuel and commodity price shocks in 2007 and 2008 and was exposed to lower demand, especially for the tourism sector, as the effects of the Global Economic Crisis extended to the economies in the Pacific. Although Samoa has used monetary policy measures in managing the economy, 1 economic policy at times of crisis or during downturns relies heavily on fiscal measures and the effectiveness of budget planning and execution. The structure of the budget has also evolved rapidly in recent years at the same time as expenditure levels expanded rapidly. Over the past years, the budget has been adjusting as it coped with natural and economic shocks, absorbed large amounts of external funding as well as some significant changes in the functions delivered by non-central government agencies. These factors have somewhat obscured the core trends that have been driving the growing recurrent and capital budgets. This note will look backwards to take stock of the factors that accounted for fiscal expansion since FY06, and look ahead to highlight the impact of cyclone Evan on the fiscal position from 2013 onwards. The first part of the note will be backward looking. It will review trends in public expenditure, revenues, grants, financing and budget execution to provide insight into which aspects resource allocation have changed the most in real terms from 2006 to 2012, and decompose trends to identify the main drivers of growth in the budget. The second part of the note is forward looking. It focuses on the impact of cyclone Evan on Samoa’s fiscal position from 2013 onwards, and presents projections of the fiscal path in long-term in light of higher deficit and debt levels. The analysis indicates that bringing the budget back on track to a sustainable footing through expenditure restraint and tighter control of debt accumulation are key policy measures looking ahead. Large external shocks, a growing budget and slow revenues have changed the fiscal landscape and placed a large strain on Samoa’s fiscal sustainability. Improving the outlook will necessarily rely on prudence in expenditure and an emphasis on value for money in both central government and public agency budgets. The emphasis on restraint in expenditure management is due in part to the potentially limited scope for an expansion in revenue given the relatively high current levels of collection. Box 1: Aggregate fiscal trends analysis data and sources This note is based on detailed analysis of disaggregated expenditure data from the FMIS provided by the Ministry of Finance covering the seven year period between FY06 and FY12. The first step was to compile a full and consistently classified dataset covering the period under review. To do this, data from the Treasury Fund account was supplemented with project grant and loan related data from MoF accounts to compile a full database of domestically and externally funded expenditure. The full data set was then sorted and reclassified to conform to economic and functional classifications, and to standardize classification across the period1. The second step involved adapting the expenditure data to structural changes in the execution of the budget as several functions and budgets that were part of central government portfolios were moved to non-central government agencies as transfers. Three main adjustments took place: the winding up of the Samoa Pacific games authority in FY08, the 1 The data was also adjusted to exclude lumpy tax expenditures recorded in ministry ledgers to better distinguish the underlying expenditure trends. 2 establishment of the Land Transport Authority (LTA) in FY08 and the National Health Service (NHS) in FY10. In order to distinguish the underlying expenditure trends from these reorganizations, the expenditure trends presented below were adjusted by reassigning transfers to the National Health Service, the Land Transport Authority and the Samoa Pacific Games Authority back to the main expenditure categories. Lastly, all of the trends discussed in this note are presented in real terms unless otherwise stated. 1.2 FISCAL TRENDS FROM 2006 TO 2012 The composition of expenditure has been stable since FY06 with two thirds of the budget funding recurrent expenditure and the remaining third being composed mostly of donor funded capital spending. The recurrent budget represents approximately two thirds of total expenditure. Transfers to non-central government agencies grew significantly in terms of their share in the budget and were the largest recurrent expenditure category in FY12, followed by the central government wage bill. The capital budget account for the remaining third of the budget and is dominated by the donor funded programs. The domestically funded capital budget has typically formed a small share of the budget. Figure 4: Composition of expenditure (FY12) Figure 5: Budget shares (% of total) in FY06 & FY12 Fiscal trends between 2006 and 2012 are characterized by increasing expenditure and modest revenue growth. Over the past seven years, government expenditure has increased markedly. Samoa’s total expenditure increased by 10 percent of GDP between FY06 and FY12 (Figure 6). Expenditure growth was driven by recurrent and capital spending in similar proportions as the government increased resources to a broad range of areas including frontline service delivery personnel, post-disaster reconstruction and other construction projects. The recurrent budget accounted for 50 percent of the expansion, 10 percent of the expansion was realized in the domestic capital budget whilst the remaining 40 percent of the increase came through foreign capital expenditure (Figure 7). At the same time, however, revenue receipts were relatively flat, growing by just 1 percent of GDP over the period. 3 The resulting funding gap has been largely financed by loans and grants from development partners, with a growing share of loan funding. Samoa’s deficit before grants grew from 4 to 13 percent of GDP since FY06, having peaked at 20 percent of GDP in the post-tsunami period. The deficit has been largely funded by donor financed loans and grants. Donor grants and loans increased from 8 to 17 percent of GDP between FY06 and FY12 and peaked at 29 percent of GDP in FY10 when the fiscal deficit before grants was also at the highest point. The spike in external resource flows helped to support the costs of the tsunami recovery operations, but unspent balances led to an increase in net domestic assets2. Loan financing has also been growing fast, particularly as large capital project and budget support loans had been contracted, and have exceeded Samoa’s debt target. Despite the concessional terms on which most of Samoa’s debt has been taken, the most recent debt sustainability analysis (DSA 2013) concluded that Samoa’s risk of debt distress has shifted from moderate to high as the debt burden breached debt sustainability thresholds3. Figure 6: Expenditure & Revenue (% GDP) Figure 7: Decomposition of expenditure growth (FY06-FY12) Figure 8: External debt (% GDP) 2 There may be a buildup in cash balances for project related accounts and programs. The team does not have access to this data. 3 According to the 2012 review of the Debt Sustainability Framework for Low Income Countries, the thresholds for “strong performers� are: 50 percent for PV of debt to GDP; 200 percent for PV of debt to exports; 300 percent for PV of debt to revenues; 25 percent debt service to exports; and 22 percent debt service to revenues. 4 1.2.1 REVENUE Unlike some of the other Pacific Island countries that receive significant revenues from fisheries or other natural resources, most of Samoa’s domestic revenues come from tax receipts. In FY12, tax receipts stood at 22 percent of GDP, which is a relatively strong rate of collection compared to similar neighboring economies. Approximately half of Samoa’s tax revenues come from taxes on goods and services, generated mostly through a 15 percent consumption tax. The remaining half is almost evenly split between income and trade taxes. Non-tax revenues are the equivalent of 3 percent of GDP. Revenues grew modestly between FY06 and FY12, growing by just 1 percent of GDP over the period. Receipts grew by 10 percent in real terms as consumption and income tax receipts increased, whilst trade related taxes contracted slightly (Figure 10). The completion of the South Pacific games and the associated drop in imports, as well as some new import duty relief measures that were introduced for raw materials in FY09 contributed to this trend. In general, revenue trends could be described as relatively flat when framed against the growing budget. Recurrent expenditure alone increased almost three times faster than domestic revenue since FY06. It is likely that a continuation of this trend would deepen Samoa’s reliance on external budget funding such as general and sector budget support to finance the majority of capital spending and possibly some recurrent expenditure in the future. Samoa’s relatively robust tax effort suggests that the scope for significant level of revenue growth, without increasing the tax burden, may be limited. Figure 9 indicates Samoa’s relatively strong tax effort in relation to the size of the economy. Recent revenue strengthening efforts have likely contributed to this trend, including tax modernization reforms and recent reviews of major tax laws. Nevertheless, although there is a basis for pursuing higher revenue receipts to support the expanded expenditure program, the scope for significant revenue expansion may be limited in the absence of policies and reforms that promote economic growth, including growth enhancing tax policy measures. Figure 9 : Tax revenue/ GDP in PICs (2011/2012) Figure 10: % change in dom. revenue (FY06-12) 5 1.2.2 THE RECURRENT BUDGET Growth in the government’s wage bill accounted for 45 percent of the increase in the recurrent budget since FY06. The wage bill, which accounts for 43 percent of recurrent spending in FY124 (10 percent of GDP), grew by an average of 6 percent a year in real terms over the period. Much of this trend reflects growth in the wage bills of the health and education sectors as additional frontline service delivery personnel were recruited to schools and clinics. The trend also reflects increases in the salary levels for civil servants, especially at the higher level grades and for the health sector. Despite the upward trend in the wage bill, the proportion of resources that are devoted to wages is not particularly high compared to regional peers or the average for lower-middle income countries. However, the greater proportion of public sector activity being performed by public agencies has led to the increased decentralization of the wage bill and fragmentation of payroll policies (for example, as reflected through the higher pay increases in the NHS staff compared to other core public service delivery personnel). This narrows government’s overall oversight and control of the public sector wage bill5. Transfers from central government to public agencies were the second largest growth area after the wage bill, accounting for 40 percent of the increase in the recurrent budget. When all public agencies are taken in to account, the volume of transfers from central government increased by 275 percent in real terms between FY06 and FY12, marking a hefty increase in the proportion of the budget executed by non-central government agencies. These transfers reflect a range of agency expenditure, including current and operational costs as well as maintenance and capital investment. As previously mentioned, a large proportion of the increase in transfers came from the redirecting of resources from central government ministries to two government agencies between FY10 and FY12: the NHS and the LTA6. Another notable change was caused by the winding down of the Samoa South Pacific games authority, causing a dip in transfers in FY09. After adjusting for these changes, the increase in transfers amounts to 87 percent in real terms. This trend is driven by a number of newly established public agencies and growing community social obligation payments to public utility companies. The establishment of a number of small public agencies such as the Samoa Qualifications Authority, Samoa Sports Facilities Authority and the Samoa Fire Services Authority has expanded the number of public agencies. Identifying the extent to which these new agencies were an expansionary factor, or whether the functions executed by the new agencies were already being budgeted and delivered through central government agencies requires further investigation. Nevertheless, it is likely that 4 Including employee compensation for the NHS and LTA. 5 See the wage bill PER note for more detailed analysis of payroll trends. 6 Transfers to the Samoa National Health Services were redirected from the Ministry of Health and the National Health Service; transfers to the Land Transport Authority came through the Ministry of Works. The Land Transport Authority (LTA) is a Public Trading Body established on 1 December 2008. It brings together the Road Asset Management and Road Use Management functions previously carried out by the Ministry of Works, Transport and Infrastructure and the Transport Control Board. 6 the additional fixed costs associated with the establishment and operations of these new agencies, including personnel, have contributed to the fixed costs of government. Similarly, CSO payments to public utilities (mainly the electric power corporation and the water authority) have increased since FY06 as operational and investment costs were subsidized by central government. As a result, public trading bodies7 continue to rely on central government funding as they received 40 percent of transfers between FY10 and FY12. Government programs to invest in expanding the infrastructure for public goods such as local roads, electricity or water supply lines have contributed to increasing the size of CSO payments. The extent to which efficiency gaps are also driving the growing CSO transfers merits further analysis given that efficiency improving measures would support the financial position of public enterprises and potentially reduce the need to subsidize their operations from the general budget. Figure 11: Transfers to public agencies (Tala millions - Figure 12: Transfers by public agency (% total in constant) FY10-12) Non-salary recurrent spending accounted for the remaining 15 percent of the recurrent budget expansion as general government consumption and interest payments increased. Expenditure on goods and services, which accounts for more than a third of recurrent spending in FY12, reflected a broad expansion across the government’s general expenditure items including general operational costs such as travel, land lease and utility bills. In addition, interest payment more than doubled in real terms since FY06 as the government’s debt maintenance obligations increased with the expansion of its liabilities, and may be set to rise further as obligations towards recently contracted loans become due. Although Tsunami related projects and grants doubled between FY09 and FY10, these expenditures were largely wound down by FY12. 7 Public beneficial bodies such as the National Health Service are non-commercial government agencies and hence, they receive an annual budget allocation to fund their operations. In contrast, public trading bodies such as the utility companies and Land Transport Authority are required to function as viable commercial entities under the Public Bodies Act (2001), and may receive community social obligation (CSO) grants from central government when they provide loss making services to the communities such as building access networks in areas with small populations. 7 1.2.3 THE CAPITAL BUDGET Capital spending grew faster than recurrent spending and gained in share up from 36 to 40 percent of total expenditure. The capital budget demonstrated robust growth as the Government of Samoa invested in a range of areas over this period to strengthen services and infrastructure, to implement post-tsunami recovery operations and to complete the construction of government buildings. Overall, 50 percent of growth in total spending over the period was realized in capital spending (10 percent domestic and 40 percent externally funded). By FY12, this trend had begun to reverse as several reconstruction works and major government construction projects had been completed. The capital budget was fuelled by external grants and loans, which doubled in real terms between FY06 and FY128 . The growing spending program over the past years would not have been feasible in the absence of grant and loans flows from development partners, especially since FY09. Although grant funding buoyed much of Samoa’s budget, loan financing has also been growing rapidly. Loan financing increased from 3 to 8 percent of GDP, having peaked at 14 percent of GDP in FY10. As a result of these trends, the external debt ratio crept up to56 percent of GDP by end FY12. A large proportion of the increase in loan financed expenditure went towards the construction of government buildings on partly-concessional terms. Even though Tsunami and economic recovery related activities drove spending levels upwards, 57 percent of loan financed expenditure between FY09 and FY12 went towards large-scale capital projects for the construction of four government buildings: the TATTE public office building, the convention center, a parliamentary complex and the national medical center HQ. Most of the external loans during the period were contracted on concessional terms, making the debt stock and the associated repayments more manageable. However, some of the largest loans (such as the government building loans) were only partly on concessional terms9. 8 The source for the data in this section is from the annual budget estimate documents. The team used the data reported for the previous year in each budget document to reflect estimates of actual grant and loan expenditure. This was the most comprehensive source of data for grants and loans to which the team had access. 9 The degree of concessionality of a loan is measured by its “grant element�. In countries with higher debt vulnerabilities, the concessionality threshold (minimum grant element) is at least 35 percent. This threshold is applied to each loan separately. Although the Exim Bank loans have an element of concessionality, this is somewhat less than 35 percent, in contrast to loans from other major creditors such as the Asian Development Bank and the World Bank whose loans to Samoa are extended on far more concessional terms. 8 Figure 13: Aid by type (% of total) Figure 14: Loans by purpose (Tala millions) 1.2.4 SECTORAL EXPENDITURE PRIORITIES10 Expenditure allocations to all of the main sectors increased in real terms, with the economic and infrastructure sectors accounting for most of the growth. Total expenditure remained almost uniformly split between infrastructure, social sectors and the administration and justice sector as public spending levels expanded. This reflects an increase in allocations to all of the main sectors. The economic and infrastructure sectors accounted for 46 percent of the increase in total expenditure between FY06 and FY12. Spending envelopes in the social sectors accounted for 27 percent of the increase in spending, with all the growth being concentrated in health and education. Although it represents a small share of the budget, women, community and social development was the only social sector where spending levels declined. Allocations to the administrative and justice sector increased at a similar rate to social sectors, taking up the remaining 27 percent of the fiscal expansion since FY06, reflecting recent investments in government buildings and higher interest payments. Figure 15: Expenditure by sector (% of budget) Figure 16: % change in expenditure by sector (FY06 -FY12) 9 1.2.5 BUDGET EXECUTION11 Budget execution at the aggregate level has been strong despite some large external shocks that affected public finances during this period. The ability to implement budgeted expenditure is an important factor in enabling the government to deliver its expenditure plan as well as public services. Samoa’s annual expenditure outturns (based on the original budget) show deviations that do not exceed ten percent on aggregate since FY06, despite some large external shocks that affected public finances during this period. Revenue execution has also been robust. Revenue outturns, based on the original budget, have averaged 96 percent on average between Figure 17: Revenue outturns (actual/ original budget) FY06 and FY12. The lowest outturns were recorded in FY08 and FY12 when actual revenue was more than ten percent lower than budgeted as taxes relating to trade and goods came in under the budgeted amounts in those years. In contrast, expenditure outturns for those same years were very strong. Revenue outturns do not seem to have been affected during the FY01-FY11 crisis years with the global down turn and the Tsunami, since the realized receipts were as high as 99 and 97 percent of the budgeted amounts. Taken together, these trends suggest that in general, the expenditure execution trends are not strongly linked to revenue outturns. The largest expenditure deviations occurred in FY10 and FY11 as public spending was stepped-up in the post-Tsunami period. The tsunami period resulted in the deterioration of budget execution, as expenditure exceeded planned levels. A milder spike in outturns during the period occurred in FY07, as Samoa hosted the South Pacific Games. However, throughout these events, budget deviations did not exceed ten percent. At the ministry level, the largest deviations were in the Ministry of Works, Transport and Infrastructure as it absorbed most of the recovery and reconstruction allocations. The Ministries of Natural Resources and Environment and Finance also experienced relatively large overspends related to clean-up operations and the administration of tsunami-recovery grants. 11 The outturn analysis is based on a comparison between the original budget and actual expenditure outcomes. The analysis excludes project loan and grant spending. 10 Figure 18: Expenditure deviations (original budget Figure 19: Expenditure deviations by ministry vs. actual) (FY06 –FY12) Tsunami peak South pacific games Tsunami wind-down Budget execution at a disaggregated level (by ministry and spending category) has also been robust when the ‘unusual’ budget years associated with the Tsunami and the South Pacific Games are taken into account. The credibility of the budget as a statement of policy is undermined when deviations from the original budget at a sub-aggregate level (composition of expenditure) are large. This is typically a result of under or overspending and the reallocation of resources between ministries or budget categories within the year, causing deviations from planned expenditure allocations. As previously mentioned, Samoa’s aggregate deviations have shown some variability between FY06 and FY12, but have remained below 5 percent of budgeted expenditure during normal years (i.e. excluding Tsunami and South Pacific Games years). In contrast, deviations within the overall envelope have been more significant. Deviations at the ministry level amounted to 6 percent of budgeted expenditure on average whilst deviations by budget category reached 9 percent of budgeted expenditure. The budgetary response to the Tsunami and preparations for the South Pacific Games (SPG) have been driving these deviations to a large extent. This can be seen in Figure 20 where the years associated with these events are separated from normal years. Deviations between ministries and budget categories are higher during the Tsunami and SPG years. In normal years, deviations between ministries fall to below 5 percent. Deviations between budget categories are also relatively low in normal years, amounting to 6.5 percent of budgeted expenditure. Figure 21 provides a breakdown of the movements underlying this figure. It shows that the movements by budget category in normal years are typically driven by under spending in the salary budgets and overspending in transfers and capital budgets. Budget reallocation controls at the output level provide flexibility for departments to redirect resources within the year. The disaggregated deviation analysis discussed so far has focused on outturns at the ministry and budget category levels in order to provide an overview of budget execution at the institutional and economic levels. It is relevant to note here that the PFM Act (2001) provides for the control for transfers between outputs and sub-outputs rather than between budget categories. It requires approval from the CEO Finance for transfers that would 11 increase the output allocation by 20 percent or more. Budget deviation analysis was carried for the Ministry of Education, Sports and Culture at the output level as an indicative exercise. It shows that the level of deviations between outputs for education stand at 8.4 percent of the original budget on average between FY06 and FY12, which is comparable in size to the level of deviations observed at the ministry and budget category levels. In general, an output by definition can include several economic categories such as salaries and capital. This, taken together with what is effectively the scope to reallocate one fifth of the departmental budget without external approval, suggests that the controls in place potentially provide generous scope for movement in the budget in all categories. Figure 20: Disaggregated expenditure deviations (FY06- Figure 21: Breakdown of deviations by budget 12) category in normal years Samoa uses a supplementary budget annually to adjust the original budget estimates upwards, but even after the in-year revision, the budget still displays similar levels of divergence from actual expenditure outcomes as the original unadjusted budget. The revised budgets have adjusted original expenditure estimates upwards for each year between FY07 and FY12. The magnitude of the upwards adjustment has typically been modest (less than ten percent of original budget), with the exception of FY10 when the original budget was adjusted upwards by 19 percent. Moreover, the revised budget is generally well executed at the aggregate level with a less than one percent difference between actual and revised budgets in FY06-12 on average. In general, the supplementary budget process provides flexibility for the budget to adjust or respond to shocks within the year in a context of risk and volatility. Nevertheless, the outturn analysis suggests that the revised budget is not more predictable than the original budget in relation to the composition of expenditure. It shows that the level of deviations by ministry and budget category is similar in to that of the original budget in both normal and Tsunami/ SPG years. In other words, even though the budget has been revised within the year, it is still displays similar levels of divergence from actual expenditure outcomes as the original unadjusted budget (Figure 23). Overall, the extent to which budgets deviate within the year is still relatively low being below 10 percent on average over the period, and being much lower when the Tsunami and SPG years are not taken in to account. Nevertheless, these results suggest that there may be scope to strengthen the planning process underlying both the original 12 budget formulation and the supplementary budget. For instance, one trend has been a regular underspend on the salaries budget, including in key sectors such as education. The salary budget was under spent by 7 percent on average against the original budgeted amount since FY06. The execution rate against the revised budget is slightly better with an under spend of 3 percent. This is a preliminary observation and there are likely to be other factors underlying the budget execution trends that could be systematically identified through further analysis to support improved budget execution. Figure 22: Size of revised Budget Adjustment (% of Figure 23: Comparison between revised and original budget) original budget deviations (FY06-12 average) 1.3 POST CYCLONE EVAN: 2013 ONWARDS The costs of recovery from cyclone Evan are estimated to be as high as 30 percent of GDP, triggering a new episode of fiscal expansion as the recovery efforts get underway. New expenditure commitments arising from the 2012 Cyclone that left large amounts of public and private property damaged mean that expenditure will need to remain heightened for the next few years as recovery and reconstruction work proceeds. The cyclone recovery plan estimated the cost of recovery operations at Tala 477 million (30 percent of GDP) over a three year period. To support the funding for these recovery efforts, development partners have proposed additional grants and concessional loans that are mostly frontloaded to the first year of the plan. The costs and financing needs of this crisis are high, and have disrupted Samoa’s post Tsunami fiscal wind down. Hence, the FY14 budgets marked a 20 percent increase on the FY12 budget as reconstruction operations got underway. The budget projects a wind down back to the target deficit of 3.5 percent of GDP or less after grants in the medium term – it is critical to achieve this goal. One of the key challenges confronting Samoa’s policy makers in the post cyclone Evan period is to restore fiscal space to be able to respond to future exogenous shocks, and re-establishing fiscal sustainability over the medium term without compromising the delivery of core services. To ensure fiscal sustainability over the medium term, Samoa will have to return to the kind of deficit profile it managed in the 13 recent past. The Government’s annual Fiscal Strategy Statement sets out a fiscal target for a deficit (after grants) of no more than 3.5 percent of GDP. Getting to this point will require a post Cyclone Evan wind down. Significant fiscal consolidation would be achieved if Samoa is able to wind back its recovery and reconstruction spending. However, an even lower deficit target Figure 24: Aggregate fiscal projections FY13-FY33 will be needed in the long-term to bring Samoa back on track vis-à-vis its target debt ratio from the current levels. Given the high levels of debt and continued low economic growth rates, it is likely that some further fiscal space would need to be identified to support a return to sustainable debt levels. The Government’s annual Fiscal Strategy Statement sets out a fiscal target that for net public external debt to be less than 50 percent of GDP. In order to achieve this target in a reasonable timeframe, the GoS would need to run budget deficits that are below its target of 3.5 percent of GDP as soon as feasible. Gradually reducing the budget deficit to 1 percent of GDP by FY23 would achieve a smooth reduction in debt ratios, which could be lowered below 50 percent of GDP only by FY30. Under this scenario, debt servicing would increase from around 2 percent of GDP now to a peak of 3.7 percent of GDP12. This indicates the significant duration that will be required to return public finances to a healthy state. It is likely to take 17 years of prudent fiscal management and a moderate fiscal consolidation to bring the debt position back to a sustainable footing13. Hence, stabilizing debt levels and then lowering them to 50% GDP will still be a long progress, requiring sustained commitments from successive governments. 12 These expenditure estimates in this scenario are based on Government’s own budget, both its recurrent and development spendin g. They do not include budget support or donor funding in the long term. 13 Setting sectoral spending depends on the Government’s s pending policy priorities as set out in their annual budgets and forward estimates. These scenarios are simply meant as indications of what degree of real fiscal space may be available to channel to priority sectors under various debt repayment regimes. 14 Even with tighter fiscal policy, there could be sufficient space for moderate increases in real resources to priority sectors while lowering debt levels. Following a spike of post- Cyclone reconstruction spending in Figure 25: Projection of fiscal space for priority sectors FY13 to FY15, domestic expenditure could be maintained at 29.3 percent of GDP14 over the period. With real GDP growth assumed to proceed at 2.5 percent15 per annum, and population growth estimated to range between 0.5 and 0.8 percent per year, real resources on a per capita basis could grow moderately. Total discretionary government spending would increase on a constant price, constant population basis by 1.44 percent per annum, which would mean that in 20 years’ time, the level of real resources per capita would be 33 percent higher than it is now. If we project justice, security and administrative spending to remain constant as a proportion of GDP, and assume that economic and infrastructure spending returns to its trend level of around 8 percent of GDP (excluding years of extraordinary capital investment), resources to social sector could be increased further. Total social sector spending could grow by 1.83 percent per year in constant price and population terms, which equates to nearly 44 percent increased resources over the 20 year periods. Even modest additional commitments of donor support would accelerate sectoral spending and service delivery improvements. The scenario presented above looks at government expenditure alone, and not how donor-financed development project spending will evolve – it is assumed to remain constant as a proportion of GDP. An alternative scenario in which development partners support the government’s development program serves to illustrate what effect this can have on fiscal space. If partners were to jointly provide an equivalent of US$ 20 million in today’s prices starting in FY15, and this was channeled to priority sectors, it could raise annual real spending growth from 1.83 percent to 2.50 percent over the next 20 years. Such resource growth over both health and education could be sufficient to make substantial progress in improved service delivery and outcomes or could further accelerate efforts to reduce the existing stock of public debt. However overall, this scenario implies a very tight balance for the budget looking ahead given Samoa’s exposure to adverse shocks, including major natural disasters, on an ongoing basis, which will periodically throw it off track. Samoa is and will continue to be highly vulnerable to both natural disasters and exogenous economic shocks, all of which have 14 Close to the FY12 pre-crisis level of domestic expenditure. 15 Average growth over the last 10 years was 2.1 percent. 15 been painfully in evidence over the last decade. Recent estimates indicate that Samoa can expect to incur annualized damages worth 1.7 percent of GDP as a result of natural disasters, and much of the burden of rebuilding will fall on the government. In light of this, the government’s plans to reduce the fiscal deficit and bring down debt levels are of critical importance in providing a buffer to respond to shocks adequately. However, the pressures on the limited resources of the budget will, at times, be overbearing, and in these cases it is likely that GoS will need the support of development partners in the recovery effort. Potential upward pressures on the budget In both the health and education sectors, significant budget pressures are likey to emerge in the medium term. The employment guarantee extended to nursing graduates and increasing student enrollments in nursing will continue expanding senior staff numbers and the health wagebill for years to come. Equally, government plans to expand secondary school enrollment will potentially increase budget costs. These trends raise important concerns for the affordability of key initiatives in the two largest sectors and may have direct implications for the sustainability of public finances. Pressure to keep salaries in line with prices may mount as most public sector employees saw a decline in salaries in real terms. Over the period, there have been two major pay awards that increased nominal pay levels across government and established some higher pay grades. Despite these adjustments, real wages have actually declined since 2006 for most sectors of the government. These trends reflect that cost of living adjustments and grade progression are not administered automatically and may drive demand for a future adjustment in the medium term. The resource needs of the education sector are likely to be augmented as the secondary school fee relief scheme gets underway. The expected increase in secondary enrollment rates will push student-teacher ratios back-up and may trigger the recruitment of additional teaching staff to maintain target class sizes. Additional resources will also be needed as donor funding of the scheme is eventually phased out to maintain school financing in the absence of fees. Hence, a medium to long term projection will highlight the sector’s resource needs in the coming years and the extent to which they are affordable within the limits of the government’s total budget framework. Increasing cost of NCD treatment is pushing out other spending that might have greater benefits. Tertiary treatment for non-communicable diseases which comes at a very high cost per patient, has continued to rise. The costs of the overseas medical treatment are high and are likely to increase as more patients are likely to seek treatment for NCDs. In line with the sector priorities, the additional resources might save more lives in the long run if focused on primordial prevention. 16 The employment guarantee for graduate nurses and the staffing gaps for junior nurses may potentially add pressures to further expand the health wage bill. The enrollment of students into nursing degrees at the National University of Samoa has increased in recent years. Since almost all graduating nurses are employed by the NHS, this implies a continued expansion is staffing levels. The Health Sector Plan notes that while staffing levels of senior nurses are now close to international standards, gaps persist at lower levels, such as for auxiliary nurses. Hence, plans to increase numbers of auxiliary nurses whilst also taking on more senior graduate nurses would imply a continued rapid expansion of payroll. 1.4 CONCLUDING REMARKS This note has provided an overview of the key trends in Samoa’s budget between FY06 and FY12 at an aggregate level, and the main factors that account for the fiscal expansion over this period. It also presented an outlook of the fiscal position from FY13 onwards taking the budgetary burden posed by cyclone Evan into account. This analysis shows how the expansion and debt accumulation over the past years, limited capacity for revenue growth and the recent cyclone’s recovery costs are placing Samoa under a significant level of fiscal strain. Given these factors, the main areas of focus for budget policy identified in this note are as follows:  Pursuing expenditure restraint, particularly in the recurrent budget and transfers to public agencies given the tight fiscal context in the post cyclone Evan period. This note highlighted the extent to which total expenditure, bolstered by donor grants and loans, has been growing in recent years. An important trend in this regard is the growth in recurrent expenditure which grew three times as fast as domestic revenue and the rapid expansion of transfers to non-central government agencies. These allocations allowed the government to invest in expanding key services including in the health and education sectors (as discussed in the health and education PER notes). However, the current budget context given recent fiscal pressures relating to cyclone Evan recovery costs and the higher debt burden emphasize the advantages of restraint and in expenditure growth. Budget strategy would benefit from being underpinned by policies that help to realize savings and that pursue value for money in spending.  Tightening debt management and improving prioritization of loan funded projects. Samoa’s most recent debt sustainability assessment concluded that Samoa’s risk of debt distress has shifted from moderate to high as the debt burden breached debt sustainability thresholds. The rapid build-up of loan stocks since FY09 contributed to this outcome despite the concessional terms on which most of Samoa’s debt has been taken. Further stress on the financing position has accompanied the cyclone Evan recovery program, which is estimated to cost as much as 30 percent of GDP (Tala 477 million over a three year period). To supplement the government of Samoa’s funding for these operations, 17 development partners have provided grants, including budget support, which was mostly frontloaded to the first year of the plan. Nevertheless, prudent debt management including medium to long term fiscal path planning will be essential for keeping debt sustainability in view. A healthy debt position will also be important if exchange rate imbalances emerged as a policy issue. Samoa has been subject to exchange rate pressures. In recent years, the Tala has been appreciating in real terms, with an accelerated real exchange rate appreciation since FY0916. This trend increases the likelihood of overvaluation and an associated loss of competitiveness in some key sectors. Hence, looking ahead, healthy fiscal ratios would place Samoa in a better position in case there is a need to address exchange rate imbalance in the future. The GoS has made much progress in this area by updating the medium term debt strategy with support from the World Bank. These reforms would be further strengthened by improving and prioritizing loan funded projects. Improving the quality and economic basis for public investments is critical. This note highlighted a concentration in investments on government buildings between FY09 and FY12 that have been a large driver Samoa’s current debt position. These investments emphasize the scope for strengthening the prioritization of loan funded projects to place a larger emphasis on investments with a clear economic rationale to underpin debt sustainability in the long term.  Increasing oversight of expenditure and efficiency of public agencies. The share of the budget executed by non-central government agencies has been growing notably since FY06 as a number of large and small agencies were established. This trend has had the effect of fragmenting the budget, including the public sector wage bill, and increasing costs associated with administrative functions. Hence, further strengthening the technical capacity to assess medium term expenditure plans and efficiency of public agencies would enhance the coherence of the national budget framework. A review of human resource management policies is also recommended to assess payroll policies of public agencies (especially large agencies such as the NHS) against those of central government. Finally, it is advisable that the creation of new agencies is only considered where it is necessary to enhance the efficiency and effectiveness of public service delivery. 16 IMF Rapid Credit Facility Staff Report, 2013. 18 GOVERNMENT PERSONNEL COSTS Summary Spending on personnel has increased markedly over the period both as a percent of GDP and as a proportion of an expanding government budget. While personnel spending growth across most of government has been moderate, the health and education sectors have been the largest contributors to higher payroll over the period. Two thirds of extra spending has gone to the health and education sectors with health increasing especially rapidly. Despite an upward trend in the wage bill, the proportion of resources that are devoted to wages is not particularly high compared to regional peers or the average for lower-middle income countries. Across most of government, increased personnel spending has financed an expansion of public sector employment, whereas salaries have barely kept in line with prices. Public sector employment has grown by 22 percent over the period, while real salaries excluding the health sector have declined slightly. An exception to this trend is the health sector, where remuneration levels have grown significantly. Changes in personnel policy in the health sector have had a marked bearing on the government’s payroll. Here, both average salaries and employment have grown rapidly, with real wages increases of 3.9 percent each year and total employment growth of 30 percent. There is a high level of reliance on overtime in the health sector, which seems to have increased over the period. This rate of increase in both average wages and employment raises questions about the sustainability and affordability of the reform processes that are underway in the health sector. Pay awards over the period have been at least partly targeted towards increasing incentives for higher-level staff. There has been some decompression of the pay scales, and staff in highly skilled and senior managerial roles have seen above average pay rises. Expansionary pressures may be building-up for government personnel costs. Key trends emphasized in this note, being rapid growth in health personnel costs and slow salary growth on average for central government employees imply a risk of accumulating pressures to expand the wage bill at a time of fiscal stress. 2.1 OVERALL TRENDS Figure 26 depicts gross personnel costs, which includes wages and salary and other payments to the employee, such as allowances17. Figure 27 shows the trend in personnel costs as a percent of total recurrent expenditure and as a percent of GDP over time. This analysis includes the NHS and LTA, which became state-owned enterprises during the period, but is otherwise confined to general government. 17 Pension fund contributions are on average 5.5 percent of gross pay and PAYE taxes 11.3 percent. 19 Figure 26: Total personnel costs, in real and Figure 27: Personnel costs as proportion of nominal terms recurrent expenditure and GDP Personnel costs have risen rapidly over the period, growing as a proportion of an expanding budget. Total personnel costs have risen by 69 percent over the period in nominal terms. Taking account of inflation, they have risen by 42 percent, equivalent to 6 percent a year on top of inflation. Personnel costs have increased from 7.4 to 9.7 percent of GDP. Payroll growth has proceeded slightly faster than the total recurrent budget, with payroll as a share of recurrent budget increasing from 37.2 to 38.8 percent. Over most of the period, personnel costs have been increasing in a steady manner. As can be seen in Figure 26, other than in FY08, growth has been steady. The short-lived increase in personnel costs in FY08 was caused by the holding of the South Pacific Games in Samoa, during which time additional staff resources were needed. The health and education sectors have accounted for the majority of the increase in personnel costs. As Figure 29 shows, payroll in the health sector made up one third of overall real payroll growth. Payroll expenditure from the Ministry of Education, Sports and Culture spiked in FY08 with the hosting of the South Pacific Games, and wages and salaries for the South Pacific Games Authority was handled by this ministry. It has grown at about the same rate as the overall wage bill, and due to its large size, made up 23 percent of overall real wage growth (Figure 29). Wages in the law & order sector have remained fairly stable as a proportion of total payroll as annual growth has been roughly in line with growth of the overall payroll. The proportion of the wage bill going to these three priority areas has increased from 55 percent of 58 percent. Of this, education has actually declined slightly, from 27.7 percent to 26.2 percent of GDP whereas health has increased rapidly from 17.0 percent to 21.5 percent of GDP. Payroll resources to core accountability functions such as the Attorney General’s Office, the Ombudsman’s Office and the Audit Office have grown rapidly, but from a low base. Payroll of line ministries such as Ministry of Agriculture, Ministry of Women, Communities and Social Development and the Electoral Voter’s Office have registered negative or negligible real wage growth (Figure 30). 20 Figure 28: Shares of personnel costs Figure 29: Growth in personnel costs by function (FY06 – FY12) Figure 30: Growth in personnel costs by ministry (FY06 – FY12) Figure 31: Share of overall growth in personnel costs by ministry (FY06 – FY12) 21 Box 2: Personnel costs analysis data and sources This analysis makes use of the public expenditure data from the Mininstry of Finance held Finance One financial management information system, supplemented with data from the payroll system. Some adjustments have been made to the standard presentation of expenditure data to ensure consistency of treatment across time for the purposes of this analysis. Two large public services, the National Health Service and the Land Transport Authority, became state-owned enterprises in FY10 and therefore no longer appear in the MoF expenditure and payroll accounts. Data on these public agencies has been collected seperately and included, so the scope of the analysis includes these services throughout. In some cases, additional data has been collected from official government sources including the annual Budget Statement, the World Bank’s World Development Indicators are used for international comparisons and outcome data and other documents. 2.2 INTERNATIONAL COMPARISONS Despite increases over the period, Samoa’s wage bill does not seem to be high compared to other countries in the region, and comparable to the average for lower-middle income countries. The wage bill as a proportion of current spending is around 39 percent, just lower than Tonga and Fiji and just above the LMIC average. Samoa’s wage bill as a proportion of GDP is just below 10 percent, placing it below Pacific island countries at a similar level of income such as Tonga and Vanuatu, but somewhat above the LMIC average. The comparison countries have been selected as those in the region with similar or larger populations (equivalent wage to GDP ratios for smaller countries like Kiribati and Tuvalu are much larger). Figure 32: International comparison, wage bill as Figure 33: International comparison, wage bill as proportion of recurrent expenditure proportion of GDP 22 2.3 STAFFING LEVELS AND AVERAGE REMUNERATION18 Across government extra resources have mostly gone to increases in staff numbers, except in health, where both staff numbers and average remuneration have risen substantially. Figure 34 presents central government employment, including the NHS and LTA. Figure 35 illustrates the growth in the two components of real payroll costs for the major health and education sectors, and the rest of government. The numbers of employees recorded on the payroll has increased steadily over the period, from 6,500 to 7,700 – an increase of 20 percent. Over one third of these additional employees were hired in FY11 alone. The balance between employment growth and salary growth has been quite different in some parts of the public sector. In the health sector, there has been roughly equal growth in both staff numbers and average remuneration. In contrast, growth in the education sector has been concentrated on staff numbers, with real wages declining. In the rest of government, there has been no change in real wages and only a slight increase in employment. The stark difference between personnel policy over the period in the NHS compared to the rest of government raises some questions as to whether aggregate fiscal policy has to some extent been undermined by sectoral policy. At a time of fiscal stress, the rapid expansion of payroll has been somewhat at odds with the government’s efforts to maintain aggregate fiscal discipline. As an independent corporation, the NHS is outside the control of PSC, and as such is not required to have new positions approved by the PSC, nor to adhere to the public service pay scales. However, as a body that is almost wholly funded by government grants, budget proposals are subject to the same budgetary scrutiny as are all public funds. In addition to the Budget Office in the Ministry of Finance reviewing all budget proposals, the State-Owned Enterprise Unit also plays an important role in liaising with public agencies and vetting budget proposals. Given the much greater independence that the NHS now enjoys, it is all the more important for the government to make effective use of the existing accountability mechanisms that are in place, to help ensure that public funds are fully aligned with the government’s overall fiscal policy and wider policy priorities. Being by far the largest public beneficial body, and one of the most important service delivery organizations in Samoa, the NHS may warrant a particular focus and dedicated resource in state owned agency monitoring department (SOEMD) in the Ministry of Finance. 18 Average salaries here are deflated by CPI to illustrate real spending power for individuals. All other data is deflated using the GDP deflator consistent with rest of the public expenditure analysis, which illustrates constant resources over the whole economic, including government spending and investment. In comparison, CPI inflation has been higher over the period that the GDP deflator. In addition, hourly waged workers are excluded from the analysis. As a result, the trends in this section are not strictly comparable to the aggregate payroll results. 23 Figure 34: Government employment Figure 35: Growth in payroll components over FY06-FY12 Average salaries in most of government have only just kept pace with inflation, and lack of wage growth may have contributed to recent wage pressures. Figure 36 illustrates the trend in constant-price average salaries over the period. The average annual remuneration for salaried workers (excluding those on hourly contracts) outside the education and health sectors was almost unchanged over the period. In the education sector, the average salary actually declined by almost 4 percent, although this might reflect high levels of recruitment of new teachers that bring the average down. The peak in education salaries in FY08 is related to additional payments for the South Pacific Games. On the other hand, there was a large increase in average salaries in the health sector. At 26 percent growth, it was equivalent to annual real pay rises of 3.9 percent a year. In the FY14 Supplementary Budget delivered in December 2013, it was announced that the government would implement a universal 3 percent cost of living adjustment, in response to growing pressure from public sector workers who have not seen real wage rises for an extended period. Across most of government, allowances and additional payments do not make up a large proportion of total pay but overtime payments are high and may have increased over the period for the health sector. According to payroll records for central government, basic pay makes up 90 percent of overall remuneration with overtime, allowances and leave payments constituting the remainder (Figure 38). However, although only incomplete data is available, there are some indications that overtime payments are high and increasing in the health sector. In the three years FY07 to FY09 that the NHS accounts were part of the government’s consolidated ledger, the proportion of overtime payments showed a rapid change, increased from 15 percent of base salary in FY07 to 22 percent in FY08 and further to 31 percent in FY09. Other sources 19 indicate that in FY12, health sector workers earned on average 34 percent more than their basic 19 NHS Workforce Development Plan, 2013 24 pay in overtime and other allowances, with medical staff earning 58 percent more and household staff 76 percent more. Pay awards seem to have been at least partly targeted towards increasing incentives for higher-level staff. Over the period, there have been two pay awards. The first was a universal multi-year pay award that amounted to a 42 percent nominal increase phased in over the three years from FY06 to FY08. At the same time, an additional higher ‘special’ pay grade was created to provide greater compensation levels to skilled professionals and Assistant Chief Executive Officer level staff. The second pay award, which was differentiated by level, was instituted in FY11. It amounted to a 9.7 percent nominal increase for standard grades and the first two special grades, and 2 percent for the highest special grade, contract workers and CEO levels. Figure 36: Average annual salary (constant prices) Figure 37: Salary to GDP per capita ratio Figure 38: Personnel costs by type of remuneration Figure 39: Salary scale range 25 Approximate benchmarks and Figure 40: Government employee Lorenz curves, FY06 and decompression ratios suggest that FY12 public service pay is competitive and incentives for higher-level staff are increasing. The average public sector wage was around 2.5 times GDP per capita in FY06 (Figure 37). This ratio has trended up and in FY12 stood at 2.8 times GDP per capita. Compression ratios for standard civil service grades (highest grade salary over lowest) have remained stable at 4.8 over the period. However, the salary scale including more highly paid ‘special grades’ has decompressed with the addition of a third level in FY08, as shown in Figure 39. The changes implemented over the period FY06 to FY08 increased decompression (including special grades), so that the compression ratio increased from 6.9 to 8.8. The second pay award led to a reduction of the compression ratio to 8.2, still substantially higher than at the beginning of the period. The highest pay grade of CEO level has remained high relative to the lowest standard grade over the period, moving from a multiple of 17.8 in FY06 to 15.8 in FY12. This overall decompression can be seen in the actual distribution of payroll in Figure 40. 2.4 CONCLUDING REMARKS This note has provided a brief overview of trends in government payroll over the last seven years. Real resources dedicated to payroll have increased significantly over the period. Most of this increase is down to an expansion on the quantity margin - the number of government employees has increased by over 20 percent. Most of these additional workers have gone to the education and health sectors. The relatively rapid and steady increase in staff numbers is therefore likely to be linked to policy objectives in these two key priority areas. The expansion comes with direct implications for fiscal commitments both now and into the future, with fiscal space becoming even more constrained as a result. Of immediate concern is whether this trend will continue, since a budget already facing severe fiscal stress would be hard-pushed to be able to accommodate further expansion, unless accompanied by additional resources such as higher revenues or additional long-term donor commitments to fund new positions. 26 Although employment growth has occurred across much of government, the health sector has experienced the most rapid growth, whilst also bucking the trend in the rest of government and realizing a large increase in the average real salary. While the exact reason for the rise in the average salary is not clear from the data, it seems likely that there has been an increased use of overtime payments over the period. It is notable that the stark difference in personnel policy outcomes for the health sector have occurred over the same period that the NHS has become independent of the PSC. Given the greater independence that the corporate body now enjoys, it will be all the more pressing for central government to make full use of the accountability and controls mechanisms that remain in place to ensure that health sector policy is not inconsistent with overall fiscal and personnel policy, and that fragmentation of responsibility does not undermine the strong control and coherency of government fiscal policy. A final point to note is related to the evolution of the pay scales and pay settlements over the period. The Samoan pay-scale exhibits relatively high levels of decompression, with the highest levels of senior staff earning up to 16 times that of the lowest standard grade. While a treatment of the incentives of the remuneration system is outside the scope of this note, it is notable that this decompression has been increased over the period, whilst average real salaries have been unchanged. This would tend to imply that salaries for high level staff have risen, while those at lower levels have seen real declines in the purchasing power of their incomes. These trends are quite likely to lead to pressure to increase wages, at least to keep up with inflation for all grades, and this does seem to have been the case in light of the cost of living adjustment that was announced in December 2013. 27 HEALTH Summary Since FY06, the health sector has seen major reforms and planning processes take place. These changes have been assosciated with a significant increase in the cost of some functions of the health service. The establishment of the National Health Service entailed substantial set-up and ongoing costs. At the same time the Ministry of Health has scaled up it bureaucratic and regulatory functions. While important for the reforms underway, these changes have led to a sharp increase in non-service delivery (or adminstrative) costs in the sector. There have been large increases in resources to nursing and integrated community healthcare services, although impacts on health outcomes are still unclear at present. Based on the information available, the build-up in primary healthcare is consistent with some of the strategic priorities set out in the Health Sector Plan 2008-2018. These changes have happened recently, and in the main, improvements in health outcomes remain to be seen. Rapid domestic spending growth has been dominated by personnel expenditure, particularly on relatively senior grade nurses. While good, trained staff are undoubedly crucial for a well-run health service, it is not clear that the build-up at particular levels has sufficiently addressed binding constraints. Expansion in some areas seems to have highlighted continued shortages at other staff levels that require addressing, while increased student enrollments in nursing will continue expanding senior staff numbers for years to come. These trends raise important concerns for the affordability of the public healthcare workforce which has direct implications for the sustainability of public finances. Health prevention has been stagnant. While primary healthcare has expanded rapidly, as per the sector plan, health prevention has not seen increases in domestic resources, and has fallen as a share of total health spend. As identified in the sector plan monitoring framework, an increased share to prevention is likely to be required to tackle major concerns such as the non-communicable disease epidemic. Establishing separate administrative and management systems are costly and should only be done if absolutely essential. Government tertiary health expenditure has not increased overall, with operating and maintainence budgets flat as personnel and NCD treatment costs continue to rise. Tertiary treatment for non- communicable diseases via the Overseas Medical Treatment programme and the National Kidney Centre comes at a very high cost per patient and has risen over the period. Hence, the resources dedicated to tertiary care continue to dwarf all other areas. Resources to hospitals and diagnostic facilities have fallen, which may be appropriate if more care is now being provided in the community. However, with the completion of the new national hospital, the NHS is likely to require additional capital and operating resources to adequately service and maintain the new facilities. 3.1 OVERVIEW OF HEALTH SYSTEM AND OUTCOMES IN SAMOA The Samoan health system comprises public, traditional and private sectors although publically funded health services make up the majority of the sector. The Ministry of Health is responsible for regulatory oversight of the health sector and provides guidance on the policy framework and health priorities of Samoa. The ministry is also responsible for: i) Monitoring overall health system performance; ii) disease surveillance; and iii) basic health promotion and prevention 28 services, including sanitation and regulation services. The National Health Service is the main publically funded provider of clinical health services to the population and includes the national referral hospital in Apia and seven district hospitals throughout the country. Outreach services are provided by Nursing and Integrated Community Health Services. These services include home-based intermediate care for patients who still need nursing and midwifery care when discharged from hospital. They can also include disease prevention and health promotion activities, such as immunizations and maternal and child health services, offered in health centers. The health centers are located within the local communities that they serve. District hospitals offer 24-hour services and serve as clinical centers in rural districts. They are staffed and managed by a multi-purpose team of nurses who are responsible for in-patient, out-patient and outreach services. The district hospitals are supported as necessary by doctors from the Upolu and Savaii main hospitals. Other important service providers under the regulatory oversight of the MoH include the National Kidney Foundation, a Government service provider, and a range of health-related NGOs20. Health outcomes in Samoa are generally good. Figures 41 to 43 describe a selection of health- related outcomes in Samoa. Samoa is on-track to meet most of its health-related Millennium Development Goals, although contraceptive use and HIV/Aids risk are a concern. Outcomes such as life expectancy at birth and infant mortality rates are better than both the Pacific Island average and the global developing country average. Almost 100 percent of births are attended by a skilled health professional, although maternal mortality at childbirth is higher than the LMIC average. An important factor in Samoa’s relatively good health outcomes is almost universal access to water and sewerage systems. 96 percent of the population has access to an improved water source21 and 98 percent to improved sanitation. Access to electricity is also high (though high costs remain an impediment). These levels of access to basic infrastructure are well above the Pacific average or that of lower middle-income countries worldwide. Figure 41: Health outcomes in Samoa 20 Information drawn from Samoa Demographic and Health Survey, 2009. 21 However, it is understood that the rate of access of potable (drinking) water is considerably lower. 29 Figure 42: Reproductive health outcomes 22 Figure 43: Access to basic infrastructure services Box 3: Health analysis data and sources The data used for this analysis is from a number of different sources. Prior to FY10, data for both the Ministry of Health (MoH) and the National Health Service (NHS) are taken from FinanceOne. From FY10 onwards, NHS accounts have been used for NHS expenditure and combined with MoH expenditure from FinanceOne. To our knowledge, this is the first time that unifed data for the public healthcare sector has been presented since the creation of the NHS. Payroll data similarly comes from different sources before and after the corporatization of the NHS. Less information is available from FY10 on numbers of staff employed by the NHS, and data from the annual budget has been combined with estimates of contract workers. 3.2 OVERALL TRENDS IN HEALTH EXPENDITURE The Health Sector Plan 2008-2018 made the case for additional resources, and this has occurred over the period. Government healthcare expenditure, excluding donor projects, has been on an increasing trend over the period FY06 to FY12 (Figure 44). Over these seven years, domestic resources have increased by 34 percent in real terms. Including donor-financed projects and capital expenditure, total resources increased by 52 percent. Public healthcare spending per capita (including donor projects) has grown from 300 to 440 Tala in constant terms over the period. From Figure 44 we can see that donor financing was a relatively modest proportion of the total in the early years, averaging 16 percent of expenditure between FY06 and FY09, but increased rapidly in the latter three years. This growth was driven partly by donor support to the new Sector Wide Approach framework, but mostly by the cost of constructing a new national medical centre in Apia. 22 Comparisons for 2010, latest year available for all countries. 30 Most of the growth in current spending over the period has been financed by domestic resources. Figure 45 shows that excluding hospital capital expenditure (CAPEX), domestic resources have accounted for two-thirds of real resource growth to the sector, and donor funding, one third. Figure 44: Healthcare expenditure Figure 45: Cumulative growth in healthcare expenditure Tertiary care continued to represent the largest share of the budget, but investments in primary care and spending on administrative functions have been growing. Starting from a low base, primary care has grown by over 50 percent since FY07. At the beginning of the period, almost the entire healthcare budget was classified as tertiary care spending, administered from the central and district hospitals reflecting the role district hospitals played in providing some level of primary care. Over time, that has changed as the government invested in shifting to a community level primary care based model. Nevertheless, tertiary expenditure has been maintained in real terms and is still by far the largest component of health spending. In FY12, 206 Tala was spent on tertiary care per capita compared to 80 Tala per person on primary care. At the same time, spending on administrative functions has nearly doubled. Almost all the increase in domestic resources (80 percent) over the period has gone on increased personnel spending. Personnel spending in primary and preventative care has continually increased, while from FY07 onwards, tertiary personnel costs have also increased significantly. Operating costs and domestic capital expenditure have remained stable, the latter being almost zero. In practice, capital spending needs are provided for by development partner programs, so it is unsurprising that domestic capital expenditure is near zero. 31 Figure 46: Health expenditure by function Figure 47: Donor project expenditure, by function Figure 48: Domestic expenditure by economic Figure 49: Domestic expenditure on personnel category Expenditure growth in the health sector has been roughly in line with the growth of budget as a whole. Figure 50 shows that over the three years from FY08 to FY10, expenditure growth in the health sector kept pace with growth in overall government spending, such that the proportion of total budget stayed constant at 12 percent. This followed a fall in FY07 when education spending has increased sharply whilst health spending remained stable. The proportion of domestic expenditures then recovered to 14 percent over the last two years as health spending growth accelerated. As a proportion of GDP, healthcare spending has increased from 4 percent to 6 percent. These ratios strip out the lumpy expenditure on hospitals in the most recent years. Growth over the period has brought public healthcare spending per capita up to a level comparable to some countries in the region, although higher than the EAP average. International comparisons of health expenditure are fraught with difficulty and so should be treated with caution. For example, differences in the mix of public-private provision of healthcare may drive differences in public health spending. There is also not provision made for the differences in health outcomes achieved in different countries. However, for some countries 32 of similar size in the region, such as Tonga and to a lesser extent, Fiji and Kiribati, health spend per capita in Samoa is at the higher end of the range, slightly lower than in Tonga. Figure 50: Public healthcare expenditure as a Figure 51: Public health sector expenditure, regional percent of GDP and total expenditure 23 comparisons The scale-up of domestic resources to primary healthcare seems to reflect the GoS priorities for reform in the sector. Under the 2008-2018 sector plan, priorities include the development of community healthcare capacity, increased program activities at the district level and increased availability of skilled professionals (see Box 2), although there is little additional information on the details of these policies or their expected costs available to the team. Similarly, although it is difficult to compare on a like-for-like basis, donor-financed project expenditure seems to have been mostly consistent with government priorities. Capital expenditure on the national referral hospital complex was identified as a priority in the sector plan, and dominates expenditure in FY11 and FY12. Other donor projects vary from year to year, but there has been a general increase in resources to a government-led sector-wide approach which has been designed to support GoS priorities. 23 Comparisons for latest year available (FY12 for Samoa, FY11 for other countries. 33 Box 4: Health sector plan 2008-2018 – strategic areas and priority actions The Government of Samoa’s current health sector plan came into effect shortly after the beginning of the period. The plan sets out six strategies to achieve the overall goal of a healthy Samoa. These are elaborated with a set of policy outputs that would help to achieve each strategic priority, along with monitoring indicators. The six strategic areas are presented below, accompanied by selected output indicators that relate to the analysis of public expenditure in this note. 1. Strengthen health promotion and primordial prevention - Increase share of resources to health promotion and primordial prevention. 2. Enhance quality health care service delivery including management of infectious diseases - Improve facilities at the National Referral Hospital and physical infrastructure - Implement professional and service standards - Develop and implement appropriate career paths for health workers - Development of community health capacity, program activities and standards in districts 3. Strengthen governance, human resources and health systems - Increased availability of appropriately qualified and skilled health workforce - Increase in qualified specialists - Professional credentialling for healh service providers for all professions - Increase numbers of midwives to meet demand - Two year review of Human Resources for Health Plan 2008-2015 4. Partnership commitment - Strengthen health systems through processes between MoH and health sector partners 5. Financing health services - Increase resources to health care & ensure health care needs for poor and vulnerable are met, and they are not disadvantaged 6. Donor assistance - Increased donor participation in health sector - Increased harmonisation of donor assistance with GoS prioritized areas Source: Health Sector Plan 2008-2018, Government of Samoa 3.3 EXPANSION IN PERSONNEL COSTS DRIVEN BY NURSING SERVICES Growth in primary services was driven by nursing and integrated community services, which made up nearly two-thirds of overall health sector growth. Figure 53 shows that the vast majority (83 percent) of the increased primary spending on nursing and community services went to personnel24. The growth in personnel spending is potentially consistent with the priorities of the sector plan to focus on attracting and retaining skilled professionals, providing in-service career paths and boosting primary care. During the period, the health service underwent a drive to improve the skills and training of public health professionals and to recruit new skilled staff. The ambition seems to be to reach WHO best practice staffing ratios, although the recent Workforce Development Plan notes that reaching these levels would be prohibitively expensive 24 The increase in resources to nursing and community services is particularly sharp in FY07, and part of this seems to be explained by reclassification of services and staff that were previously grouped as part of central hospital services becoming decentralized. 34 at this stage. These reforms, which are expected to improve standards of service and coverage, led to a significant increase in salary costs. Figure 52: Domestic expenditure growth by output Figure 53: Domestic nursing & community services (FY06 – FY12) expenditure There are risks that rapid increases in staffing are not addressing shortages at all levels, and further recruitment may not be affordable. Although we do not have sufficient information available to discern which grades or professions spending has gone on exactly, the NHS 2013 Workforce Development Plan notes that there has been an increase in staffing levels of Senior Nurses. The enrollment of students into nursing degrees at the National University of Samoa has increased from about 30 in recent years to 90 in 2012. Since almost all graduating nurses are guaranteed employment by the NHS, this implies a continued expansion in the coming years as students graduate. The Plan notes that while staffing levels of Senior Nurses are now close to international standards, gaps persist at lower levels, such as for Auxiliary Nurses. Plans to increase numbers of Auxiliary Nurses whilst also taking on more Senior Nurses would imply a continued rapid expansion of payroll. The average salary in the health sector has increased by 25 percent in real terms over the 7 years. This compares a 4 percent decline in real wages in the education sector and zero growth in the rest of government. Since the NHS became an SOE, it is no longer obliged to use the PSC public sector pay-scales, so compensation rates may have diverged. However in practice there has reportedly not been any pay settlements specifically for the NHS over the period (although there was one that was subsequently reversed). There are two possible explanations for the increase in average wages. First, as employment growth was concentrated at more senior levels, the average wage is likely to have moved up somewhat to reflect the changed mix of grades in the NHS. Secondly, evidence suggests that there is high and increasing use of overtime, such that average (and total) wages are now 34 percent higher than the basic salary. 35 Especially in view of the limited availability of data, it is too early to say whether improvements in outcomes have been observed following these reforms. While shortages of well-supported, professional nurses were identified in the sector plan, so too were shortages of support staff. Sector reviews to date indicate mixed progress against service delivery and health outcome metrics. 3.4 ADMINISTRATIVE COSTS HAVE GROWN RAPIDLY Administrative costs have increased rapidly in the public health sector over the period, so much so that they are second only to nursing in the amount of increased resources they have taken up. This increase is a product of extra costs from running separate agencies, such as the NHS and National Kidney Foundation, as well as increased resources devoted to strategic development planning and regulatory functions in the Ministry of Health. Increased administration costs are linked to reforms underway, and may be effective, but have clearly come at a cost. Additional Figure 54: Administrative costs breakdown administrative spending reflects a number of factors. First, the establishment of the NHS entails the setting up of separate management and administrative functions (for example, the NHS has a different FMIS system to the rest of government and therefore requires its own support team). Figure 54 shows that specified admin costs relate to increases in spending in the Ministry of Health on activities such as strategic planning and advice, credentialing and other regulatory functions, but that the majority of growth in admin costs are general costs not assigned to any specific output. Both the MoH and NHS now have units to provide advice to the minister so the cost of this function has grown by over 50 percent, while strategy and performance management units have been scaled up and a registrar of health professionals established in the MoH. All these changes have been implemented in the context of the Health Sector Plan and may bring improvements in health outcomes over time. Nevertheless, it is worth emphasizing that the costs are substantial and in the context of a very tight fiscal environment, it will be all the more important for government to monitor the progress of reforms closely to ensure that they are bringing the intended results. 36 3.5 PREVENTATIVE CARE SPENDING STAGNANT In contrast to the rest of primary spending, domestic preventative health care spending25 (Figure 55) has not grown in real terms over the period, and fallen as a share of health spending from 3.7 to 2.8 percent. The lack of extra government resources may to some extent reflect the fact that donor projects are addressing preventative health care priorities. Although categorization of donor projects can be difficult, Figure 56 aims to reflect donor projects related to this area over the period, which have generally increased. Not all of these resources may have been spent directly on health promotion and prevention activities and might also have gone to admin, infrastructure and other primary healthcare. The spike in FY10 reflects high utilization of funds for a diabetes-screening program and Ausaid’s ‘enhanced primary healthcare’ program. Some areas of preventative care have lagged behind other countries although the situation is improving. Because of changes to prevention programs, gaps emerged and measles and DPT vaccination rates dropped drastically in the early 2000s to below the EAP and global developing country average (Figure 57 and Figure 58). However, an increased focus in recent years has yielded steady progress in improving vaccination rates, as shown by increased spending on vaccination programs (Figure 56), and DPT vaccination rate are now on par with the EAP developing average. To meet the stated policy goals and to increase the effectiveness of overall health spending, this is an area that would require significant focus in the coming years. Figure 55: Preventative health care expenditure Figure 56: Vaccination expenditure (inc. donor programs) 25 The category of preventative health care in this section includes health promotion programs. 37 Figure 57: Vaccination rate, measles Figure 58: Vaccination rate, diphtheria-pertussis- tetanus (DPT) 3.6 INCREASED NCDS COSTS AND TERTIARY PERSONNEL SPEND Although volatile from year to year, domestic tertiary spending has been roughly stable over the period. However, the costs of the Overseas Medical Treatment program and the Kidney Foundation have continued to increase in real terms. Figure 60 shows that from FY07 onwards, personnel costs have also increased in real terms. The same trend seen in the primary level of an up-scaling in tertiary personnel resources can also be seen at the tertiary level. At the same time, domestic resources to the two national hospitals and diagnostics have declined in real terms, although some of this is due to the reallocation of personnel from hospitals to community nursing in FY07. As with preventative care, the trends in domestic hospital spending reflects complementary trends in donor project funding, of which a large amount has been focused on tertiary care. As Figure 60 shows, including the two major donor projects supporting tertiary functions, total spending has increased in real terms, particularly from FY07 onwards. However, with declining budgets and growing personnel costs, there is likely to be increasing pressure on operating and capital costs to adequately run and maintain equipment, especially with the completion of the new National Medical Centre. 38 Figure 59: Domestic tertiary expenditure Figure 60: Total tertiary expenditure 3.7 CONCLUDING REMARKS It is clear from the preceding analysis that there have been major shifts in both the bureaucratric structure and resource prioritisation to the health sector in Samoa. These reforms reflect a level of government commitment to a critical sector that is to be applauded. The challenge for the health sector now is to ensure that the reforms and increased resources are translated into a better healthcare system, and better health outcomes on the ground. The reforms to the health sector have entailed significant increases in resources to bureaucratic functions, partly as a result of the establishment of a new and separate organisation, the NHS, and partly due to a scale up in planning and management functions in the MoH. In a highly constrained fiscal environment, it is all the more important that the government scrutinises all expenditure to ensure they are as effective as possible. Personnel costs in the sector have increased very rapidly, in a way that has decoupled from trends in the rest of government. This major scale up raises a number of questions. While people are undoubtedly the bedrock of healthcare services, the government is simply unlikely to be able to afford the levels of staff that it may be targeting under its present resource envelope. In order to maintain overall fiscal discipline, tough decisions have to be made on rationing scarce resources both on the part of the agencies and the MoF. If the rate of spending seen in the NHS continues in the present fiscal environment, it is hard to see how this will be accomodated without accumlating additional debt or pushing out other critical expenditure. It is also important to ensure that personnel growth is led by demand. Some evidence has suggested that recruitment has been concentrated in a small number of professions at a certain level, and while filling these gaps, people constraints in other areas have emerged. 39 The focus on scaling up of primary care services is an important achievement of GoS, and an effective primary care system would help reduce the strain on tertiary services. However, primordial preventative care is another area in dire need of attitional funds, and this has not seen any major increase in resources over the period. Going forward into the next phase of their health sector plan, supporting health prevention, whilst making best use of existing resources, is likely to become a priority. 40 EDUCATION Summary Education is both the single largest expenditure item in the domestic budget and the sector that has experienced the largest increase in expenditure since FY06. The education sector was largely protected during the tsunami period, and its funding remained steady whilst total expenditure was being wound-down subsequently. Most of the increase in primary and secondary education spending went towards the teaching wage bill as size of the teaching workforce was expanded. Teacher salaries were also increased on average, although they did not fully keep up with prices, and have hence declined on average in real terms. Transfers to the tertiary sector did not increase in real terms, whilst sports and culture funding increased by 20 percent. The education sector invested in increasing the number of teachers even though student-teacher ratios were close to national target levels. Hence, the average ratios for both primary and secondary schools have decreased to well below the target levels by 2012. Although the education sector has invested the majority of its growing resources towards direct service delivery, it is not clear that the additional investments, especially the reduction of class sizes, have led to improving outcomes at the primary level as indicated by the lower progression rates from primary to secondary and weaker scores in some key subjects. Based on this analysis, some key questions emerge that would benefit from further investigation , including the value for money in one of the sector’s key program for improving learning outcomes: smaller classrooms, when compared with other interventions particularly those related to teacher incentives. 4.1 SAMOA’S EDUCATION SECTOR Primary and secondary education in Samoa is provided mostly through government, with a significant role played by mission, village based and to a lesser extent, private schools. Primary education continues to be compulsory until the end of the year the child turns fourteen26. It consists of eight years of schooling provided by a total of 168 schools (26 of which are mission/ private schools). Secondary education consists of 5 years of schooling provided by 40 schools (16 of which are mission/ private schools). Overall, the government budget directly funds 80 percent of primary and secondary schools. The government also partially funds mission and private schools through grants. In contrast, early childhood education (ECE) is exclusively provided by fee-charging mission and community27 schools with grant funding from the Ministry of Education, Sports and Culture (MoESC). Similarly, university level education is provided by fee charging tertiary institutions (mainly the National University of Samoa) that are also partially funded through operational grants from the ministry budget. 26 MOESC statistical digest, 2012. 27 Early childhood education centers are typically owned and run by pastor’s wives, school boards of the local community and affiliated to NGOs. 41 Education is one of the central priorities of the Strategy for the Development of Samoa (SDS). The SDS explicitly acknowledges that improved education underpins progress in all other areas of development including nutrition, livelihood opportunities and law and order. In the area of education, the strategy focuses on improving the quality of teaching services as well as the learning environment. It also aims to increase and broaden access to education, ranging from ECE to vocational and tertiary training. Samoa also has an Education Sector Plan that lays out high-level goals and an explicit recognition of the need to educate citizens who can be productively engaged in the economy. It identifies as set of objectives including: a) enhanced quality of education at all levels; b) enhanced educational access and opportunities at all levels; c) enhanced relevance of education and training at all levels; d) eliminating gender disparities; and e) improving early childhood education and adult literacy. Almost 90 percent of the MoESC budget Figure 61: Composition of MOESC spending – FY06 has been allocated to direct service to FY12 average (share of total) delivery at the primary, secondary and tertiary levels. MoESC data does not allow for a concise break down of expenditure by the four main levels of education: early childhood, primary, secondary and tertiary. Nevertheless, the data does show that the largest share of MoESC spending is allocated to the provision of primary and secondary education, followed by the tertiary sector. Expenditure levels on vocational education and early childhood development28 are estimated to present a small proportion of total MoESC expenditure. Samoa is one of the best performing countries in the Pacific region in terms of enrollment rates and female participation. Samoa’s primary and secondary education enrollment rates tend to be higher than the average in neighboring Pacific countries. Girls are not disadvantaged in terms of access to education, and have increasing participation rates when compared to boys as they progress from primary through to tertiary education29. Moreover, a regional comparison of Samoa’s education budget indicates that it delivers these results with relatively fewer resources allocated to education (as a share of the budget) compared with the average for neighboring countries. 28 SABER country report: Early Childhood Development, 2013 29 Relatively low levels of male participation in tertiary education have been identified as a concern to the Samoan authorities. 42 Figure 62: Selected education indicators compared Figure 63: Education spend as a share of total budget 30 31 to PICs compared to PIC average Enrollment Girls’ participation Box 5: Education analysis data and sources The data used for the expenditure analysis in this note is from the government’s FinanceOne database. The data was adjusted for the spike in trends relating to the South Pacific Games in FY07 and for lumpy tax expenditures to establish the underlying structural expenditre trends. The outcome data are mostly based on the Ministry of Education’s statistical database. Changes are presented in real terms taking FY06 as the base unless otherwise stated. 4.2 WHAT WERE THE MAIN EXPENDITURE TRENDS SINCE 2006? Education is the single largest expenditure item in the domestic budget each year, and it is the sector that has experienced the largest increase in expenditure since FY06. The education budget increased on average by 6 percent in real terms each year between FY06 and FY12, going up from 5.3 to 6.7 percent of GDP over the period. Expenditure on education also increased in relation to the population. Education expenditure per person increased from Tala 368 in FY06 to Tala 461 by FY12. These trends represent a notable increase in resource allocation to education over the past seven years. Donor project related expenditure played an important role in supplementing the domestic education budget as it amounted to 26 percent of total education expenditure on average over the period. These initiatives have been mostly supporting sector wide approaches for investment in education infrastructure and well as teaching and curriculum development. The project portfolio also includes the funding to support the primary school fee relief scheme. 30 PICs average includes Fiji, Kiribati, Solomon Islands, Tonga, Tuvalu, Vanuatu, RMI, PNG and FSM. Source: South Pacific Community - NMDI database. 31 PIC average includes Fiji, Tonga, Vanuatu, Solomon Islands, Kiribati and Tuvalu. 43 Figure 64: Education budget (2006 prices) Figure 65: Education expenditure per person (2006 prices) Two thirds of the increase in domestic education spending went towards teacher salaries in primary and secondary education. Approximately two thirds (63 percent) of the increase in MoESC spending was directed towards primary and secondary education (Figure 66)32. Most of this increase went towards a growing teacher wage bill (Figure 67). Teacher salaries increased in nominal terms as the sector was part of a major salary restructuring for the whole public service between FY07 and FY09, and a cost of living adjustment for the public service in FY11. These managed pay increases helped to increase teacher salaries, but not enough to keep in line with prices over the period33, meaning that teacher salaries decreased on average in real terms since 200634. The growing education wage bill was also driven by increases in the number of teachers employed. Approximately 300 additional school teachers (188 teachers for primary schools and 116 for secondary) were recruited since 2006, bringing both primary and secondary student- teacher ratios down below the national target levels35. Non-salary related expenditure also increased as additional resources were directed to school infrastructure and operational costs associated with the school rehabilitation and the primary school fee removal programs. Funding for sports and culture accounted for the remaining third of the education budget, whereas government funding for the tertiary education sector did not increase over the period. Overall, MoESC funding to the tertiary sector did not increase in real terms36. MoESC provides grants to partly fund tertiary institutions, mainly the National University of Samoa and the University of the South Pacific. An increasing allocation to the National University of Samoa was balanced by reduced funding to the University of the South Pacific between FY06 and FY12 resulting overall in a flat tertiary budget. Since the universities are both fee charging institutions, 32 MOESC expenditure data classifications do not distinguish between primary and secondary education related expenditure making it difficult to separate the trends for these two sub sectors. In addition, the data does not distinguish allocations for early childhood education. 33 Salary increases have been measured against the GDP deflator to maintain consistency with rest of the public expenditure analysis. CPI inflation has been slightly higher over the period than the GDP deflator. Hence, teacher salaries would not have kept up fully with prices by the end of the period even when measured against the CPI. 34 See the PER note on the wage bill for more detail on government remuneration expenditures. 35 Only one new government school was opened since 2006, a primary school in Apia Urban. 36 The MOESC budget data does not directly distinguish expenditures for skills training or vocational education. 44 grants from the ministry account only for a part of the tertiary sector’s financing. Hence, an increase in tuition fees and growing student numbers indicate that the overall financing envelope for tertiary education may have increased even if MoESC grants remained constant37. The ministry also funds the Samoa Qualifications Authority, which was established in FY07 as part of the tertiary sector. In contrast, spending on sports and culture accounted for 29 percent of the increase in the education budget between FY06 and FY12 as funding for the Samoa Rugby Union and the Samoa Sports Facility Authority commenced in FY08. The Samoa Rugby Union manages Samoa’s World Cup Squad development and campaign while the Samoa Sports Facility Authority was established following hosting of the South Pacific Games, effectively merging the South Pacific Games Authority and the Apia Park Board into one entity that looks after all the sporting facilities in Samoa. In total, three new authorities were established under MOESC since 2006: the Samoa Qualifications Authority, the Samoa Rugby Union and the Sport Facility Authority. Figure 66: Decomposition of expenditure growth by Figure 67: Decomposition of expenditure growth by sub-sector(domestic) economic category (domestic) The education sector was largely protected during the tsunami period, and its funding remained steady whilst total expenditure was being wound-down subsequently. The 2009 tsunami had a large effect on Samoa’s fiscal framework with a significant expansion in FY10 and the wind down in the two subsequent years to FY12. Education expenditure increased in the immediate post tsunami period, going up by 7 percent in FY10 as some of the recovery effort related to damage in schools and school related infrastructure. In addition to government resources, several projects contributed to the recovery effort in the sector, helping to quickly re- establish local services. As a result, the education sector budget grew at the lower rate of 5 percent in FY11. Subsequently, the education budget remained constant in real terms as the wind down trend continued for Samoa’s budget in FY12. However, overall, the expansion in the 37 For instance, the University of Samoa approved a phased fee increase in 2006 and has also recorded growing student numbers (Samoa Review of NTR Report, May 2013). The PER team does not have further information on the resources directly generated by tertiary institutions that are form part of the funding envelope for the tertiary sector. 45 education budget has been somewhat slower than the total budget. Hence, education’s share in total spending declined slightly from 20 to 18 percent of the total between FY06 and FY12. Figure 68: % change in education compared to total Figure 69: Education budget growth compared to budget total budget 4.3 HOW DO THESE EXPENDITURE TRENDS RELATE TO SECTOR OUTCOMES? Much of the expansion in the education budget was dedicated to direct service delivery by increasing the number of teachers. As previously mentioned, a large proportion of the growing education budget was allocated to increasing the number of teachers in primary and secondary schools. As a result, the share of direct primary and secondary service delivery expenditure 38 increased a little from 68 percent to 70 percent of the total education expenditure. The education sector invested in increasing the number of teachers even though enrollment rates have been growing at a modest pace. Gross primary school enrollment rates stood at 105 percent in 2006. Since then, the remaining percentage of children not enrolled in primary education was reduced slightly as gross enrollment rates increased to 106 percent in 2012. However, even though gross enrollment rates increased a little, the actual number of primary students enrolled in 2012 was steady at the 2006 level at around 39 thousand. Secondary gross enrollment rates also increased slightly from 76 percent in 2006 to 77 percent in 2012. Hence overall, the education sector improved the enrollment levels albeit at a modest rate. The primary school fee relief scheme has already taken effect on enrollment rates. It is likely that the planned elimination of secondary school enrollment fees in 2013 with funding from development partners should help boost enrollment rates at the secondary level and provide welcome relief to the poorest. 38 Direct primary and secondary service delivery expenditure is estimated by excluding administrative, sports and culture expenditure from the education budget. It includes transfers and third party outputs relating to primary and secondary education. 46 Student-teacher ratios were close to national target levels, and have been reduced further for both primary and secondary schools as the number of teachers increased. Samoa’s student-teacher ratios (STRs) had been relatively high in the mid-2000s in relation to its own national STR targets for primary and secondary education (30:1 for primary and 20:1 for secondary). Modest increases in enrollment and the recruitment of around 300 additional teachers since 2006 translated in to smaller class sizes, bringing both average primary and secondary ratios well below the national target levels. At the primary level, the highest ratios had previously been in the urban Apia area and for secondary schools, in the more rural Savaii. Primary and secondary STRs fell in all areas with the exception of secondary schools in Savaii, where the ratios increased. Figures 72 and 73 show the most recent STRs by district for primary and secondary schools. They indicate that most districts have average STRs below the target rate. The highest ratios remain in Apia Urban for primary schools and in selected districts outside of Apia at the secondary level. Rural areas tend to have smaller school sizes. This is a factor that may tend to keep the STRs low in rural schools (possibly below the national target level) if the full curriculum is to be provided, especially at the secondary level where a wider range of subjects are taught. Hence, the reduction of ratios to below the target levels may be a necessary reform to maintain the quality of learning. Nevertheless, it is not clear why the recruitment of additional secondary teachers did not result in lower secondary STRs in Savaii, where the ratios have in fact increased since 200839. Figure 70: Net student enrollment ratios in 41 Figure 71: Student-teacher ratios 40 government schools 39 This trend might reflect specific challenges in attracting and retaining secondary school teachers in Savaii. The PER team does not have information about the human resource management issues specific to Savaii or any special measures led by the education sectors to attract secondary teachers to rural areas. 40 Source: MoESC Education Statistical Digest, 2013. 41 Source: MoESC Education Statistical Digest, 2012. 47 Figure 72: Student-teacher ratios by district Figure 73: Student-teacher ratios by district (primary) 2013 (secondary) It is likely that the increased investment in the education sector contributed to improved performance at the secondary level, whereas additional results at the primary level are less apparent. Performance at national examinations in secondary school has improved significantly since 2006 for most subjects, including mathematics, English and the sciences. Moreover, dropout rates in the final year of secondary school have been almost halved since 2006. These outcomes have contributed to the higher transition rates from secondary to tertiary education. Performance improvements in primary schools have not been as apparent as in secondary education. Transition rates from primary to secondary school decreased slightly since 2006 and national examination scores declined for English, Samoan and mathematics42. Figure 74: National year 8 (primary) exam results Figure 75: National year 12 (secondary) exam results 42 Education statistical digests, 2012. 48 Figure 76: Primary and secondary progression rates Figure 77: Primary and secondary drop-out rates 43 (percent) Primary Secondary The removal of primary school fees is also an important contribution from the sector to household welfare, especially amongst underprivileged communities . Education is a key factor in determining whether a household will be able to rise out of poverty in Samoa. Evidence from the 2008 household survey shows that those who have only achieved primary education have a greater chance of being in the bottom three deciles because of difficulty in finding income earning opportunities44. Samoa commenced the primary school fee relief program in 2008 with support from development partners, and plans are underway to also remove fees for secondary schooling in 2014. The significant costs of these initiatives have been partially supported by development partners and will eventually be fully borne by the domestic education budget, marking a significant lowering of education costs borne by households. Box 6: Global Evidence on Cost Effectiveness of Reducing Class Sizes and Other Education Interventions45 Research globally suggests that simply increasing the number of teachers is not always that best value for money intervention for improving learning outcomes. Although the right class size and curriculum subject coverage are important, a series of other interventions can potentially improve learning outcomes and may provide better value for money. Here is a summary of some lessons learnt from an analysis of recent and rigorous education program evaluations around the world. Providing “more of the same� resources is generally insufficient to improve learning w hen unaccompanied by other reforms. Evaluations of programs that focused on providing additional teachers to reduce class sizes showed that this investment had no effect on student test scores in India and Kenya. Non-teacher inputs, such as flip charts or textbooks, similarly had no impact on average test scores in Kenya. Even programs in which schools are given discretionary grants to purchase the inputs they feel the students need most have little impact, if any, on student learning. 43 Source: MESC Education Statistical Digest, 2012. 44 Samoa HIES Report 2010: 5. 45 Box two serves to share some international experiences in contexts different to that of Samoa. It indicates the potential improvements in learning outcomes that can be achieved from policy innovations, including strengthening teacher incentives. The appropriate mix of policies for Samoa would be determined through country specific diagnostics and context appropriate policies. 49 Interventions that focus on providing incentives for teachers to increase effort and performance can lead to significant learning gains if they are objectively administered. For example, adding an extra teacher on a contract basis can produce significant learning gains at a relatively low cost. Many teachers in developing countries have poor incentives and absenteeism rates are high. Contract teachers — who are hired and held accountable by the local community and whose contracts can be terminated if they perform poorly—are often more likely to attend school and extend more effort when in the classroom than their civil service counterparts. Contract teachers are often paid only a fraction of the salary of civil-service teachers, which makes such programs extremely cost-effective. Other incentive approaches link teacher pay with student performance. In India, whereas simply increasing the number of teachers did not demonstrate gains in learning, a program that linked teachers’ pay with performance in their own class provided effective results. Interventions that direct instruction toward children’s actual learning levels have also been shown to be consistently effective at improving learning outcomes whilst being very cost-effective. For example, in Kenya, streaming grades or classes in to different levels of abilities improved test scores for school children. Individually paced computer programs also proved to be effective in targeting students in the lower half of their class in India as they facilitated self-paced learning. Providing resources and information to communities and families can be effective when they have means for affecting change in the school or local education system. Programs in Indonesia that have sought to increase the authority of local school committees and that have provided funds for the community to manage at the same time as empowering them to affect change demonstrated results at low costs. Similarly, a program for incentivizing students and their families by providing information on the higher wages earned by those with more years of schooling proved to be a very cost-effective way to improve student attendance and test scores in Madagascar. Source: Abdul Latif Jameel Poverty Action Lab (J-PAL) analysis: “Improving Student Learning: Cost-effectiveness of education programs�. http://www.povertyactionlab.org/policy-lessons/education/student-learning?tab=tab-cost 4.4 CONCLUDING REMARKS The analysis clearly indicates that the education sector has invested the majority of its growing resources towards direct service delivery. The recruitment of new teaching personnel and the reduction of student-teacher ratios emerge as one of the main budget policies of the sector. These investments in the number of teaching personnel have been complemented by the externally funded (or co-funded) projects and programs in the sector that have been implementing school infrastructure improvements, teacher skills and curriculum development amongst other interventions. This combination of inputs may be complementary in that they combine investments in both human resources and the learning environment. These are positive trends that point to a focus on strengthening the direct learning environment and associated services. At the same time, some key questions emerge from a review of the outcome indicators of the sector when considered against the expenditure priorities. 50 Firstly, it is not clear that the additional investments in the sector, especially the reductions of student-teacher ratios, have led to improving outcomes at the primary level as indicated by the lower progression rates from primary to secondary and weaker scores in key subjects. In contrast, secondary schooling trends, half of which are provided by private/ mission schools, seem to present a healthier progression. It would therefore be informative from a policy perspective to establish what drives these trends, especially at the primary level, and whether the combination of inputs prioritized by the education sector, including increasing the number of teachers, present the best value for money and the best results in terms of learning outcomes. Key questions include the extent to which the average fall in teacher wages in real terms over the past few years may have affected incentives and performance, and whether differences in remuneration and management schemes between private or mission schools and their public counterparts have contributed to outcomes at the secondary level. Secondly, the resource needs of the education sector are likely to be augmented as the secondary school fee relief scheme gets underway. The expected increase in secondary enrollment rates will push student-teacher ratios back-up and may trigger the recruitment of additional teaching staff to maintain target class sizes. Additional Figure 78: Potential cost pressures for MOESC resources will also be needed as donor funding of the scheme is eventually phased out to maintain school financing in the absence of fees. Hence, a detailed medium to long term projection would highlight the sector’s resource needs in the coming years and the extent to which they are affordable within the limits of the government’s total budget framework. In the meantime, Figure 78 presents indicative estimates of the costs of recruiting additional secondary teachers and government take-up of donor funded share of the fee relief schemes46. The estimates suggest that these items may amount to almost a fifth of the education budget (18 percent) by FY17, representing a significant sources of upwards pressure for sending in the sector. 46 The estimates for the recruitment of teachers are based on the assumption that the secondary fee relief scheme will have an enrollment effect of 30 percent in the first year, 20 percent in the second year and 10 percent in the third year and the recruitment of one teacher for every twenty students newly enrolled. The estimates for the government take-up of the fee relief schemes is based on the winding-up of donor funds after three year and a phased transfer to the government budget thereafter. 51 Data Annex The PER is primarily based on analysis of disaggregated public financial data covering the seven year period between FY06 and FY12, taking advantage of the FinanceOne FMIS system that has been in place since FY06. This analysis represents the first time that the FinanceOne data has been used to produce a consistent public expenditure analysis. To facilitate the analysis of expenditure based on consistent functional and economic classifications some adjustment to the data were made for the purposes of the analysis. While in more recent years, FinanceOne has full coverage based on GFS classifications, these were not fully in place in FY06 and FY07, so the team remapped expenditure by line item using GFS86 as a guide. While the data from FinanceOne has full coverage of domestic expenditure, externally-funded expenditure was recorded most consistently in other systems held by the Ministry of Finance, so data on estimated grant and loan utilization by ministry was added separately to the analysis. Because this information was only available at a higher level, an analysis of donor-funded expenditure by economic classification or detailed functional classification was not possible. Some further adjustments were made to the FinanceOne data to treat expenditure classifications consistently over the period, to help to illustrate the underlying trends. Firstly, tax expenditure incurred by government ministries that was subsequently rebated was netted off both expenditure and revenue, as it’s treatment in the financial accounts changed over the period which would otherwise give the impression both expenditure and revenue has risen. Secondly, some capital items that were recorded above the line were moved into financing in the budget frame. Thirdly, and most significantly, newly incorporated public beneficial bodies, the National Health Service and the Land Transport Authority were reincorporated into the public accounts. This was necessarily since their new status as corporate bodies meant that halfway through the period, their accounts were no longer consolidated in the public accounts ledger, but instead expenditure was in the form of a public grant. The team used the bodies’ corporate accounts to reintegrate disaggregate expenditure trends for the period since they became corporatized. The South Pacific Games Authority was also reincorporated into the public accounts to smooth out a temporary spike in transfers to that agency relating to the hosting of the South Pacific games. The analysis of the public wage bill combines FMIS data with data from the payroll system. It also incorporates payroll data from the accounts of the largest public agencies to present approximate estimates of payroll trends for the whole government for the first time. Similarly, the health analysis represents the first time that unified data for the public healthcare sector has been presented since the creation of the National Health Service as an autonomous agency. For the education analysis, the data was adjusted for the spike in trends relating to the South Pacific Games in FY07 and for lumpy tax expenditures to establish the underlying structural 52 expenditure trends. The outcome data are mostly based on the Ministry of Education’s statistical database. Changes are presented in real terms taking FY06 as the base unless otherwise stated. In some cases, additional data has been collected from official government sources including the annual Budget Statement, the World Bank’s World Development Indicators are used for international comparisons and outcome data and other documents. 53 ANNEX TABLES Summary trends (% change FY06-FY12) % Change in Change in Change in expenditure (Tala constant) share of GDP share of total Receipts 19% 4% n/a Domestic revenue 10% 1% -6% Grants 56% 3% 6% Exepenses 43% 10% n/a Recurrent expenditure 33% 5% -4% Compensation of employees 36% 2% -1% Transfers 87% 2% 3% Other recurrent 12% 0% -6% of which Social benefits 45% 0% 0% of which Goods and services 79% 2% 3% of which Interest payments 136% 0% 1% of which Other -66% -2% -9% Development expenditure (domestic) 60% 1% 1% Development expenditure (foreign project) 61% 4% 4% 54 Budget frame (% GDP) FY06 FY07 FY08 FY09 FY10 FY11 FY12 Receipts 31% 37% 31% 34% 42% 36% 35% Domestic revenue 25% 29% 26% 26% 26% 28% 26% Grants 6% 7% 5% 8% 15% 8% 8% Exepenses 29% 34% 34% 40% 47% 44% 39% Recurrent expenditure 19% 19% 21% 21% 23% 24% 24% Compensation of employees 8% 8% 9% 9% 10% 10% 10% Transfers 3% 4% 5% 5% 4% 4% 5% Other recurrent 8% 7% 7% 7% 10% 9% 9% of which Social benefits 1% 1% 1% 1% 2% 2% 2% of which Goods and services 3% 4% 4% 5% 5% 6% 5% of which Interest payments 0% 0% 1% 1% 1% 1% 1% of which Other 4% 2% 1% 1% 1% 1% 1% Development expenditure (domestic) 2% 5% 5% 4% 5% 6% 3% Development expenditure (foreign project) 8% 10% 7% 15% 19% 15% 13% Deficit before grants (left axis) -4% -5% -8% -14% -20% -16% -13% Surplus/ (deficit) after grants 2% 2% -3% -6% -5% -8% -5% Financing -2% -2% 3% 6% 5% 8% 5% Total expenditure 4% 4% 2% 8% 14% 10% 9% Disbursements 3% 3% 2% 7% 14% 10% 8% Amortisation 1% 1% 0% 0% 0% 0% 0% Domestic financing -5% -6% 0% -1% -9% -2% -4% Domestic non bank (net) -3% -1% 0% 0% 0% 2% 0% Domestic bank (net) -2% -5% 0% -1% -8% -4% -4% Domestic revenue (% GDP) FY06 FY07 FY08 FY09 FY10 FY11 FY12 Domestic revenue 25% 29% 26% 26% 26% 28% 26% Taxes on income 4% 6% 5% 5% 5% 5% 5% Taxes on international trade 6% 6% 5% 5% 6% 6% 5% Taxes on goods and services 11% 12% 11% 12% 12% 13% 11% Non-tax revenues 4% 5% 4% 4% 4% 4% 4% 55 Expenditure by Economic Category* (Tala) *Adjusted estimates: NHS, LTA and SPG spending extracted from Transfers and integrated in to main expenditure categories. Current FY06 FY07 FY08 FY09 FY10 FY11 FY12 Total expenditure 362,826,802 454,622,575 491,413,543 574,182,182 679,484,386 664,377,523 618,491,340 Recurrent expenditure 233,102,495 256,853,645 304,175,075 302,978,665 334,915,202 354,994,123 370,095,633 Compensation of employees 97,409,194 109,406,062 125,829,305 131,130,298 141,565,243 152,504,200 157,720,825 Transfers 34,974,560 55,420,222 78,055,808 65,114,855 54,604,783 66,516,302 78,014,368 Other recurrent 100,718,740 92,027,362 100,289,962 106,733,513 138,745,177 135,973,621 134,360,440 of which Social benefits 14,004,563 15,122,304 15,775,741 16,414,121 32,150,097 23,403,717 24,086,886 of which Goods and services 37,465,256 49,881,364 55,656,904 67,464,949 75,407,203 83,282,622 79,836,688 of which Interest payments 4,371,096 4,319,469 10,614,896 10,141,699 11,640,866 13,042,664 12,277,978 of which Other 44,877,825 22,704,224 18,242,421 12,712,744 19,547,011 16,244,618 18,158,888 Capital expenditure (domestic) 25,055,840 66,036,975 79,062,240 55,114,270 73,370,329 84,554,122 47,678,998 Capital expenditure (foreign) 104,668,467 131,731,954 108,176,228 216,089,246 271,198,854 224,829,278 200,716,709 Constant FY06 FY07 FY08 FY09 FY10 FY11 FY12 Total expenditure 362,826,802 436,067,308 448,197,922 508,288,227 592,918,617 567,382,355 519,714,801 Recurrent expenditure 233,102,495 246,370,251 277,425,477 268,208,408 292,247,273 303,167,092 310,989,283 Compensation of employees 97,409,194 104,940,690 114,763,693 116,081,600 123,529,944 130,239,494 132,531,924 Transfers 34,974,560 53,158,264 71,191,467 57,642,183 47,648,177 56,805,318 65,555,036 Other recurrent 100,718,740 88,271,296 91,470,317 94,484,625 121,069,152 116,122,281 112,902,324 of which Social benefits 14,004,563 14,505,093 14,388,400 14,530,413 28,054,201 19,986,914 20,240,075 of which Goods and services 37,465,256 47,845,473 50,762,355 59,722,577 65,800,385 71,123,854 67,086,321 of which Interest payments 4,371,096 4,143,171 9,681,407 8,977,823 10,157,829 11,138,513 10,317,116 of which Other 44,877,825 21,777,559 16,638,156 11,253,811 17,056,737 13,873,000 15,258,811 Capital expenditure (domestic) 25,055,840 63,341,698 72,109,392 48,789,279 64,023,008 72,209,723 40,064,394 Capital expenditure (foreign) 104,668,467 126,355,359 98,663,053 191,290,540 236,648,336 192,005,540 168,661,124 56 % of total FY06 FY07 FY08 FY09 FY10 FY11 FY12 Total expenditure 100% 100% 100% 100% 100% 100% 100% Recurrent expenditure 64% 56% 62% 53% 49% 53% 60% Compensation of employees 27% 24% 26% 23% 21% 23% 26% Transfers 10% 12% 16% 11% 8% 10% 13% Other recurrent 28% 20% 20% 19% 20% 20% 22% of which Social benefits 4% 3% 3% 3% 5% 4% 4% of which Goods and services 10% 11% 11% 12% 11% 13% 13% of which Interest payments 1% 1% 2% 2% 2% 2% 2% of which Other 12% 5% 4% 2% 3% 2% 3% Capital expenditure (domestic) 7% 15% 16% 10% 11% 13% 8% Capital expenditure (foreign) 29% 29% 22% 38% 40% 34% 32% 57 Expenditure by Function (Tala) Current FY06 FY07 FY08 FY09 FY10 FY11 FY12 Health 54,128,954 57,216,133 59,475,771 71,853,525 81,288,672 87,339,540 75,281,113 Education 66,947,415 126,003,296 119,220,152 87,963,786 105,138,991 101,249,681 100,468,787 Women and social development 13,463,942 15,355,165 14,629,903 14,222,161 12,251,122 13,412,575 13,303,908 Social sectors 134,540,311 198,574,594 193,325,826 174,039,472 198,678,786 202,001,796 189,053,808 Economic and infrastructure 114,162,618 144,826,175 170,838,207 191,165,081 237,836,435 208,398,404 194,647,191 Justice and security 21,474,197 31,602,999 38,157,109 34,018,801 43,071,890 42,064,745 45,332,139 Administrative 100,290,645 89,831,923 102,103,319 187,189,043 219,951,441 156,486,325 120,816,862 TOTAL 370,467,771 464,835,692 504,424,461 586,412,397 699,538,553 608,951,270 549,850,000 Constant FY06 FY07 FY08 FY09 FY10 FY11 FY12 Health 54,128,954 54,880,876 54,245,385 63,607,513 70,932,560 74,588,486 63,258,297 Education & Sport 66,947,415 120,860,514 108,735,759 77,868,938 91,744,368 86,467,830 84,423,358 Women, community and social 13,463,942 14,728,449 13,343,328 12,590,006 10,690,339 11,454,419 11,179,199 Social sectors 134,540,311 190,469,840 176,324,472 154,066,458 173,367,267 172,510,735 158,860,854 Economic and infrastructure 114,162,618 138,915,144 155,814,447 169,226,707 207,536,263 177,973,476 163,560,942 Justice and security 21,474,197 30,313,133 34,801,518 30,114,756 37,584,565 35,923,542 38,092,342 Administrative 100,290,645 86,165,464 93,124,205 165,706,965 191,929,803 133,640,252 101,521,731 TOTAL 370,467,771 445,863,581 460,064,641 519,114,886 610,417,899 520,048,005 462,035,869 % of total FY06 FY07 FY08 FY09 FY10 FY11 FY12 Health 15% 12% 12% 12% 12% 14% 14% Education & Sport 18% 27% 24% 15% 15% 17% 18% Women, community and social 4% 3% 3% 2% 2% 2% 2% Social sectors 36% 43% 38% 30% 28% 33% 34% Economic and infrastructure 31% 31% 34% 33% 34% 34% 35% Justice and security 6% 7% 8% 6% 6% 7% 8% Administrative 27% 19% 20% 32% 31% 26% 22% TOTAL 100% 100% 100% 100% 100% 100% 100% 58 Expenditure by Function (Tala) Original Budget Revised Budget FY06 -1.6% 2.6% 8.1% 8.1% FY07 4.6% 16.3% -0.4% 17.9% FY08 0.7% 4.9% -0.5% 3.4% FY09 -1.6% 6.5% -3.6% 5.0% FY10 8.7% 10.5% -8.8% 16.4% FY11 4.6% 12.8% 3.3% 7.7% FY12 0.6% 12.1% -3.9% 7.9% Expenditure by Function (Tala) Original Budget Revised Budget FY06 -1.7% 2.6% 6.8% 8.2% FY07 2.7% 7.9% -3.4% 3.8% FY08 -1.3% 6.7% -2.6% 4.8% FY09 0.4% 3.9% -1.5% 4.0% FY10 7.6% 11.8% -8.5% 11.3% FY11 5.0% 7.1% 1.8% 8.3% FY12 2.5% 4.2% -4.7% 6.5% 59