Report No. 39043-LA Lao PDR Public Expenditure Tracking Survey in Primary Education and Primary Health Making Services Reach Poor People March 2008 Poverty Reduction and Economic Management Sector Unit East Asia and Pacific Region FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. FOR OFFICIAL USEONLY ACKNOWLEDGEMENTS The report was prepared in partnership with the Ministry o f Finance, Ministry o f Education, MinistryofHealth, Department ofPublic Administration at the Prime Minister's Office, National Statistics Centre. and the World Bank. It was written by a team which includes Kaspar Richter (World Bank, Team Leader), Xubei Luo (World Bank, Team Leader), Tomomi Kurokawa (education questionnaire design), Yoshimi Nishino (United Nations Children's Fund, health questionnaire design and training), Yasuyuki Sawada (University o f Tokyo, analysis), Bounmy Vilaychith (National Statistics Center o f the Lao PDR, PETS project manager) and Hiroyuki Yamada (Chicago University, analysis). On the side o f the GOL, the PETS Committee, its technical staff and its secretariat provided the overall guidance. The PETS Committee included H.E.Mr.Somdy Douangdy (Minister o f Finance, President o f the Committee), Dr. Samaychan Boupha (Director General o f NSC, Vice President o f the Committee), Mr. Khamphet Manivong (Director General o f Planning & Budget Dept., MOH), Dr. Prasongsidh Boupha (Deputy Director Department of Planning and Budgeting, MOH), Mr. Nisith Koepanya (Acting Director General of Administration Dept., PMO), Mr. Niphon ManuKoon (Deputy Director General o f Finance Dept., MOE), and Mrs. Khamphay Vithaxay (Deputy Director General o f Treasury, MOF). Technical staff included Dr. Samaychane Boupha (Director General o f NSC, leader), Mr. Bounmy Vilaychith (Acting Director o f Data Service & Information Technology, NSC), Dr. Swady Kingkeo (Director o f Statistic Division, MOH), Mr. Vimonh Sisouva (Director o f Budget Division, MOE), Mr. Khampane Sivongxay (Director o f local administration division, PMO), Mrs. Boudsaba Xaypasit (Deputy Director o f Budget Division, MOF), Dr. Maitri Senchantixay (Deputy Director o f CBHI, MOH), Mr. Phanhthong Buasavan (officer, MOH), Mr. Bounkhong Chakvongsa (officer DOFNOE), Mr. Pasomphet Khamtanh (Deputy Director, Fiscal Policy Dept, MOF), Mr. Sengaloun Nhotleuxay (Officer, PMO), Mr.Thonglien Praseth (Technical Staff -Administration division, MOH), Mrs.Manivone Phonh-Amath (Officer, External Finance Dept, MOF), and Mr. Nilandon Xayyaphet (Officer, ExternalFinance Dept, MOF). On the side o f the World Bank, Ian Porter (Country Director), Deepak Bhattasali (Acting Sector Director), Indermit S. Gill (Sector Manager), Patchamuthu Illangovan (Country Manager), and Kazi Mahbub-A1 Matin (Lead Economist) provided oversight. The World Bank peer reviewers for this report were Brendan Horton (Lead Economist) and Luis Benveniste (Country Sector Coordinator). Susan Middaugh (Consultant) did the copyediting. LynnGross provided assistance on the preparation o f the report. The Lao PETS team benefited from consultations at various stages o f the preparation of the report, including during PETS steering committee meetings and workshops on background studies. The report builds on the findings o f the Lao PDR Poverty Assessment (2006a) and the Lao PDR Public Expenditure Review (2006b), including a background study on the central-local fiscal system (World Bank 2006~).I t also draws upon approaches in the 2004 World Development Report on "Malung Services Work for Poor People". We are thankful to Alessandro Magnoli Bocchi, Somneuk Davading, Shabih Ali Mohib, Emiko Naka, Toomas Palu, Ekaterina Vostroknutova, Jeffrey Waite and Roland White for their ready support. We are also indebted to the staff at National Statistics Center, Ministry of Finance, Ministry o f Education, Ministry o f Health, Department of Public Administration at the Prime Minister's Office and World Bank resident mission, as well as the many families invillages across Lao PDR who gave their time to teach us about their lives. This document has a restricted distribution and may be used by recipients only in the performance o f their official duties. Its contents may not be otherwise disclosed without World Bank authorization. TABLE OF CONTENTS EXECUTIVESUMMARY .............................................................................................................. i Achievements and challenges...................................................................................................... ii Key constraints to service delivery............................................................................................. ... Frontline services: access and quality fall short ......................................................................... 111 iv INTRODUCTION........................................................................................................................... Way Forward.............................................................................................................................. vi 1 The Purpose ofthe Rep0rt........................................................................................................... 1 Accountability-Long- and Short-Route Relationships................................................................ The highly decentralized fiscal system........................................................................................ 4 Education and Health Sector Context.......................................................................................... 3 Public Expenditure Tracking Survey-Design, Methodology and Limits.................................... 7 9 Conclusions ............................................................................................................................... 14 EDUCATION................................................................................................................................ 15 15 Public Funding.......................................................................................................................... Outcomes................................................................................................................................... Expenditures Management ....................................................................................................... 17 -20 Budget Preparation............................................................................................................... -20 26 Frontline Services...................................................................................................................... Executionand Reporting....................................................................................................... 28 Teachers ................................................................................................................................ 29 41 Accountability....................................................................................................................... Operating funds and school quality....................................................................................... 48 Conclusions ............................................................................................................................... 50 HEALTH....................................................................................................................................... 52 Outcomes................................................................................................................................... 52 Public Funding.......................................................................................................................... 55 Expenditures Management ........................................................................................................ 59 Budget Preparation................................................................................................................ 59 64 Frontline Services...................................................................................................................... Execution and Reporting....................................................................................................... 66 Health workers ...................................................................................................................... 66 Resource flows and Revolving DrugFunds.......................................................................... 71 Operating funds and health center quality............................................................................. Accountability ....................................................................................................................... 75 79 Conclusions ............................................................................................................................... 79 FINDINGS AND RECOMMENDATIONS.................................................................................. 81 Main findings o f PETS.............................................................................................................. 82 Finance Management ................................................................................................................ 83 Public Finance Management................................................................................................. 83 External FinancingManagement .......................................................................................... 86 Budget Law Implementation................................................................................................. 84 Bolster Public ExpenditureManagement Strengthening Program............................................ 86 88 Conclusions ............................................................................................................................... Steps forwards ........................................................................................................................... 90 REFERENCES .............................................................................................................................. 91 FIGURES Figure 1:Lao PDR's subnational authorities account for almost one-half o f total government spending. a large share by East Asian standards.............................................................................. 5 8 Figure 3: District-Level Map ofthe PETS Survey ........................................................................ Figure 2: The framework o f accountabilityrelationships ................................................................ 11 Figure 4: There are large disparities ineducation outcomes ......................................................... Figure 5: Poor parentskudents have little clout with policymakers ............................................. 16 17 Figure6: Lao PDR stands out ininternationalcomparison for its low spending on education.....18 Figure7: While public educational expenditures have increased since the late 1990s, it still remains below levels from the mid 1990s..................................................................................... 18 Figure 8: As education spending has shifted away from primary education. it puts poor people at a further disadvantage.................................................................................................................... 19 Figure 9: Policymakers may not be able to ensure schools deliver efficient services...................20 Figure 10: While all PES receive government funds, surplus provinces are more likely to obtain donor fundingand deficit provinces are more likely to rely on community contributions............22 Figure 11: Reliance on foreign fundingand a lack o f domestic funds are associated with insufficient recurrent spending...................................................................................................... 23 Figure 12: The share o f capital spending as a percentage o ftotal education spending varies by year and by province...................................................................................................................... 23 Figure 13: Due to the small budget envelope for recurrent education spending. the rise in the share for salaries has squeezed operating expenditures................................................................. 24 Figure 14: The share o fnon-wage spending as a percentage o f total recurrent education spending varies widely across provinces...................................................................................................... 25 Figure 15: Schools inless well-off areas participate less inbudget preparations.......................... 26 Figure 16: DFO innon-priority districts are better equipped than those inpriority districts........28 Figure 17: The accountability o f schools to policymakersand parents is crucial for education ...29 Figure 18: More thannine out o f 10teachers are quota teachers.................................................. 30 Figure 19: Salaries vary little innominal terms; they are higher inless well-off areas inreal terms. ....................................................................................................................................................... 31 Figure 20: Salary payment delays are more serious outside o f cities; most teachers have experienced salaries arrears inthe past.......................................................................................... 33 Figure 21: Differences inteachers' salary payment are small betweenreported by DEB and Figure 22: Fundingshortage inDEB and DFO creates a bottleneck that leads to salary delays...35 schools ........................................................................................................................................... 37 Figure 23: Inspite o f the difficult circumstances. teachers like their work and became teachers for 38 Figure 24: Teachers are dedicated to improve education service.................................................. the right reasons............................................................................................................................. 38 Figure25: Most school directors and teachers are local or know the local language.................... 39 Figure 26: Teachers' absenteeismrates do not appear to be high................................................. 40 Figure 27: About two out o f five teachers who were absent from work were absent for reasons other than official reasons or sickness........................................................................................... 40 Figure28: Teachers inurban, non-poor, and Lao-Tai areas are more likely to holdmultiplejobs . 41 Figure 29: Most schools receive fundingfrom parents and communities..................................... ....................................................................................................................................................... 41 Figure30: Schools inwell-off areas can mobilize moreresourcesthan schools inless well-off areas............................................................................................................................................... 42 Figure 31:Three out of four schools collect tuitionhegistration fees............................................ Figure 32: Inover two out o f three cases. parents are incharge o f setting tuition fees.................42 43 Figure33: Tuition fees are higher per student inwell-off areas.................................................... 43 Figure34: Schools inpriority districts have worse infrastructure................................................. 44 Figure 35: Many students cannot afford textbooks........................................................................ 45 Figure 36: Non-Lao-Tai villages have fewer textbooks than other villages.................................. 45 Figure 37: Teachers and school directors inwell-off areas are better educated and more experienced.................................................................................................................................... 46 Figure 38: Multi-grade classes are common inless well-off areas................................................ 47 Figure 39: Teachers spend on average 45 minutes to prepare for each class................................ 47 Figure 40: Student-to-teacher ratios are higher inless well-off areas than inwell-off areas ........48 Figure 41: More school directors meet with parents on a weekly basis than with officials from Figure 42: More children are kept out o f school for reasons other than tuitiodregistration fees..49 DEB............................................................................................................................................... 50 Figure 43: Poor people have poorer health outcomes.................................................................... Figure 44: Overall per capita health expenditures are increasing, but they started at a low level 56 . 54 Figure 45: Public health expenditures as a percent o f GDP inLao PDR are low compared to other low income countries..................................................................................................................... 57 Figure 46: Concentrationcurves o f consumption and health care subsidy by health provider and inpatientloutpatient care................................................................................................................. 58 Figure 47: Poor patients have little clout with policymakers on budget decisions........................ Figure 48: Policymakersmay not be able to ensure health centers deliver efficient service .........59 59 Figure 49: Actual health spending compared to plannedhealth spending i s lower indeficit provinces........................................................................................................................................ 61 62 Figure51:Healthnon-wage spendingvaries across provinces..................................................... Figure 50: Health capital spending varies across provinces and by year....................................... 63 Figure52: Healthrecurrent expenditures per capita vary across provinces.................................. 63 Figure 53: DHO innon-priority districts have better office facilities ............................................ 65 Figure 54: Accountability o f Health Centers to policymakers and patients i s crucial for health service delivery.............................................................................................................................. 66 Figure55: H C workers' salaries vary little innominal terms; they are higher inrural areas inreal terms.............................................................................................................................................. Figure 56: Health workers inhealth centers get paid less frequently than they are supposed to 68 ...67 Figure 57: One out o f three health workers had outstanding balances due for their salaries and overtime at the time o f the survey ................................................................................................. 68 Figure 58: Overall difference in salary payments reportedby DHO and HC i s small, but some areas have greater discrepancies.................................................................................................... 69 Figure 59: Delays inrevenue collection at DFO are the primary reason for salaries being in arrears............................................................................................................................................ 70 Figure 60: Less than one out o f five health workers are absent from work, just two out o f five are absent without a legitimate reason ................................................................................................. 71 Figure 61: Less than 20 percent DHO receivedfinancial supports from PHO.............................. 72 Figure 62: Donors and NGOs are the main financiers o frenovation costs for HC....................... 73 Figure 63: Health centers inpoor, rural, or non Lao-Tai areas are more likely to rely on off- budget funding............................................................................................................................... Figure 64: RDF profit ratios cluster around 25 percent, but they differ across provinces .............73 74 Figure65: Health centers innon-poor, urban, and Lao-Tai areas are better equipped than those in poor, rural, and non Lao-Tai areas................................................................................................. 76 Figure 66: The poor have less access to health care, less access to safe water, and less access to local hospitals................................................................................................................................ 77 Figure 67: People inless well-off areas spend a lower share o f household consumption on health althoughtheir householdconsumption i s already low................................................................... 78 Figure 68: HC health workers inpoor rural areas are less experienced........................................ 78 TABLES Table 1: PETS Timeline .................................................................................................................. 9 11 Table 3: Analytic Domains............................................................................................................ Table 2: Modules of the PETS Questionnaire ............................................................................... 12 BOXES Box 1: Educationand Health-a Public Responsibility .................................................................... 2 Box 2: Central-Local Relationships................................................................................................. 6 Box 3: Subnational Administration ................................................................................................. 7 Box 4: BudgetProcess inthe Education Sector............................................................................. 21 Box 5: Why Poor People Have Poor Health.................................................................................. 53 Box 6: Budgetprocess inthe Health Sector .................................................................................. 60 Box 9: New Organic Budget Law inLao PDR-Changes inIntergovernmental Fiscal Relations.84 Box 8: Public Expenditures inPriority Districts............................................................................ Box 7: Revolving DrugFunds....................................................................................................... 74 85 EXECUTIVESUMMARY Lao People's DemocraticRepublic (Lao PDR) has come a long way since launching its economic reforms two decades ago. The country can take pride in its solid track record o f delivering basic services and improvements in social indicators. Primary education and health outcomes in Lao PDR have improved inthe past decade. Inspite o f these achievements, Lao PDR remains one o f the poorest countries in East Asia with some o f the worst social outcomes. In order to improve upon present conditions, Lao PDR has set itself the goal of graduating from its status as a least developed country by 2020. The Government of Lao (GOL) has articulated a bold social strategy in the 6th National Socio-Economic Development Plan (NSEDP) for 2006 to 2010. This plan puts improving service delivery at the center of its agenda. To improve the effectiveness of public spending on education and health, it is important to understand how resources flow from the center through layers of subnationalauthorities,then what factors affect frontline services. Eventhoughthere are many well-functioning social facilities in the country, lack o f information makes it difficult to learn what works and what does not work, to draw clear lessons, and to improve policies. Closing part o f this knowledge gap with information about current public expenditure management practices in the education and health sectors at the district and facility levels was the motivation for this Lao PDR Public Expenditure Tracking Survey (PETS). The report looks at how public financial resources, human resources, and budget processes combined to deliver services at the facility level for students and patients inLao PDR. Its objectives are to: (a) analyze management practices relevant to current public expenditures, then advise Lao PDR about how to improve basic service delivery in education and health; and (b) serve as an independent monitoring tool to assist government in improving accountability. PETS links downstream with the analysis from the Lao PDR Poverty Assessment (World Bank, 2006a) on education and health outcomes at the household level, and upstream with the analysis from the Lao PDR Public Expenditure Review (World Bank, 2006b) on public expenditures at the national and subnational levels. The report presents a snapshot of financial flows and human resources at the provincial, district, and facility levels. It aims to complement and test some of the findings from Public Expenditure Review and Poverty Assessment mainly at the district and facility levels. PETS tracks salary payments from districts to facilities by comparing salary records o f individual teachers and health workers; this i s a first for Lao PDR. It also offers information about the timeliness of salary payments and correlates performance indicators with service delivery. The survey results show reasonable frontline service in Lao PDR, given tight budget constraints in the education and health sectors. Among the positives: (1) teachers' and health workers' job morale i s high; (2) teachers' and health workers' absenteeism i s low and ghost-workers are rare; (3) salary leakages from the district to the facilities are limited; and (4) parentdpatients and community participation is active. However, challenges remain: (1) education and health services are poor in Lao PDR; there i s considerable variation between surplus provinces and deficit provinces, between priority districts and non-priority districts, between cities and villages, between non-poor and poor areas, and between localities with a high concentration o f Lao-Tai and elsewhere; (2) the three major contributors to the overall low level o f primary education and health service delivery are insufficient resources, inefficient allocation o f finds, and weak expenditure management; and (3) effectively delivering services to poor people requires better targeting and improved management o f public expenditures. This executive summary: (1) reviews achievements and challenges inthe education and health sectors; (2) examines key constraints on service delivery; (3) explains why frontline services fall short on access and quality; and (4) offers recommendations for the way forward. Achievementsand challenges In the last decade Lao PDR has made improvements in primary education and health. According to the 1995 and 2005 Population Censuses and the 1992/93 and 2002/03 Lao PDR Expenditure and Consumption Surveys (LECS), the achievements in education and health have made a positive difference inthe lives o f many citizens. Primary education and health service delivery in Lao PDR is reasonably good, given the limited resources available. 0 School teacherdhealthworkers are satisfied with their jobs; morale is high. 0 Teachers'health workers' absenteeism is low; ghost-teachers or ghost-health workers are rare. 0 Salary leakages from the district to the facilities are limited. Parents / patients and the community actively participate infrontline services. However, achieving the MDGs and the national development objectives on education and health will not be easy. Education and health outcomes in Lao PDR are low compared with East Asian standards. Adult literacy rate i s 69 percent and life expectancy at birth i s 61 years in Lao PDR; while average levels in East Asia are 91 percent and 71 years respectively. Within the country, there are also gaps in service that highlight differences between cities and villages, non-poor and poor areas, Lao-Tai concentrated localities and elsewhere. .. 11 K e y constraints to service delivery In Lao PDR, low public funding, inefficient allocation of resources, and weak managementof expenditures are key constraints to primary education and health service delivery. The findings from PETS and those from the Public Expenditure Review (PER) reinforce each other. Low Public Funding Based on internationalcomparisons,Lao PDR is unusualfor the scant resources it allocates to education and health. In 2004/05, total public expenditures for education were 2.3 percent o f GDP; domestic outlays were just 1.1 percent; education's share o f overall government spending was about 11percent. As a result, the education system has to operate very frugally. In Lao PDR the unit spending per student in primary school i s just two percent o f GDPper capita, compared to seven to 14 percent inmany developing countries (World Bank, 2006b). In 2002/03, public health spending was 1.2 percent of GDP; as a share o f overall government spending, it was about five percent. Total public and private expenditures for health were 8.7 USD per capita in 2002. That figure puts Lao PDR at the very low end o f the spectrum compared with other countries. For example, low income countries spend an average o f 29 USD per capita compared to East Asian countries which spend an average o f 64 USDper capita. Givenlimitedpublicspending, it is notsurprisingthat insufficiencyof resources was the single most pressing problem facing the education and health sectors. The majority o f PES, DEB, PHO, and DHO consider insufficient resources as the reason for budget cuts. Education and health departments in deficit provinces and priority districts often face even stiffer adjustments. Nine out o f 10 schools and health centers have to rely on off-budget resources fiom donors, NGOs, parentdpatients and communities to cover their operating costs. Three out o f four schools collect tuitiodregistration fees. The profits from Revolving Drug Funds (RDF) serve as an important source of revenue for health centers to cover the cost o ftheir office supplies, transportation, and utilities. Ineficient and inequitable allocation of funds A large part of public spending for education and health, limited as it is, goes for capital expenditures. In Lao PDR during 2003/04 and 2004/05, less than half o f the national education and health budgetswent for recurrent expenditures, such as wages and operating costs. This imbalance leads to delays in the payment o f salaries and an acute shortage o f funds for routine operating expenses. It also results in poor education and health services. For example, schools get built, but the supply o f teachers and materials is inadequate; hospitals have beds, but the staffing and training o f medical personnel are insufficient to the need. Public education and health spending gets diverted from the people who need it most. A shift in funding to secondary schools and above has put a dent in the share o f funds for primary education. Only about one-sixth o f public education spending accrues 111 to the poorest quintile o f the population. Likewise, much o f government spending for health favors the central level; public subsidies to central hospitals are heavily biased in favor o f the non-poor. The richest quintile captures more than one-fourth o f public subsidies for health services compared with one-eighth for the poorest quintile. Just one- fifthoftotal public healthexpenditures went to district health facilities and village health centers, which primarilyserve the poor. Weak expenditure management Weakness in expenditure management reducesthe effectivenessof public spending on education and health service delivery. InLao PDR, the budget i s a weak planning tool for several reasons. First, budget preparation suffers from a lack o f realism. Expenditures are inflated and there i s a lack o f follow-through because o f insufficient funds. Second, the process i s compromised by off-budget funds; as a result, budget allocations do not reflect actual expenditures. Third, a lack o f integration leads to a mismatch o f capital and recurrent spending. Fourth, schools, health centers and districts usually do not have a role in budget preparations. Fifth, budget preparation runs from M a y to September; this compressed budget-cycle pays short shrift to budget planning and leaves little time for input from the districts and provinces or for a dialogue among the MOF, MOE, MOH, and the provinces. Budget details are often not final until four months after the fiscal year starts inOctober. Problemswith expenditure managementare compounded by weaknesses in budget execution, reporting,and capacity.Until2005, treasury reports were submittedalmost entirely on paper. The lack o f automation and agreement on a reporting format makes it difficult to generate regular, timely, and accurate reports that can serve as a check on budget execution. Moreover, capacity is low at the provincial anddistrict levels. In a highly decentralized context, the central government of Lao PDR has limited control over the finance department. Systematic links are weak between provincial finance departments and the Ministry o f Finance. InLao PDR, budget allocations depend on negotiations between provincial governors and the centre-not explicit rules. The treasury offices at the provincial and district levels lack authority. In case o f a funding shortfall, PES and PHO are more likely to turn to provincial government than to Provincial Financial Offices (PFO). Provinces that collect tax revenues negotiate how much tax revenue they will remit to the central government and how much they will retain. Provincial authorities have the power to transfer funds within and across expenditure items; they also have the power to determine spending at the sectoral level. As a result o f this discretionary power by the provincial governors, there is a disparity in resource allocationbetweenprovinces and a lack o f alignment with national priorities. Frontlineservices: access and quality fall short Low public funding, inefficient and inequitable allocation of funds, and weak managementof expenditureshavehadthe followingnegativeeffects:low salaries and frequent delays in payment, inadequate access to schools and health centers with iv reasonable equipment, and teachers and health workers who receive short shrift on training. Low salaries andpayment delays In Lao PDR, teachers'/health workers' salaries are very low. The average monthly salary for teachers, including base salary, bonus, and allowance i s about 390,000 kips or 39 USD. O f that, about 320,000 kips represents the base salary. For health workers, it i s about 450,000 kips or 45 USD; o f that, about 350,000 kips i s base salary. Teachers' and health workers' salaries represent 130 percent and 150 percent GDP per capita, respectively; these salaries are very low compared with international standards. Teachers'/health workers' salaries are all based on the same payment scale; as a result, there i s little variation across cities and villages. However, in real terms, purchasing power is higher in less well-off areas because living expenses are lower there. Central government budget i s supposed to cover teachers'/health workers' salaries. PFO transfers resources to cover teachers' and health workers' salaries to DFO, which then transfers the resources to DEB and DHO. DEB and DHO actually deliver salary payments to schools and health centers for teachers and health workers, respectively. More than 90 percent o f teachers and health workers surveyed reported that their salary is inadequate to cover minimum household living expenses. Most teachers and health workers have to hold multiple jobs, usually in agriculture. Some reported they have to borrow money occasionally to make ends meet. Frequent payment delays make low salaries even less appealing, especially in rural areas. Throughout the country, fewer than one out o f two teachers get paid timely every month; nearly two out o f every five teachers gets paid irregularly. Seventy percent o f teachers in urban areas get paid on time every month compared with 30 percent o f teachers inrural areas. Throughout the country, only one out o f two health workers gets paid on time every month; one out o f three receives a salary every three months; and about one out of eight gets paid irregularly. Between October 2004 and September 2005, two out o f three health workers reported delays in salary payments; at the time o f the surveyabout one out o f three health workers had an outstandingbalance that was due. In urban areas during this same period, one out o f four health workers had their salaries in arrears; the number was even higher in rural areas-up to two out o f five. Delays intax collection and in money transfers from upper layers o f authorities are among major reasons for delays insalary payment. Low salaries and frequent delays in payments negatively affect the delivery of education and health services. Inadequate and irregular salary payments are a major cause o f concern for teachers and health workers. Teachers and health workers like to contribute to the community development and they have highjob morale. Nevertheless, teachers who experience long and frequent payment delays are more likely to spend less time on lesson preparations and to be absent from work for unofficial reasons. Likewise, low-paid health workers tend to spend less time at work when their salary i s inarrears Schools and health centers: short on access and equipment V Students and patients lack access to primary schools and health centers.According to LECS 2002/03, one out o f six families had to send their children to neighboring villages for education. According to the Lao Population Census (2005), only four out o f five villages have primary schools. Because o f the lack of teachers, low student enrollment, and limited facilities, many primary schools are incomplete; they only have the first two or three grades. Nationally, only about three out o f four households have access to skilled medical personnel at a dispensary; in Luang Namtha and Bokeo, the number is one out o f three and two out of five, respectively (World Bank, 2004). Poor people are more likely to rely on self-healing or self-medication when they are sick compared with the non-poor for two reasons. The poor cannot afford to pay for health services and they may think the quality o f care they have access to i s not worth paying for. Most schools and health centers are poorly equipped. The disparities are especially pronounced between priority and non-priority districts, and urban and rural areas. Many schools and health centers do not have basic facilities. For example, schools have an inadequate supply o f student textbooks; health centers a dearth o f medical laboratories. Teachers and health workers: short on training and qualijkations Teachers and health workers are not skilled enough or trained to deliver good service. This i s especially true for personnel in less well-off areas. Teachers and health workers need more training. Almost 30 percent o f teachers in rural areas instruct more than one grade at the same time, compared to only five percent inurbanareas. Yet, many teachers in rural schools are either untrained or trained only in monograde pedagogy. In addition, they have few teaching or learning resources. Many health centers do not have staff who are capable o f performing laboratory tests. Throughout the country, health workers generally receive training from the PHO and DHO two to three times a year, but that number drops to once a year inpriority districts. Way Forward In order to build on the progress in educationand health outcomes, it is important that Lao PDR allocates more resources to these two priority sectors and improve targeting of the neediestpeople. To improve service delivery at the facility level, it is important to remove/minimize teachers' and health workers' salary delays, improve access to equipments, and ameliorate public finance management practice at the subnational levels. Given scarce budget resources, it i s important to strengthen accountability among service users, service providers, and government. In 2006, the GOL adopted a detailed five-year Public Expenditure Management StrengtheningProgram(PEMSP). Inaddition to enhancing the overall budgetprocess, this initiative also supports adequate fiscal arrangements for the Nam Theun 2 project. This hydro power project is expected to generate revenues from 2009 onwards; the revenueis supposed to be directedto basic social services. The findings o f this report, as vi well as o f the recently completed Lao PDR Public Expenditure Review (World Bank 2006b), suggest that the PEMSP reform agenda lies at the heart o f improving basic education and health services. At the core of the financial management system are institutional mechanisms for fiscal transfers and monitoring. The fiscal framework i s highly decentralized. As a result, priorities in allocation of funds at the subnational level are often different from those at the central level. Effective implementation o f the new Budget Law to improve intergovernmental fiscal relations and develop a new revenue sharing framework i s critical for improving the revenue sharing system and treasury. To minimize a mismatch between capital expenditures and recurrent needs, it i s important to improve local capacity and to enhance the effectiveness o f donor assistance. The findings of the PETS have shown that this survey tool can serve as a useful monitoring device for improvingpublic service delivery.This PETS collects national representative information on facility characteristics and human resources at the district and facility levels. Given its pilot nature, this PETS focuses on tracking teachers' and health workers' salary payments from districts to facilities. With the strengthening o f local capacity and improvement o f survey design and execution, it would be useful that future PETS aims at obtaining broader quantitative information on leakages o f general public resources and identifying binding constraints o f education and health service delivery in the overall public expenditure systems. The GOL should implement future PETS to expand on the lessons learned from this pilot experience and to track progress in public financial management and frontline service delivery in education and health sectors. vii INTRODUCTION The Government of Lao PDR (GOL) has a policy objective of making major improvements over the next few years in health and education outcomes through improved delivery of public service. Theprimary objectives of the PETS report are (i) to advise GOL how to improve basic service delivery in the health and education sectors; and (ii) serve as independent monitoring tool to assist government in improving accountability. The analysis is based on a PETS survey which monitors the movement of budget resourcesfrom the center tofacilities. However, in the Lao case, thefiscal system is highly decentralized. As a result, monitoring is at the subnational level because primary services are the exclusive responsibility of sub-national authorities. Tracking focuses on salary payments from districts to schools/health centers. The PETS report presentsfindings at three levels (province, district, andfacility). These three levels reflect the enormous variety in the nation's economy, geography and ethnicity. The purpose of this chapter is to present an overview of the report. It consists of six sections: (1) the purpose of the report; (2) achievements and challenges in the education and health sectors; (3) the decentralized fiscal system; (4) the chain of accountability in service delivery-long route and short-route relationships; (5) the design,, methodology, and limites of the Public Expenditure Tracking Survey; and (6) conclusions. The Purposeofthe Report The Lao Government's 6th National Socio-Economic Development Plan (NSEDP) for 2006 to 2010 lays out its overall development framework, its poverty reduction objectives, and a road map to reach these objectives in the medium term. Service delivery i s at the center o f the development agenda. Part o f GOL's development vision i s to graduate permanently from its least developed country status and to achieve that by 2020. By 2015, Lao PDR aims to achieve universal primary school enrollment, which would represent a significant increase from approximately 83 percent in 2005. In that same year, Lao PDRwants to lower the mortality rate for childrenunder five years o f age to 55 deaths per 1,000 live births, an improvement over 98 deaths per 1,000 live births in 2005. Meeting this ambitious agenda will only be possible if Lao PDR's economy continues to grow rapidly. However, family income generated by growth will only have a modest impact on school enrollment and utilization o f health services. Basic health and education are public responsibilities (Box 1).For this vision to be achieved, it will be essential to bring about a greater transfer o f resources to these sectors by government, donors, and communities alike. It i s difficult to calculate precisely the required funding levels. Nevertheless, it is fair to say that policy and institutional reforms will have to accompany any increase infunding to ensure the money is well spent. Box 1:Education and Health-a Public Responsibility By financing, providing, or regulating the education and health services, the GOL demonstrates its responsibility for the health and education o f their people. Why? First, these services are replete with market failures-with externalities, as when an infected child spreads a disease to playmates or a farmer benefits from a neighbor's ability to read. So the private sector, left to its devices, will not reach the level of health and education that society desires. Second, basic health and basic education are considered fundamental Constitutional rights: "The state implements compulsory primary education" and "...attends to providing health care to all people" (Articles 22 and 25 o f the 1991 Constitution). No matter how daunting the problems o f delivery may be, the government cannot walk away from this fundamental responsibility. The challenge i s to see how the public sector -in collaboration with the private sector, communities, and outside partners-can ensure basic education and health services for all their citizens. Source: Adopted from World Bank (2003). The objective of the present PETS report is to inform policymakers and citizens about whether the budget process and frontline service providers succeed in delivering primary services in education and health to the population, particularly poor people.It has two mainobjectives: (1) to analyze management practices relevant to current public expenditures and subsequently advise Lao PDR how to improve the delivery of basic education and health services; and (2) to serve as an independent monitoring tool for the purpose o f assisting government inimprovingaccountability. Two recent World Bank reports lay out the context for PETS to examine the relationshipbetweenpublic expenditures and education and healthoutcomesinLao PDR. The Lao PDR Public ExpenditureReview focuses on public-based revenues and expenditures at the central and provincial levels as expressions o f public policy and public participation in the economy. The Lao PDR Poverty Assessment reviews poverty indicators at the micro level over time and by region. It also assesses the impact o f growth and public actions on the extent and causes o f poverty and inequality. Yet, information on actual budgetary spending o f the center, provinces, and districts in Lao PDR is scarce. Budget allocations are a poor proxy for how services reach intended beneficiaries. Consequently, little i s known about the budget process that exists between the frontline facilities and the center as well as the impact o f public spending on education and health outcomes. PETS aims to close such gaps by tracking the flow of resources at the subnational level and by providing information on frontline service delivery. It links downstream with an analysis from Poverty Assessment on education and health outcomes at the household level, and upstream with an analysis from Public Expenditure Review on public expenditures o f the layers o f the nation, provinces and districts. 2 Typically, PETS tracks how public resources flow from the center to the frontline, then identifies how much of the original allocations reach each level and how effectively the services are delivered. In Lao PDR, the fiscal structure i s highly decentralized. Provinces and districts are responsible for delivering primary social services to households. Due to the weakness in financial controls, accounting, and reporting systems, records are not always available or reliable on the amount and timing o f the flow o f finds from each province to each district within that same province and from each district to the facilities within it. The main source o f capital investmentinLao PDR comes from donor support. To track government expenditure, PETS focuses on recurrent spending. In the education and health sectors, human resources are the principle vehicle for the delivery o f social services; as a result, salaries account for the bulk o f recurrent expenditures. For tracking purposes, PETS focuses on the linkage between districts and facilities, particularly the flow o f salary payments to teachedhealth workers from districts to schoolshealth centers. This PETS complements the 2002/3 household survey Lao Expenditure and Consumption Survey I11 (LECS), which offered a comprehensive and nationally representative picture of social service delivery. It documents staffing and equipment at frontline services, including what goes on in schools and health centers; what teachers and nurseddoctors do; and how textbooks and drugs reach people. This report has three distinct features: It is based on the 2002/3 LECS sample which provides nationally representative findings; the sample is based on 17 provinces, 54 districts, and 252 villages; This link implied the need to focus on lower levels o f service delivery, namely districts and facilities; and Government priorities suggested focusing on human resource issues, including salary levels andpayment delays. Education and Health Sector Context Lao PDR has made strides since the launch of economic reforms. As one o f the poorest countries in East Asia, Lao PDR has a sense o f urgency about its development. The NSEDP lays out the specific targets and indicators for poverty and social sectors for 2010, it also endorses the 2015 MillenniumDevelopment Goals (MDG) for Lao PDR as set forth in the 2004 MDG-Report. They reflect an unprecedented global commitment for reducing poverty, presented as time-bound, outcome-based targets to focus strategies. Low income, illiteracy, poor health, gender inequality, and environmental degradation are all aspects o fpoverty. Five o fthe eight goals concern health and education. 3 Lao PDR has made progress towards achievin'g education and health MDGs; nevertheless,challenges remain. Progress has been positive inthe following areas: 0 gender equality: the ratio o f girls to boys inprimary and secondary school has increased; so has the number o f literate women to men from the 15 to 24 age group; 0 death rates from malaria and tuberculosis have decreased; protection in malaria risk has increased, and the proportion o f detected and cured tuberculosis cases have gone up; and 0 access to water and sanitationhas also improved. Progress i s mixed on the following: 0 there has been a reduction inthe infant mortality rates and for children under- five; however, the proportion o f children immunized against measles has declined; and on a positive note, there has been a drop in maternal mortality rates and an increase inthe use o f contraceptives, but the share o f births assisted by skilled personnel has increased only very slowly. Progress i s lacking: 0 achieving universal primary education i s a distant goal because increases innet primary school enrollment rates have been slow according to LECS; the same i s true o f completion rates for primary school and literacy rates for young adults aged 15 to 24 years. However, the assessments should be qualified in four ways. First, there are discrepancies between administrative and survey data for some indicators, especially school enrollment. Second, the data refer to national trends only and do not shed light on rising or falling disparities across subgroups. Third, some information is up to five years out-of-date and does not reflect recent changes. Finally, meeting the targets will be a stiff challenge. For example, the targets for poverty, gender equality, malaria, tuberculosis and water will be met only if current trends persist. Maintaining progress i s different from launching it. As easy gains are exploited, further advances are increasingly more difficult. The highly decentralized fiscal system Lao PDR is one of the most decentralized countries in East Asia. Subnational administrations raise around three-fifths o f total revenues and spend almost one-half o f total expenditures (Figure 1). Provinces have defacto autonomy in collecting revenue, in allocating it within envelopes set by the center, and in establishing their budget process (Box 2). Provinces and districts are responsible for the delivery o f most services in 4 education, health, transport infrastructure and rural development. These are the sectors GOL identified as priorities for meetingNSEDP's goals. Figure I:Lao PDR 's subnational authorities account for almost one-half of total government spending, a large share by East Asian standards Subnational Spending (Percent of National Spending) Latest Available Year I I UChina Olndonesia 1 I 0Phillipines OCambodia OThailand 0 10 20 30 40 50 60 70 Data source: World Bank (2006b). Provincial governors play a very important role in public financial management. Budget resource flows are horizontal instead o f vertical. District and provincial offices collect and retain all revenues within the areas assigned to them (income taxes and some tax on natural resource from all payers, taxes on profits, excise and turnover taxes from medium-sized payers). Then these offices use the revenues to fund their budget expenditures. Provinces that collect tax revenues have the ability to negotiate how much tax revenue they will remit to the central government and how much they will retain. In Lao PDR, the budget allocations depend mainly on negotiations between provincial governors and the central government instead of explicit rules. At the provincial or district levels, the allocation o f resources to different sectors, including education and health, i s up to the provincial governors rather than the central ministries (World Bank, 2006d). Therefore, the amounts spent on each sector (or as a share o f total budget expenditures) vary drastically across provinces. As to the sub-national administration, the provinces essentially mimic the center (Box 3). Provincial finance offices comprise the following: treasury, customs, tax, and budget departments. These departments are controlled by the office o f the provincial governor. A similar structure exists at the central level and i s under the Ministry o f Finance. Administrations with deficits (districts or provinces) can receive a transfer o f funds from administrations with surpluses. These transfers, which go from the higher level to a lower one, are block grants. Such transfers come through the Provincial Financial Offices (PFO) and District Financial Offices (DFO) of the Ministry o f Finance; they are not earmarked by sector or service. 5 Box 2: Central-Local Relationships The 1991 Constitution laid the foundation for a national budget based o n the concept o f a unified, decentralized state. Article 75 o f the Constitution specifies that there are three layers o f subnational administration. Provinces and districts formulate budgets. The budget, authorized by the National Assembly, defines revenue targets and allocates expenditures for the center and for the provinces. In 2000, Prime Ministerial Decree 01 (PM 01) extended this framework by granting provinces wide responsibility for fiscal management. This decree established "provinces as the strategic unit, districts as the planning unit, and villages as the implementing unit." The decree granted the provinces defacto autonomy incollecting revenue, inallocating it within envelopes set by the center, and inestablishing their budget process. The decentralized fiscal structure is based o n an "upward revenue sharing" system inwhich the provinces collect most revenues. Richprovinces are to transfer surplus revenues to the center to fundboth central government expenditures and transfers for the "deficit" provinces. However, rich provinces have little incentive to give to poor provinces. An equivalent transfer system from surplus to deficit districts, with similar incentive problems is in place within provinces. Lao PDR's central-provincial fiscal system is portrayed in simple terms in the diagram below. Provinces collect about 60 percent o f total revenues, 90 percent o f which they retain. These funds account for about one- half o f total provincial expenditures. Provinces with a surplus after expenditures remit the difference to the center; they also collect their own assigned revenues. Provinces that have budget deficits receive unconditional block grants from the central government. There is no provincial borrowing. Provinces can and do incur liabilities on behalf o f the center by runningup arrears on commitments to private contractors or civil service salaries. But they cannot borrow from financial institutions intheir own right. h Central revenues Remittances Block grants Subnational authorities have extensive fiscal and functional responsibilities; they also control the exercise o f these responsibilities. For example, they are incharge o f delivering infrastructure and services related to: primaryand secondary education primary and secondary health water supply and reticulation agriculture support and extension; non-national roads and other transport infrastructure; and environmental management. Source: World Bank (2006~). 6 Box 3: SubnationalAdministration When it comes to administration, the provinces essentially mimic the center. The Governor's Office i s equivalent to the Prime Minister Office. Every line ministry has a provincial line agency that reports to the line minister and the Governor's Office. Similar to the central government, the Provincial Planning Office and to a lesser extent, the Provincial Finance Office, play a key role in making decisions about expenditures. But they are always under the supervision and approval o f the Governor's Office. Provincial governors have important but not exclusive controls over senior appointments in provincial and district level offices o f line ministries. However, the role o f the districts inthe provincial structure is less significant than that o f the provinces in the national context. District governors do not have significant autonomy in sectoral allocations or in funding their districts. Policy malung is concentrated I at the provincial level. Source: World Bank (2006d). The provision of primary services in Lao PDR is the exclusive responsibility of subnational authorities. Provincial expenditures for education and healthrepresent more than two-thirds o ftotal spending. This makes the effectiveness o f subnational spendinga crucial concern o f the central government. GOL i s having an internal dialogue to develop consensus on these issues - and on the actions that may be taken to address them. Accountability-Long- and Short-RouteRelationships This PETS report refers to a framework of long-route and short-route accountability as set forth in the 2004 World Development Report-Making Services Work for Poor People. The framework i s a flexible device for discussing a range o f issues that impact the delivery o f basic services for education and health in Lao PDR. The service delivery chain can be unbundled into the accountability relationships between three sets of actors-the policymakers; the frontline providers o f services; and the citizens or clients who utilize the services-and two routes o f accountability - "short route" from clients to service providers and "long route" from citizens to policymakers to service providers (Figure 2). "Short route" o f accountability-in a competitive market transaction, studentdparents and patients as clients hold service providers schoolshealth centers accountable through the power o f the purse; they pay for satisfactory service or take their business elsewhere. "Long route" o f accountability applies to many basic services, such as education, health, water, electricity, sanitation. Inthis situation, service providers are not directly accountable to clients. The societies have decided, frequently for good reason, that the government will provide, finance or regulate these services. Under the long-route, clients as citizens must influence policymakers; policymakers in turnmust influenceproviders to bringabout effective service delivery. 7 Figure 2: Theframework of accountability relationships Long Route o f Accountability J I Short Route of Accountability The triangle in figure 2 highlights what is required for effective public action: a state that is strong in setting objectives, enforcing regulations, and providing adequate resources; providers that have a clear vision and mission of service, who act with professional autonomy and initiative; and empowered citizens who demand services. These three actors should be embedded in strong relationships o f accountability, but establishing such relationships in primary education and primary health care is difficult. Private and collective objectives have to be aligned. When relationships along either "long-route" or "short-route" of accountability break down, service delivery suffers. Inthe education and health chapters, on the long- route, this report focuses on: (i) public funding to discuss the relationship between citizens and policymakers; and (ii) expenditure management to discuss the relationship between policymakers and service providers. On the short-route, this report focuses on frontline services to discuss the relationship between clients and service providers. However, it is not easy to impose and enforce outcomes on service providers because of the multiple and complex objectives of education and health. First, primary education and health services are discretionary and transaction-intensive. It i s difficult to know whether they worked well. Second, the day-to-day pressure o f local demand for health care can sometimes compromise efforts in disease prevention and other public health activities that are not demand-driven, inparticular when resources are tight. Third, education and health outcomes cannot be solely attributed to the actions o f service providers; there are important "co-producers". For example, health and education outcomes are closely associatedwith household and community behavior. 8 PublicExpenditureTracking Survey-Design, Methodology and Limits The Lao PDR Public ExpenditureTracking Survey Project beganin June 2004 as a partnership between the Government of Lao PDR and the World Bank. Ten government officials initiated it at an international training workshop inCambodia (Table 1). Based on the rapid assessment for determining the availability o f records at various layers o f government, the scope o f PETS was defined and approved by its steering committee. After designing and field-testing the questionnaires in October 2005, the PETS steering committee gave a go-ahead for the field survey in December 2005. The trainees consisted o f 40 enumerators includingnine supervisors. Based on a request by the PETS steering committee, the enumerators were experienced employees o f the provincial health and education offices and provincial statistical departments. The field work took place from January 31to March 11, 2006. After data entry and cleaning, NSC submitted the dataset to the World Bank for analysis in September 2006. It is worth noting that the Ministry o f Finance, Ministry o f Education, Ministry o f Health, Department o f Public Administration at the Prime Minister's Office, the National Statistics Centre, and the World Bank participated throughout the project cycle. The NSC, which i s widely recognized for its unique capacity for data collection and analysis, was in charge of the survey work. Its official status as a statistical agency i s a requirement for collecting sensitive information on public resources and staffing across layers o f government. A steering committee, working group, and a secretariat coordinated and guidedthe work. Table 1:PETS Timeline 1 International PETSWorkshop, Phnom Penh Jun-04 2 RapidAssessments Sep 04 to Dec 04 3 Questionnaire Design and Field Testing Jun 05 to Oct 05 4 EnumeratorTraining Jan-06 5 Field Survey Jan 06 to Mar 06 6 Data cleaning Apr 06 to Sep 06 7 Analysis and ReportWriting Oct 06 to Mar 07 Given the pilot nature of the project and the complexity associated with its implementation, PETS wanted to keep the evaluation simple and focused on high level priorities. Of the three GOL fiscal reform areas-allocation o f public spending; government budget planning and processing; and revenue mobilization-PETS offers information on the first two. Specifically, PETS explores the impact of public expenditure management and human resources on education and health service delivery by: 0 documenting financial and humanresources at subnational levels; tracking salary payments from districts to facilities; 9 correlating performance with actual service delivery; and proposing policy reforms on expenditure management practices. The survey approach emulates the methodology of the World Bank's public expenditure tracking survey (www.pub1icspending.org). Within East Asia, such surveys have occurred in Cambodia, East Timor, Papua New Guinea and Thailand. However, the issues o f public service delivery in Lao PDR are different from other countries. The content o f the PETS came about based on the findings and recommendations o f the rapid assessment reports for education and health, and after consultation with the government and other stakeholders. The PETS questionnaire consists of three parts (Table 2). First, it includes a series of quantitative assessments o f budget flows at the various levels (province, district, and facility levels). The first goal was to explicitly record the funds that are allocated to primary schools and health centers and to track the funds for salary payments from the district to facility levels. Second, a facility level survey collected detailed information on inputs, including public funds, outputs and outcomes for primary schools and health centers. Information was collected in questionnaires and data sheets. Third, data on community characteristics was recorded through a scaled-down version o f the village questionnaire o f the 2002/3 Lao Expenditure and Consumption Survey. The focus of the survey is to collect information at the facility level to relate it upstream to the province and district levels, and downstream to the household levels. That is because primary education and health are the responsibility o f provinces and districts; and because the users o f schools and health centers are households. The PETS sample was based on schools and health centers invillages surveyed by the 2002/3 Lao Expenditure and Consumption Survey ("LECS villages"). One operational advantage o f this sampling was that the NSC was familiar with the location and village representatives. An analytical advantage o f this sampling design was that this PETS was derived from a nationally representative household survey that covered 17 out o f 18 provinces and 56 out o f the 141 districts (Figure 3).'In contrast, most PETS take place in selected areas o f a country to conserve spending on the cost o f field work. As o f 2003, there were 8,451 primary schools and 717 health centers in Lao PDR. Based on this number and the experience o f other PETS surveys, the target number of primary schools was 259, the number o f health centers was 114. The target was o f f by seven for both the schools and the health centers. The results presented in this report are weighted to replicate the same population as the LECS 2002/03; weighted and unweighted statistics generally do not differ substantially though. Similarly, nominal variables, such as budgetary spending, are deflated with the spatial price index embedded in the 2002/03 poverty line. These steps ensure that comparisons across localities adequately reflect cost differences. The PETS did not interview users o f facilities separately. Village level household information from the LECS 2002/03 was merged with the PETS data to evaluate how characteristics o f facilities and the management o f public expenditures correlate with health and education outcomes. 10 Table 2: Modules of the PETS Questionnaire 1 Province Financial Office 2 District Financial Office 3 Province Education Service 4 District Education Bureau 5 Primary School 6 Province Health Office 7 District Health Bureau 8 Health Center 9 LECS Village Survey Figure 3: District-Level Map of the PETS Survey (Survey Districts are Marked in Dark Green) I .".. * " -I The PETS report presents findings at three levels (province, district, and facility). The aim was to capture the enormous variety in economy, geography, and ethnicity o f Lao PDR that reflects the context and conditions under which service delivery operates (Table 3). At the provincial level, findings are presented separately for surplus provinces as distinct from deficit provinces. Deficit provinces have a weaker economic base and less financial autonomy than surplus provinces. According to their financial status inthe 2003/04 Official Gazatte, five provinces, Borikhamxay, Champasak, Luangnamtha, Savanakhet, and Vientiane Capital are considered surplus provinces; the remaining 12 are 11 considered deficit provinces. There are separate findings for "priority" districts and "non- priority" districts, as defined by the government. O f the 56 districts inthe survey, 10 are priority districts and 46 non-priority districts. Within districts, living standards differ from village to village. There are se arate findings presented for urban areas and rural areas, non-poor areas and poor areas ,Lao-Tai areas and non Lao-Tai areas3.Out o f the P 252 schools covered by the survey, 66 are in urban areas, 186 inrural areas; 168 innon- poor areas, 84 in poor areas; and 191 in Lao-Tai areas, 61 in non Lao-Tai areas. O f the 107 health centers covered by the survey, 33 are in urban areas, 74 in rural areas; 71 in non-poor areas, 36 inpoor areas, and 91 in Lao-Tai areas, and 16 in non Lao-Tai areas. Due to geographic remoteness and the lack o f development o f humanresources, the rural areas, poor areas, and non Lao-Tai areas are generally less developed than urban areas, non-poor areas, and Lao-Tai areas. Because the services in the former are not as satisfactory as the latter, the report refers to the former as "less well-off areas" and the latter as "well-off' areas. Table 3: Analytic Domains I LAYER DOMAIN OBSERVATIONS (#) Province Surplus 5 Deficit 12 ~ District Non-Priority 46 10 Health Schools Centers Urban 66 33 Rural 186 74 Facility Non-Poor 168 71 Poor 84 36 I I Lao-Tai 191 91 Non-Lao-Tai 61 16 Within provinces, there is a total o f 143 districts. Of these, 47 are considered priority districts inneed o f public investments to attain their poverty reduction targets. These 47 districts were identified based on a set *o The f household, village, and district level indicators for basic minimumneeds. poverty rate o f a village reflects the average level o f poverty across 15 households in that village, based on interviews for the 2002/03 LECS. If the poverty rate o f a village i s hlgher than the median o f the national average, the village i s considered poor; otherwise, non-poor. The population o f Lao PDR can be grouped into four broad ethno-linguistic categories: Lao-Tai, M o n Khmer, Hmong-Iu M i e n and Chine-Tibet. If the share o f Lao-Tai ethnicity in a village i s higher than the median o f the national population average, the village i s considered Lao-Tai; otherwise, it is non Lao-Tai. 12 Ideally, tracking the flow of public expenditures from the central level to the provincial level, from the provincial level to the district level, and from the district level to the facility level would be important. The same i s true about the need to quantify the efficiency/leakage o f this process and to identify possible weak links. However, in Lao PDR, reliable data i s not always available on the timing and amount o f resources that flow from each source to the intended recipients. Reliable data is also lacking on the timing and amount o f resources that recipients receive from each source. This PETS has a narrow focus; it tracks salary payments which flowed to schools and health centers from the districts. It does not address or quantify the leakage o f general public resources from the center to the provinces, from the provinces to the districts, and from the districts to the facilities for any expenditure other than salary payments. Nor i s it intended to pinpoint where and to what extent leakage exists in the overall system. Reliable information on resource inflows from sources other than the government budget is not always available. Therefore, results mustbe interpretedwith caution. This PETS tracked and compared salary payments made to individual teacherdhealth workers, based on records at the district and facility levels. The findings show that leakages o f teachers' and health workers' salaries from districts to facilities is not a big problem. Differences in salaries reported at the facility level and district level arejust a few percentage points. Furtherinformation, such as the amount o f money that comes to the schools or health centers from sources other than the government budget and how it i s actually spent, would be helpful to support this conclusion and to better understand financial management at the facility level. Survey results indicate that schools and health centers rely heavily on off-budget resources from donors, NGOs, parentdpatients and community contributions to finance their operating costs. The quality of survey data is sometimes unsatisfactory. Some respondents at the facility, district, and provincial levels are unfamiliar with accounting. For example, some individual responses to questions about financial surpluses and deficits o f PES/PHO and DEB/DHOwere inconsistent with their responses to another section o f the questionnaire. These responses reflect a lack o f understanding o f basic accounting principles, such as revenues and expenditures. These concepts are fundamental to the profession and pivotal to a public expenditure tracking study. The fact that some respondents failed to grasp these principles undermines the scope and deptho f this PETS report. Given the weaknesses in financial control, accounting, and reporting systems in Lao PDR and the pilot nature of this first PETS, it is unrealistic to expect perfect information. Currently, documentation is almost entirely paper-based. Formats and nomenclatures are often inconsistent. Records are kept in various books and rarely reconciled. Basic accounting concepts are not always understood. These realities hampered PETS inthe collection o f accurate information. 13 Conclusions This PETS aims to provide information on primary education and frontline health services. It also sets out to track salary payments to teachers and health workers from districts to facilities. The survey i s designed to link downstream with an analysis from Lao PDR Poverty Assessment; and upstream with an analysis from Lao PDR Public Expenditure Review to track the flow o f resources from sub national levels and to provide information on frontline education and health service delivery from a national perspective. The report sets out to document how resources flow through layers of subnational authorities to reach schools and health centers. It examines what factors affect frontline services by referring to a three-way accountability relationship between clients/citizens, policymakers, and service providers. It presents a snapshot o f financial flows and human resources at the provincial, district, and facility levels separately for surplus provinces and deficit provinces, priority districts andnon-priority districts, urban areas and rural areas, poor areas andnon-poor areas, and Lao-Tai areas and elsewhere. 14 EDUCATION Education is a vital part of a person's capacity to lead a life he or she values. More education is linked to better family health, greater participation in society, higher productivity, and thus a reduced risk ofpoverty. Education outcomes improved in Lao PDR in thepast decade, although in the aggregate they are still below the East Asian standard. Teachers are satisfied with their jobs and are highly motivated; their rate of absenteeism is low; ghost-teacher phenomenon is uncommon; parents ' and communities participation is active; leakage of teachers' salariesfrom the district level to the schools is low. However, to achieve the level of education identijied by the MDGs and the National Educationfor All Action Plan, moreprogress is needed.I n the last decade improvements in education have not been evenhanded.Disparities between well-off areas and less well- offareas are large. The level ofpublic funding is low; in 2004/2005 it represented only 1 percent GDP. The inequitable distribution of resources and inefficient management of public expenditures have reduced the effectiveness of education spending, which is limited at best. Two factors constrain service delivery at the facility level-low salaries and frequent delays in thepayment of teachers' salaries as well as inadequate operatingfunds. Both of these are the result of insufficient public expendituresfor education. Poor children in less well-off areasare more likely to suffer even more because they have less access to school, particularly schools of decent quality. This chapter: (1) reviews the changes in education outcomes; (2) examines the level and allocation ofpublic educationfunding; (3) analyzes the weaknesses in the management of public expenditures; (4) presents a snapshot of education service delivery at thefrontline and an analysis of its key constraints; and (5) conclusions. Outcomes The overall education outcomes in Lao PDR improved in the past decade. Both survey data and administrative data suggest that Lao PDR is making progress towards meeting the education MDGs, although administrative data displays a more favorable picture. According to the LECS, the net primary enrollment rates increased from 61 percent in 1992/3 to 65 percent in2002/3; incontrast, 2005 Population Census indicates a 15 net primary enrolment rate o f 83 percent in 2005. However, enrollment in first grade does not guarantee completion o f the primary cycle. According to the LECS, no more than 70 percent ofmale students and fewer than 60 percent o f female students are still in school at the end o f the fifth grade. Just three out o f four boys and two out o f three girls who completed five years o fprimary education go on to secondary school. However, disparities in school enrollment widened between priorityhon-priority districts, poorhon-poor and Lao-TaUnon-Lao-Tai areas. According to LECS data, the net enrolment gap for primary education between poor and non-poor children widened from 16 percent in 1992/3 to 24 percent in 2002/3 (Figure 4). The disparity from priority to other districts increased by five percent. Similarly, there are stark differences across the four broad ethno-linguistic categories. Net primary enrolment rates were 76 percent among the Lao-Tai, compared to 49 percent among the Mon-Khmer, 35 percent among the Chine-Tibet, and 47 percent among the Hmong-Iu Mien in 2002/03. Administrative data confirm these differences. Gender disparities have narrowed but slowly. The number o f girls enrolled per 100 boys rose from 82 in 1992/93 to 86 in 2002/3. In 2005, the female net primary enrollment rate was 6 percentage points below the one for male. According to the Lao Population Census (2005), the 2004/05 net primary enrolment rates were 73 percent inpriority districts and 90 percent inother districts; literacy among 11to 16 year-old was 84 percent for Lao-Tai and 61 percent for non-Lao-Tai. To achieve the objective o f the education policy o f NSEDP and that o f the National Education for All Action Plan, universal primary education by 2015, more progress needs to be made, especially inless well-off areas. Figure 4: There are large disparities in education outcomes by poverty, area, and ethnicity. "" i I Poor Non-Poor 47 Pnority Other Lao.Tai Mon-Khrner Chinese- Hrnong-lu Districts Distncts Tibetan Mien Data source: LECS 2002103. 4 The discrepancies in the net primary enrollment rate between the 2005 administrative data and the 2002103 survey data appear large. However, small discrepancies are not uncommon. One reason i s that the admmistrative data tend to capture beginning-of-year enrollment while the survey data tend to capture school attendance. 16 Public Funding The low level and inequitable allocation of public education funding are major constraints in Lao PDR. The less satisfactory education outcomes for poor people and for people inless well-off areas are closely associated with their inferior access to schools and the unsatisfactory quality o f schools they can access. The limited public education funding disproportionately accrues to the non-poor as poor people have little clout with policymakers on budget decisions. They may not be well informed about the quality o f public service or place less weight on public services (Figure 5). Figure 5: Poorparentshtudents have little clout withpolicymakers on budget decisions Based on internationalcomparisons of low-incomecountries, Lao PDR stands out for its scant budget allocations for education (Figure 6). While government spending on education has increased since 1998/99 (Figure 7), it rose mainly because of higher donor assistance. Between 1998/1999 and 2004/05, total public expenditures on education rose from 1.2 percent to 2.3 percent o f GDP while domestic outlays fell from 1.1 percent to 1.0 percent of GDP. Funding is still below its peak in the mid-1990s. Public education spending as a share of overall government spending was about 11 percent in 2004/05 compared to about 15 percent inthe mid-1990s; as a percent o f GDP was about 2.3 percent in 2004/05 compared to around 3 percent in the mid-1990s. Consequently, the education system at the primary level operates with very frugal resources. The unit spending o f a pupil inprimary school i s only 2 percent o f per capita GDP in Lao PDR compared to around 7 to 14 percent in most developing countries (World Bank 2006b). 17 Figure 6: Lao PDR stands out in international comparisonfor its low spending on education. Public Spending on Education (Percent of GDP) 2004/05 or Latest Available Year 0Lesotho OYemen. Rep OMongolia OKenya OMalawl I OBhulan OMoldova OEthlOpla 0 India OGhana OBUNndi OErilrea OComoros OSlerra Leone OSenegal OMauritania OSolomonIslands ONepal OB?"," OK gylRewblic Oh21 OMadaQascar OTaJiklstan OGambia, The ORwanda OTogO OUganda OBangladesh 0Mozambique DNiger OPapua New Guinea .Lao PDR OGuinea-Bissau OChad OZambia OGuinea OCambodia OPaklstan 0 2 4 6 8 10 12 Data source: WDI 2005 and the World Bank (2006b). At each sub-nationallevel, insufficiencyof publicfundingis the single most pressing problemfor education departments, especially in less well-off areas. In2004/05, the approvedbudgets o f all PES were lower than their proposed amounts. For 10 out of the -- 16 PES surveyedthe reason providedwas insufficient funds at the PFO. The bulk o f the education budget adjustment fell on deficit provinces and priority districts. Overall, close to two-thirds o f the DEBs received lower allocation of funds than proposed; the number raisedto four out o f five DEBs inpriority districts. For two-thirds of DEBs, the reason of budget cut i s the lack of funds at the DFO. Figure 7: Whilepublic educational expenditures have increased since the late I990s, it still remains below levelsfrom the mid 1990s. 18 ,I Government Spending on Education as percentages of total government spending and GDP, 1991-2004lO5 I I Government Spending 12 6 Data source: The World Bank (2006b). 18 A shift of public education funding from primary to secondary schools and above diverts the already limited resources away from the people who need it most. A sharp rise in secondary, technical, vocational and higher education enrollments has put a dent in the share o f primary education spending (Figure 8). Primary education spending fell from 49 percent in 1998/99to less than 40 percent in2004/05 out of overall recurrent expenditures on education (World Bank 2006b). Since poor people have a lower stake in secondary and tertiary education than in primary education, this change has made the incidence o f public education spending more inequitable and less consistent with the declared policy focus on achieving universal primaryeducation. Only one-sixth o f public education spending on all programs accrued to the poorest quintile o f population in 2002/03; compared to about 22 percent of public spending on primary education accrued to the poorest quintile. Figure 8: As education spending has shifted awayfrom primary education, it puts poor people at afurther disadvantage. Allocatlon of Education Expendltures by Level ("A), 199912000 to 2003/04 199912000 2001102 2002103 2003104 IEarlyChildCareandEducationlPrlmaryEducatlon Secondary Education and Above .Administration Share of PublicEducationSpending Accruingto each Quintile (%), 2002103 I Rcnes' -) A Programs 4th 3ra I 2nd Poorest 0 5 10 15 20 25 Data source: The World Bank (2006b). 19 ExpendituresManagement Lao PDR's management of public expenditures for education is weak. For policymakers to ensure schools to deliver efficient service there i s still a long way to go to improve budget preparation, execution and reporting, and to strengthen local capacity (Figure 9). Figure 9: Policymakers may not be able to ensure schools deliver efJicient services. BudgetPreparation Five weaknesses-lack of realism, a significant amount of off-budget funds, fragmentation, a weak bottom-up process, and a compressed budget cycle- characterizethe budgetpreparationprocess.Onthe plus side, the process is largely in line with regulations (Box 4). 1.Lack of Realism The provincial budgets approved by the National Assembly as part of Lao PDR's unified budget are often lower than the proposed budget. The latter is based on revenue forecasts that are unrealistically high. Sound fiscal practice demands that expenditure plans should be based on realistic forecasts o f fiscal revenue. The lack o f budget planning realism leads to ineffective budget execution. Highrevenue projections serve the center's agenda in terms of mobilizing highrevenues from the provinces, while surplus provinces have little repercussions to fear when they do not meet those targets. When a large share o f these revenues fails to materialize, budget becomes a weak planning tool. GOL has to reduce budgeted spending amounts by cash rationing. As budgetingtends to be incremental,unrealistic expenditure envelopes perpetuate themselves. For the 2005/06 budget, 15 out of 17 PFO that provided budget ceilings to PES based them on last year's (planned) expenditures. This process can seriously distort the priorities embedded in the budget, as education and health expenditures can often be squeezed below budgeted amounts when revenues fail to materialize. It also weakens the policymakers' authority as treasury officers-rather than the National Assembly-end up deciding which invoices to pay and which to add to accounts payable. 20 Box 4: Budget Process in theEducation Sector 1. The Office o f the Prime Minister (PM), the Committee o f Planning and Investment (CPI). and the . Ministry o f Finance prepare generalframework o fprocedures and guidelines, which are sent to MOE and the other ministries. 2. MOE produces guidelines outlining nationalpriority and targets for the education and training sector, which are sent to all Provincial Education Services and higher education institutions (HEIs - NUOL, TTC, and others) in April, and the PESs send the Ministry o f Education guidelines to District Education Bureaus. 3 . The DEB prepare their draft budgets, which are returned to the PES. 4. The PES summarize and consolidate the DEB draft budgets and their own budgets, w h c h are submitted to the Provincial Planning Service (PPS) and Provincial Finance Office (PFO). The draft budgets prepared by PES and HEIs in M a y contain both the investment and recurrent needs o f the eachprovince and the various institutions. The PPS and PFO submit the draft consolidated budgets to the Department o f Planning and Cooperation (DOPC) and Department o f Finance (DOF) inthe central MOEoffice. 5. DOPC and DOF review the draft budget plans submitted and consolidate them into a single MOE draft budget. 6. The MOE draft budget i s represented at a meeting, chaired by the Minister o f Education in June. At this meeting, MOE priorities are fixed based on the CPI and MOF guidelines and the provincial budget requests. The outcome o f this meeting i s the consolidated first draft o f MOE's budget, which is submitted to CPI and MOF inJuly. 7. When all ministries have submitted their draft budgets, CPI and MOF chair several seminars to discuss the financial report, anticipated revenue, and the anticipated loan-grants activities for the forthcoming year. Based o n the outcomes o f the discussions and the submitted draft budgets from the various ministries, CPI and MOF prepare a final budget plan to be submitted in a July government's meeting. 8. The National Assembly then reviews and approves the plan inlate September. 9. When the plan i s approved, the Prime Minister decrees the execution o f the budget plans and CPI and MOF distribute them to the various ministries andprovinces. 10. Since 2000, MOE upon receiving the approved allocated budget from MOF, distributes it as the "school attachment"; MOE i s responsible for Technical Schools, Teacher Training Colleges and the Ethnic Minority Schools, a total o f 30 schools and institutions. 11. At the province level, PFO distributes allocated Education Sector budget to PES and DFO. Source: Noonan (2001). 2. Off-Budgetary Funds Heavy reliance on off-budget funds compromises the effectiveness of the budget process.About two-thirds ofPES receive funding from donors and about two-fifths from communities. Surplus provinces are more likely to obtain donor funding and deficit provinces are more likely to rely on community contributions (Figure 10). The absence of off-budgetary resources in the budget means that the budget does not capture the whole picture o f actual allocation o f resources between national priorities or across provinces. It i s unclear whether and how these receipts are captured in their budgetary financial reporting. 21 Figure 10: While all PES receive governmentfunds, surplus provinces are more likely to obtain donorfunding and deficit provinces are more likely to rely on community contributions. Revenue Sources of PES Budgets (Percent) 100 80 60 40 20 0 Lao PDR Surplus Province Deficit Province Data source: PETS. 3. Fragmentution Finance and planning functions at the central and local levels are fragmented and coordination between capital and recurrent expenditures is weak. Due to low domestic resources, a large part o f education funding comes from abroad. The share o f foreign financing increased from 30 percent in2001/02 to close to 40 percent in2003/04, allowing for a rise in government spending on education. In 2003/04, around four-fifth of capital expenditure was financed by international donors, and almost half of total spending benefited investments. Compared to around 30 percent for most neighboring countries in the region, education spending in Lao PDR is tilted towards capital expenditures. While donors have started since 2002 to finance the recurrent costs implicit in their projects, there are concerns about inadequate domestic funding of recurrent expenditures once donor projects are completed. Otherwise, schools will be built but cannot teach well for lack o f funds to pay teachers, buy textbooks, or carry out maintenance. About 40-50 percent of national education budget is allocated to recurrent expenditures; provincial figures vary significantly. From 2002/03 to 2003/04, the share o f recurrent spending increased from 40 percent to 53 percent, but the 2004105 budget plan reverted to about 46 percent (Figure 11). However, the share of capital expenditures varies widely from one province to another and from one year to another. Investment in capital expenditures can also be erratic within the same province (Figure 12). For example, in 2002/03, Saravane spent 36 percent o f its education budget on capital investments and Sekong spent 76 percent. The following year, Saravane spent only 4 percent o f its education budget on capital investments and Sekong spent 33 percent. This reflects to a large measure the lumpiness and selectivity o f donor funding, 22 but it also raises question marks about the planning and management capacity at the subnational level. Figure 11:Reliance onforeign funding and a lack of domesticfunds are associated with insufficient recurrent spending. Composition of Education Expenditures (%), 2002103 to 2004105 2002/03 2003104 2004105 Plan =Recurrent Capital Domestic Capital Foreign Data source: Lao PDR Official Gazette. Figure 12: The share of capital spending as apercentage of total education spending varies by year and by province. EducationCapitalSpendingby Province(%of Total EducationSpending),2002103 and 2003104 WATT XAY SR DSEK CHA nSAV SAR nVIE BORP KHA 2003104 DXIE nXAY nLUP HUA ~~ nOUD BOK LUN #VIE M PHO DXAY SR nATT CHA 1 DSAR SEK I SAV KHA #VIE P BOR # XAY XIE 2002103 nHUA Lup DOUD BOK nVIE LUN PHOM 0 10 20 30 40 50 60 70 80 Data source: Lao PDR Official Gazette. Within the small envelope of recurrent expenditures, budgetary resources are inadequate to cover salaries and operational expenditures. Although the share of teachers' salaries in recurrent expenditures increased in the early 2000s (Figure 13), the 23 level of teachers' salaries in Lao PDR remain low. Teachers' salaries average 390,000 kips, or 39 USD, a month. Salaries are approximately 130 percent o f GDP per capita in Lao PDR, compared to an Asian average of 240 percent. Many teachers reported that they have to hold multiplejobs to supplement their incomes. However, an increase inthe wage-bill has squeezed the already limited resources from non-wage spending. This situation has led to domestic arrears and to acute shortages o f funds for routine operating expenses, including textbooks. There i s remarkable variation inthe allocation of recurrent budgets across provinces. In 2003/04, 15 provinces out o f 17 spent less than the internationally recommended 33 percent on non-wage costs. Vientiane Municipality spent only 14percent (Figure 14). Figure 13: Due to the small budget envelopefor recurrent education spending, the rise in the sharefor salaries has squeezed operating expenditures. Composition of Recurrent Expenditures in Education (Oh),2001/02 to 2003104 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2001/02 2002103 2003/04 MSalaries Teaching Materials IIOperationsandMaintenance Data source: Lao PDR Official Gazette. 4. Weakbottom-up process Although the process works reasonably well at the provincial level, budget preparation at the district and school levels is weak. This puts constraints on the bottom-up process. At the district level, only three out o f five DEB received DFO guidance on their budget preparations before submitting their proposals to DFO. Three- fifths of the DEB submit their budget proposals quarterly; their annual proposals get revised anyhow. Similarly, only 43 out o f the 55 DFO received guidance from PFO; 42 o f them obtained a budget ceiling and received a DEBbudget proposal. Fifty-one of them submitted their annual budget proposals to PFO. At the school level, budget preparation process i s still weaker. Less than one-third o f schools in 2005/06 submitted an annual budget proposal to DEB. Distance to DEB and the transportation convenience can be major reasons: more than two out o f five schools in urban areas submitted proposals, compared to one out o f four in rural areas; more than three out o f five schools in urban areas discussed with DEB, compared to less than two out o f five in rural areas 24 Figure 15). Just about one-half o f the schools had discussions with the DEB after their budgets were approved. Again, there are wide disparities by areas: over three-fifths of urban schools met with DEB, compared to one-third o f the rural schools without access to roads. Similarly, just one- quarter o f the school directors knew whether the DEB had revisedtheir 2004/05 budget proposal. The budget process works reasonably well at the provincial level-for example, in preparing the 2005/06 budget, all PFO provided guidance to PES; in 15 out o f 17 provinces, the PFO indicated a budget ceiling to PHO and PES. All PFO submitted their budget proposals to the MOF for approval, and the MOF provided guidance to all PFO. The MOF also gave a budget ceiling to 16 out o f 17 PES. All PFO submitted their budget proposals to the MOF for approval. However, the weakness at the frontline and district level lowers the overall effectiveness o f budget preparations. Figure 14: Theshare of non-wage spending as apercentage of total recurrent education spending varies widely across provinces. Non-Wage Education Spending by Province (Ohof Recurrent Education Spending) 2002/03 and 2003/04 XAY SR ATT CHA SEK SAR SAV KHA 2003104 VIE P BOR IXIE XAY HUA LUP OUD BOK LUN BVIE M PHO BXAY SR ATT CHA SEK SAR SAV KHA I VXiE P BOR i E XAY 2002103 HUA LUP IOUD BOK LUN VIE M PHO 0 10 20 30 40 50 60 Data source: Lao PDR Official Gazette. 25 Figure 15: Schools in less well-off areas participate less in budget preparations. 50 , Schools Submitting 2005/06 Budget to DEB ( O h ) Lao PDR Urban Rural Non-Poor Poor Lao-Tal Other Schools Discussing 2005/06 Budget with DEB ( O h ) ! 70 1 " 50 40 4 ""I * O t Lao PDR Urban Rural Non-Poor Poor Lao-Tal Other Data source: PETS. 5. Compressed Budget Cycle An overly compressed budget preparation cycle weakens the budget's effectiveness. Budget preparation runs only from May to September. Provinces receive their target revenues and expenditure envelops inJune, and submit their budget outlines inJuly. This leaves almost no time for dialogue among the Ministry o f Finance, sector ministries, and provinces. Budgets are often published late, and the structure and coverage o f both the budget and end-year accounts is deficient. The National Assembly approves only a very high-level budget plan o f two pages ahead o f the fiscal year start date October lst, and budget details are not finalized untilFebruary and are not publisheduntil near the end of the fiscal year or even after it i s over. ExecutionandReporting Budget execution and reporting systems need to be improved. Treasury offices at each level o f GOL lack o f authority. Formally, the treasury offices at provincial level are departments o f the central treasury and report both to the director o f treasury and to the 26 provincial governor. Inpractice, the provincial governor has much greater influence over the provincial treasury office, not least because the staff i s hired at the provincial level. The provincial governor has also a strong influence on provincial education officers. For example, in case o f a funding deficit during the year, 15 out o f the 17 PES would send a request to the provincial authority, compared to only 12 to the PFO. The governor's influence appears less pronounced at the district level. Out of 53 DEB, 35 would request additional funds in case o f funding shortfall from PFO, 12 from PES, 3 from DFO and 3 from the district authority. However, most DEB that face financing shortfalls for salaries turnto the provincial governor for additional funding. The financial accounting and reporting systems of the treasury department of the Ministry of Finance are weak. Until 2005, treasury reports have been almost entirely paper based, with provincial information available late if at all. The lack o f automation, together with the lack o f agreement on a reporting format, made it difficult to generate regular, timely, and accurate reports on budget execution. By early 2006, all PFO were equipped with computer, telephone and fax, and all but one had a software program for accounting. Still, only one PFO sent its reports to M O F by email, and 12 PFO used fax. Nine in 17 PFO report to MOF only on a monthly basis, and another 2 even less frequently. The situation was worse at the district level. Some 10 percent o f the DFO did not have computers. Overall, one in five DFO used a software program for accounting, butnone o fthe DFOinpriority districts. While allbut one DFO receive PFO supervision visits, they occur no more often than every four months in23 DFO out o f 54 DFO, and 5 DFO report their financial situation to the PFO only on a less than monthly basis. Currently, the new Treasury Budget Systemi s being rolled out over the country under the management o f treasury department o f the Ministry o f Finance. This system uses a single, standard Chart o f Accounts (COA) and budget nomenclature. It i s an integrated computerized network system that currently links 39 line ministries, 18 provincial treasury offices, four provincial tax offices and 17 customs offices. While this i s a very positive development, the COA should both be simplified and extended, as they do not allow a functional breakdown. The current system does not allow separating spendingat the primary level from other levels. The accounting and reporting systems in PES and DEB are ineffective when it comes to accountability. All PES own computers and telephones, but only three-fifth could communicate via email with PFO or the MOE. While all PES keep accounting records and issue reports, they use different formats and guidelines according to instruction from MOE, MOF, or PFO. Only four out o f five DEB have computers, and nine in ten rely on telephone to communicate with PES, DFO or MOE. All DEB keep records; however, just 49 out o f 55 DEB surveyed produce accounting reports and just six o f them release these reports to the public. Weaknesses in the budget process and financial accounting and reporting have muchto do with the lack of human and materialresourcesin provincialand district offices. 27 At the provincial level, PES has on average 41 staff, or around one staffper 40 primary teachers. Overall, staff increased in the last five years intwo-thirds of the PES, but in deficit provinces it was only one out o f two. In surplus provinces, three quarters o f the staff interviewed have at least 15 years o f education, compared to only two-fifths in deficit provinces. All PES have electricity, water supply and toilets, and-as a major change over the last five years-computers. Partly due to donor projects, PES in surplus provinces are better equipped than PES in deficit provinces with cars (6.3 versus 5.0) and mobile phones (4.3 versus 3.7). PFO employs on average 226 regular staff and 39 contract workers. Staffing i s inper capita terms around 10 percent higher in surplus provinces than in deficit provinces. All PFO have telephones, computers, and faxes. At the district level, DEBhas an average o f23 staff, or around one staffper 10 primary teachers. This compares to around 25 staff, in addition to 10 contract workers, in DFO. Three-fifths o f the DEB staff interviewed in non-priority districts have at least 12 years o f education, compared to only one out o f two in priority districts. DEB inpriority districts are less well equipped than DEB in other districts (Figure 16). Innon-priority districts, close to 90 percent of DEB have sanitary water supply, and an average o f 1.5 computers, 0.43 cars, 6.7 motorbikes; in priority districts, no more than 75 percent have sanitary water supply, and an average o f only 0.9 computers, 0.19 cars, but more motorbikes (about 9.0 in average). Figure 16:DFO in non-priority districts are better equipped than those inpriority districts. Office Equipment in DFO and DEB (Ratio of Priority District to Other Districts, K) 150 , 125 7 100 - 75 - DFO Phone DFO Computer DEB Water DEB Computers DEB Cars (#) DEB Motorbikes supply (#) (#) Data source:PETS. Frontline Services Making efficient use of funding at the frontline level is critical to the effective delivery of educational services. How well teachers are paid and motivated, how well classrooms are equipped, and how well schools are accessible are keys as the ultimate objective o f funding and expenditure management arrangement i s to contribute to 28 learning inwell-functioning schools with well-motivated teachers. This section elaborates on how parents/communities can complement the long route o f accountability o f schools vis-a-vis policymakers through direct accountability (Figure 17). From the perspective o f schools, this section focuses on teachers' salary, operational resources, accountability, and school access and quality. Figure 17: The accountability of schools topolicymakers andparents is crucialfor education. Teachers Low salaries and frequent delays in salary payments are a major constraint on teachers' performance and the overall delivery of education services in Lao PDR. PETS allows tracking the salary payments fi-om DEB to the schools from the 1346 teachers working inthe survey schools. No evidence from the survey suggests that salary leakage from the district level to the facility level or ghost teachers i s a major concern; the system o f teachers' salary payment is rather standard; but inadequacy o f resources remains a keybottleneck. In Lao PDR, about 95 percent of the teaching profession are civil servants; just a handful are contract or temporary/volunteer teachers (Figure 18). Such ratio does not differ much in less well-off areas and in well-off areas. Quota teachers are regular civil servants; contract teachers receive salaries in line with the government's pay-scale but do not receive the social security benefits; temporary/volunteer teachers are usually hired by the community at negotiated wages, which are lower than those o f quota teachers. 29 Figure 18: More than nine out of 10teachers are quota teachers. Compositionof Teacher By Contract Type (Percent) Lao PDR Urban Rural Noneoor Poor Lao-Tai Non-Lao. Tai rnQuota Contract # TemporaryNolunteer Datasource: PETS. In Lao PDR teachers' salaries are low. In absolute terms, average overall monthly salary i s around Kip 390,000, or 39 USD. Kip 320,000 i s the base salary, and Kip 70,000 i s bonuses, family allowance, and supplements. In relative terms, teachers' salaries are around 130percent o f GDP per capita, which i s low based on international standards. In many East Asian countries, teachers' salaries are about two and a half times GDP per capita as teachers are better educated relative to average workers. The level of teachers' overall nominal salary i s similar in different areas (Figure 19). While the better qualification o f teachers in urban areas lead according to the government pay-scale to a higher base salary than in rural areas, the difference is roughly compensated by higher incentives and family allowances for teachers working in remote and less well-off areas. For example, close to four infive schools innon-Lao-Tai villages pay benefits to teachers for multi-gradeteaching, compared to only two infive schools inLao-Tai villages. Inreal terms, since life i s cheaper in less well-off areas, salaries are higher there. 30 Figure 19: Salaries vary little in nominal terms; they are higher in less well-off areas in real terms. Nominal Monthly Teacher Salary Including Allowance (Kip, 2002 nomlnal terms) 500,000 j 450,000 400,000 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 b o PDR Urban Rural Non-Poor Poor Lao-Tal Other m Base Salary Incentives * Family Allowance rn Supplementary Real Monlhly Teacher Salary Including Allowance (Kip, 2002 Vientlane Urban Prices) 500000 1 ...... 450,000 400,000 350,000 300,000 250,000 200.000 150.000 100,000 50,000 0 Lao PDR Urban Rural Non-Poor Poor Lao-Tal Other Base Salary rn incentives 8 ,Family Allowance rn Supplementaly Data source: PETS. The delays in payment of teachers' salaries are serious, particularly in less well-off areas. The data from Ministry o f Education suggests in the last quarter o f FY05 and the first quarter o f FY06 the GOL has paid out the salaries of teachers such that the stock of salaries outstanding more than 1 month came below 6 percent of a month's payroll; and that severe arrears in salaries (over three months) have been eliminated (World Bank, 2007a). However, the PETS results indicate a less favorable p i c t ~ r eIn . ~ Lao PDR, less than 50 percent o f teachers are paid on time every month, almost 40 percent teachers got paid irregularly. While salaries may have a greater purchasing power in less well-off areas, other aspects o f the wage package are less appealing (Figure 20). First o f all, salary payment i s less frequent. For example, the survey suggests that while 70 percent o f teachers get paid monthly in urban areas, this i s just over 30 percent in rural areas. In The numbers are not directly comparable, Discrepancies in salary payment between different sources o f information can come from, for example, sampling and timing. 31 particular, over one intwo teachers inrural areas, poor villages or non-Lao-Tai localities receive wages irregularly. Overall, 70 percent teachers had experienced salary payment delays inthe past year and a half. Salary arrears are more often in less well-off areas: six out o f seven teachers innon-Lao-Tai villages had their salaries delayed at least once since October 2004. One intwo school directors inurban areas report paying salaries on time, compared to only one out o f four in rural areas. About eight percent o f teachers' salaries are delayed every month in cities compared to 18 percent invillages. While five percent o f the delays last three months inurban areas, the number doubled to about 10 percent in rural areas. These findings are also confirmed at the district and province level. DEB report that only one out o f ten salaries i s paid every month on time in priority districts, compared to almost one out of four innon-priority districts. PES state that less than one out o f ten salaries are paid every month on time indeficit provinces, compared to two out o f five in surplus provinces. Given the problems o f low and infrequent payment, it i s little wonder that only five percent of teachers interviewed say that their salary i s sufficient to cover living expenses. 32 Figure 20: Salary payment delays are more serious outside of cities; most teachers have experienced salaries arrears in thepast. Frequency of Salary Payment8 100 80 80 70 eo so 40 30 20 10 o i L.0 PDrn urban Rural Non.Poor Poor L.0.T.I ah.. IMonthly IQuafiefly !regular Other Salary Delays between October 2004 and February2006 (%) 100 - no J -1I Lao PDR Urban Rural Non-Poor Poor School Director Reporting Frequency of Salary Delays 100% 80% 80% 70% 60% 50% 40% 30% 20% 10% 0% Lao PDR Urban Rural Non-Poor Poor Lao-Tal Other rn Every month Irregularly *IOnly at the start of the fiscal year School Director Reportlng Duratlon of Salary Delays (Percent) loo%, I 60% 50% 40'% 30% 2OVO 10% 0% Lao PDR Urban Rural Nom-Poor Pool Lao-1.1 Other IFewweeks IOnemonth 2-3 months rn Mote than 3 months Irregularly Data source: PETS. 33 The survey results suggest that the system of salary payments is standard. However, the system itself is not a major factor in the delays o f salary payments. In most cases (over 80 percent), school director collects the teachers' salaries from DEB in return for signing a DEB'Sregistration form, and hands out the salaries to the teachers, once again inreturn for them signing a receipt form. Only in some five percent of cases no receipt forms are signed by either school director or teachers. Virtually all salaries (97 percent) are paid in cash. In the other cases, school director receives the salaries from DFO, or DEB itself delivers the salaries to school, or some other arrangement is in place. However, inremote areas without road access, travel to the DEB takes time and i s costly. Almost one out o f four schools pay a fee for a messenger to fetch salaries from DEB every few months rather than every month. Anecdote suggeststhat the fee for messenger to cover transport costs can be as highas one-third o f a teacher's salary. This also can be an important factor that leads to salary payment delay, inparticular inremote areas. There is little evidence to suggest that salary leakages between DEB and schools are a major concern. Deviations in the amount of salary payment occur in both directions; however, average difference i s small between what i s reported at the district level and what i s reported by school (Figure 21). For overall salaries, the DEB and school reports coincide exactly for about two out o f five survey cases; the DEB reports lower payments than the school for about one out o f four survey cases; and DEB reports higher payments than the school for about one out o f three survey cases. The average difference between payments reported by DEB and school is less than three percent of the school reported overall salary. When we consider teachers' base salary only, then over 55 percent o f salary reports coincide exactly between DEB and the school, and the average gap i s no more than one percent o f the school reported base salary. There does not appear to be a systematic salary leakage at a significant scale between DEB and schools. These findings confirm the impression obtained during the preparatory qualitative field work for the PETS. They also match the findings from a nationally representative survey conducted in May/June 2006 in Lao PDR, which suggests that salary leakage o f secondary school teachers i s low-among the 155 lower secondary schools surveyed, 58 percent o f teachers responded that they paid a small "facilitation fee", which amounted on average to 11,000 Kip out o f a 437,000 Kip payment or 2-3 percent in relative terms, to receive their salaries (World Bank 2006e). 34 Figure 21: Differences in teachers' salary payment are small between reported by DEB and schools Teaohwd Ovrrall Salary Dlsprnsrd by DEB and Recdved by Toachrrr (%) 100% SO% 80% 70% 80% 50% 40% 30% 20% 10% 0Yo Lao PDR Urban Rural Non-Poor Poor bo-Tal 0th.r m School=DEB m SshoolcDEB SchoolDDEB Dlfference between Overall Salary reported by DEB and Schools (as Percent of Overall Salary reported by Schools) 4 0 3 5 3 0 2 5 2 0 1 5 1 0 0 5 0 0 Teachers' Base Salary Dispensed by DEB and Rocolvod by Teachom (%) 100% 00% 80% 70% 80% 50% 40% 30% 20% 10% 0% Leo PDR Urben Rural Non-Poor Poor Lao-Tal 0th.r ISchool=DEB ISChOOlDEB Difference between Base Salary reported by DEB and Schools (as Percent of Base Salary reported by Schools) 4.5 - I 4.0 -fl 1.5 - II 1 0 - I Lao POR Urbmn Rural Nan-Poor P D O I Lmo.T.1 Other Data source: PETS. 35 There is also no evidence of widespread ghost-teachers. The survey suggests that only 16 out o f 1360 names recorded inthe DEB forms could not be matched to active teachers in schools; only 40 teachers registered on school records did not appear on DEB lists. These minor differences could reflect reporting discrepancies, time lags in the updating o f teacher registrations in case o f new appointments or promotions, or community volunteer teachers. Only one out o f three schools report the number o f teachers to DEB and DFO less frequent than every month. Insufficiency of funds is more likely to create a bottleneck that leads to delays in salary payments. Teachers' salaries are supposed to be fully covered by central budget. However, more than 50 percent o f DEB reported government funds are insufficient for teachers' salaries; and the number i s almost 70 percent for DEB in priority districts (Figure 22). When facing insufficiency o f resources, overall about 50 percent DEB rely on funds from the provincial governor and 30 percent rely on community and parents to make up the difference; however, the numbers are about 85 percent and 15 percent respectively for DEB in priority districts. Correspondingly, DFO are also faced with salary delays. Overall, two out o f five DFO consider that one reason for salary delays i s the delays in district tax collection; however, the number i s about seven out o f ten in priority districts. Discrepancies in timing o f resource availability and relatively high transaction costs also contribute to salary payment delays, particularly in remote poor rural areas. 36 Figure 22: Funding shortage in DEB and DFO creates a bottleneck that leads to salary delays. DEBSwith Insufficient Government Fundsfor Teachers' Salaries ('4 70 LBOPDR Other Districts Priority Di6tmts Sources of DEB Funding to Supplement Government Funding for Teachers' Salaries (%I Lao PDR Other Districts Priority Districts a Province Governor international donors Community 8 Parents NOther DFO Reasonsfor Salary Delays ( O h ) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Lao PDR Other Districts Prioritv Districts IL District tax collection dela ed NationalAssembly budgerdelayed NDEWDHO requestdelayed Province transfer delayed Data source: PETS. Most teachers have high job morale despite their low salaries, frequent delays in getting paid, lack of training opportunities, and a heavy workload due to multi- grade teaching. The last was particularly true inrural areas. Nineteen out o f 20 teachers surveyed reported they were satisfied with their overall work at school (Figure 23). Three factors may be in favor of a good education service delivery. First, many teachers, often the majority, are driven by an intrinsic motivation to serve. B e it through professional 37 pride or a genuine commitment to help poor people (or both), many teachers are dedicated to deliver timely, efficient, and courteous services, often in difficult circumstances (Figure 24). Second, a higher share o f female teaching staff may also have contributed to the equalization o f enrolment rates. Third, most school directors and teachers work in their home village or districts, or know the local language (Figure 25) Even in non-Lao-Tai villages, only one out o f sixth teachers does not speak the local language. This has clear advantages. Because local teachers know the local language and customs, they are likely to be better able to communicate with students andparents. Figure 23: In spite of the dijjcult circumstances, teachers like their work and became teachersfor the right reasons. Teacher's Satisfaction with Work (%) Lao PDR Urbmn Rurd Non-Poor Poor Lao-Tal Non-Lmo- T I , IVerymuch IWellenough More or less INotmuch Not at all Teacher's OccuDational Choice PA) Lmo PDR Urban Rum1 Non-Poor Poor Lco-Tal Non-Lao- TI, Ilikeworkingwithstudents I ITherewerenootherjobs My parents suggested it ITeashers are hlghlvrespectad Other Data source: PETS. Figure 24: Teachers are dedicated to improve education service. Teacher's Job Motivation (%) 100% SO% 80% 70% 00% 50% 40% 30% 20% 10% 0% Lao PDR Urban Rural Non-Poor Poor Lao.T.1 Non-Lao- 1 - 1 IlmprovseducationInvlllaae Salary 18 good INootherjobavailable Data source: PETS. 38 Figure 25: Most school directors and teachers are local or know the local language. Place of Birth of School Directors (%) 100% 90% 80% 70% 60% 50% 40% 30?`o 20% 10% 0% Lao PDR Urban Rural Non-Poor Poor Lao-Tal Non-Lao- Tal IVillage IDistrict Province IElsewhere LSOPDR Urban Rurml Non-Poor Poor Lao-Tal Non-Lao- TmI INative rn Local language Elsewhere Data source: PETS. Teachers' absenteeism rate is low in Lao PDR. Overall, just over eight percent o f teachers in survey schools were absent at the time o f an unannounced visit (Figure 26).6 Absenteeism rate is slightly higher inrural areas (almost 10 percent) and innon Lao-Tai areas (about 13 percent), but still low compared with other developing countries where teachers' absence rates range from 11percent (Peru) to 27 percent (Uganda) (Chaudhury et al., 2005). In addition, only about two out o f five teachers who were absent from work were for reasons other than official reasons or sickness (Figure 27). However, most teachers live near schools (within a ten minute walking distance). The low rate of absenteeism may be an underestimate. Teachers who were not working at school at the time o f the visit could come back to school to be "present" ina short time if notified. 39 Figure 26: Teachers' absenteeismrates do not appear to be high. Absenteeism Rate of Teachers (%) I I Lao PDR Urban Rural Non-Poor Poor Lao-1.1 Other Data source: PETS. Figure 27: About two out offive teachers who were absentfrom work were absentfor reasons other than official reasons or sickness. Reasons Whv Teacher is Absent (%) Lao PDR Urban Rural Nan-Poor Poor Lao.Tal Other S Official Reasons W Sickness Other Data source: PETS. However, modest salaries and frequent delays in payment result in teachers' having to hold multiple jobs. This negatively affects education service delivery. Overall, about one out o f three teachers surveyed reported they hold multiplejobs. Multiplejob-holding i s more common in urban areas, non-poor areas, and Lao-Tai areas, partly because teachers' salaries there are lower inreal terms and partly because teachers there are more qualified and may have better opportunity to find secondary (or tertiary) job to work on (Figure28). Most teachers reportedthat they have to hold two or morejobs to make ends meet. Many secondary jobs, however, are agriculturaljobs. For example, among teachers who are absent for personal reasons, four out of five are absent only at the busiest season for the farming. Although most teachers are eager to contribute to the community and have highmorale inteaching, the low level o f salary and frequency o f salary delays does 40 affect teaching negatively. Multivariate analysis suggests that teachers with long and frequent delay are likely to spend less time on lesson-preparation and to be absent from work for non-official reasons. Figure 28: Teachers in urban, non-poor, and Lao-Tai areas are more likely to hold multiplejobs. 1 Percentage of Teachers Who Hold Multlple Jobs so - 40 - Lao PDR Urban Rural Non-Poor Poor LLIO-TLII Other Data source: PETS. Operatingfunds and school quality The limited amount of resources for school operating costs is inequitably distributed.Funding for teachers' salaries is supposed to come from the central budget. In contrast, funding for operating costs, such as equipment, textbooks, and teaching materials, mainly comes from resources that schools and DEB can mobilize. Nine out o f 10 schools raise revenues from parents and communities. As a result, the economic circumstances of the locality directly impact the school's finances (Figure 29). Schools in less well-off areas are more likely to face tighter resource constraints than schools from well-off areas. For example, operating revenues for schools in rural, poor, and non-Lao- Tai areas i s between Kip5,OOO to Kip15,OOO per student. For schools inurban, non-poor, and Lao-Tai areas, it is close to Kip 50,000 per student (Figure 30). Figure 29: Most schools receivefunding from parents and communities. Percentage of Schools Receive Funding From Various Sources for Operational Costs 100 I 1Lao PDR Urban Rural NowPoor Poor Lao-Tal Other UDistrict W NGOs Community & Parents Other 1 Data source: PETS. 41 Figure 30: Schools in well-offareas can mobilize more resources than schools in less well-off areas. School Operating Revenues 2004i05 (Kip Per Student, 50,000 2002103 Prices) 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 Lao Urban Rural Non- Poor Lao-Tal Other PDR Poor Data source: PETS. A large part of the difference is accounted for through the collection of tuition/registration fees. Three out o f four schools inLao PDR collect such fees (Figure 31).Inover two-thirds o f the cases, parents are incharge o f settingthe fee ( Figure 32). Although tuitionhegistration fees stand for a larger percentage o f resources covering operating costs for schools in less well-off areas, schools in less well-off areas are less likely to be able to collect such fees and the absolute amount collected per student there is much lower than that in well-off areas (Figure 33). Tuitiodregistration fees are typically collected at the beginning of the school year, and on average increased from Kip 13,000 in 2003/04 to Kip17,OOO in 2004/05 per student (2002/03 prices). According to the survey, almost universally, such fees are not used to prop up teachers' salaries, but instead to purchase teaching materials, fundrenovations etc. Figure 31: Threeout of four schools collect tuitionhegistrationfees. Tution Fees (Percent of Schools) 1 Lao PDR Urban Rural Non-Poor Poor La0.Ta1 Other Data source: PETS. 42 Figure 32: I n over two out of three cases,parents are in charge of setting tuitionfees. Responsibility for Setting Tuition Fee (Percent) 100% 90% 60% 70% 60% 50% 40% 30% 20% 10% 0% Lao PDR Urban Rural Non-Poor Poor Lao-Tal Other a School Director DEB Parents Other 1 Data source: PETS. Figure 33: Tuitionfees are higher per student in well-off areas. School Tuition Fees (Kip Per Student, 2002103 Prices) 35,000 30,000 Lao PDR Urban Rural Non-Poor Poor Lao-Tal Other Data source: PETS. Many students, especially those in less well-off areas, do not have good access to primary schools with a complete cycle of grades. According to LECS, in 2002/3, one out o f six families had to send their children to neighboring villages for education. According to the 2005 Population Census, only four out o f five villages have primary schools. While some villages have no schools at all, other have only incomplete primary schools. Due to the small number o f children, lack o f teachers or limited facilities, some primary schools offer only the first two or three grades. According to LECS 2002/03 and Lao Population Census (2005), about one out o f three villages had complete schools. The percentages are especially low for the poor, priority districts, and non-Lao-Tai ethnic groups. 43 Many schools, especially those in less well-off areas, do not have adequate basic facilities. Figure 34 portrays the infrastructure characteristics o f the primary schools in 2002/3. Some o f these attributes, such as the share o f classrooms with a blackboard and the presence of a school library, are directly related to pedagogical processes, while others may affect instruction and learning mainly through making the school more comfortable for students (e.g., availability o f drinkingwater, a toilet for students) and for school personnel (e.g., availability o f a principal's office, teachers' room). Less than 10 percent o f schools in priority districts have electricity, compared to about 20 percent in non-priority districts. In most aspects (except library and no-leaky roof), facilities in schools inpriority districts are inworse condition than those in other districts. Similarly, schools in poor areas and non-Lao-Tai areas are less well-equipped than those in non- poor areas and Lao-Tai concentrated areas, although the differences are less pronounced than across priority and non-priority district categories. Figure 34: Schools in priority districts have worse infrastructure. School Infrastructure in Priority and Other Districts (YO) Electrlcltv Data source: PETS. The lack of student textbooks is a big concern throughout the country. Students from less well-off areas face more constraints. On average, two students have to share one textbook. The situation i s slightly better in urban areas, where three textbooks are divided among five students, and worse in non-Lao-Tai communities, where one textbook i s shared among three students. It is not surprising that non-Lao-Tai villages fare poorly in terms o f teaching materials. The last time, schools received textbooks was about 7 % years for teaching guides, and close to six years for textbooks. DEB provide close to three-quarter o f the teaching materials, and, as we have already seen, they function less well in remote areas. In one out o f six non-Lao-Tai villages, there are no textbooks sold at the market, compared to one out o f twenty elsewhere. And if they are available, in non-Lao-Tai village less than one out o f ten children can afford to buy textbooks, compared to close to one out o f three in Lao-Tai villages (Figure 35). The provision o f teaching guides i s better for grades one to grade four. Most schools provide 44 teaching guides on Lao Language, Mathematics and The World Around U s to each teacher. The textbook to student ratio innon Lao-Tai areas is lower than those inLao-Tai areas (Figure36). Figure 35: Many students cannot afford textbooks. Affordability of Textbooks (%) Lao PDR Urban Rural Non-Poor Poor Lao-Tal Other HAlmost everyone rn Only some Not at all Datasource: PETS. Figure 36: Non-Lao-Tai villages havefewer textbooks than other villages. 70 ......_l_.lll .".l..".l_." _._..I.." Textbook-Student Ratio (%).____.__.l._...ll......l.... ill-." .... 1111I I " 11111I 60 50 40 30 20 Ian PDR I eo-Tai Other aGrade 1 Lao Language =Grade 1 Mathematics W Grade 3 Mathematics Grade 2 Lao Language Grade 2 Mathematics IGrade4Mathematics Grade 3 Lao Language Grade 4 Lao Language Grade 4 World Around Us Data source: PETS. Teachers' education and experience are important indicators of school quality. Teachers in well-off areas are in average better qualified. The variation in the education and experience profile of teachers and schools directors i s pronounced (Figure 37). Overall, teachers have about 16 years o f experience and school directors have about 11 years. In Lao-Tai localities, two out o f three teachers have completed upper secondary education; compared to less than one out o f two in non-Lao-Tai villages. The gap i s wider for school directors: two out of five have post-upper-secondary education in cities, andabout one out of eight invillages. 45 Figure 37: Teachers and school directors in well-off areas are better educated and more experienced. Work Experience of School Directorsand Teachers (Years) mSchool Director Teacher Lao POR Urban Rural Non-Poor Poor Lao-Tal Non-Lao- T.1 Teacher Education (%) Lao PDR Urban Rural NowPoor Poor LBO-Tal Non-Lao- --, I 5 + 1 I 5 + 3 8+3 111+1 11+3 IOther School Director Education ( O h ) Lao PDR Urban Rural Non-Poor Poor Lao-Tal Non-Lao- I. a j r 5 + 1 I 5 + 3 8+3 I 1 1 + 1 11+3 IOther 1 Data source: PETS. Multi-grade teaching is more prevalent in rural areas, but teachers there are not as well trained. Many teachers who find themselves in a multi-grade environment, however, are either untrained or trained only in monograde pedagogy and have few teaching resources or learning resources. While only one out o f 20 teachers in urban areas instructs multi-grade classes, one out o f four teachers in rural areas have to do so 46 (Figure 38). Multi-grade classes arise through necessity rather than choice. Parents in rural areas often face the choice between sending their child to a school with multi-grade or to leave the child uneducated. Figure 38: Multi-grade classes are common in less well-off areas. Multigrade Classes (Percent of Teacher) -.... 1 Lao PDR Urban Rural Non-Poor Poor Lao-Tal Non-Lao- Tal Data source: PETS. Teachers' demand for professional training is largely unmet. Nineteen out o f 20 teachers surveyedreported they would like to receive more training to improve teaching skills and to meet other teachers and exchange information. Only about one out o f two teachers attended a teacher training duringthe last 15 months, about halfo f which offered by PES or DEB. Teachers show remarkable effort in getting ready for instruction. In average, teachers in Lao PDR spend 45 minute to prepare for each class; teachers in less well-off areas spend even more time partly because they face more teaching difficulties, such as multigrade teaching; partly because they may have less education or experience (Figure 39). Figure 39: Teachers spend on average 45 minutes topreparefor each class. 60 50 40 30 20 10 0 Lao PDR Urban Rural Non-Poor Poor Lao-Tal Non-Lao- Tal Data source: PETS. The supply of teachers is greater in urban areas than in rural areas. Less well-off areas face more severe constraints o f teacher shortage (World Bank, 2007b). Overall, the average pupil-teacher ratio for primary schools i s about 30 to one (Figure 40). The ratio 47 ranges from around 25 to one in cities to 34 to one in non-Lao-Tai villages. This i s a remarkable variation inview o f the official government policy that allocates an additional teacher to a school when it enrolls another 33 students. The difference emerges largely on account o f demographic trends-while fertility rates have come down sharply in urban areas, they have remained highinrural areas-and the inadequacy o f the quota system to allocate teachers to less well-off areas. Figure 40: Student-to-teacher ratios are higher in less well-off areas than in well-off Lao PDR Urban Rural Non-Poor Poor Lao-Tal Non-Lao- TaI Data source: PETS. Accountability The inequitabledistributionof resourcesfor operating costs betweenpoor and non- poor areas is an important concern. But a certain degree o f financial dependence on communities and parents may help holding schools directly accountable to parents. There may be a tradeoff between the need for greater central authority to capture society-wide benefits, such as social cohesion, and the need for greater local influence because student learning i s difficult to monitor at the central level. Some argue that as long as locally collected fees are retained by the school, fees are a good thing. While centrally controlled resources are exclusively devoted to payroll, resources collected at the school level can increase the total amount o f resources available to school^.^ Also, by financially contributing to schools (in monetary terms or in kind), parents can have a better say in affecting the resource utilization to help tailor the service to their needs. They can also be effective monitors o f providers, since they are at the point o f service delivery. The survey shows that one out o f four school directors meet with parents every week; while only one out o f seven school directors meet with officials from DEB this often (Figure 41). 'Such accountability i s crucial. For many parents there i s no choice o f schools, hence no "voting with your feet." For families inrural areas, there i s usuallyjust one school inthe neighborhood. 48 Figure 41: More school directors meet withparents on a weekly basis than with officials from DEB. Percentage of School DirectorsMeeting with Parents EveryWeek 40% 30% - 20% -- _ _ - ~ I O %-- _ _ - ~ 0% 7 Lao Urban Rural Non- Poor Lao-Tal Non- PDR Poor Lao-Tai Percentage of School Directors Meetingwith DEB Every Week 20% - Lao Urban Rural Non- Poor Lao-Tal Non- PDR Poor Lao-Tai Data source: PETS. Free primary education rests on social, equity and efficiency considerations; the negative impact of a tuitionhegistration fee on enrollment may be limited. First, tuition fees are a small part o f household expenditures on education. According to the 2002/03 LECS, tuition and other fees account only four percent o f household education expenditures in rural areas, and about nine percent in urban areas (King and van de Walle, 2005). The bulk o f private expenditures on education go towards uniforms (45 percent), textbooks and materials (20 percent), transportation (15 percent), and other items. Second, about two out o f three schools exempt pupils from tuition fees when they are unable to pay, and others request inputs in-kind. Only three percent o f the schools 49 dismiss student in case they cannot afford tuition fees. Third, more children fail to enroll inschool for different reasons than direct expenses (Figure 42). Among children aged 10 to 18 years who have never attended school, the direct cost o f schooling (rather than the opportunity cost such as lost income from work) i s preventing no more than two percent to go to school. Instead, nearly two out o f five are not interestedinschool8; over one out o f four consider the schools as too far way; one out o f seven have to work instead; and less than one out o f ten claim to be too young. Hurdles for school enrollment vary across areas. In urban areas, less than one out o f ten i s not enrolling due to the distance to school, while it is almost three out o f ten in rural areas. Among the urban poor, almost three out o f ten never attended school reported that they have to work, compared to about one out o f seven among the rural poor. Figure 42: More children are kept out of schoolfor reasons other than tuitionh-egistration,fees. Children Aged 10 to 18, 2002/3 W No Interest Too Far Had to Work Too Young WOther Illness =No TeacherslSupplies Too Expensive 1D Language Data source: PETS. Conclusions To achieve education MDGs and the objectivesof the NSEDP education policy, it is important for the Lao government to devote more resources to education.It is also critical to more efficiently allocate public funding to primary education and to target less well-off areas. Although teachers have high job morale and low absenteeism, the low level and delays o f teachers' salaries, insufficiency o f operation funding, lack o f schools with basic facilities, shortage o f teaching materials and textbooks, and inadequacy o f professional training for teachers combined negatively affects the quality o f primary education. Inequitable distribution o f resources between well-off areas and less well-off areas, mismatch o f capital spending and recurrent spending, and shift o f public education expenditure from primary to secondary further lower the effectiveness o f the limited However, reasons behind are to be examined. 50 public spending. Compared to teachers in well-off areas, teachers in less well-off areas face more frequent salary delays, work in schools with poorer infrastructure and smaller operational budgets, have fewer teaching materials available, have larger class sizes, and receive less training especially for multi-grade teaching. Improvingresource availability and timeliness and lowering transaction costs are important for reducing delays in teachers' salary payment and enhancing education service delivery. There is little evidence to suggest that salary leakages from the district to school, absenteeism or ghost teachers are a major concern in Lao PDR. To better deliver service to poor people, it i s important to increase education spending to bolster teachers' salaries and to better cover operating costs, streamline budget planning and execution process, strengthen reporting system and capacity, and improve the accountability o f schools and policymakers to parents/citizens. 51 HEALTH Health defines the ability to work, feed and sustain a family, to savefor the future, to grow and develop, and to lead a fulJilling life. A healthy population translates into greater national wealth by improving the ability of children to learn, of families to plan their lives, and of workers to beproductive. Key health outcomes improved in Lao PDR over thepast decade, even though in absolute terms, they are still low. Life expectancy got better, infant and child mortality did too. Health workers are satisfied with their jobs and housing conditions. Their rate of absenteeism is low and they are willing to contribute to community development. NGOs and local community actively support frontline services. Leakage of health workers' salariesfrom the district level to healthcare centers is also low. However, to achieve the level of health identiped by the MDGs and the NGPES, more progress is needed. For example, there is still signipcant roomfor improvement on childhood immunizations and maternal health andfor health outcomes in general. Spending on health is low and the allocation of public resources is inequitable. Public expenditures tilt towards the center, which means poor people have less access to care. There is a mismatch between capital spending and recurrent spending, which reduces efficiency evenfurther. Unrealistically high revenueforecasts and shortfalls in public funding translate into budget cutbacks. These cutbacks tend to perpetuate themselves and weaken the management of expenditures. Severalfactors constrain service delivery at thefacility level; low salaries andfrequent delays in thepayment of health workers' salaries are two glaring examples. Healthcare centers are also poorly equipped; they rely heavily on local support and Revolving Drug Funds. Poor people also suffer from limited access to healthcare services of acceptable quality. This chapter: (1) reviews the changes in health outcomes; (2) examines the level and allocation of public health funding and analyzes the weaknesses in public expenditure management; (3) presents a snapshot of health service delivery at thefrontline, examines the sources of funds, and analyzes key constraints; and (4) conclusions. Outcomes Key health outcomes have been improving over the last decade for Lao PDR's population. The most important achievements are the reductions in infant and child 52 mortality rates. Infantmortality rate, defined as the number of newborn children who die during the first year of their life, decreased from 104 per 1000 live births in 1995 to 70 per 1000 live births in 2005. The under-five mortality rate, defined as the number of children who die betweenbirthand the age o f five, dropped from 170per 1000 live births in 1995 to 98 per 1000 live births in 2005. The fall in mortality rates helped to lift average life expectancy from 50 years in 1990 to 55 years in2003. According to the 2005 Population Census, life expectancy increased from 52 years for women and 50 years for menin 1995 to 63 years for women and 59 years for menin2005. Progress in lowering mortality rates is remarkable, but important challenges remain. While life expectancy has increased, Lao PDR's average life expectancy of 61 years is low compares to about 70 years in neighboring countries, such as China, the Philippines, Thailand and Vietnam. The proportion o f children immunized against measles and the share of births assisted by skilled personnel are still low. Poor health outcomes result from many factors, such as inadequate access to health care, poor nutrition practices, and careless handling of water and waste; unfortunately, health outcomes are worse for poor people (Box 5). Infant mortality rates are higher and life expectancy i s lower in provinces with higher poverty rates. For example, in 2002/03, Vientiane Municipality i s the province with the lowest poverty headcount, lowest infant mortality rate, and highest life expectancy (Figure 43). Sekong has a poverty headcount more than 150 percent in excess o f Vientiane Municipality, and it has the highestinfant mortality rate and lowest life expectancy. Box 5: why Poor People HavePoor Health Poor people in most countries have the worst health outcomes. They are inpoor health because they are poor; and they are pushed further into poverty due to illhealth. Unlike education, health outcomes are produced by households-with contributions from many services. Inadequate care for illness, lack o f access to good health care services, poor nutrition practices and careless handling o f water and waste are major contributors to poor health. Illiteracy, women's ignorance o f health issues, and lack o f decision- making are often the causes. Moreover, the difference in reactions to the same level o f health input between the poor and the non-poor o n health outcomes is large. Studies suggest that poor people are less efficient than rich people in translating inputs, such as drinking water, sanitation, and health services, into better health outcomes. While the availability o f inputs is already different, the l o w ability to transform those inputs into health further results in poor health outcomes o f poor people. The gap in 1 health outcomes between poor people and richpeople have been widening over time. Source: WorldBank (2003). 53 Figure 43: Poor people havepoorer health outcomes. OSt-K 0t)C>K OXAY SR OPtlO OIIUA O C > L J U 0 . W i A 0 LUN 0ATT 10 surplus provinces 0 deficit provincesI -VIE M 0ATT OSEK 0 surplus provinces deficit provinces Data source: Lao PopulationCensus 2005. 54 PublicFunding Public and privatespendingon healthis low in Lao PDR.Although the sum o fpublic and private per capita health expenditures increased by about one-third since the late 1990s, it was just 8.7 USDper capita in2002, which was equivalent to over 2 percent of GDP (World Bank, 2006b). Compared with other low income countries with an average o f 29 USD health expenditureper capita, 8.7 USDkeeps Lao PDR inthe very low end of the spectrum. If compared with the East Asia average, 64 USD, the gap i s even larger. The low level o f spending i s closely associated with the poor health outcomes in Lao PDR. The proportionof publichealthexpendituresas a share of totalhealth expenditures i s low; an increase in health expendituresover the last few years was due mainly to an increasein privatehealthexpenditures.Public health spending as a share o f overall government spending averaged over 4 percent in the mid-l990s, collapsed to below 2 percent towards the end o f the 1990s, and increased steadily since then to reach 6 percent in 2003/04, reflecting a renewed emphasis on social sectors (Figure 44). Correspondingly, public health spending as a percent o f GDP tripled from about 0.4 percent in 1997/98 to 1.2 percent in 2002/03. While these are encouraging trends, the most salient feature o f the public health finance remains the scarcity o f resources. Compared with other low income countries, the share o f public spending on health in GDP is very low inLao PDR, which is only one-half o fthat o f Vietnam and two-fifths of Cambodia (Figure 45). 55 Figure 44: Overallper capita health expenditures are increasing, but they started at a low level. Public and Private Health Spending (Percent of GDP), 1992193 to 2002103 2.5 IPublic Private 2.0 1.5 1.o 0.5 0.0 199213 199718 200213 Public and Private Health Spending (Percent of GDP), 1992193 to 2002103 8 fPercentofGDP(LHS) Percent of Government Expenditure(LHS) 6 4 2 0 I 1991 199213 199314 199415 199516 199617 199718 199819 199910 200011 200112 200213 200314 Data source: World Bank (2006b). 56 Figure 45: Public health expenditures as apercent of GDP in Lao PDR are low compared to other low income countries. Public SDending on Health (Percent of GDP) lable Ye: vblr 0 2 4 6 8 10 12 Data source: WDI (2005) andWorld Bank (2006b). A shortfall in public funding at the subnationallevel translates into cutbacks on of expenditures. Only one PHO (PHO in Oudomxay) out o f the 17 provinces surveyed reported the proposed budget was approved with no changes made in 2004/05. In the understanding of the 16 PHO whose approved budget was lower than proposed budget, 13 PHO considered the insufficiency o f funds at the provincial financial office i s the key reason why the approved amount was lower than the proposed amount; three PHO (Huaphanh, Savannakhet, andXekong) considered provincial governor's discretion i s the key reason. Explanation on the difference generally comes from PFO. Large difference between proposed budget and approved budget also exists at the district level. But, only three out o f five DHO receive explanation on the difference from DFO. At the district level, for the 85 percent DHO surveyed with data available, 21 percent received approved budget with no change made, 64 percent received approvedbudget lower than proposed amount.' Overall, almost 60 percent o f D H O that received approved budget lower than proposed budget received explanations; for D H O in priority districts that received approved budget lower thanthe amount proposed, 75 percent received explanations. Public health spending is not only low, but it is also allocated in an inequitable manner. Much o f government health spending is biased in favor o f the central level; 'The difference in the budget acceptance rate is not significant between DHO in priority and non-priority districts. 57 public subsidies to central hospitals are heavily biased in favor o f the richest quintile. This richest group alone receives nearly one-half o f total subsidies for central hospitals because the rich can better afford using health services when needed, while the poorest group gets only seven percent (Figure 46). Overall, the richest quintile captures 27 percent o fpublic subsidies for health services comparedto 13 percent o f the poorest. This leaves only limited resources for frontline services: just one-fifth o f public health expenditures in 2002103 was spent in district health facilities and village health centers. These are the same facilities that should play a key role in delivering essential health services to the poor people. As in many other countries, this leads to the de facto allocation o fpublic health spending infavor o f the less poor. Figure 46: Concentration curves of consumption and health care subsidy by health provider and inpatient/outpatient care Outpatient services 1;100 t 90 80 70 10 0 0 20 40 60 80 100 Cumu1ative popu1ation Inpatient services 0 20 40 60 80 100 Cumulative population Data source: LECS 2002/03. 58 Inferior access to healthcare services, especially to services that offer good quality, means poor people have less satisfactory outcomes. As poor people may fail to exert their influence on public policy, governments spend a large part o f health budget on "free" public hospitals used by the non-poor in urban areas, rather than population- oriented outreach services, such as immunization, that provide large benefits to the poor (Figure 47). Figure 47: Poor patients have little clout withpolicymakers on budget decisions. ExpendituresManagement Management of public expenditures in the health sector is weak in Lao PDR. Ineffective budget planning, execution, monitoring, and reporting aggravate the difficulties o f insufficient funding at the subnational level; it also hinders efficient health service delivery (Figure48). Figure 48: Policymakers may not be able to ensure health centers deliver efficient service. BudgetPreparation Budget preparations for the health sector are characterized by a lack of realism in revenue forecasts and fragmentation between the finance and planning functions at the central and local levels. Budget planning receives short shrift due to a compressed budget cycle (Box 6). 59 Box 6: Budgetprocess in the Health Sector 1.Budget preparation starts at the district level, as HC only make informal requests for resources but are not actively involved in formal budget proposal procedures. Most HC do not prepare or submit budget proposals to DHO or discuss annual budget with DHO. The links between HC and DHO i s rather weak. 2.DHO receive guidance and instructions from PHO o n budget planning and use estimates o f current year fundingneeds, expenditure o fprevious fiscal year, andproposedbudget o fprevious fiscal year as criteria to decide the total amount o f annual budget proposal. However, DHO understand that the approved budget i s likely to be different from what was proposed. DHO prepare annual budget plans and submit proposals to DFOfor approval andto PHO for official acknowledgement. 3. PHO consolidate budgets received from DHO without adjusting the proposed budgets and submit budget proposals to PFO. PHO also submitbudget proposal to the MOH, althoughnot for approval. 4. PFO receive health budget proposals from PHO and DFO, and submit provincial budget proposals, which include health budget proposals, to MOF. When PFO get budget approvals from the MOF, they approve budgets to PHO according to what MOF accepts and approve budgets to DFO, which in turn approve budget to DHO. However, due to the large difference in approved budget and proposed budget, after receiving approval o f annual budget by PFO, PHO officers submit quarterly budget proposals to request non-salary funds. 1.Lack of Realism Budget planningin the healthsector is unrealistic;it is often based on high revenue forecast.Nineout o f 10 PFO provide budget ceilings to PHO, while the ceiling is based on the unrealistic planned expenditure o f PHO o f last fiscal year. The budget acceptance rate i s particularly low for non-salary items. No clear criteria determine whether and how much a budget proposal gets approved. According to the field interviews, the acceptance rate ranged from 13 to 29 percent for the capital costs and from 29 to 72 percent for the administration costs at the provincial level. Deficit provinces faced tighter budget constraint than surplus provinces. In 2003/04, throughout the country the actual health spending was about 84 percent o f the planned spending nationwide; 94 percent insurplus provinces; and 81percent indeficit provinces (Figure 49). 60 Figure 49: Actual health spending compared toplanned health spending is lower in deJicit provinces. Acutal as percentage of planned health spending (2003104) 90 85 80 75 1 70 Lao PDR Surplus provinces Deficit provinces Data source: Official Gazette. Unrealistic proposals perpetuate themselves in budget planning. Provinces and districts have no incentives inmaking exact budgetproposals but they do try to maximize the non-salary budget. Many staff working inhealth system take low budget acceptance rates for granted as it i s a yearly event and do not make efforts to actively seek explanations. The role that budget planning plays i s actually weak. DHO and PHO have to readjust budget and resubmit quarterly budget proposals for non-salary expenditures even after their budgets got approved. Such budget re-submissions and re-approvals usuallytake a few weeks with a few steps, which waste extra staff-time, increase logistic costs, and generate delays indecision making and payment. 2. Fragmentation Expendituremanagementin the healthsector is fragmented at the centraland sub- national levels; mismatch of capital investment and recurrent expenditures is a serious problem.Dueto the lack of coordination betweencapitalandrecurrent spending, and the lack o f resources for matching funds from donor contributions, capital investment represents a high share o f total health expenditures in Lao PDR.For example, the number of Lao PDR's hospital beds compares favorably with that o f richer countries, but its health workers do not receive their already very modest level o f salaries on time. The share o f capital investment intotal public health expenditures was 79 percent in 2002/03, and 56 percent in 2003/04. The share o f capital expenditures out o f total health expenditures varies dramatically across provinces and from one year to another (Figure 50). In 2002/03, Vientiane Municipality spent 8 percent o f total health expenditures on capital investments; in 2003/04, the share increased to 12 percent. By comparison, in 2002/03, Luangprabang province spent 95 percent; in 2003/04, the share dropped to 48 percent. The changes reflect partly the lumpiness o f capital investment, but also adjustmenthow donors allocate funds to provinces. 61 Figure 50: Health capital spending varies acrossprovinces and by year. Health capital spending by province (% of total healthspending) , , m i 1 2 1 0 3 1 0 4 , I , , IATT XAY SR BSEK CHA ISAV SAR IBOR KHA 2003104 ~ IXIE VIE P IHUA XAY IBOK LUP I ILUN OUD I VPHOM I E BXAYSR ICHA ATT ISAR SEK IKHA SAV BVlE P BOR IXAY XIE ILUP HUA IOUD BOK IPHO LUN VIE M 0 10 20 30 40 50 60 70 80 90 100 Data source: Official Gazette. The amount of resources allocated to recurrent health expenditures are low and volatile as a disproportionately large share of the already limited amount of public resources has been devoted to capital spending. Overall, in 2002/03, 29 percent o f recurrent health care expenditures went to non-wage expenditures; in 2003104, it was 22 percent. The share varies widely across provinces (Figure 51). For example, Phongsaly province spent about 10 percent o f recurrent expenditures on non-wage costs; some provinces, such as Luangnamtha, Xayabury, Vientiane and Attapeu spent more than 35 percent in 2002/03. Within each province, the variation from one year to the next i s relatively small as wage allocations remains fairly constant. In per capita terms, public health recurrent expenditures also vary dramatically across provinces (Figure 52). For example, in 2003/04, per capita health recurrent expenditures in Sekong province were more than three times that o f Vientiane Municipality. However, per capita recurrent health expenditures show no systematic correlation to income, poverty, or health outcomes. 62 Figure 51:Health non-wage spending varies acrossprovinces. Health non-wage spending as ?LOof health recurrent expenditure XAY SR I ATT I CHA SEK SAR I SAV , KHA I 2003/04 VIE P BOR I XIE XAY I HUA I LUP mOUD BOK I VPHOM LUN I E I SR ATT XAY I CHA ISAR SEK IKHA SAV I VXIE P BOR I E 2002/03 DXAY I LUP HUA OUD BOK I I LUN VIE M PHO 0 10 20 30 40 50 Data source: Official Gazette. Figure 52: Health recurrent expendituresper capita vary acrossprovinces. Health recurrent expenditureper capita 2003104 (kips) . . .... ............. Xaisornboun special region Anapeu Champassak Sekong Sarabane Sabannakhet Khammouan Bolikhamxay Vientiane Xengkhouang Sayabouly Houaphan Louangphabang Bokeo Oudornxay Luangnamtha Phongsaly Vientiane Municipality 0 5000 10000 15000 20000 25000 Data source: Official Gazette. 63 3. CompressedBudget Cycle The tight budget preparation schedule, which runs from May to September, leaves little time for input from districts or health centers. Coordination between the center and provinces i s poor; budget preparation cycle i s always late. MOF i s the institution that determines overall budget allocation among provinces and across sectors, and redistributing resources from surplus PFO to deficit PFO, but M O F does not control how PFO further distribute the payments across sectors and/or to different DFO. Provincial governor makes decision in allocating resources to sectors within the envelope, and the M O H i s only acknowledged in the process. Criteria for determining whether a budget proposal gets approved are not transparent. At the provincial level, PFO provide in most cases an explanation when PHO budget submissions get revised. But only three out o f five DHO received in2004/05 an explanation from DFO. Execution and Reporting Budgetexecutionis weak for the healthsector.An unclear division o fresponsibility by ministries leads to delays in executing the budget. In average, it takes about a month for DFO to approve payment requests from DHO; in priority districts, it takes up to six weeks. When making purchases, the DHO has first to request a purchase authorization for the proposeditem from DFO; then DHO gives the authorization number to the retailer after it i s issues by DFO. Finally, the retailer requests payment from the DFO by providing the authorization number. Fourteen out o f 17 PHO have a dedicated bank account for money transferred from the PFO to PHO. Eight PHO also have cash accounts for government funds. However, only one out o f two DHO have bank accounts; and only one out of four have a dedicated bank account for money transferred from DFO. Few D H O have cash accounts for governmental funds. The long and cumbersome authorization and processing procedures and the unclear division o f responsibilities between MOF and MOH and their officers at the subnational level lead to delays and inflated procurement costs. A reporting and monitoring system exists for subnational governments and line ministries, but their respective responsibilities are unclear and institutional mechanisms are weak. Reporting system mainly starts at the district level as reporting from H C i s generally missing. DHO report the total number o f health workers at the district level to the District Personnel Department quarterly, and make a request in written with defined format for salary payment approval. PHO closely supervise DHO, for example, PHO visits D H O for supervision every month; DHO staffs also have monthly visits to PHO for official reasons. PHO provide trainings to health workers o f D H O and HC. H C receive supervisions from DHO every three months, and supervisions from PHO twice a year. In two out o f three cases, H C health workers are formally assessed by H C director or DHO a few times a year. Due to the variety o f reporting standards, it is difficult to monitor health expenditures and to evaluate whether or not national priorities are being followed. Basic systems for report keeping and accounting are in place. Vertical reporting and supervision varies little across the provinces and districts. The reporting system usually starts at the district level, as there i s little reporting 64 from health centers. PHO have standardized and well-documented financial records, including computerized accounting books, money execution authorization forms, and receipts. PHO have separate accounting books for governmental and project budgets as well as drug funds. However, D H O lack the resources and capacity to maintain spending records in the same detail formally providedby PHO. As a result, functional data for the health sector is not provided to MOH or MOF. Health centers get supervised by D H O every three months and by PHO twice a year; but do little reporting. Healthdepartmentsat the districtlevel, especially DHO in priority districts, are not adequately equipped to carry out their execution and reporting functions. For example, only one out o f three D H O has fax machines. Relatively speaking, DHO innon- priority districts have better office facilities than DHO in priority districts (Figure 53). Almost 90 percent o f DHO in non-priority districts have computers compared to 70 percent of D H O in priority districts. Almost 40 percent o f DHO in non-prioritydistricts use accounting programs compared to 20 percent o f DHO in priority districts. About 85 percent of DHO innon-priority districts have telephones compared to 67 percent o f D H O in priority districts. Compared with DHO, PHO are better equipped-each PHO has computers, telephones, and fax machines, use accounting programs; 14 out o f 17 PHO surveyed have copy machines. Figure 53: DHO in non-priority districts have better ofice facilities. DHOs in non-priority districts have better office facilities b 100% 80% 60% 40% 20% I 0% Total Priority districts Non-priority districts B%of DHO havecomputers .%of DHO use programs for accountinglfinancial records %of DHO havetelephones I%ofDHOhavefaxmachines Data source: PETS. 65 Frontline Services Efficient functioning of health centers at the frontline is important for service delivery. As the entire health sector has inadequate resources, H C to which poor people can access are more likely to face severe constraints. This section elaborates on how communities/NGOs can complement the long route o f accountability o f healthcare centers vis-a-vis policymakers through direct accountability towards patients (Figure 54). From the prospective o f Health Centers, this section focuses on health workers' salaries, operation costs, Revolving Drug Funds (RDF), H C access and quality, and accountability. Figure 54: Accountability of Health Centers topolicymakers andpatients is crucialfor health service delivery. Health workers In Lao PDR health workers' salaries are low. The average level o f health workers' overall monthly salary i s about Kip 450,000, or USD 45, including base salary, overtime, bonus, and family allowance. On average, about Kip 350,000 is for base salary; and about Kip 100,000 is for overtime, bonus and allowance. The total salary o f health workers is about 150 percent o f the GDP per capita o f Lao PDR, which i s low by international standards. InLao PDR, more than nine out o f 10 health workers surveyed reported their salaries were too low to cover the cost o f a minimum standard o f living. In nominal terms, health workers' base salary varies little for workers in poor areas and non-poor areas, those in rural and urban communities, Lao-Tai concentrated areas and elsewhere (Figure 55). However, health workers in poor areas and rural areas have higher benefits and family allowance payments. Inreal terms, as living expenses are lower in less well- o f f areas, the salaries o f health workers there have higher purchasing power than those in well-off areas. 66 Figure 55: HC workers' salaries vary little in nominal terms; they are higher in rural areas in real terms. Nominal monthly HC workel rrlrryIncluding riIoWrnc0 (Kip. 2002 nomlnrl tormr) ................................................................. .___ ... ..... 400,000 ................ 300,000 .............. ................. 200.000 100.000 n Lao PDR urban Rural Nom-Poor Poor Lao-Tal Other IBaseSalary*overtime.bonusandfamilyallowance Ron1monthly HC workor .rlrry lncIudln(l rIIOWllnce (Kip, 2002 Vlontlrno urbrn prisor) 500.0007 400,000 300.000 ............. 200.000 100,000 ............ ................... n Lao PDR Urban Rural Non-Poor Poor Lao-Tal Other I Salary Base overtime. bonus and family allowance Data source: PETS. Delays in payment of salaries are frequent for health workers, especially those in less well-off areas. The data from Ministryof Health suggests inthe last quarter of FY05 and the first quarter o f FY06 GOL has paid out the salaries of health workers such that the stock of salaries outstanding more than 1month came below 22 percent o f a month's payroll; and that severe arrears in salaries (over three months) have essentially been eliminated at less than one percent o f a month's payroll (World Bank, 2007a). However, PETS results indicate a less favorable picture." According to the survey, two out o f three H C workers experienced delays in payment o f salary or benefits between October 2004 and September 2005. Although health workers inrural areas have higher total salary in real terms than those inurbanareas, they are almost 10 percent more likely to experience payment delays. In addition, health workers receive their salary less frequently than they are supposed to. Three out of four health workers inhealth centers reported they should receive their salary every month (Figure 56). However, inpractice, only one intwo health workers receive their salaries every month; 37 percent receive payment every three months; and about 14 percent are paid irregularly. Monthly payment schedules are more common in urban, non-poor and Lao-Tai areas. Health workers in poor, rural, and non lo numbersarenotdirectlycomparable.Discrepanciesinsalarypaymentbetweendifferentsourcesof The information can come from, for example, sampling and timing. 67 Lao-Tai areas received payment less fiequently than those in non-poor, urban, and Lao- Tai areas. At the time o f the survey, one out o f three health workers had an outstanding balance o f salary or overtime, bonus or allowance (Figure 57). Even in urban areas, one out o f four health workers was owed salaries. Figure 56: Health workers in health centers get paid lessfrequently than they are supposed to. How often HC workers supposed to receive salary payment (%) Non Lao-Tho1 Lao-Thal RuroI Urban poor Rlch Total 0 10 20 30 40 50 SO 70 80 90 100 Monthly I IIIrreaulmrly Quarterly How often HC workers actually receive salary payment (%) Non Lao-Thal Lao-Th-l Rural Urban Poor RlCh Total 0 20 40 60 80 100 Monthly I IQuerterly Irrepularly Data Source: PETS. Figure 57: One out of three health workers had outstanding balances duefor their salaries and overtime at the time of the survey. % of HC workers who had outstanding salarylovertime payment balance at the time of survey 450h 1 I 40% ........................................................................................................................................ ........... ...... T0t.l RlCh poor Urban Rural Lao-Thd Non Lao-Thai Data Source: PETS. There is little evidence to suggest that salary leakage between district and health centers is a key concern in Lao PDR. Based on the salary payment records of the 267 health workers surveyed in PETS, the discrepancy in reported salary between DHO and HC is small: amounts only to about 3 percent of total salaries and 1 percent of base 68 salaries (Figure 58). Deviation takes place inboth positive and negative directions. DHO and health centers report payments coincide for about two out o f five cases for total salaries. DHO reports higher payments than the health centers for about one out o f three cases; and DHO reports lower payments than health centers for about one in four cases. The numbers are similar with regard to base salary. Discrepancies are greater in urban area, non-poor, and Lao-Tai areas, but the average magnitude remains low in both directions. In poor non-Lao-Tai areas, salaries received by health centers often exceed salaries released by the DHO, which coincides with the anecdote observations that facilities obtain support fiom communities and donors to supplement health workers' salaries. Figure 58: Overall difference in salarypayments reported by DHO and HC is small, but some areas have greater discrepancies. nc workers' overall salary dispensed by DHO and recelved by HC workers (%) 100% 90% 80% 70% 50% 50% 40% 30% ZO?4 'I0?4 0% Lao PDR Urban Rural Non-Poor Pool Lao-Thai Other 1.I HC=DHO IHCcDHO HCsDHO I Dlfference between overall salary reported by DHO mnd HC (a- percent of ov-rill salary rrportrd by HC) 10% ............................................................................................. -5% 0% :I, ........................................................................................ L.O P D I ........................................................................................... L.D.Th., HC workers' base salary dispensed by DHO and WNed by HC workers (%) Difference betweenbase salary reportedby DHO and HC (as percantof base 100% 43% 80% 6 % ........................................................... 70% 40,................................................................... M)% 50% 2% ................................. 40% 0%. I' 30% La0PDR U M Rvnl .2% .................................... 20% 10% .41 .................................... 0% *y .................................. 69 There are good reasons why ghost health workers are not a concern. Health workers need to sign when receiving their payments, as all salaries are paid in cash. DHO keep salary records for each health worker. D H O report quarterly the total number o f health workers at the district level, and make a written request, based on a prescribed format, for salary payment approval. PHO closely supervise DHO every month; and D H O staffs also have official monthly visits to PHO. Insufficient funds are apt to be a key contributor to salary delays in the health sector. Health workers surveyed have little doubt about the cause o f payment delays. All o f them stated that insufficient funds are the main reason. Two out o f three DHO considered weak revenue collection from DFO to be the reason for delay in fund transfer for salary payment (Figure 59). Indeed, when DHO receive the salary funding from DFO, it takes only a few days for DHO to pay ip cash to health workers at the district level. As inthe education sector, hightransactioncosts contribute to the delays insalary payment, Figure 59: Delays in revenue collection at DFO are theprimary reasonfor salaries being in arrears. Reasonsfor delay in fund transfer for salary payment in DHO 1 ,OO% . Lao PDR priority districts non-priority districts IDelayinofficialrequestfromDHOtoDFO IDelayinrevenuecollectionatDFO Delay of official approval of money transfer within DFO IDecision of governor of district Decision of governor of province Iunknownreasons Data source: PETS. Despite low salaries and frequent delays inpayment, most health workers have high job morale and are seldom absent from work. Three out o f five o f the health workers are natives to the area where the health center i s located, and most understand the language o f the community. Most health workers inhealth centers surveyed reportedthey are satisfied with their job and housing conditions. They believe that they provide satisfactory services at the HC. They are satisfied with the staffing at H C and with the availability o f drugs. They are very satisfied with the overall work at HC, and willing to contribute to community development. A high degree o f job satisfaction and job motivation in Lao PDR helps to limit job absenteeism. The absence rate for health workers i s 18 percent in health centers for a PETS unannounced visit (Figure 60). In addition, only two out o f five o f absent health workers were absent without legitimate reasons, such as official visit or sickness. The absenteeism rate i s low in international comparison, Similar survey measured absence rates o f 25 percent in Peru, 35 percent in Bangladesh, 37 percent inUganda and 40 percent inIndonesia and India. 70 However, low salaries and frequent delays in payment have a negative impact on health service delivery. Many health workers have to take multiplejobs, most often in agriculture, to supplement their incomes; some also have to borrow money. On average, a health worker spends 14 hours a week doing work other than their jobs at health centers. Among those who have income from a non-medical practice, the average monthly non- medical income i s about 250,000 kips. Providing private health care services, selling medicines from home, or doing commercial works are, however, uncommon among health workers. Fewer than one out o f two health workers consider it acceptable to receive money from patients for the health services they provide, including consultations and drugs. Figure 60: Less than one out offive health workers are absentfrom work,just two out of five are absent without a legitimate reason. Absenteeism Rate of Health Workers (%) 20 18 12 a 4 0 Reasons Why Health Worker Is Absent (YO) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% Lao PDR Urban Rural Non-Poor Poor Lao-Tal Other IOfflciaiReasons ISickness Other Data Source: PETS. Resource flows and Revolving Drug Funds Resources from donors, NGOs, communities, and patients are often important sources of funds for the health sector. However, providing budget for PHO andDHO i s responsibilities o f provincial and district authorities. Resources from governmental budgets mainly cover salary payment, but that are in general inadequate, PHO are mostly 71 funded by resources from the provincial government. PHO pay the salaries o f PHO staff and the provincial hospital staff. DHO rely on resources from governments, donors and NGOs as well as profits from drug sales to fund their expenditures. DHO pay the salaries o f DHO staff and the district hospital staff. Health workers' salaries and overtime compensations are mainly financed by governmental budget. Resources from district authorities, resources from donors and NGOs, and profits from drug sales are main sources o f funds o f DHO. Although budget received from DHO is an important source, Health centers rely heavily on RDF funds and contributions from donors, NGOs, and communities to supplement operation costs. DHO may also distribute free drugs, provide free means o f transportation, and give free equipments to HC. PHO, DHO, and H C located inmore affluent places face less severe budget constraints. Vertical financial support from PHO to DHO and to HC for salary and overtime payments and for maintenance costs is limited. Less than 20 percent DHO received money transfer for salary, overtime, or family allowance from PHO (Figure 61). A high percentage o f central governmentalbudgets are spent on salary payment, and only a small share is spent on other items. PHO pay salaries o f PHO staff and provincial hospital staff. DHO pay salaries o f DHO staff, district hospital staff, and health workers o f H C in the district. H C may request additional budget from DHO for urgent needs, though most o f the time DHO may not grant additional fundingto the requests. Figure 61: Less than 20percent DHO receivedjinancial supportsfrom PHO % of DHOs received moneytransfer for salary, overtime, or family allowancesfrom PHOs 2004105 Both Overtirne/family allowance only Salaries only 0 I O 20 30 40 50 60 70 80 90 Total Non-priority district IIPriority district Data Source: PETS. Health centers often rely on local support. Almost two out o f five health centers surveyed reported donors and NGOs were the main financiers o f their recent renovations (Figure 62). Overall, 30 percent o f health centers receive contributions o f money and labor from the villages and communities, often for construction and renovation o f health centers. In urban areas, this ratio is about 15 percent; while in rural areas, this ratio is about 35 percent (Figure 63). About one out of three HCs innon-poor areas have plots o f land for farming; while this ratio i s over one out o f two in poor areas. However, healthcare facilities in less well-off areas are less able to collect fees from users as a 72 higher percentage o f their patients are eligible to exemptions and donors and NGOs tend to be less active there. Figure 62: Donors and NGOs are the mainJinanciers of renovation costsfor HC. Main financier of recent renovation Cost for HCs ................" .....--....I ................ .............. ................... ......... total non-poor poor urban rural Lao-Tal other lmDHO MPHO ICornrnunities DononINGOs mothers 1 Data source: PETS. Figure 63: Health centers inpoor, rural, or non Lao-Tai areas are more likely to rely on off-budget funding. % of HCs received money gifts or labor contribution from villages served by the HC In 2004-05 45% 1 I dno,- .-,-1................................................................................................................................ 1 35% ........................................... ............................ .......................................... 30% ........................................... ............................ ............................ ..... 25% ------ ............................ ............................ .......... .......... ..... 20% - - ~ - ~ - .......... .......... ............................ .......... .......... 15% ------ .......... .......... ............................ .......... .......... ..... 10% - ~ - - - ~ .......... .......... .......... .......... .......... .......... ..... 5% ...... .......... .......... .......... .......... .......... .......... ..... 0% Total Non-Poor Poor Urban Rural Lao-Tai Non Lao-Tai % of HCs have plot of land for farming 6oy0 ................................................................. ... ............................................................ 50% ........................................... .............................................................................. Total Non-Poor Poor Urban Rural Lao-Tai N o n Lao-Tal Data source: PETS. 73 Revolving Drug Funds are an important source of revenue for health centers. These hnds supplement their operating costs for office supplies, transportation, and utilitycosts (Box 7). Guidelines stipulate that healthcenters can earn a 25 percent margin o fthe retail price over the purchase price, even though there are no clear regularities. While PETS indicates that most health centers have adopted policy, RDFs appear to have different profit ratios (Figure 64). Drug sale profits varied across provinces - health centers in Vientiane province had almost Kip100,000 profits from selling drugs between October 2004 and September 2005, while health centers in 15 out o f 18 provinces had less than Kip 20,000. Box 7: Revolving Drug Funds The RDF system was begun in the 1990s as regional pilot projects in conjunction with community health programs; they were supported by non-governmental, bilateral and multilateral development assistance. A national drug policy was approved in 1993 to improve access to essential drugs and to attain more rational utilization o f these drugs. A user fee was introduced in 1995 as a way to recover costs for the health sector. According to PETS, in2005, more than 95 percent o f the health centers had RDF. Almost 60 percent o f RDFs in health centers were established after 2001. RDFs at the frontline health facilities have become one o f the basic drug supply routes. Although RDFs are supposed to be used only to replenish drugs, PETS indicates that profits from drug sales are being spent on administrative costs, outreach activities, transportation and other cost heads of health centers. One concern i s that such revenues create an incentive for health centers to over-prescribe antibiotic and injection. Figure 64: RL3FproJit ratios cluster around 25percent, but they differ across provinces. "1 0 0 .2 .4 .6 .a profit ratio=dr'ug sale profiffdrugsale amount -B8.Er- * \ V E P '9- ge 9 - a -al ;s- F * L A N O 9 - 2 *" " t u A*No I:,"0 K E *#E = B 0 C K PI n"I3 Aw %iA ~~ R* A 1 N 0 - * P H O N - n I, r e A - S u r p l u s P r o v i n c e s * D e f i c i t P r o v i n c e s Data Source: PETS. 74 RDFs are managed differently in some districts and provinces. The DHO or PHO determines purchasing, financial accounting, and management of inventory o f RDFs in health centers. There is no standardized financial reporting system inplace. For example, health centers in Vientiane Municipality have almost an exclusive right to purchase their own drug supply directly from private pharmaceutical companies. InLuang Prabang and Champsasak, health centers usually obtain drug supplies from DHO, as they do not have the autonomy to procure drugs independently. Only one out o fthree health centers has a separate bank account for RDFs. One third o f health centers experienced in-stock drug shortage in 2004-05. The longest period for which a drug shortage lasted was about two weeks. The occasional shortage o f drugs more often occurs in health centers that have less autonomy. Operatingfunds and healthcenter quality Health centers have to rely on multiple sources to cover operational costs as funds from DHO are limited. While the DHO are the principle source o f salary payments for health workers, financial support o f operational costs to the health center i s inadequate. Requests from H C to DHO for cash to fund recurrent costs are typically rejected. Although health centers might receive free drugs, offer transportation at no charge, and donated equipment from DHO, most H C rely on multiple revenue sources, including government, donors, NGOs, communities, and patients. Because of a lack of operating funds, most healthcare centers in Lao PDR are not well-equipped. This i s especially true in less well-off areas. Although each province has a provincial or regional hospital and each district has at least one district hospital and an average o f six health centers, most health centers are public and more than 10 years old. Due to insufficient funds, reconstruction or renovation o f facilities i s rare. Health centers are open for business eight hours a day Monday through Friday; some are also open on weekends. Closure for non-official reasons i s rare. However, if closure does happen, the health center will report it to DHO. Most health centers take patients outside o f regular hours. Some health workers see patients in their homes or in patients' homes. Since equipment and facilities are inadequate and staff knowledge is limited, health centers often have to refer patients to district or provincial hospitals for laboratory tests. Only 14 percent o f health centers have an official telephone; 30 percent have a laboratory. Health centers innon-poor, urban, and Lao-Tai areas are relatively better equipped than those in poor, rural, and non Lao-Tai areas (Figure 65). They are more likely to have an electric grid, accessto safe drinkingwater, a functioning refrigerator, and an official telephone. 75 Figure 65: Health centers in non-poor, urban, and Lao-Tai areas are better equipped than those inpoor, rural, and non Lao-Tai areas. % of HCs connected to the electric grid 1 I 70% ....................... .......................... .......................... ..................... 60% ------ .......... .......................... .......................... ..................... 50% - ~ ~ - ~ .......... - .......................... ......... ......... ..................... 40% ------ .......... ......... ......... ......... ......... ......... ..... 30% ------ .......... ......... ......... ......... ......... ......... 20% ------ .......... ......... ......... ......... ......... ......... 10% ------ .......... ......... ......... ......... ......... ..... n% + -.- . Total Non-Poor Poor Urban Rural Lao-Tal Non Lao-Tal % of HCs have a refrigeratorthat is functioning 70% .- 60% ....................... .......................... ....................................................... 50% - - - - - - .......... .......................... .......................... ......... ----, 40% - ~ - - - - .......... .......................... ......... .......... ......... ....< 30% ------ .......... ......... ......... ......... .......... ......... ..~.< 20% - ~ ~ - ~ .......... ......... - ......... ......... .......... ......... ~~~~. 10% - ~ - ~ - ~.......... ......... ......... ......... .......... ......... 0% _ _ - Total Non-Poor Poor Urban Rural Lao-Tat Non Lao-Tal % of HCs have access to safe drinkable water source &Jy0 ..... ~ ..... ...................... .......................... .......................... ................... ......... .......................... ......... ......... ................... ......... ......... ......... ......... ......... ................... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... ......... Total Non-Poor Poor Urban Rural Lao-Tai Non Lao-Tal % of HCs have an official telephone 25% 1 ................................. ._ "._ - ........ ................. 20% ........................................................ ..................................................... 15% ...................... .......................... .......................... .................... 10% -- - - - - ......... .......................... ......... ......... .................... 5 % ...... ......... .......................... ......... ......... .................... no," - - Total Non-Poor poor Urban Rural Lao-TaI Non Lao-Tal Data source: PETS. 76 Poor people have less access to healthcare, to safe water, and to hospitals within distance (Figure 66)." The non-poor use more and better public health facilities than the poor, as out-of-pocket expenditures i s a serious constraint for poor people to seek health care. Poor people are much more likely to rely on self-healing or self-medication when they are sick compared with the non-poor. There are two reasons: one the poor cannot afford to pay for health services; second, the poor do not think the quality o f care they can receive i s worth paying for (unskilled health workers, unavailability of medical equipments, etc.). Although their level o f household consumption is low, the poor spend an even lower share on health compared to that o f the non-poor. To extend the comparison, people in priority districts spend an even lower share on their health compared with people in non-priority districts; and non Lao-Tai people spend a lower share on health compared with Lao-Tai people (Figure 67). The richest quintile captures twice as much public subsidy for health services as the poorest quintile. As inmany other countries, this imbalance leads to the de facto pro-rich allocation of public health spending. Figure 66: Thepoor have less access to health care, less access to safe water, and less access to local hospitals. M bHlJ4 OOUD Data source: World Bank (2004). .. I 1Access to health care i s defined as access to skilled medical personnel such as medical practitioners, trained nurses, and community health workers at a dispensary. Access to safe water i s defined as access to piped water or protected wells. 77 Figure 67: People in less well-off areas spend a lower share of household consumption on health although their household consumption is already low. Health Expendituresas Percentageof Household (2002/2003) 2 0 1 5 1 0 0 5 0 0 Total Poor Non.Poor 47 Priority Other Lao.TaI Mon. Chinese. Hrnonplu District8 Districts Khmer Tibetan Mien Data source: LECS 2002/03. Health service that poor people have access to is of lower quality. Health workers in poor or rural areas are less experienced (Figure 68). Health workers receive training from the PHO and DHO two to three times a year on average, but only once a year inpriority districts. Health workers in less well-off areas may have less incentive to go to H C working because o f the unpleasant environment and inadequate facilities. H C in less well-off areas are more likely to face severe funding constraints, which limit their village outreach activities and health educational programs. Figure 68: HC health workers in poor rural areas are less experienced. Years of experience of HC workers surveyed in governmental health care services 1 20 ........................ " - Total Non-Poor Poor Urban Rural Lao-Tai Non Lao-Tai Data source: PETS. 78 Accountability Community participation may help enhance accountability, but inequitable distribution of resources remains a key concern. The short route from client to provider in the health sector i s blunted by asymmetries o f information and conflicts of interest (World Bank 2003). Getting highly transaction-intensive and discretionary individual-oriented clinical services to poor people i s most challenging. To influence quality, poor clients should have greater power-through third-party payments, information, and greater oversight o f health workers and facilities. A variety o f market failures-disease related externalities and fragmented insurance markets-and concerns for equity suggest public intervention i s necessary in financing health and nutrition services. Poor people lack the money for market transactions, but still often need to pay for health service directly. Insurance market failures, asymmetries o f knowledge, and conflicts o f interest mean that governments need to invest in purchasing key services to protect poor households and foster a pro-poor professional ethos. To effectively use public health resources, it is important to improve outreach services and community and family-oriented services. Policymakers need to be accountable for health outcomes-which means greater investment in monitoring and evaluation mechanisms that capture disparities inhealth (World Bank 2003). Population-oriented outreach services-standardized services that can include vector control, immunization, or vitamin A supplementation-are easier for policymakers to monitor (World Bank 2003). Even governments with limited capacity can provide these servicesnr write contracts with public or private entities to provide them. Community and family-oriented services that support self-care-such as information and social support for promoting breastfeeding or safe sex- are often well placed for community and civil society organizations and commercial networks to provide services close to poor households. Governments can establish partnerships and provide information and targeted subsidies. Conclusions Health outcomes in Lao PDR have been improving, but they still lag behind most EastAsian countries.Peoplewho have the best access to quality care are those who can afford it most. Besides income, wide disparities also exist based on geography and ethnicity. Although health workers' job morale i s high and absenteeism i s low, insufficiency o f resources and inequitable allocation between capital and recurrent spending are key constraints of health service delivery. Health workers' salary level is low, and delay in payment i s frequent. Maintenance at health centers is inadequate. Public health spending tilts toward the central level. Health facilities rely on contributions from donors and their communities and from Revolving Drug Funds to cover their operating costs, which widens disparities between well-off areas and less well-off areas. 79 Poor people suffer worse health service as they have limited access to health facilities with reasonable qualities and low capacity to pay. Little evidence suggests that salary leakages from the districts to health facilities are a major concern in Lao PDR. The same i s true o f absenteeism and ghost health workers. To achieve health MDGs and delivery better health service to poor people, it i s important for the government to increase overall public health resources, raise the share o f recurrent expenditures, increase the resources devoted to front-line services in rural areas, as well as to streamline and harmonize the expenditure management. 80 FINDINGSAND RECOMMENDATIONS PETS provides useful information on financial flows at the subnational levels and on characteristics of frontline education and health services. The survey results provide support that leakage of salay payments between districts andfacilities is small; ghost- workers and absenteeism by teachedhealth workers are rare. There is also high j o b satisfaction among teacherdhealth workers, and active participation by parentdpatients and community at thefrontline level. All of thesefactors may contribute to the reasonable education and health outcomes in Lao PDR, particularly since it operates under tight fiscal resources. The survey results, however, indicate that insufficient resources, ineficient and inequitable allocation of funds, weak public financial management and local capacity have a negative impact on service delivey , particularly to poor people in less well-off areas. Strengtheningfinancial management to make effective use of limited budget resources is important for improving service delivery. So is the eficient allocation of resources for capital and recurrent expenditures, implementation of the new Budget Law, and improving the effectiveness of donor aid. Along with the recent Lao PDR Public Expenditure Review, the PETS results suggest the following as important steps in bolstering the Public Expenditure Management Strengthening Program (PEMSP) reform agenda: aligning subnational spending programs with the national education and health goal; allocating adequate resources for basic services across all localities; and improving the budget process to ensure effective use of public resources. Thispilot PETS succeeded in making afirst attempt in documenting the characteristics of frontline services, tracking salay payments from district to facility, correlating performance indicators with service delivey , and recommending policies to strengthen public expenditure management and service delivey. With improvements in local capacity and survey techniques, future PETS reports may provide more accurate information onfinancial flows andfrontline services, and increase the scope and depth of analysis on the tracking of public expenditures. This chapter: (1) summarizes the findings of PETS; (2) reviews the financial management system; (3) provides policy recommendations on strengthening Public Expenditure Management Strengthening Program; and (4) presents suggestions for future PETS. 81 Mainfindings of PETS This PETS provides Lao PDR with information on financial and human resources of the education and health sectors at the provincial, district, and facility levels. It i s the first to track salary payments from districts to facilities by comparing salary records o f individual teachers andhealthworkers. The survey results indicate some positive points: (1) the evidence does not suggest that leakage o f salary payment from district to facilities i s a big problem; (2) ghost-workers and absenteeism are not a significant constraint to education and health service delivery; (3) teachers and health workers report high job satisfaction despite low salaries and delays in payment; and (4) parentdpatients and communities actively participate at the frontline level. These factors contribute to reasonable outcomes and improvements in education and health service deliveryinLao PDR over the past few years. However, the survey results show three major constraints in a highly decentralized Lao system wrapped in a weak PFM framework: (1) insufficient resources; (2) inefficient and inequitable allocation o f funds; and (3) weak public financial management and low capacity at the subnationallevels. Shortage o f funds i s the most common problem faced by the education and health systems. Lao PDR's public expenditures for education and health are low compared with other low income countries. Nine out o f 10 schools and health centers have to rely on off-budget resources, which makes the disparity inservice delivery even larger. 0 Limited public spending on education and health is disproportionately allocated to capital expenditures and diverted away from the people who need it most. A mismatch between capital and recurrent spending reduces overall efficiency o f service. Coordination between the central and sub-national levels i s inadequate. Public spending benefits more of the non-poor than the poor. Expanding services to reach poor people requires better targeting o f public expenditures. 0 Budget planning i s weak and revenue forecasts are unrealistic. Finance and planning finctions are fragmented at the central and subnational levels; planning receives short shrift and i s compressed into a comparatively short period o f time. As a result, planning seems irrelevant and budget execution suffers because o f it. This cycle perpetuates itself. Higher transaction costs are one outcome o f an inefficient system that can illafford it and whose limited resources for education and health are already stretched thin. Capacity at the subnational level needs to be strengthened. Accuracy o f record keeping i s often lacking, knowledge o f accounting and revenue/expenditure tracking by staff members need to be improved. 82 Due to these constraints: (1) Money for human resources and equipment i s in shortage. Schools/health centers in less well-off areas have less experienced teachers/health workers; there i s also a severe shortage o f teaching materials/ medical equipments. Furthermore, multi-grade classes are more common; so are longer commutes to education and health facilities. (2) Salaries are low for teachers and health workers; payment delays are frequent and long. Differences in allowance and living costs both work to the advantage o f teacherdhealth workers in less well-off areas. That's because teacherdhealth workers inless well-off areas receive higher allowances for working there and the cost o f living is less. Inone sense that puts them at a competitive advantage with their counterparts in well-off areas. However, teachedhealth workers in less well-off areas are more likely to experience delays inthe payment o f their salaries and the delays tend to be longer. Many teachedhealth workers have to hold multiple jobs (mainly in agriculture) to pay their bills. (3) Schools/health centers must rely on off-budget resources to cover their routine operating expenses. Schools/ health centers in less well- off areas have it even worse. That i s because their local communities and parentdpatients are unable to make up the shortfall. Donors andNGOs are often less active inthese areas. FinanceManagement Strengthening financial management to make effective use of limited resources is critical for improving service delivery. This session focuses on: (1) reviewing the PFM system in the decentralized central-local framework; (2) offering suggestions on implementing the new Budget Law to improve the PFM system; and (3) harmonizing external financial management to improve the effectiveness o f donor aid. Public FinanceManagement The central government of Lao PDR exerts little control over provincial finance departments. N o systems link provincial finance departments with the Ministry o f Finance. Lao PDR i s one o f the most decentralized countries in the world. Provincial authorities have the power to transfer funds within and across expenditure items; they also have the authority to determine spending at the sectoral level. National priority sectors, such as education and health, do not always get the same funding priority at the subnational levels that they are supposed to. As a result, these sectors often face tight resources constraints. Because of the weak systems, a lack of enforcement and of internal controls, the fiscal transfer process is unpredictable and erratic. Provinces whose revenues exceed expenditures are mandated to transfer their surpluses to the central government. The central government inturn i s mandated to transfer funds to provinces whose revenues fall short o f expenditures. And yet, projected transfers from surplus provinces often fall short o f actual transfers. No institutional mechanism exists in the national budget process for targeting priority districts (Box 8). Furthermore, it i s not possible to determine the cause for these shortfalls because there i s no revenue monitoring system at the central level. In addition, unpredictabletransfers undermine the effectiveness of limitedpublic resources. 83 Box 8: Public Expenditures in Priority Districts No institutionalmechanismexists inthe national budget for targeting priority districts. The government channels resources to priority districts: (a) favoring projects inthese districts at the provincial level; and (b) directing donor funding to these areas. However, the central government has set no minimum allocations for priority districts. And yet, an analysis o f public expenditures in the priority districts i s not possible because o f the poor quality o f local information systems. The current system allows the collection and analysis o f sectoral expenditures at the provincial level, but not at the district level. Inthe next few years, it will be necessary to enhance the monitoring o f district expenditures to assess the effectiveness o f the government's anti-poverty strategy. Provincial line agencies should start screening their expenditures by district to better target their poverty reduction programs. The new Budget Law, which i s under preparation, will set bindingminimumallocations for the provinces to spend on priority sectors. Itremains unclear whether the new budget will make it easier to monitor district expenditures. Source: World Bank (2006b). The current decentralized public finance management practices along with inadequate accountability and controls has led to: (i)an uneven application o f national policies; (ii) an accumulation o f arrears on goods and services; (iii)a funds shortfall for deficit provinces; and (iv) an inability o f the central government to ensure execution o f the budget, as approved by the National Assembly. Strengthening public financial management i s key to improving primary education and health service delivery. BudgetLaw Implementation The National Assembly passed a new organic Budget Law in November 2006. The President signed it into Law in February 2007. The new Budget Law aims to fundamentally transform intergovernmental fiscal relations. It will centralize the Treasury, Tax, and the Customs Departments and develop a new revenue sharing framework that will better align policies with the budget. The changes mandated by the new Budget Law aim to significantly improve the planning, execution and reporting of public finances. Box 9 highlights the implications o f the new Budget Law on intergovernmental fiscal relations. 84 Box 9: New OrganicBudget Law in Lao PDR-Changes in Intergovernmental Fiscal Relations The new organic Budget Law aims to address some o f the weaknesses in the current intergovernmental fiscal relations. These deficiencies stem from decentralized management o f public finances, inadequate monitoring and controls. In essence, the new Budget Law, together with the Public Expenditure Management Strengthening Program, aims to put in place the fundamentals o f a well functioning public finance management system. The key elements o f the new Budget L a w are: (i) the centralization o f the Treasury Department; (ii) the centralization o f the customs and tax departments; and (iii) development o f a robust revenue-sharing framework that would align policies with budget allocations. If successful, these new measures will significantly improve intergovernmental fiscal relations and help the government establish greater control over public finances and more clearly alignpolicies with the budget. Effective implementation o f the Budget L a w will require significant capacity, continued political commitment, technical assistance, and a well developed and sequenced 3-5 year Budget Law Implementation Plan. This Plan will centralize the three departments and develop the revenue sharing framework. In this regard it will be particularly important to ensure that inter-linkages between these elements o f the plan are carefully calibrated. Another important issue is the implementation o f the Value Added Tax-currently planned for October 2008. In order for this measure to be successful, close coordination will be required between the tax department, customs department and the Treasury. Particular attention will be needed to ensure that the VAT refund system is properly instituted and tested prior to the introduction o f the VAT. Of particular importance i s the effective centralization o f the Treasury. Without an effective Treasury, all other reform elements may not succeed. Implementation o f the Budget Law has started with the restructuring o f the Ministry o f Finance via the Prime Ministerial Decree Number 80 o f February 28, 2007. In this regard the Minister o f Finance has constituted a high level committee within the Ministry o f Finance with five subcommittees to oversee implementation o f key aspects o f Budget L a w implementation. These are: (i) preparing the implementation decree for the Budget Law and related guidelines; (ii) devising a framework for introduction o f budget norms for some sectors; (iii) developing budget execution report formats consistent with the new Budget Law; and (iv) revising the budget nomenclature. These subcommittees have started functioning. The government i s also working o n developing a detailed 3-5 years Budget Law implementationplanthat looks at each o f the mainelements o f the Budget Law, along with their inter-linkages. This plan is expected to be ready by end-May 2007. There will be need for significant and untied financial support to government for developing information systems, human resource management skills, and related technical assistance, for which greater harmonization o f donor support will be crucial. The effective implementationof the Budget Law is criticalfor improvingthe public financial management system. If the Budget Law is implemented effectively and the reform elements proposed under the PEMSP are also inplace, then Lao PDR will have a robust and efficient public financial management system: (i) budget allocations will be much better aligned with policies; (ii) government will have the ability to uniformly execute policies throughout the country; and (iii) central government will be able to the ensure that budgeted funds get spent on those items; and (iv) improved reporting and monitoring of budgetwill be possible. However, implementation o f this new law does not imply a sudden increase in resources, but it will improve revenue sharing. In case the Budget Law is not implemented in an orderly manner, there is significant risk that the current system will be in worse shape, due to the additional uncertainty in the intergovernmental fiscal framework. For example, if revenues are centralized before the 85 national treasury is capable o f sending the funds back to the provinces, then service deliverywill suffer. The whole reform programmay get derailed. Therefore it is very important to make sure that the budget law implementationis properly sequenced. For that to happen, it is important to have: (i) continued political support, (ii)significant implementation capacity and systems development; (iii) coordinated resources; and (iv) proper sequencing o f reforms and consideration of the links between revenues and expenditures; and (v) expansion o f capacity to mitigate key personnel risks. External Financing Management Misalignment of capital expenditure with recurrent needs is a key constraint of effectiveresourcemanagement.An issue of importance is that external loans andgrants that Lao PDR received from overseas development assistance (ODA) in the form o f project support finance a significant amount o f capital spending. In FY2004/05 project loans and grants amounted to 28.0 percent of total expenditures, or 74.8 percent o f the total public investment plan (PIP); budget support amounted to 9.8 percent o f total expenditures. Between 2002/03 and 2004/05, overall expenditure has been tightly squeezed: PIP went down, donor funding decreased, and non-wage recurrent started to increase slowly. It i s important to better align capital expenditures with recurrent needs. Information on grant aid is sketchy. Nearly all project loans and grants are executed outside the general budget. The money comes through special accounts operated by project units that follow donor procurement and financial management practices. In-kind supportby development partners is not effectively recordedby the Ministryo f Finance or the recipient agency because the financial flows take place largely outside Lao PDR.As grant aid is in general not reported to the system, it i s difficult to access the total magnitude and composition of assistance off-budget. However, donors do show such support as receivedby Lao PDR. The effectivenessof donor assistance is often compromisedby an inaccuratematch betweenthe donor's wish and the government's need.It is important to strengthen the capacity o f public finance management at the sector level to effectively execute donor funding in the system. To improve the effectiveness o f donor support, the government has recently welcomed steps to establish multi-donor trust f h d type of arrangements. The Avian Influenza trust fund was established. There are plans to establish another multi-donor trust fund arrangement for the public finance management strengthening program, including implementation o f the Budget Law. BolsterPublic ExpenditureManagement Strengthening Program To improve public finance management, the Lao government implemented the Public Expenditure Management Strengthening Program (PEMSP) in September 86 2005. The PEMSP has six components: (i) planning and budget preparation; (ii) fiscal budget execution, accounting and financial reporting; (iii) government financial local management; (iv) financial legislation and regulatory framework; and (v) capacity building. Tension between Ministry o f Finance and sector ministries often comes from a difference inperception: the budgets "needs" o f the sector ministries and the inability o f the MoF to finance these "needs." MoF often reverts to the previous year's budget and makes an incremental increase, which is not inline with the sector budgetproposal. To bolster the PEMSP reform agenda for basic education and health services, the findings of this report and that of the recently completed Public Expenditure Review (World Bank 2006) suggest that it is importantto closely align subnational spending programs with the national education and health goals; to allocate adequate resources for basic services across all localities; and to improve budget process to ensure effectiveuse of public resources.This is also important for ensuring effective fiscal arrangements for revenue generated by the hydro power project Nam Theun 2. From 2009 onwards, these revenues will go to fundingbasic social services. In this connection, there are three priorities: First, subnationalspendingprogramsshould be alignedwith the nationaleducation and health goals articulatedin the NSEDP. The following policies could support this principle: The annual budget process should be started earlier in the year to ensure sufficient time for input from districts and provinces. This could provide the space for frontline services to be integratedinto a bottom-up budgetprocess. Broad substantive guidelines and norms for expenditures could help -the provinces and districts allocate funds to those areas where the government has set clear national priorities. While different needs, conditions, and institutions will mean some variation among localities, a set o f uniform procedures could inform sub-national allocations at the beginning o f the budget year. These procedures could also improve transparency and limit ad hoc cash management, Second, subnational authorities should have adequate resources for basic services across all localities. The following policies could support this principle: GOL could more appropriately allocate public resources to the education and health sectors to balance the need for capital investment and recurrent spending. An increase in resources to cover wages and operating costs could help make better use o f existing and new investments and ultimately deliver better services. GOL could raise compensation for teachers and health workers and reduce payment delays. There could be incentives for teachers and health workers to work inremote areas. 87 Tie salary increases for teachers and health workers to a management policy that covers recruitment, training, andperformance accountability. Community-based programs in less well-off areas could attract and retain non- Lao-Tai students into basic education, and ensure basic maternal and child health care. Third, the budget process should ensure the effective use of public resources. The following policies could support this principle: Reassert the authority o f the central treasury over provincial and district treasuries to capture any off-budget revenues at the subnational level. Put the roll-out o f the Government Financial Information System on a fast track to the provinces and districts. Standardize the reporting formats and payment procedures at the subnational level and base them on the new budget execution reports and the Chart of Accounts. Require the provinces to publish their budgets and end-year accounts andpost them on public notice boards. 0 Strengthen the education and health management information systems at the subnational levels by giving them effective planning and monitoring tools. Accelerate capacity building in financial management at the subnational level bypairing a needs assessment with animplementation strategy. Steps forwards A PETS offers a technique for tracking public expenditures and their impact on service. This pilot Lao PETS succeeded in obtaining rich and valuable information in most areas, such as facility characteristics, financial flows, and accountability arrangements, thanks to expert data collection by the NSC and the enumerators. Thanks also go to the Ministries who ably assisted inthe preparation o fthe questionnaire. Buildinglocal capacityis importantfor improvingpublicfinancialmanagementand service delivery in Lao PDR. It i s also important for conducting an effective survey. It may be helpful to clarify upfront in the survey what budget items the questionnaire covers, as these items may be different from what the respondents envision. Improving survey design to cross-check questionddata and to strengthen interim survey data quality control are also technically advisable. With the strengthening of local capacity and improvement of survey design and techniques, future PETS should aim at providingbroader and deeper information and analysis on public expenditurestracking and service delivery. 88 0To better track public expenditures and understand the constraints that hinder service delivery, for the next PETS, it would be helpfulto have the following information more accurately: 0 Expenditures and revenues-How much educatiodhealth budget resources are allocated from the center/province/district to province/district/facility? For what purpose? When? How much o f it actually reaches its intended audience? When do they receive it and how do they use these resources? What share o f budget and off-budget resources does it represent out o f total revenues and expenditures? 0 Budget process and financial management-When are schools/DEB/PES and HC/DFO/PFO supposed to submit budgetproposals to their respective upper layers? D o they submit them on time? What are the procedures? How much do they propose? How much i s approved and allocated? How much do they actually spend, for what? How do funds flow within and betweenthe layers o fthe MoF and the lineministries MoE/MoH? To strengthen the analysis at the frontline level, it would be useful to have the following information more accurately: 0 What is the approximate share o f total resources that schools or health centers receive from each source, namely the central budget, provincial/district authorities, DEB, DFO, donors, NGOs, communities, and parentdpatients? 0 How did they spend these resources? Does this match the intended purpose? 0 What is the timing o f the inflow o f resources and the outflow o f expenditures? Are there gaps between the two? If so, how are they filled andby what source? What resources are spent most efficiently? Under what conditions and where? What contributes to that efficiency? What bottlenecks do provinces/districts/facilities face? If a brief user survey can be included, what is the parents'/students' and patients' perception of the quality o f service they receive? Does the service match the potential demand? 0Giventhe frequent and common delays inpayment o f salaries to teachers and health workers, it would be useful to have the following information more accurately: 89 0 D o teachers get paid in full when there is a delay? Are salary delays a result o f limited resources? Ifso, has it become standard procedure to wait for next year's budget to satisfy the delayed salary payments incurred this year? If total resources are insufficient for operating costs and salaries, would salaries get a higher priority over funds originally assigned to operating costs? Ifresources are not enough to pay all teachers'health workers' salaries in full, did some teachershealth workers get paid in full on time while some did not get anything? Did all get paid a certain percentage o f their salaries on time? Or did they all have to wait before resources became available to pay them infull?How did this affect the amount o f arrears and the length o f the delay? Conclusions Primary education and health service in Lao PDR have improved over the last decade with reasonable outcomes. The PETS results suggests that salary leakages from the district to facility levels are low, ghost-workers and absenteeism are not big concerns, the morale o f teachers and health workers is high, and parent/patients and community participation i s active. However, insufficient resources and capacity, especially at the local level, inefficient and inequitable allocation of funds, and weak public financial management undermines education and health services, particularly for poor people in less well- off areas. Teachers' and health workers' salaries are not only low but these workers also frequently experience delays in payment. Schools and health centers face an acute shortage o f funds to cover their operating costs. They have to rely on off-budget resources. Donors' aid does not always match local needs. To improve education and health service delivery, it is important to allocate more resources to these national priority sectors and strengthen financial management. Effectively implementing the new Budget Law effectively i s a high priority. That single step should help to transform the intergovernmental fiscal relations and align policy with budgets. It should also bolster the Public Expenditure Management Strengthening Programreform agenda. This pilot PETS succeeded in providing rich and valuable information, such as documenting frontline service characteristics, financial flows and accountability arrangements. It was the first attempt to track salary payments to teachers and health workers from the districts to the facilities. The findings o f this PETS serve as useful baselines for future studies. With improvements in local capacity and survey techniques, future PETS will provide broader and deeper information and analysis on the tracking o f public expenditures and the delivery o f services. 90 REFERENCES Chaudhury, Nazmul et al. 2005. Teacher and HealthWorker Absence inDeveloping Countries. Government o f Lao PDR, 2000-2006, Official Gazette. Government of Lao PDR, 2006, National Statistics Center, Lao PDR Expenditureand Consumption Survey I11(2002-03). Government o f Lao PDR, 2007, Population Census 2005. King, Elizabeth and Dominique van de Walle. 2005. Schoolingand Poverty inLao PDR. Background Study for the Lao PDR Poverty Assessment Report 2006. Noonan, Richard. 2001. EducationFinancing inLao PDR. Noonan, Richard. 2007. Ethnicity and Participation inPrimary EducationinLao PDR: Some Statistical Results from the 2005 Census. World Bank. 2003. World Development Report 2004. World Bank. 2004. Lao PDR Country Economic Memorandum2004. World Bank. 2005. World Development Indicators. World Bank. 2006a. Lao PDR Poverty Assessment Report: FromValleys to Hilltops Fifteen - Years of Poverty Reduction. World Bank. 2006b. Public Expenditure Reviewand IntegratedFiduciary Assessment. World Bank. 2006c. Central-Local Fiscal System. Background Study for the Lao PDR Public Expenditure Review. World Bank. 2006d. Notes for the Poverty and Social Impact Analysis. Technical Report. World Bank. 2006e. Baseline Survey 2006: Lower Secondary Schooling inLao PDR. World Bank. 2007a. Programdocument, Lao PDR Poverty Reduction Support Operation 111. World Bank. 2007b. Lao PDR Civil Service Reform for teachers: Teaching inLao PDR. 91 IBRD 33431 100°E 102°E 104°E 106°E 108°E This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. C H I N A LAO P.D.R. To Lincang To Gejiu To Daluo Gnot-Ou 22°N Ou 22°N Phongsaly Boun-Nua PHONGSALY MYANMAR VIETNAM LUANG NAMTHA Muang Luang Khoa To Namtha Hanoi Meung Muang Xiangkho MekongBOKEO Xai Nambak Ou Sam-Neva Ban Beng LUANG Viangxai Huaisai Ta OUDOMSAI PHRABANG HOUAPHAN Rai Xiangkhoang Plateau 20°N Pakbeng Luang ChiangoT Phrabang SAYABOURY Plain of Jars Kham Nan Phokhoun Xiang Khoang XIANG Sayaboury Kasi KHOANG Phou Bia Gulf of Vangviang (2,817 m) Ban Nalé Nalé Xaisomboun Tonkin VIENTIANE Nam Ngum Reservoir Pone BOLIKHAMSAI Paksane Khamkeut Paklay Me kong Hong Mekong 18°N Xanakham VIENTIANE Kading 18°N PREFECTURE OF KHAMMOUANE VIENTIANE MUN. Cammon Plateau Ban Na Phan Thakhek To Vinh To Khon Kaen Noy XeBangfai To Khon Kaen Xebangfai 0 50 100 Kilometers SAVANNAKHET Xéno Xéno Xepon 0 25 50 75 100 Miles THAILAND Savannakhet Phin To Qui Nhon Se Banghiong Samouay LAO PEOPLE'S 16°N 16°N DEMOCRATIC SARAVANE Don Saravane REPUBLIC Khongxedon SEKONG Sekong SELECTED CITIES AND TOWNS Pakxong To PROVINCE CAPITALS Ubon RatchathaniCHAMPASSAK Bolovens NATIONAL CAPITAL Champassak Plateau Attapeu Sanamxai RIVERS ATTAPEU TT MAIN ROADS RAILROADS Khong PROVINCE BOUNDARIES 14°N 14°N CAMBODIA INTERNATIONAL BOUNDARIES 104°E 106°E NOVEMBER 2004