ACKNOWLEDGEMENTS This report was produced by a team led by Sanjay Dhar and comprised Laura Bailey, Malcolm Bale, SoJia Bettencourt, Benu Biclnni, Josk Bmz, Julia Devlin, Peter Fallon, Jean Foerster, Adrian Fozzard, Francis Ghesquiere, Ronald Hood, Christopher Joubert, Michael Lewin, Patrick Meagher, Kasper Richter, Michael Stevens and Craig Wilson. Muriel Greaves assisted with document processing. Sarah Cliffeprovided helpful sst~ggestionsat various stages. Saumya Mitra and Michael Stevens sewed as peer reviewers. The report was produced under the gtlihnce of Homi Kharas and Klaus Rohland. A CIYClft of the report was provided to the East Timorese authorities in December 2001 and was discussed with the authorities in March 2002. The team grateft~llyacknowledges commentsfrom government ofJicials in East Timor. CURRENCY EQUIVALENTS Currency Unit: US Dollar FISCAL YEAR July 1 - June 30 WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS ADB Asian Development Bank ARP Agricultural Recovery Project AusAID Australian Agency for International Development BE Business Environment BPA Banking and Payments Authority CA Constituent Assembly CEP Community Empowerment Project CFA Central Fiscal Authority CFET Consolidated Fund for East Timor CGIAR Consultative Group on International Agricultural Research CPO Central Payments Office EEZ Exclusive Economic Zone ETTA East Timor Transitional Administration EIRP Emergency Infrastructure Rehabilitation Project FA0 Food and Agriculture Organization FDI Foreign Direct Investment FIAS Foreign Investment Advisory Service FTP First Tranche Petroleum GDP Gross Domestic Product IAS International Accounting Standards ILO International Labor Organization ITPT Information, Technology, Post and Telecommunications JAM Joint Assessment Mission JICA Japanese International Cooperation Agency JV Joint Venture LPU Land and Property Unit MAFF Ministry of Agriculture, Fisheries and Forestry NCC National Consultative Council NGOs Non-Governmental Organizations OECD Organization of Economic Cooperation and Development SEP Small Enterprise Project TFET Trust Fund for East Timor TLSMS Timorese Living Standard Measurement Survey USAID United States Agency for International Development UNTAET United Nations Transitional Administration in East Timor VAT Value Added Tax TABLE OF CONTENTS Page No . EXECUTIVE SLMMARY.............................................................................................................. i INTRODUCTION ........................................................................................................................ 1 CHAPTER 1:THE TROUBLED PATH TO INDEPENDENCE ............................................... 3 Historical Background .......................................................................................................... 3 Prelude to Independence ....................................................................................................... 5 CHAPTER 2: ECONOMIC AND POVERTY DEVELOPMENTSAND ISS'LTES................... 7 Economic Repercussions of the Ballot and Transition Period ............................................. 7 Characteristics of Poverty .................................................................................................... 11 Economic Issues of Transition.............................................................................................. 17 CHAPTER 3: IMPROVING THE BUSINESSENVIRONMENT ............................................. 21 Infrastructure ..................................................................................................................... 2 2 The Legal Environment......................................................................................................... 26 The Legal Framework for Land Ownership.......................................................................... 31 CHAPTER 4: RAISINGAGRICULTURAL INCOMES AND PRODUCTIVITY .................35 The Importance of Agriculture in the Economy ................................................................... 35 Medium-Term Framework and Principles ............................................................................ 39 Policy Issues in Agriculture .................................................................................................. 41 Donor Assistance and Funding Requirements ...................................................................... 50 CHAPTER 5: ADDRESSING HUMAN DEVELOPMENTNEEDS......................................... 54 The Indonesian Legacy ......................................................................................................... 54 The Transitional Strategy...................................................................................................... 58 Towards a Medium Term Framework ..................................................................................62 CHAPTER 6: TIMOR SEA REVENUE : ISSUES OF MANAGEMENT AND GOVERNANCE.................................................................................................................. 70 Oil: Blessing or Curse? ........................................................................................................ 70 Expenditure Management ..................................................................................................... 71 Management of Revenue and Savings .................................................................................. 73 CHAPTER 7: MAINTAINING PRUDENT FISCAL POLICY.................................................. 80 Fiscal Developments: An Overview ....................................................................................80 Tax Policy .............................................................................................................................85 Financing Options .................................................................................................................87 . Fiscal Sustainability..............................................................................................................88 CHAPTER 8: ENSURING EFFECTIVE PUBLIC EXPENDITUREMANAGEMENT.........90 Budgeting. Expenditure Programming and Performance ..................................................... 90 External Assistance Management .......................................................................................100 Capacity in Public Expenditure Management..................................................................... 102 Potential Threats and Opportunities .............................................................................. 104 CHAPTER 9: PROMOTING EFFECTIVE GOVERNANCE.................................................106 Building the Civil Service...................................................................................................106 Promoting Effective Governance and Reducing Corniption Opportunities .......................115 Empowering Local Communities........................................................................................ 125 BOXES IN TEXT Box 2.1: Declining Competitiveness in Coffee Processing ............................................................ 1 8 Box 2.2. -Bruneiand the Crowding Out Effect.................................................................................. 20 Box 3.1. The Judiciary ......................................................................................................................28 Box 4.1: Principles of Donor Assistance for Agricultural Sector....................................................51 Box 6.1: The Tuvalu Trust Fund.......................................................................................................77 Box 8.1: The Budget Execution Process.......................................................................................... 95 .. Box 8.2. Botswana: Lessons in Managing Mineral-Led Development ............................................ 103 FIGURES IN TEXT Figure 2.1: Indicators by District of Impact of the Violence of 1999............................................... 16 Figure 7.1: Expenditure Allocations. 2001-02................................................................................... 83 Figure 7.2. Capital Spending Allocations. 2001-02........................................................................... 83 TABLES IN TEXT Table 2.1. GDP of 2000 by Industrial Origin at Current Market Prices ........................................... 7 Table 2.2. Key Economic Indicators. 1998 .2002 ........................................................................... 9 Table 2.3. Selected Social Indicators ................................................................................................13 Table 2.4. Self-Assessed Change Since Before the Violence in 1999............................................1 6 Table 2.5. Perceived Changes by Sector..........................................................................................17 . . . .Table 4.1: Agriculture's Contribution to Rural GDP, 1993-1999 ...............................................3 6 Table 4.2. Agricultural Production. 1997.and2001 .........................................................................39 Table 4.3. Current Estimated Funding for Agriculture and Rural Development, 2001-2005 ..........53 Table 5.1. Cross-country Comparisons of Per Capita spending in Health and Education..............59 Table 5.2. Key Education Statistics Before and After the Referendum (Public and Private) ..........59 Table 7.1: Expenditure and ETTA Revenue ....................................................................................81 Table 7.2. Combined Sources Budget 2001-02 ................................................................................82 Table 7.3. Consolidated Fund Accounts and Estimates....................................................................84 Table 7.4. Actual and Projected Revenues .......................................................................................86 Table 9.1: Civil Service Employment by Level (2000-2001).........................................................107 Table 9.2. A Comparison of Monthly Salaries ..............................................................................110 Table 9.3. Civil Service Pay Compression Ratios for Various Countries ..................................... 112 Table 9.4. International Comparisons of Civil Service -Wages and Employment .......................113 ANNEXES Annex 1: Characteristics of Poverty ................................................................................................. 128 Annex 2: Rice Production and Markets ............................................................................................ 133 Annex 3: Decompressing Civil Service Salaries .............................................................................. 148 . . REFERENCES ................................................................................................................................. 150 EXECUTIVE SUMMARY 1. East Tilnor is due to become the first new state of the 21'' century on May 20, 2002 after inore than four centuries of colonial rule and a quarter century of occupation and conflict. Following the tragic events of 1999, East Timor, with generous financial and technical assistance from the donor community, has made enormous progress in rehabilitating its economy, reconstructing its infrastructure, reintegrating its refugees, and building key elements of a sustainable political process in an environlnent marked by inteinal peace-the latter in sharp contrast with other post-conflict societies. 2. Since 1999, when real GDP may have fallen by about a third, economic recovery has been strong-a working assumption of over 15 percent average annual GDP growth during 2000 and 2001 has been used. While output was primarily boosted by the presence of the large expatriate community and donor-financed reconstruction efforts, the recovery was not limited to the urban sector; agricultural output over the past two years is also estimated to have recovered significantly. Overall growth is, however, expected to slow sharply in 2002 and 2003 as large numbers of expatriate personnel leave the country. 3. The main challenge facing East Timor is how to reconcile a simultaneous existence of acute poverty and severe shortage of human management skills wit11 solid prospects of future flows from the country's natural resource wealth. Policies to meet these two priorities-sustained poverty reduction and sound management of natural resource wealth-are the focus of this report. 4. Tra~zsitionalIssues. A pressing concern in the next two years is that of managing the ecoilomic transition from LINTAET administration, which will involve the departure of much of the expatriate community and the associated loss of employment and income for Timorese citizens-and the loss of critical technical expertise. In the short term, East Timor will not have the option of allowing its exchange rate to depreciate in response to the adverse demand shock of departing expatriates, given the use of the U.S. dollar as the official currency. A further transitional issue is how to manage public resources to bridge the period until adequate offshore revenue can be realized. 5. Wealtlz Creatio~z.In the medium term, the expected increase in offsl~orerevenue constitutes an added threat to the coinpetitiveness of the non-oil traded goods sector. In order to develop an economy not wl~ollydependent on aid and oil, the nation needs to enhance the private investment climate, foster agricultural incomes and productivity, promote a sustained improvement in human development, and ensure that the nation's natural resource wealth is invested wisely. The impediments to improve the business environlnent are serious given East Timor's historical legacy of conflict and isolation and the ellonnous gaps this has created wit11 the rest of Southeast Asia in terms of infrastructure, institutions, legal framework, education and productivity. The manner in which these gaps are reduced will be key to raising einployllent and productivity and reducing poverty. 6. Hzunan Development. In a situation of multiple demands competing for limited resources, the iinportance of raising human development standards for the quality and pace of long-term growth cannot be overstated. Basic education and training needs are fornlidable and together with core health sei-vices must remain the predominant claimant of public resources in any development strategy. The design of effective and affordable measures to assist the poor is also essential; hence the need to better understand the characteristics of poverty and the capacity of families to pay for some services. 7 . Gover~za~zcemid Pziblic Sector Managenze~zt. Post-colonial and post-conflict experience amongst poor countries is unfortunately replete with examples of resource misallocation, mismanagement and corruption. Heavy reliance on aid and natural resources exacerbates these concerns. In East Timor's case, a limited number of qualified personnel to tackle the institutional challenges of building a nation from scratch compounds these risks. The ability to formulate and adhere to effective means of allocating resources towards high priority development objectives will be critical. Initially, the honesty and dedication of the small core of existing leaders will be essential; over time, a larger cadre of capable civil servants must be developed. Building and sustaining effective governance will thus require a range of institutional reforms, which will be difficult. 8. Poverty Redzictio~z. Preliminary results from the recently completed household survey (TLSMS) indicate a 40 percent poverty incidence in 2001. The incidence of poverty is lower in urban areas (25 percent) than in rural areas (44 percent). Poverty is overwhelmingly a rural phenomenon: 85 percent of all the poor live in rural areas. Poverty incidence is lowest in DiliBaucau (14 percent) and highest in the rural center and the rural west (49 percent). Agriculture, not surprisingly, is the primary occupation for the population and particularly so among the poor. In East Timor as elsewhere, there is a strong correlation between consumption poverty and low levels of education. Ownership of assets, including access to land, is also correlated with lower poverty. 9. Estimates from a survey conducted in all the country's villages (sucos) show a significant destruction of the housing stock and loss of livestock. Among the more striking results of the suco survey is the extent of food insecurity: over 40 percent of sucos report not having enough to eat between November and February. This share peaks in January, when 80 percent of sucos report facing food shortages. 10. The Tinlorese authorities are well aware of the challenges they face: following the August 30, 2001 Constitueilt Assembly elections, a process to design a medium-term strategic framework to articulate and implement development objectives has been initiated, and a Planning Coinmission has been established to coordinate this work, with the objective of producing a National Developn~entPlan by independence. A key objective of an earlier draft of this report was to serve as an input into that effort. 11. The objective of sustained growth and poverty reduction would clearly be facilitated by fostering a business environme~ltthat is conducive to private investment and tlvough continued ag~iculturalrehabilitation and assistance to farmers in the fornl of capital, knowledge and marketing expertise-thereby generating greater private employment and higher rural productivity and incomes. At the same time, there are serious impediments to building a significant manufacturing base, and the prospects for rapid agricultural growth on a sustained basis are also uncertain. 12. These factors highlight the importance of continued donor support in the short term, sound management of East Timor's oil and gas wealth over the long term, and the effectiveness with which aid and offshore wealth can be channeled towards poverty reduction, in particular to build the human capital base. This in turn requires: the development of sound savings and investment strategies to extend the benefits of offshore wealth beyond the windfall revenue period; the adoption of transparent public expenditure management and budgeting tools to utilize aid and revenues effectively; the development of an efficient and service-oriented civil service; and fostering a culture of accountability and results orientation that minimizes the scope for corruption. Addressing Distortions from the Large Expatriate Presence and Urban Bias 13. The presence of the large and wealthy expatriate community has raised wages and prices well above those in neighboring regions or in countries with comparable levels of productivity despite the prevalence of widespread underemployment. This has already constrained the development of small-scale enterprise and tourism and reduced employment in other productive activities (e.g., in coffee processing), nonetheless raising Dili's population significantly from its pre-1999 ballot level of 120,000. As the expatriate community begins to depart, it is essential not to embed through regulation or legislation a price structure that reflects the current artificial dual economy, but instead to allow the flexibility that would be conducive to private sector development and associated job creation. The risk of wage and price rigidity is heightened by the inability to allow the exchange rate to depreciate in response to the demand shock of departing expatriates. With little capacity to manage its own exchange rate or monetary policy in the short term, the option of East Tinlor adopting its own currency would appear ill-advised in the near term.' 14. The combination of higher than sustainable wages and exchange rate rigidity implies that wages and prices must be allowed to adjust in response to economic conditions rather than bureaucratic regulation. Given the artificial nature of current urban conditions, a floor should not be set for minimum wages, at least until the impact of the expatriate departure has been fully felt. This would not preclude promoting other aspects of labor market regulation, although special effort will be needed to align such regulation to the realities of East Timor's prospective labor market situation following the expatriate departure. 1 Under UNTAET, the US dollar was adopted as East Timor's official currency, though the Indonesian rupiah continues to circulate in the rural economy. 15. The large but temporary expatriate preseilce call have significant economic and social costs, for example, in ternls of encouraging uilsustainable rural-urban migration and misallocated investment to service expatriate needs. The incentive to remain in agriculture and rural areas will depend inost iinpoi-tantly on iilcoine prospects and the quality of rural life. The report addresses these objectives through highlighting the importance of maintaining the prevailing effort at raising farm incomes and productivity, improving the quality of rural educatioil and health facilities, and promoting an appropriate governance structure at the comnlunity level. The donor community itself can mitigate the repercussions of its departure by phasing the withdrawal of LWTAET staff- ensuring also that skills that are difficult to replace locally are not abruptly withdrawn- and providing favorable severance packages to local staff made redundant, which could facilitate startup of small-scale enterprise. 16. Tax policy can also play a role to avoid exacerbating urban bias: High priority needs to be accorded to connecting urban customers with electricity and water meters and enforcing utility payments, thereby removing the free use that has prevailed through most of the UNTAET period. Taxation of urban services (hotels and restaurants) will need to be maintained and taxation of rural commodities should be avoided. Improving the Business Environment 17. East Timor's current development has been dominated by donor-financed projects, but fostering an extended period of labor-intensive growth will require sustained private investment, both domestic and foreign. Yet, there are major impediments to private sector development reflecting: the violence in 1999 which destroyed power generation capacity, telecommunications infrastructure, real estate and land records; the severance of the legal and regulatory system from that of Indonesia in 1999; as well as longstanding deficiencies in the legal framework for land ownership, broader legal environment and judicial capacity. Business services have also suffered from the exodus in 1999 of Indonesians and Chinese who dominated these fields, whereas local entrepreneurial capability has suffered due to the historical legacy. 18. Capacity and resource constraints imply that not all impediments to investment can be resolved at once; priorities thus need to be defined. High on the list of priorities is to resolve the complex set of conflicting land claims, originating from the Portuguese and Indonesian periods, that has already inhibited commercial activity and investment and spawned various forms of land fraud. Lack of clarity over ownership of land and real estate assets also cuts off the most important source of collateral that could be used for commercial lending. Work on establishing a formal land administration system that confirms ownership of registered land should hence proceed with urgency, with priority given to urban areas where conflicting claims and disputed ownership are most acute and pressures for security of tenure are greatest. 19. The inability to collateralize assets also reflects the lack of a functioning Company Law, which inhibits creditors from making enforceable claims against corporate assets, and the lack of a legal framework for bankruptcy. Rectifying these gaps is needed to facilitate lending (at their own risk) by the existing two foreign banks to commercial and individual clients, and to encourage prospective foreign banks to begin operations in East Timor. Given the nascent bank supervisory capacity and inability to insure deposits or respoild to bank runs, the authorities should avoid granting commercial banking licenses to local investors, and limit future liceilses to respectable foreign banks that are subject to regulation and backed financially by their own regulatory authorities. 20. Microfinance programs could play an important role in developing both the private and rural sectors. In recognition of this, several programs have been established with the support of donors, though start up has been slow. Moreover, the focus has been on the general business sector due to that sector's relatively greater loss of physical capital following the 1999 destruction. None of the available credit programs however tailor lending to the particular needs of fanners. There is thus a need to ensure that farmers and rural-based enterprises have sufficient access to credit, offered in conjunction with technical initiatives and training in responsible use of credit. 21. There is now clear recognition within government of the importance of an effective and efficient telecommunications system for the business sector and indeed for the broader objectives of economic development: the recently adopted approach has been to bundle into one contract all the major required system elements, and expressions of interest from international telecommui~ications companies have been sought and received. It is essential that the coming transition from the current UNTAET-supported telecommunications arrangements to the proposed private structure occur with minimum disruption-hence the need to expeditiously tender the telecommunications network contract. 22. A stable supply of power provided at competitive rates is also an important consideration for potential investors. While electric power generation has recovered to 90 percent of its pre-referendum operating capacity, the period since the 1999 referendum has been marked by severe power shortages and outages, substantial losses by the Power Authority and consequent claims on the budget, delayed introduction of metering, and a relatively high cost structure. Current objectives of attaining financial self-sufficiency for the Power Authority through careful rate setting, increased metering and lowering the billing threshold, are appropriate. 23. The road network is extensive relative to the size of the economy, but constructed with low standards of design and material-the cost of maintenance and upgrading the network has been estimated at $24 million per year, which is larger than the Government's onshore revenue in FY02. Moreover, the roads department faces a severe shortage of qualified local staff, necessitating the continued use of foreign staff. Hence difficult decisions are needed to clarify the maintenance respoilsibility of central government; in such prioritization, con~mercialand social factors both need to be incorporated. Raising Agricultural Incomes and Productivity 24. Piior to the 1999referendum, agriculture and its support activities employed more than 75 percent of the population, contributed 32 percent of GDP, and accounted for 90 percent of foreign exchange. With most fainlers engaged in subsistence agriculture often iilvolving multiple crops, enhancing fann incomes and productivity is clearly central to any poverty reduction strategy and to the goal of food security. Yet agriculture was not spared from the destructive frenzy of 1999; in the afternlath of the 1999 referendum, starvation was perhaps oilly avoided through the tradition of solidarity and sharing in communities, the availability of traditional standby crops (cassava) and wild foods (sago) and the free distribution of rice, cooking oil and other food supplies (often by helicopter into remote areas). 25. Agricultural production is estimated to have recovered significantly during 2000 and 2001 for most crops-aided by concerted rehabilitation and reconstruction efforts and the provision of extension services by the donor community and NGOs, and despite the departure of many farmers of Indonesian origin. The recovery has occurred in conjunction with major adjustments. The previous Indonesian practice of regulated prices for inputs and outputs, control over the distribution of inputs, and government purchase of surplus production ended abruptly-reflecting the sudden departure of the overseeing bureaucracy-and was replaced by exposure to international prices (with imports subject to an across-the-board 10% tariff). At the same time, farmers have benefited from a level of external assistance that will be difficult to maintain indefinitely. 26. The adjustment for rice farmers appears to have been particularly traumatic, or at least better publicized. Though rice production in 2001 is estimated to have recovered to about 90 percent of its 1998 level and well exceeds production in 1999 or 2000, farmers have faced difficulties in selling their surplus, as distribution networks and marketing for imported rice in Dili and upland communities appear better developed. Local rice producers face difficulties in accessing markets due in particular to a lack of post harvest processing, especially packaging for transport, and unfamiliarity with distribution and marketing given the previous regulated system. Lack of stores to purchase inputs and inadequate access to credit also represent significant hurdles. By contrast, the price of imported rice at the retail level does not appear to be lower than local prices-though this may reflect higher mark-ups and profits for distributors of imported rice-and free donated rice does reduce the ability of local farmers to sell their surplus. 27. A sustainable policy response should include addressing the specific problems ailing rice farmers-post harvest processing, packaging, marketing, credit access, and facilitating the availability of inputs coupled with tariff exemption. But these problems cannot be resolved instantaneously, giving rise to considerable pressure for increased protection in the form of higher tariffs on rice. Such a policy could perhaps be warranted if the higher tariff was modest and temporary-to allow a reasonable time frame for the above adjustments to be developed. 28. But there are also serious drawbacks to raising the tariff on rice that need to be carefully considered before any decisioil is reached. First, rice farmers comprise some 26 percent of households (not all of wl~onlare net sellers of rice); ally boost to their incoines from higher protection nlust be weighed against the loss of purchasing power to the rest of the population, in particular the reduced consun~ptionof rice by poor consumers in urban and upland areas (the latter is where poverty is most acute). Second, as it is a basic consumer good, raising rice prices would exacerbate the problems with competitiveness noted earlier. Third, even if a higher tariff is intended to be temporary, international experience indicates that reversing the higher protection would be difficult in practice. This in tuin could promote shifts in cropping patterns not necessarily beneficial or sustainable in the long term. These factors suggest that a hasty decision on higher protection should be avoided, and any such decision should be preceded by developing feasible means to assist poor rice consumers. 29. Beyond rice policy, East Timor will need to decide on the broad policy framework it wishes to follow for agriculture. This report's assessment is that after a transition period, the country would be best served by allowing maximum flexibility to farmers in production and marketing decisions, with the government's role focused on supervising and financing its own agricultural investments, establishing and enforcing a regulatory framework and providing services, research and education. This limited role (relative to the Indonesian era) is suggested not only because of the government's limited capacity to intervene more aggressively and its scarcity of resources for doing so on an ongoing basis, but especially because of evidence from other small economies of the benefits of relying on market signals and international prices for guiding farmer decisions. 30. Food security. In the long term, food security will be achieved when households are productive enough to grow or purchase adequate food for their families. But short- run food contingencies are better dealt with by establishing a buffer fund instead of aiming for self-sufficiency in food production through managing public stocks, which in other countries has distorted farmer incentives and proved expensive and difficult to sustain. A buffer fund would be earmarked to provide assistance directly to the most needy (in cash or vouchers) during exceptionally poor harvests, thus obviating the need for government to purchase staples, store grain or control prices itself. 31. Coffee is East Timor's major onshore export and an important source of livelihood for an estimated 40,000 families. Despite the current very low level of world prices, there still appears to be considerable scope for increasing farmer income from coffee. Particular benefits call derive from improvements in the processing of coffee beans and several alternatives for increasing the value added in East Timor need to be explored. It is also important to provide assistance to farmers in improving the coffee tree stock (replanting coffee and shade trees and pruning) and in organizing more competitive marketing facilities. The organic nature of East Timor's coffee is a factor that can give it a premium over the average world coffee price and needs to be safeguarded in effective and cost-efficient ways. Addressing Human Development Needs 32. A core feature of any poverty reduction strategy nlust include efforts to ensure everyone-pal-titularly the poor-has effective access to an affordable basic education and package of essential health care services. A healthy and educated population will enable a more productive workforce to respond to the changing enviroi~nentand develop a population better able to contribute to the cost of services. 33. Tlze I~zdo~zesia~zLegacy. While provision of education and health services was significantly expanded under Indonesian rule, East Timor nonetheless inherits poor education and health outcomes relative to either Indonesia or the region more generally. Overall illiteracy remains high-about 50 percent of the population-and even universal primary educatioil has not been achieved. Net enrolment rates in 2001 (i.e. the proportion of the relevant age cohort attending each level of schooling) were only 74 percent for primary aged children, 22 percent for junior secondary aged children and 16 percent for senior secondary school aged children. Rural-urban, income and sex inequalities in access and completion of school are also evident. 34. Health outcomes during the late 1990s remained unacceptably poor. Life expectancy remained very low, by one estimate below 50 years. The infant mortality rate was very high at 135 per 1000 live births in 1995, the reproductive health status of the population was of serious concern, and the total fertility rate was 4.43 (compared to 2.78 in Indonesia). The natural rate of population increase-of about 2.2 percent is thus fairly high. Further, with only 22 percent of births in 1998 attended by a doctor or midwife, maternal mortality rates were also very high. Malaria and TB constitute the two most serious diseases. But despite high shares of outpatient visits linked to malaria and TB, outpatient visits were very low atjust over 1visit per capita. 35. Tra~zsitio~zalStrategy. UNTAET and the donor community have accorded very high priority to education and health services. Considerable effort has been directed at improving the efficiency of public expenditure on health and education by dealing with chronic overstaffing issues, increasing the previously low civil service wages, replacing infrastructure only where there is a demonstrated need, ensuring needed quality inputs are financed and trying to focus attention on service delivery and outcomes. Both sectors have had to confront massive skill shortages caused by the departure of Indonesian secondary school teachers, doctors and administrators. 36. Aggregate school enrolment rates were restored by 2000, although junior secondary and senior secondary enrolments remain below pre-disturbance levels. Planned investments will ensure capacity for junior secondary school enrolments to rise above pre-disturbance levels. The secondary school and tertiary education system, however, face significant skill shortages, creating the need for in-service training. 37. In health, efforts were made to reestablish district health services and essential inpatient services while establishing a central health authority and a capacity to develop hunlan resources. From the outset, there was a recognition of the need to redevelop the health system within a sustainable policy framework. From a programmatic perspective, emphasis was placed on the delivery of a core program that inet the major public health problenls facing the country-prevention, inmunization, TB control, basic clinical care (including for malaria) and the development of a refei-ral mechanism for enlergeilcy care. 38. Towards a Medium-Temz Franzework. The FY02 coinbined sources budget and the foiward projections for the operating budget for the period to FY05 constitute iinportant steps towards placing the budget within a medium-term public expenditure framework. There remains some uncertainty, however, as to whether all the medium- term pressures on the education and health budgets-which comprise 38 percent of the total CFET budget-have been captured. To ensure an appropriate public expenditure planning process, it is crucial that a "whole of policy costing approach" is established for all expenditure, however financed. It is also vital that sustained operating costs are clearly identified and fully incorporated before decisions are made to implement new policy. This has significant repercussions for donors and how they cooperate with government. The accountability process within agencies and the associated procurement, accounting and auditing procedures also require ongoing focus. 39. Mediunz-Term Issues Faci~zgEducatio~z. A crucial question is how fast to expand each level of education. Given resource and teacher supply constraints and the relative benefits of expanded education, a commitment to ensuring that all children complete at least primary school should be the top education priority-and is achievable in the first years of independence. Achieving universal junior secondary education in the medium term will be more difficult in light of added costs and the limited supply of trained teachers. Thus increasing the skills and competencies of the teaching force should be given high priority; the pace at which junior secondary schooling can be expanded will depend on this and the availability of additional resources, while maintaining current continuation rates for senior secondary schooling appears appropriate. Private returns to tertiary education are likely to be very high in the short term, and the reintroduction of fees for families who can afford them is warranted. Generous donor support will continue to be required; but it is important that such support is optimized to meet the immediate skill and educational needs of the economy; and that education and training capacity is developed consistent with long-term needs. 40. Mediunz-Ternz Issues Facing Health. The Ministry of Health has made significant advances in the development of a program budget for FY2001-02 consistent with the principles of public expenditure planning. The range of issues where attention will need to be focused in the medium term include finalizing: a hospital development strategy that contains hospital costs to 35-40 percent of the total health budget; the transition of responsibilities for district health programs and the implementation of district health plans to district health management teams; a human resource development strategy including the developnlent of a nurse practitioner cadre to partially replace some of the activities perfoimed by doctors; a health policy framework including a pricing policy which takes account of the needs of the poor; and a need to continue to refocus resources and efforts on the core health care package which can have the greatest impact on health outcomes-immunization, communicable and vector-borne diseases including malaria, TB and HIVIAIDS, child health, reproductive health and preventive health programs. Managing Timor Sea Oil and Gas Revenue 41. 11 July 2001, East Tiillor negotiated ail arrangement with Australia, which is subject to ratification by the East Tiillor and Australian parliainents, on the division of oil and gas resources from the Joint Petroleun~Development Area (JPDA) of the Tinlor Sea, with 90 percent accruing to East Timor. Subsequently, a tentative agreement on the development of the Bayu Uildan LNG project in the JPDA was reached, which could yield revenue to East Timor over a 20-year period beginning in 2004. Negotiations on the development of another large field (Greater Sunrise) are ongoing, and production is not expected to begin before the second half of the decade. While the legal basis of these agreements remains to be finalized and their revenue impacts are subject to uncertainty, the prospect that East Timor will be able to drastically reduce its dependence on external assistance within three years appears promising. 42. Based on the experience of other natural resource-rich countries, the prospect of a significant oil and gas revenue windfall call be a mixed blessing as there are many dangers and pitfalls. For example, over-spending by governments during the boom can raise the relative price of non-traded goods and draw resources away from the non-oil traded goods sector, retarding its development. Government spending can easily exceed absorptive capacity, reducing investment efficiency, increasing waste, and bequeatlung unsustainable recurrent expenditure. Volatility of revenues promotes unstable investment patterns making the entire economy more prone to business cycles. Finally, corruption and poor governance have been closely associated with oil-exporting developing countries. 43. To avoid the past errors of other oil exporters and to ensure that offshore wealth is wisely invested for the benefit of current as well as future generations, a strategy on the sustainable use of this wealth needs to be developed before revenues begin to exc,eed desired spending levels. Two key elements of this strategy are: (i) devising guidelines for saving versus spending the revenue windfall-to permit smoothing of government expenditure to extend beyond the windfall period; and (ii) ensuring that investment of the financial savings accumulated is managed in a transparent manner and entrusted to top- rated international fund managers who can maximize the financial returns to East Timor consistent with well defined investment objectives. 44. There are no unique rules for smoothing and a precise government spending strategy cannot be finalized before the status of commercial agreements is clear. Moreover, assumptions regarding prices, production and rates of return on accumulated savings will need to be updated on a regular basis. Nevertheless, it is important to develop the capacity to rigorously assess the implications of alternative scenarios, to accurately inform the public debate and to develop reliable parameters for fiscal policy. 45. Allocating a constant nominal stream in perpetuity to government expenditure once fiscal surpluses are attained is a useful starting point in the analysis of utilizing the revenue windfall. Depending on assessnleilts of absorptive capacity, investment needs, prospects for the development of the non-oil economy and the extent of availability of concessional financing, the authorities may wish to either increase or decrease the pace at which the oil revenue windfall is utilized. Whatever strategy is chosen, a rules-based approach to smoothing government spending call serve as an essential tool to avoid the pitfalls from windfall revenues that so inany other countries have succumbed to. It will also be importailt to make available accurate information about revenue flows to foster a transparent savings fonnulation process, and to minimize the scope for diverting resources froin public priorities. Maintaining Prudent Fiscal Policy 46. In the medium term, East Timor's public finances will be heavily shaped by the extent of Timor Sea revenues, which, while subject to considerable uncertainty, will be by far the largest revenue source. But until more significant oil and gas revenue comes on stream, the nation will remain heavily dependent on donor financing: in the current fiscal year, for example, projected revenue of $31 millioil (including offshore revenue) comprises less than 60 percent of Consolidated Fund expenditure and only 14 percent of overall expenditure including spending financed by multilateral and bilateral agencies. 47. Developing robust, non-corrupt administrative systems for collecting taxes is an importailt discipline from the outset. At the same time, the temptation to raise taxes to reduce a massive budget deficit in the short term will have to be tempered by the poverty and lack of purchasing power of the bulk of the population, the paucity of private enterprise, the scarcity of skilled administration officials (and hence their high opportunity cost), and the high cost (relative to collection proceeds) of maintaining an effective tax administration. In practice this may call for a regime of modest taxation of productive activity, at least during an initial phase as an incentive for the establishment of such activity. By contrast, additional effort is needed to address the problem of poor collection of user fees and charges-not only to reduce subsidies for utilities, but also to promote econoinic use of scarce resources and avoid waste. And administrative vigilance will need to be maintained in the collection of import tariffs, which account for more than half of onshore revenue. 48. East Timor will continue to require significant levels of funding from donors to finance its budget deficits in FY03-FY05. -1n large measure, this reflects the recurrent cost repercussions of investment projects financed by the donors themselves, the cost of a substantially pared-down civil service and the need to maintain essential capital expenditure. With little productive activity and income to tax in the short term, no means of deficit financing domestically (through money or debt finance) and unlikely access to external commercial finance, there is little alternative but for the donor community to continue its generous level of financial and technical assistance. 49. Continued donor support for the next three years is made more palatable by the prospect of a rapidly diminishing need for such support thereafter-when donor assistance can be limited primarily to technical assistailce and possibly some concessional lending. While considerable uncertainty remains regarding offshore oil and gas revenue, current estimates suggest that East Tiinor will be able to meet its public expenditure requirements after FY05 from its own resources, primarily offshore revenue flows, and fiscal surpluses would grow significantly by the end of the decade. Effective management and oversight over public expenditures will need to be maintained, however, particularly if reliance on debt-creating flows is to be minimized as intended by the authorities. Ensuring Effective Public Expenditure Management 50. ETTA has put in place the key eleine~ltsof a transparent, perfonllance oriented public expenditure management system applying international best practice. The principal features of the present system are: an annual budget process that generates forward estimates of expenditures at agency level, linked to performance targets, and reconciles these proposals with hard aggregate resource constraints; a centralized payment system that requires Treasury authorization of most financial commitments and executes most payments from the Consolidated Fund on the agencies' behalf; and a centralized recruitment process for establishmeilt posts and payroll management system that relates staffing to budget limits. This system has allowed rigorous expenditure control and timely reporting on core government expenditures. Final accounts on the 2000101 Consolidated Budget have been audited by an international finn and certified without qualification. 51. Six challenges are identified for the future to a system that is naturally at an early stage of development: Developmerrt of a medium-term franrework for expenditure marragenzent, corrsisterzt witlz nzacroecorzomic stability, tlrat car1 allocate resources in line witlz priorities arrd deliver a predictable flow of resources to agencies. This requires a formal expenditure policy review as part of the budget process, supported by periodic fundamental reviews of public expenditure policy and its impact. To support this policy process, further work is needed in determining appropriate service delivery standards and analysis of unit costs. Strengtlzerrirrg tlze perfornrarrce arzd service delivery orierrtatiorz of marragenrerrt systenrs. Budget regulations should ensure that line and district managers are involved in the formulation of agency budgets. Ideally, budgets would specify allocations at the district level, so as to track and facilitate the transfer of resources towards the field. A strong case may also be made for the institutionalization of a system of community grants, to ensure the longer-term sustainability of community empowerment initiatives. Mairrterrarzce of experrditure corztrols, to erzsure tlrat resources are applied in line with policy directives arrd comply witlr experrditure lintits. This can best be achieved by retaining the centralized expenditure execution and civil service recruitment system. Pressures are likely to build for a more decentralized-but potentially weaker-management system, particularly if the performance of the centralized systems deteriorate. Creatiorr of goverrrarrce structures tlrnt guarantee trarrspnrerrcy arrd hold tlre executive to nccourrtfor itsperforr~rarzce.An independent auditor, reporting to the legislature, is crucial to ensure adequate oversight over the executive. Ideally, expenditure reporting mechanisms should be complemented by reporting on performance against agreed service delivery targets at the agency level. xii e Forl~lzilatio~zof realistic, resource co~zstrai~ledplalls to guide decisiolz-makilzg throtigkotit goverlznzelzt. East Tiinor is in the process of preparing a National Development Plan, supported by wide-ranging consultations with civil society. An important element of this plan will be a visioning exercise that will identify priorities, long-term developmental goals and strategies. An appropriate institutional and procedural framework is required to ensure that this process supports the annual operational planning and the budget process. Developt~zelzt of alz external assista~zcenla~zage~rzelztsystenl that will align external resources with Gover~zme~zt's developnzelztal priorities and facilitate soulzd expenditure lnanagemelzt. Existing systems provide ample opportunity for donor-government consultation but do not include mechanisms that enable the government to manage external financing effectively. Coordination could be improved by the formulation of an external assistance strategy, laying out the broad principles for aid delivery. Basic systems for tracking disbursements against commitments should also be put in place, together with controls to ensure that agencies' register all external assistance on budget. 52. In the immediate post-Independence period, the maintenance and efficient operation of existing, Treasury expenditure control mechanisms should be considered a priority. This will require both political commitment to budgetary discipline and considerable investment in experienced expatriate personnel and capacity building. Where countries have tried to delegate responsibility for expenditure control to agencies, discipline has often weakened, leading to excess expenditures, arrears and un- programmed deficits. 53. It should be stressed that these expenditure control mechanisms are not incompatible with deconcentration and decentralization within the public sector. Close coordination between territorial and sectoral planning, the involvement of district managers in agency budget formulation and the identification of district allocations within agency budgets, can strengthen the role of local administrations in budgetary decision making. A system of block grants direct to communities, such as those implemented through the Community Einpowerment Projects (CEP), can provide an importailt channel for financing local initiatives. Building an Effective Civil Service 54. An honest, efficient and service-oriented public administration will be essential to maintain confidence in the government and the goodwill of the donor community. Key issues in this context relate to the size, composition, compensation and training needs of the civil service that will support the objective of sound administration while remaining consistent with fiscal constraints. 55. Much progress has in fact occurred in building the civil service and restoring its functions following the departure of its upper echelon in 1999. The authorities intend to mailltail1 the size of the civil service to less than half its pre-referendum level, and staffing is also modest relative to comparator countries. But the civil service is afflicted by two major problems: significant shortfalls at higher grades (particularly Levels 5 and 6); and the reci-uitillent of middle-to-higher level staff, apai-t fi-om teachers and nurses, whose qualifications and experience are lower than required. These problems warrant decompression of salaries of Levels 5 through 7 and Level 4 excludiilg education and medical personnel relative to a shadow wage for lowest skilled employees, and mapping staff at the decompressed salary levels illto appropriate salary scales. Any decompression would, however, need to remain consistent with fiscal constraints. To the extent that recruitment problems reflect the shortage of qualified staff, expatriate staff with the requisite skills will need to be retained, possibly for an extended period. 56. An important parallel response to the problem of qualificatioils and experience is the availability of relevant training. The UNTAET Civil Service Academy provides a good basis for developing appropriate training programs, which could, however, be enhanced following a comprehensive training needs analysis. Promoting Effective Governance and Reducing Corruption Opportunities 57. Fostering a culture of accountability and results orientation within the civil service and judicial and bureaucratic efficiency are important objectives that can be enhanced through the active involveinent of civil society. The promotion of transparent and accountable structures that avoid the development of corrupt practices hence merit careful consideration. 58. Administrative priorities include: Enhance citizen monitoring and complaint mechanisms for monitoring of and reporting on government performance such as in the delivery of public services. Build citizen input into the public service commission to improve transparency and integrity in public service recruitment. Enhance administrative transparency through requiring publication in the Official Gazette as a precondition before a regulation or directive can coine into legal effect. Review procuremeilt rules and practices to ensure strong compliance and enforcement. Community Empowerment as a Means to Enhance Service Delivery 59. A potentially importailt decision with repercussions for both governance and service delivery relates to the extent of administrative decentralization. There is no off- the-shelf model of local government that can be recommended; rather, the solution most appropriate to East Timor should emerge from an assessment of cost, effectiveness and capacity to provide public services at the central versus local levels. The issue of local government is one of the themes for discussion in the context of a consultative process that will provide input for the nation's first National Developnlent Plan. A number of studies are undeiway and should provide guidance on the extent of appropriate decentralization. Given there may be significant budgetary repercussions, decisions should not be rushed. It is preferable to start with a simple structure and expand it xiv gradually if needed, as it is easier to create new civil service positions later than to reduce them if an initial structure proves to be unllecessarily burdensome. 60. To the extent that strong local government institutions become a reality, it may be possible to rethink the structures of public service fonnulation and provision in ways that enhance their quality and cost-effectiveness. The initial experience with the Coinmuility Empowerment Projects suggests that empowering local cornn~unities-the end-users of the services concerned-can achieve significant benefits in teims of governance and cost- efficiency of service delivery. If the debate on decentralization were to point to the establishment of representative community councils with decision-making responsibilities, for example, it would be possible to focus the role of the central ministries to setting standards and overall policy, rather than service delivery. Local community involvement could also mobilize revenues that are harder to collect by central government-including community contributions in kind. EAST TIMOR: POLICY CHALLENGES FOR A NEW NATION Introduction East Timor is soon to become the first new state of the 21'' century. On August 30,1999, with a tunlout of 98 percent, 78.5 percent of the registered electorate voted for a transition to independence, after 24 years of occupation by Indonesia. Inlplementation of the results of the ballot should have been ail occasion for celebration. Instead, the people of East Timor were forced to watch the birth of their new country through a haze of smoke and tears, as the announcement of the ballot results was followed by orchestrated violence that wreaked physical destruction and human terror throughout the territory. It is estimated that over 75 percent of the population was displaced in the weeks following the ballot results, and almost 70 percent of physical infrastructure destroyed or rendered inoperable. On August 30,2001, exactly two years after the independence ballot, voting took place to elect a constituent assembly entrusted with drafting the country's constitution and possibly becoming its first parliament. The UN transitional authority, UNTAET, had gradually transferred govemnent functions to Timorese nationals during the run-up to the elections and, although still formally responsible for the administration of the tenitory until the expected election of a president, had nominated a Cabinet on the basis of the results of the election of members for the Constituent Assembly. Compared with other situations of post-conflict resolution, the transition to political independence in East Timor has been remarkably swift and problem-free. The transition to economic independence will take much longer. East Timor is one of the poorest areas in Southeast Asia, with severe problems of illiteracy, malnutrition, malaria and tuberculosis. According to official statistics, life expectancy in 1996 averaged 52 years, 10 years less than the Indonesian national average; infant mortality rates were amongst the highest in the world (135/1,000 in 1995). Poverty incidence in 2001 is estimated at 40 percent. East TimorYsinlpoverishrnent is particularly acute in the area of technical and administrative capabilities. During four-and-a-half centuries of colonization by Portugal and two-and-a-half decades of occupation by Indonesia, foreigners filled almost all positions of administrative responsibility or of economic opportunity. The violent transitions of 1975 and 1999 were accompanied by the departure of most skilled personnel and many traders, leaving the new ilation with a severe shortage of persons with the skills needed for nation building. In July 2001, East Timor negotiated an arrangement with Australia regarding the division of the oil and gas resources of the Timor Sea, and it has subsequently negotiated a tentative agreement on a new tax and investment regime to develop a major new field. Assumiilg these agreements are ratified, East Timor is expected to receive a significant revenue windfall over 20 years beginning in the middle of the decade. The challenge facing East Timor is how to reconcile this simultaneous existence of acute poverty and grave shortage of human lnanagelnent skills together with solid prospects of large annual cash flows from the country's natural resource wealth. The new govenme~lt will need to devise aild implelllent policies that effectively combat the new nation's widespread poverty and relative backwardness in telms of low levels of education and productivity. At the same time, institutional safeguards need to be put in place to ensure that the oil and gas wealth is not lost to mismanagement and corruption. Policies to meet these two priorities are the focus of this report. 1. THE TROUBLED PATH TO INDEPENDENCE Poverty and Slow Growth Until the lgthCentury 1.1 The eastenl Lesser Sunda Islands, including Timor, were always the poorest regions of the Indonesian archipelago. Timor's poverty was due to poor soil (volcanic soils in Java and Bali were much more fertile) and irregular weather (as the monsoons are affected by the proximity to the Australian desert). Its only valuable product was sandalwood, but the Dutch finally occupied the main port, Kupang, and the regions where sandalwood was plentiful, in the western part of the island. 1.2 For ceilturies East Timor was governed by the Dominican friars, wanior- missionaries that guaranteed its defense and formal Portuguese sovereignty. The main exports were sandalwood (increasingly scarce), wax and slaves (to Dutch plantations in Java). During the Napoleonic campaigns, the British occupied the Dutch East Indies in order to prevent a French intervention, but left aside the Portuguese possessions in the region. A Coffee Economy 1.3 At the end of the lgth century, in the process of imperialist competition among European nations, Portugal reconsidered the value of its remaining overseas territories, and made significant efforts to "pacify" East Timor and promote exportable crops. Using fiscal incentives and promoting the creation of commercial firms, some large plantations were established by the end of the century, expanding commercial and smallholders' production of coffee, a crop that rapidly became (and remains) the main export product2. Indigenous rural populations were at times required to pay taxes in kind (mainly coffee). The tax increased coffee production, and exports rose in this way. 1.4 In 1896, Timor began reporting directly to Lisbon (instead of Goa or Macao). After military campaigns of "pacification", large patches of land were converted into commercial plantations. Public finance in Timor was always in deficit and the annual subsidy sent from Lisbon, sometimes supplemented by subsidies from Goa or Macao, was very important to the financial equilibrium of the colony, as well as for offsetting the trade deficit. 1.5 Until the end of the colonial period, the admillistration of the territory was achieved through traditional authoiities, entrusted with keeping public order, promoting ' Much of this section is taken from Braga de Macedo, Braz and Sousa Monteiro, National development and economic transition under ilzternational governance: the case of East Tintor, presented at the Washington Workshop on East Timor, August 3-6,2000. The share of coffee in total exports was between 58 percent and 82 percent according to Fortuna (1971, p.2 10). public works and collecting taxes. But the history of Portuguese Tiinor is also marked by recurrent tribal wars and revolts against the Dili govenunent. 1.6 During the 20"' centuiy, the economy endured several iilsurrections (and a Japanese invasion). Meager subsidies fi-om Lisbon meant that development plans were never fully implemented. In the late 1960s and early 1970s, public expenditure grew, allowing basic education to be generalized to the whole territory and the first tarmac roads to be built3. Mineral resource exploration was promoted and rural exteilsion services implemented. 1.7 Following its 1974 revolution, Portugal abandoned its colonial empire and hurriedly gave independence to its overseas territories. In its poorest possession, the eastein half of Timor, a decolonization process was started, in a difficult coiltext of low levels of income and schooliilg and a generalized scarcity of infrastructure. 1.8 Two principal political parties emerged in East Timor, the Timorese Democratic Union (UDT) and the Tiinorese Social Democratic Association (ASDT), in addition to a smaller party with strong links to Indonesia, the Timorese Popular Democratic Association (APODETI). On August 10, 1975, UDT launched a coup in Dili and Portugal confined its troops to the barracks. Fighting broke out between UDT and ASDT, and the latter transformed into the Revolutionary Front for an Independent East Timor (Fretilin). Fretilin controlled most of the territory by late September, but incursions over the border with Indonesia continued during October and November. On November 28, 1975, Fretilin proclaimed the independence of the Democratic Republic of East Timor; on December 7, Indonesia invaded East Timor. Indonesia and Timor 1.9 After the Portuguese departure in 1975, the legal status of East Timor remained different from the defacto situation of Indonesian occupation and annexation. Although Iildonesia considered East Timor its ~ 7 state, this was never accepted by the United ' ~ Nations, where Portugal remained the "administrative authority". Internationally, the Indonesian annexation was ratified only by Australia. 1.10 The Timorese resistance continued active, despite the capture of successive leaders and amid international indifference for most of the next two decades. In the 24 years of Indonesian rule and resistance, nearly a quarter of the Timorese population are thought to have died from the fighting and forced migration, and associated malnutrition and other public health failures. In the 1990s, with the Santa Cruz massacre and the award of the Nobel Peace Prize to two East Timorese leaders, international pressure intensified on Indonesia to allow East Timor to return to self-determination. This process was facilitated by the internal fragmentation of the Suharto regime in Indonesia and by 5 Fortuna (1971) provides extensive data on colonial Timor, iilcluding a physical production index obtained by an average of production data for coffee, copra, rubber, wax, leather, soap, cattle, rice, corn, sweet potato, beans and manioc. This simplistic index, with base 1950=100, registered a minimum of 85 in 1952, a maximum of 250 in 1958 and values around 130 in the early 1960s. He also presents a growth index made up of the same products in 1950 prices, plus trade, govenlnlent revenue, construction, money stock and automobiles in circulation. The index peaks at 506 in 1959, and drops to 284 in 1962. the consolidation of the various independence parties in East Tinlor under the umbrella of the CNRT-the National Council of the Timorese Resistance. The UN Transitional Administration 1.11 In May 1999, agreement was reached at the United Nations between Portugal and Indonesia with a view to the holding of a referendum that would measure Timorese support for an Indonesian proposal for greater autonomy within the Indonesian nation. The referendum was held in August and an overwhelming majority of East Timorese voted against "autonomy" and for independence. 1.12 The announceinent of the results was followed by widespread looting, intimidation and killings, leading the international community to organize an emergency military intervention under the aegis of the United Nations to insure the replacement of Indonesian forces by an international military force. This was followed by the establishment of a United Nations transitional administration for East Timor (UNTAET) in November 1999, which was given wide powers to act as the legal state during the transition to independence and to create national institutions. 1.13 It soon became clear that East Timor was different fiom other post-conflict situations. The Joint Assessment Mission (JAM) of November 1999 stated, in its Macroeconomics Background Paper (World Bank 1999a, page 2): "East Timor is different fiom other post-conflict situations in one very important aspect. There is no apparent need for pacification between different ethnic, cultural or religious segments of the population. The need-and it is enormous because the destruction was premeditated, thorough and massive-is for reconstruction. In a way, it would be more useful to treat the problem as a post-natural-disaster situation, where a vicious hurricane destroyed all buildings and most crops and removed all records and institutional memory". The final JAM report (World Bank 1999b, page 1) also stated: "Security is a critical pre-requisite for reconstruction. However, East Timor is fortunate in having avoided a situation of full-blown internal conflict, and security operations should be kept to a level appropriate to maintaining law and order, whilst minimizing the economic and social distortions consequent on a large scale military presence". 1.14 The first meeting of donors for East Timor was held in Tokyo in December 1999 and the financing requirements for humanitarian assistance and the initial phase of ' reconstruction were amply covered by donor commitments. A UN Trust Fund was set up to cover most functions typical of a government administration (capital and recurrent costs and capacity building), a separate Trust Fund, administered by the World Bank and the Asian Development Bank, provided financing for reconstruction and developmeilt projects, while UNTAET's administrative costs were covered by an assessed contribution budget. Initially, the Transitional Authority set about recruiting the personnel needed to fill the positions in administration, practically all expatriates in view of the absence of Timorese with the required skills. 1.15 Simultaneously, UNTAET established a Timorese counterpart structure, the National Consultative Council (NCC), with representation of the different political forces in the territory. The CNRT, an unlbrella inoveineilt of pro-independence parties that received 79 percent of the votes cast in the ballot, was allotted 7 of the 14 positions on the NCC. During the early months of the UNTAET mandate, many positions in .the transitioilal adiniilistratioil remained unfilled while recruiting efforts continued, progress in solving immediate problems of the population such as housing and school rebuilding was very slow and there was increasing popular dissatisfaction with the lack of local influence in decision-making. 111 mid-2000, UNTAET responded by replacing the NCC with a larger and more formally organized National Council that began to function as a precursor of a national parliament. At the same time, half of the ministerial positions were filled by Tiinorese nationals and a stronger effort was made to recruit locals for positions in the administration. 1.16 As the year progressed, Timorese pressure for more rapid self-rule combined with the UN's assessment that the threat of domestic instability was low, led to the decision to accelerate the transition process. Elections for a Constituent Assembly (CA) were set for the end of August 2001. The elections registered a high voter turnout (91.3 percent), were virtually problem-free and resulted in a majority victory for Fretilin, with 55 of the 88 seats, followed by the Democratic Party with 7 seats and the Social Democratic Party and ASDT with 6 seats each. The remaining 14 seats were distributed amongst 8 smaller parties and an independent candidate. On the basis of these results, the SRSG nominated a cabinet of 10 Timorese, with 6 from Fretilin and 4 independents. The CA has recently finalized a constitution following a period of consultation, and a presidential election has been set for April 14, 2002. East Timor has scheduled its full independence for May 20, 2002, upon which the UN's role will become one of assistance with peacekeeping and development activities. 1.17 The transition from devastation to an elected Constituent Assembly in two years, without major internal conflict, can be considered exemplary as an exercise in political management. From an economic perspective, however, the difficulties that were apparent during the transition, especially in terms of recruiting Timorese candidates with the technical and administrative skills required to run a government, remain an important hurdle to be overcome by the new nation. The transition process itself placed heavy demand on limited Timorese human resources, promoting migration to the larger towns and creating many jobs that may disappear with the UN expatriates. The resultant boost to wages, coupled with low skill levels leaves the new nation's potential output uncompetitive and the authorities with little margin for adjustment. 2. ECONOMIC AND POVERTY DEVELOPMENTS AND ISSUES ECONOMIC REPERCUSSIONSHE BALLOTAND TRANSITION O F T PERIOD Output Decline in 1999 2.1 East Timor was and remains a primarily agricultural society, with about three- quarters of its people deriving their primary means of livelihood from the land. Prior to 1999, about a third of GDP derived from agriculture. Government services accounted for another fifth of GDP reflecting the significant provincial administrative apparatus under the Indonesian system. Construction was the next largest category, although a significant part of this was associated with public sector capital projects. Activities in the modem industrial sectors including manufacturing, utilities, banking and finance accounted for less than 8 percent of non-oil GDP. Table 2.1: GDP of 2000 by Industrial Origin at Current Market Prices 2000 Value Share in Share in Added GDP non-oil GDP"' excl. Y Oil & oil GDP $ million % YO UNTAET % Agriculture, Forestry & Fishery 83.3 21.2 25.9 35.0 32.2 Mining & Quarrying 74.5 19.0 1.O 1.3 0.9 Manufacturing 8.7 2.2 2.7 3.7 3.0 Electricity, Gas and Water 2.6 0.7 0.8 1.1 0.7 Construction 45.9 11.7 14.3 19.3 18.5 Trade, Hotels and Restaurants 25.1 6.4 7.8 10.6 9.3 Transport and Communication 23.2 5.9 7.2 9.8 9.7 Finance and Business Services 21.1 5.4 6.6 8.9 3.9 Services (Govt. & Private) 108.2 27.6 33.7 10.4 21.7 a. Government 106.3 27.1 33.1 9.6 20.7 UNTAET 83.4 21.3 26.0 Other ~ o v t . ~ 22.9 5.8 7.1 9.6 b. Private 1.9 0.5 0.6 0.8 1.1 TOTAL 392.6 100.0 100.0 100.0 100.0 - a/ Non-oil GDP is estimated at $321.2 million; GDP excluding oil and 1 NTAET is estimated at $309. million. bl Includes ETTA, TFET, bilateral and international NGOs. Source: Ministry of Finance 2.2 Recently constructed estimates of GDP and its components measured in U.S. dollars for 2000 indicate a sharp increase to 19 percent of GDP in the mining and quarrying category share due to increased oil production in the Timor Sea. The other notable increase relative to the 1990s is for government services, and reflects the relatively high pay scales of the expatriate UNTAET staff.4 The increase in these two categories tends to lower the percentage contributioil to GDP of most other sectors relative to the earlier period. The contribution of these other sectors progressively rises if The estimates in Table 2.1 include living allowances/per diems paid to expatriate UNTAET staff, though not their full salaries. 7 value added shares are estimated, first excluding oil and then excluding both oil and UIVTAET (Table 2.1). However measured, value added shares for the various sectors are destined to continue to change, as the colltributioil of oil rises and expatriate UNTAET staff depart. 2.3 Relative to Indonesia's economy, the local economy in East Tiinor was less affected by the Asian financial crisis: its real GDP expanded by about 4 percent in 1997 and the decline in 1998 was limited to about 2 percent-versus a 13 percent decline in Indonesian GDP in 1998. The slowdown continued in the first half of 1999, exacerbated by the uncertainty regarding the tenitory's future, particularly after the early May accord at the UN to hold a ballot on the choice between autonomy and independence at the end of August. Many Indonesian nationals (the higher income earners) left East Timor in advance of the ballot. Assessing the economic consequences of the ballot and the subsequent devastation, including the forced displacement of about one-quarter of the territory's population, the Joint Assessment Mission (JAM) in November concluded that productive activity had virtually ceased in the final four months of 1999. Adding to this the significant slowdown prior to the ballot, the overall decline in GDP in 1999 was estimated at 38 percent. A subsequent JMF estimate, based on a fuller breakdown of GDP components, placed the decline at 38.5 percent in 1999, with agricultural output falling by 48.4 percent.5 2.4 More recent information indicates that these estimates probably overstated the decline in economic activity in 1999. Estimates by the FAOIWFP in April 2000 place the 1999 decline in farm food crop (mainly rice and maize) production at about 35 percent. For nonfood crops (mainly coffee) the original estimates of significant output decline are not confirmed by a recent study of a representative sample of about 700 coffee farmers, which showed an 18 percent decline in overall volume produced in 1999. The estimates for declines in utilities and services (50 percent and 54 percent, respectively) in 1999 also appear to be overestimates, even if zero output is assumed during the last four months. Taking all these indicators into account, and pending a more thorough statistical exercise, it would seem reasonable to conclude that the real GDP decline in 1999, though very substantial, was less than 38 percent. The brunt of the slowdown fell on domestic savings, which fell from 26 percent of GDP in 1998 to 5 percent in 1999, and on investment, which fell to little more than half the 1998 level. Strong Recovery in 2000 and 2001 2.5 In the absence of a reliable time series of national accounts, real GDP is estimated to have grown by about 15 percent in 2000 and accelerated to 18 percent in 2001 (Table 2.2).6 The strong pace of growth in 2000 and 2001 was primarily driven by a jump in consumption to well above the level of GDP, mirrored by a commensurate rise in external savings and decline in governnlent savings. h~vestmentby the public sector also increased in 2000-01 as a result of increased spending froin donor-financed projects. See Table A2 of Enst Tintor: Eslnblishirtg he Fourtdn~io,~~ofSound Mncroeconor~~icMnnngerne~tt.IMF, 2000 I M F estimates are utilized in this section. Table 2.2. East Timor: Selected Economic Indicators Estimates Proj. (In n~illionsof U.S. dollars) GNP (at current prices) GDP Oil income (Annual percentage change) GNP growth (at constant prices) GDP Oil income Inflation (end of period) 11 80 140 3 0 -2 (In percent of GDP) Investment-saving balance Gross investment Non-goven~ment Government Gross domestic savings Non-government Government External savings Money, credits, and interest rates Broad money Currency Demand deposits 21 Saving and time deposits 21 Net domestic assets 21 Interest rates (percent per amum) Onlending 31 Consumer loans Business loans (In millions of U.S. dollars) External sector Current account excl. official transfers -121 -92 -254 -290 -235 Current account incl. official transfers -2 1 6 53 -1 1 -70 Trade balance Merchandise exports Merchandise imports Services (net) Incomes (net) Current transfers (net) Capital and financial account (net) Overall balance Gross foreign assets (end-period) Olw: Timor Sea Account Sources: Data provided by the UNTAET and the Indonesian autl~orities;and IMF staff estimates. I/ Rupiah-based CPI for Dili through 2000 and, thereafter, U.S. dollar-based CPI for Dili. 21 Figures for 2002 refer to end- March (partial estimates). 31 Refers to the rate of interest on loans extended under the Small Enterprise Project fi~ndedby the TFET. 2.6 As the couiltryside stabilized and weather reinailled relatively favorable, agricultural output rebounded. Food crop production (mainly maize and rice) in 2000 rose by some 15 percent, and production of most of the inajor crops recovered hrther in 2001. Coffee production in 2000 rose by some 40 percent, but, in tenl~sof income, most of the increase was lost to coffee price declines, which coiltiilued tluough 2001. In 2000, reconstructioil activity was initially directed at road repairs and school rebuilding. The large foreign presence led to a rapid resumption of commerce and services such as restaurants and hotels and the rebuilding of residential and commercial properties increased activity in the constiuction sector, especially in 2001. The number of registered finns also rose by 13 percent in 2001. 2.7 The 1999 disruptions produced shortages that led to a surge in prices, with inflation, in terms of Indonesian rupiah, reaching an estimated 140 percent for that year. But inflation subsided rapidly since mid-2000. And in 2001, prices, now measured in tenns of US dollars, were flat for the full year and declining by year end, reflecting increased food availability and perhaps some adjustment of prices and wages to a falling international presence. Measured inflation could increase, however, if the rupiah were to appreciate significantly against the US dollar, as virtually all non-expatriate consumption items are imported from Indonesia and sold in rupiah. 2.8 The balance of payments since 2000 has been heavily influenced by the grant- financed reconstruction effort, implying large differences in the measured current account depending on whether official transfers are excluded or included. Imports rose sharply during 2000 and 2001 owing to the increased demand from reconstruction projects and consun~ptionby the expatriate community. Exports (which have been reclassified as of 2000 to exclude unrecorded cross-border trade) are estimated to have fallen in 2001 due to falling coffee prices which reduced coffee exports by $1!h million to $4 million. 2.9 The rapid growth of the past two years, is, however, not expected to be sustained in 2002 and 2003; indeed, staff working on the National Development Plan envisage zero or negative growth during this period followed by a more moderate and sustainable pace than witnessed recently. Several factors are expected to coiltribute to slower (or negative) growth in the next two years: the partial withdrawal of the international community with the associated decline in demand and repercussions for the services that primarily cater to them; and reduced spending through the assessed contribution budget, which could lead to a decline in income of about 5 percent of non-oil GDP in FY2002- 03. Rapid agricultural growth in 2002 can also not be counted on given that crop production appears to have recovered strongly in 2000-01 aided by favorable harvests. Balancing these factors, rural areas are expected to benefit froin the coming on stream of expanded employment from development and reconstruction projects, capital expenditures under TFET are expected to be sustained through the next year, and the activities associated with independence celebrations can also be expected to boost demand. 2.10 The violeilce that devastated East Tiinor in 1999 has led to profouild changes in the lives of people. The impact of the destruction and violence after the August 30, 1999 referendum could only be properly assessed with new sources of infornlation. As a building block for the development of the poverty reduction strategy, a joint partnership of ETTA, the ADB, JICA, UNDP and World Bank iinplemented the 2001 poverty assessment. The activities included: (i) Szico Survey. A survey conducted in all 498 sucos7 in East Timor and completed between February and April 2001. It provides an inventory of existing social and physical infrastructure, of the econonlic characteristics of each suco, and of the coverage of public services. Data were collected on: population and number of families; access to infrastructure (electricity, water supply, irrigation, roads and markets); availability of education and health services; humanitarian and reconstruction prograins operating in each suco; features of the local economy and prices of key commodities. (ii) Houselzold Survey. A multi-purpose Timorese Living Standard Measurement Survey (TLSMS). Data collection was completed in December 2001. A sample of 1,800 households fiom 100 sucos was administered a conlprehensive household questionnaire covering household demographics, housing and assets, household expenditures, agriculture and the labor market, basic health and education, subjective perceptions of poverty and incomes and social capital. (iii) Participatory Poverty Assessme~zt. An in-depth qualitative assessineilt to complement data collected from the quantitative surveys. This assessment was undertaken in November and December 2001 in 24 sucos, selected from the 100 sucos covered in the household survey and covering a sample of two aldeias in each district. It helped local people to compile information on the causes of poverty and the potential of the people to help themselves, with modest assistance, from outside. 7The total number of sucos is based on interviews with suco chiefs all over the country. According to Internal Administration there are 446 sucos in the country. How the 446 sucos relate to the 498 sucos is a matter which still needs to be resolved. Main results from the Suco and Household Surveys Poverty Profile 2.11 According to preliiniilary findings of the 2001 TLSMS, the incidence of poverty8in East Timor, as measured by the head count index (or the percentage of the East Timorese population that is poor), is 40 amountiilg to 330,000 individuals (Table 2.3). It is difficult to assess the trends in poverty since 1999, as the surveys and methodologies to calculate the poverty rates are not strictly comparable. However, it is clear that poverty in East Timor was widespread in 1999, and estimates indicate that the East Tinlor poverty rate was twice the national Indonesian poverty rate (Annex I).'' Poverty in urban areas is lower (25 percent) than in rural areas (44 percent). It is lowest in DiliBaucau (14 percent), and highest in the rural center and the rural west (49 percent). Since over three quarters of the population (76.5 percent) resides in rural areas, it is clear that poverty is overwhelmingly a rural phenomenon: 85 percent of the poor live in rural areas. Agriculture, not surprisingly, is the primary occupation for the population and particularly so among the poor. These findings are confim~edif one tums to other commonly used poverty measures, like the poverty gap or the severity of poverty: rural poverty is both deeper and more severe than urban poverty. Equally, the ranking across geographical dimensions is unaffected within a wide range of levels for the poverty line. 2.12 Poverty is not just related to consumption levels but also encompasses dimensions such as access to education and health. These dimensions capture capabilities that are important in their own right and in empowering individuals to take advantage of opportunities to improve their welfare. In East Timor as elsewhere, there is a strong correlation between consumption poverty and low levels of education. Poverty is also strongly correlated with household size: larger families tend to be poorer. However, this result should be viewed with caution as the analysis does not account for economies of scale. Ownership of assets, including access to land, is correlated with lower poverty. As in other East Asian countries, female headed housel~oldsare less poor than male headed households. There is no discernible relationship between poverty and age of the housel~oldhead. 8 -x[?la The most commonly used poveity measures belong to a class of measures proposed by Foster, Greer and ' Thorbeck (FGT). The FGT measures are given by: Pa = where, n=population, n q=number of poor people, z=poverty line, yi is the income of individual I. a =O gives the headcount index which is the proportion of the population for which consunlption is less than the poverty line and thus gives the proportion (or percentage) of the population that is poor. a =I gives the poverty gap index which provides a good indication of the depth of poverty and depends on how much below the poverty line the consumption of the poor is. a =2 gives the severity of poveity and the poveity gap of the poor is weighted by the poverty gap to provide an aggregate picture. A household is deemed poor if it has insufficient expenditures to afford a basic bundle of foods that gives 2,100 kcals per person per day, and an essential bundle of non-food goods. 10 The poverty rates are not strictly comparable to the 1999 SUSENAS figures for a number of reasons, including differences in survey design, the construction of the poverty measure, and the survey period (February 1999 versus August to December 2002). Table 2.3. Selected Social Indicators, (%) Popzrlatioil Share of urban population Share age group 0 - 14 Share age group 15- 64 Share age group 65 plus Poverty Headcount Poverty Gap Severity Educatioiz - Net Eizrolntent Ratio (2000/2001) Primary (7 - 12 years) Junior Secondary (13 - 15years) Senior Secondary (16 - 18years) CIiildHealth Immunization againstmeasles (1 to 2 years) 17 Immunization against DPT3 (1 to 2 years) 7 Use of bednets last night (childrenunder 5 years) 48 Source: 2001 Timorese Living Standard Measurement Survey. Food security 2.13 The suco survey asked: "Are there nzontlzs whenfanzilies in tlze sucos don't lzave enough to eat?". The results suggest food insecurity is high from November to February, with over two fifth of the sucos reporting not having enough food. The share peaks in January, when four out of five sucos claim to face food shortages. The food shortages appear linked to the harvest cycle, which shows that food shortages are greatest at the end of the rice harvest and before the maize harvest." 2.14 The household survey also asked a similar question of all households. The results corroborate the findings generated at the suco level. A large fraction of the population reports that their family did not have enough food to eat, during the months of December and January, followed by November and February. Educatioiz 2.15 The suco survey indicates that school participation is relatively high, with almost 80 percent of sucos reporting that all or almost all children aged 6-10 are in school. Approximately 900 scl~oolsare operating curreiltly and 80 percent of these are operated by the government, with the majority of the remainder operated by the church. Student- teacher ratios on average are 52 but vary significantly across districts. I IThe normal maize harvest is concentrated in the months of February, March and April. Rice is normally harvested from April to October. 2.16 Preliminary results from the ho~~seholdsurvey show that the gross piimary enrolment rate'' during the school years 2000-2001 and 2001-2002 exceeded 100 percent, compared to only 83 ercent in 1998-99 and 73 percent in 1999-2000. Equally, net prinlary enrolment rates1' increased from 65 percent in 1998-1999 to around 75 percent in 2000-01 and 2001-02. Furthermore, the drop of net enrollment rates relative to gross eilroll~nentrates indicates that there is a large proportion of children who are outside of the appropriate age brackets for priinary school. Primary enrolment rates (gross and net) for girls are higher than for boys. Enrollment rates are higher in urban areas than in rural areas. The increases in enrollment rates for girls and the poor has narrowed the gaps in school participation rates between the richest and the poorest quintiles, boys and girls and urban and rural areas. Healtlz 2.17 There are over 200 operating health facilities-this includes more than 100 cornmunity health centers and around 60 health posts. Most health care is provided by community health centers and hospitals. Access to health facilities remains a potential problem. For example, while a community health center operates in most sub-districts, none are reported in Cova Lima and there was no facility with beds reported in Manufahi. The most common means of reaching a community health center is walking and the average time to a community health center is 70 minutes. The closest hospital is in the district capital for 48 percent of sucos and in Dili for 33 percent of sucos. The main mode of transport to these locations is a minibus-the time and money costs of reaching hospitals is significant. For example, the average travel time to a district capital is 2 hours and the average one way cost is Rp. 13,000. 2.18 The preliminary results from the household survey show that immunization rates for children are very low. Only 17 percent of children of 1 to 2 years of age are vaccinated against measles, and only 7 percent received three shots of DPT. More than one fifth of the population reported a health problem in the past month, including about 2.5 percent who report having malaria. lllness disrupts the main daily activity for 60 percent of those fallen sick, with an average of 7.5 days lost. About 45 percent of the sick sought health care, with a larger fraction seeking care in urban areas (63 percent) than in rural areas (48 percent). Almost 40 percent sought health care in a community health center, and one quarter went to a public hospital. Electricity and Water 2.19 Only about one fifth of all the aldeias in East Timor are electrified, in contrast to 30 percent of aldeias prior to the violence. The TLSMS shows that about 630,000 Timorese, or three quarters of the population, live without electricity. A third of the 12The gross primary enrolment rate is calculated as the ratio of the total number of children enrolled in the primary grades (1-6) to the total number of children in the primary school age group, which is 7 to 12 years of age. Hence if there are overage or underage children ellrolled in primary school this ratio can exceed 100. 13The net primary enrolment rate is calculated as the ratio of the total number of 7-12 year old children enrolled in primary school to the total number of children in the primary school age group (7-12 years). aldeias report that their main source of drinking water is a pipe or pump in a public place and 7 percent reported water piped directly to the house. The only significant difference relative to before the violence is that around 29 percent reported public pipes in contrast to 25 percent before the violence as their main water source. The TLSMS shows that about 430,000 Timorese, or over half of the population, rely on unsafe sources for drinking water. Tlie I~tipnctof tlre Violence of 199914 2.20 Data from the suco survey allow one to assess the number of families still displaced, the number of houses made uninhabitable by the violence, and the number of livestock lost as a result of the violence. Many households lost their main econon~ic assets-housing and livestock. The evidence points to a broad pattern of destruction in the country with the areas closest to the West Timor border being more affected than those in the eastern part of East Timor. Districts with the greatest number of families still displaced are also the districts in which there are the greatest number of houses damaged (as indicated by the ratio of destroyed houses to the present number of families). These district are Bobonaro, Cova Lima, Manatuto, Oecussi and Liquica (Figure 2.1). The household survey confirms this picture. 2.21 One fifth of sucos reported that no houses were destroyed. Most of these were from the districts of Baucau, Viqueque and Manufahi. Of the houses that were damaged, 53 percent were reported as partially or fully rehabilitated. 2.22 The sucos also reported a significant loss of live~tock'~,with Oecussi, Cova Lima and Bobonaro reporting the highest losses, and Baucau, Viqueque and Manufahi reporting significantly lower losses. Subjective Weyare 2.23 The household survey includes a section on subjective measures of well-being, which complements the information on the change in living standards since Indonesian times. It asks household members over the age of 15 about their assessment of their welfare today compared to their situation before the violence in 1999. These questions shed light on how the population perceives the change they experienced since the violence in 1999. l 4The results on the suco survey are reported from "The Survey of Sucos in Timor Lorosa'e-Summary of Initial Results and Implications for Poverty Reduction", prepared by Craig Sugden, Consultant, ADB. 15 Different types of livestock were coilverted to animal units based on the standard international definition of an animal unit. That is, large livestock (i.e. buffalo, cattle and horses) assunled equal to 1 animal unit, while smaller livestock (i.e. pigs and goats) are assumed to equal half an animal unit. Chickens are not included. Figure 2.1: Indicators by District of Impact of the Violence of 1999 2.24 When asked about the change in economic living standards since before the violence in 1999 (Table 2.4), more people feel their situation has improved than deteriorated (35 percent relative 23 percent), but the bulk believe little has changed (43 percent). More striking is peoples' views about the change in their power. Before the violence in 1999, two-thirds felt they were completely without any power. Currently, only 5 percent feel this way. Overall, 85 percent believe they have more power now than before the violence in 1999. Table 2.4: Self-Assessed Change Since Before the Violence in 1999 (%) Economic Position Power Better 34.7 84.5 The Same 42.6 9.9 Worse 22.7 5.6 Source: 2001 Timorese Living Standard Measurement Survey. 2.25 The l~ousel~oldsurvey also asked which areas have iillproved or deteriorated since before the violence in 1999 (Table 2.5). The respondents feel that progress has been achieved in safety, political participation, education and status in community, while housing, demand for products, enlployment and infrastructure have worsened. This pattern is coillpatible with the evidence reported in the previous paragraph. As East Timor prepares itself for independence, its population feels more empowered compared to Indonesian times, but less secure about the progress achieved in economic well-being. Table 2.5: Perceived Changes by Sector (%) Improvement Deterioration Safety 61 22 Political Participation 33 7 Education 32 13 Health 22 20 Employment 16 35 Status in Community 12 2 Demand 8 34 Housing 6 41 Infrastructure 5 18 Land 4 9 Note: Individuals were asked to indicate up to two areas. The numbers sum over the first and second choices. Source: 2001 Timorese Living Standard Measurement Survey. 2.26 The overriding concern of East Timor's new government will be the need to devise and implement policies that effectively begin to reduce the new nation's widespread poverty and low levels of education and productivity. Subsequent chapters of this report address several of the factors that will need to be considered. 2.27 An immediate eco'nomicissue relates to the significance of the expected reduction of international staff after the transition. The economic impact of the influx of some 15,000 expatriates, which is large relative to the economy (particularly given their high purchasing power relative to the local population), has been to temporarily raise income and employment, concentrate expenditure in Dili, raise the relative price of non-tradable to tradable goods and services, and hence skew the flow of resources into activities that may not be viable after the departure of the foreigners. The repercussions could be costly, for example, in the case of uncompetitive real wage levels, excessive levels of urban concentration and migration, and misallocated investment in service industries. 2.28 The large influx of international agencies, NGOs and other foreign employers- and the shortage of skilled local staff-resulted in an estimated quadrupling of wages, measured in US dollars. In the coffee industry, for example, unskilled farm labor earned the equivalent of $0.90 per day in mid-1999 and now eanls $3.50. Civil seivice wages were set at rates starting at about $85 per month, which was about three times the average in Indonesia. In compensation, the total number of civil servants is to be held to less than half the pre-ballot figure, so the overall budgetary impact is not excessive. The more serious problem with relatively high public sector wages (at the lower grade levels) is their deinonstratioil effect for the private sector. Already there is evidence of jobs being lost to mechanization and to workers abroad (Box 2.1) and many import-substituting activities that could be viable at lower wages inay simply not be established, resulting in the loss of potential employment opportunities. Box 2.1. Declining Competitiveness in Coffee Processing The coffee sector in East Timor has traditionally been a large employer, albeit on a seasonal basis. Employment takes place both in the production areas, to transform the cherries into green beans, and in Dili, to prepare the beans for exporting, and total seasonal employment has been estimated at about 11,000 people, excluding transportation workers16. According to a major processing and exporting firm, events since the 1999 ballot have made East Timor the most expensive producer of coffee in the world. Infrastructure construction and transport are very expensive; security costs are high and labor is more expensive than in Brazil. Daily wages are now USs3.50, as opposed to 90 cents in 1999. By comparison, daily rates are about US$1.00 a day in Indonesia and even less in Vietnam. In relatively high-cost Brazil, coffee workers earn US$60 per month. In response, exporters have reacted in two ways, both resulting in a reduction of jobs in East Timor. Where possible, mechanization has been increased to reduce dependence on labor. A major processor has brought in sorting equipment and this year employed only 150 workers for sorting, as opposed to the usual 400 to 500 workers. The largest exporter (in volume terms) has started exporting parchment coffee to Indonesia for processing, instead of performing that function in East Timor. Again, the result is fewer jobs in East Timor. 2.29 The question of employment creation needs to be placed in perspective. Since the public sector will presumably refrain from expanding simply to create employment for the service workers made redundant by the impending departure of foreigners, these workers will need to find alternative employment in the private sector or develop some form of self-enlployment. Before the 1999 ballot, there was very little manufacturing activity, so most of the newly unemployed would have been employed in agriculture, lower and middle level public sectorjobs that have not been maintained or in commerce, much of which has not yet been reactivated. To the extent that new sources for jobs become available, wage levels should be allowed to adjust to the new conditions of demand and supply, and regulatory or legislative impediments to such adjustments-such as setting a minimum wage based on current artificial market conditions-should not be created. 2.30 The cost of labor is clearly not the only nor necessarily most important issue impeding the growth of private enteiprise. Even if wage levels were the lowest in the 16 Source: Pomeroy, J. "Coffee and the Economy in East Timor" presented at ADB-ETSG Econonlic Policy Fon~m,Dili, 26-28 March 2001. region, there inay not be much job-creating private investment until the longer-ten11 political situation is defined (i.e. some time after independence, when investors can gauge the new government's position on property rights, taxation policies, investment protection and labor legislation, for example-as discussed in the following chapter). Nonetheless, as the economy stabilizes, labor costs could become an importailt restraint on private investment and employment. Currently, practically all manufactured goods consumed in East Timor are imported; if effective labor costs are not brought into line with regional levels, the new nation runs the risk of perpetuating this situation. 2.31 An issue closely related to wage levels is that of labor productivity. In the view of some employers in East Timor, the problem is not only relatively high nominal wages but also the lack of worker productivity. Skill levels are low but so is worker motivation, reflected in high levels of absenteeism, lack of discipline and conflictive workedmanager relations. Some of this may be explained by the sudden increase in wages and the independence euphoria more broadly but, for the longer term, it is important that labor legislation and its application be sufficieiltly flexible in relation to worker rights so as not to repress employment. 2.32 The risk of perpetuating non-competitive labor costs is heightened by the fact that exchange rate depreciation in response to the demand shock of expatriate departure is not feasible while the US dollar remains the unit of account. On balance, the option of introducing its own currency may not be advisable for East Timor, at least until economic and financial management capacity is enhanced (and may not necessarily be advisable even with adequate management capacity). But if the dollar remains East Timor's official currency, the need for wage and price flexibility to adjust to changing demand conditions becomes all the more pressing. 2.33 Events since mid-1999 have led to significant rural-urban migration, especially to Dili, where the population has almost doubled. Some of this flow can be explained by the disruption of rural life and by the availability of jobs in Dili, but other factors contribute to migration even when prospects of employment are low. Services such as clinics and schools are more readily available in urban centers as are utilities, namely electricity. Additionally, problems associated with metering have resulted in electricity being free of charge to household users. Measures need to be taken quickly to prevent such incentives from becoming permanent features of urban bias. Such measures should include ensuring that user fees are charged for all utilities and extending the provision of services and utilities to rural areas. Ultimately, the incentive to remain in rural areas will depend on income prospects in agriculture and the quality of rural life-topics that are addressed in Chapter 4. Finally, the implementation of labor-intensive projects to fulfill objectives already prioritized through the budget process, and the availability of start-up credit for small-scale entrepreneurial activities could be timed to coincide with the departure of expatriates to lessen the expected demand shock. 2.34 The issue of job creation is linked to that of the anticipated doinination of the economy by the oil sector within a few years. For the next few years, East Timor will coiltinue to be dependent on donor funding, but oil and gas revenues is soon expected to exceed current needs. The creation of new jobs in the oil sector is not expected to be significant, as the activity is offshore and not labor-intensive, but the availability of such large fiscal flows undernlines the perceived need for other sectors to be competitive. The resultant risk is that wage levels will remain high, with little einploylnent outside the public sector, oil flows will fund the budget and domestic production will be limited to subsistence agriculture and non-tradable services. As has happened in other small oil-rich countries (Box 2.2), pressures will mount for increasing transfer payments and, in time, even the services sector will be staffed mainly by immigrants. In East Timor's case, the prospective jump in oil revenues is currently expected to be temporary, which means that after its demise the economy could be left with little in terms of a competitive foundation for agriculture, manufacturing or services, and hence with no means to support the prior over-reliance on offshore resources. Box 2.2. Brunei and the Crowding Out Effect Brunei is the third largest oil producer in Southeast Asia and the fourth largest producer of liquefied natural gas in the world. Substantial fiscal revenues from oil and gas production have fueled growth of a large public sector that provides for generous remuneration, a comprehensive welfare system, and a high level of investment in infrastructure. While the country has been able to maintain high living standard thanks to its oil windfall, the development of a private sector has been precluded by increasingly heavy government structures and high public sector remuneration. Despite government efforts to diversify the economy, non-oil related productive activities have gradually disappeared and the private sector has been reduced to the construction sector and a number of service industries, which are heavily dependant on government development projects'7. The country currently faces increasing unemployment problems despite having more than 50 percent of its work force employed by the government. The disappearance of a fiscal surplus in the 1 9 9 0 ~together ~ with the crowding out of the private sector activities has made Brunei's economy vulnerable to petroleum price fluctuations. 2.35 These risks point to the need to enhance competitiveness on a sustained basis by: maintaining wage flexibility; strengthening the legal and regulatory framework for private enterprise and tackling the major impediments to private investment; avoid letting the tax system become an impediment to such investment; and continuing the assistance program to fanners in the form of capital, knowledge and marketing expertise. At the same time, a strategy is needed to ensure that East Timor's oil and gas wealth can be effectively channeled towards poverty reduction, in particular to build the human capital base. This in turn requires: the development of sound savings and investment strategies to extend the benefits of offshore wealth beyond the windfall revenue period; the adoption of transparent public expenditure management and budgeting tools to utilize aid and revenues effectively; the development of a service-oriented civil service; the harnessing of community participation towards development objectives; and fostering a culture of accountability that minimizes the scope for corruption. The remainder of this report is devoted to the discussion of these issues. "Agriculturerepresentsapproximately 5percent ofGDPandevenbasicvegetablesarenowimported. 20 3. IMPROVING THE BUSINESS ENVIRONMENT 3.1 Business eilvironmeilt (BE) analysis traditioilally focuses on the policies that deternine how successfully the private sector can thrive and provide the growth necessary to create jobs and widely accessible income generating opportunities. This includes the legal and regulatory environment in which business operates, the physical infrastructure on which it depends for production and distribution of goods and services, the availability of business services such as accounting, finance, insurance, legal services and logistics, and the quality of human capital. This chapter focuses on the first two of these factors. Business services can be expected to develop as private activity expands. An immediate governmeilt role is not foreseen, though a regulatory environment will ultimately have to be developed. Human capital issues are addressed in Chapter 5. The endowment of natural resources also enters the picture but rather as a factor that determines growth potential, rather than something that can be directly influenced by policy. Initial Conditions 3.2 East Timor faces several challenges in building a solid BE largely because of the unique initial coilditions linked to the severing of ties with Indonesia. The Indonesian government invested fairly heavily in some forms of infrastructure. The network consisting of 1200 krn of main roads, 2000 Ism of district and 1800 km of feeder roads, is very dense for East Timor's population and per capita income. These roads have deteriorated sharply in the period since the 1999 referendum. Other infrastructure was substantially destroyed in the violence of 1999. Electrical power generation was virtually wiped out, although operating capacity has since risen to 22 MW at end 2000 as compared with a 26.5 MW prior to the referendum. It is estimated that 38 percent of the population had access to power in 2001 compared to 46 percent in 1997. Most of the telecommunications infrastructure was destroyed in 1999. Presently, Dili has a reasonable mobile phone service but services are limited and intermittent in other district capitals, and use of satellite phones is generally necessary for businesses outside Dili. 3.3 After the referendum the links to the whole legal and regulatory system of Indonesia were cut. The passage of new laws has been slow, creating risk and uncertainty for domestic and foreign investors. Resolving this problem is not simply a matter of passing replacement legislation. Even though Indonesian law has been retained where not revoked by UNTAET regulations, a whole set of locally based institutions must be established and staffed in order to give force and substance to the law. This is a costly overhead of nationhood that takes time to put in place. 3.4 Business services have suffered from the exodus of Indonesians and Chinese, the permanent closure of Indonesian business service enterprises and the destruction of records. Most financial services are now only offered through branches of foreign firms with offices in Dili only. Accounting and audit services are very limited and there are no local insurance facilities. Restoration of these services depends on re-establishment of a legal framework, traiiliilg programs to increase human capital and a demonstrated period of stability which will entice new investments. 3.5 There are hard policy choices to make in deciding how far and how fast to rebuild. The infrastnlcture needs of East Timor in the medium tern1 inust be measured against the rate of ecoiloillic growth the country is likely to be able to achieve. For instance maintenance of the extensive road network established under the Indonesian administration may not be econon~ic given the limited public resources even if the initial refurbishment is done with grants. Prior to the refereilduin East Timor received net fiscal transfers from Indonesia. But even with additional financial resources there are real resource constraints in both rehrbishment and maintenance. In the power sector for instance, all staff in middle and upper level management at the power utility were lldonesians who left after the referendum. Staff is only 70 percent of the pre-referendum level and critical technical and managerial positions remain unfilled. Capacity to operate the existing system is already strained. Financial and real constraints apply also in the public provision of business services. Land titling offices, extension services and commercial courts are critically important to business and agriculture, but they require buildings and supplies with significant recurrent cost implications for the budget, as well as demands for qualified staff that may be scarce. In short, there are costs to establishing a favorable BE. While the BE helps determine investment and growth, the targeted growth must be commensurate with the costs and supply constraints involved in creating the necessary supporting BE. 3.6 The road system affects several aspects of the business environment. First, it is critical to the distribution of agricultural goods and the export of coffee. Second, in the immediate vicinity of the capital it provides connections to the port and the airport and thus determines the attractiveness of the most probable sites for the development of light manufacturing and export- oriented industry. Third, the internal network provides distribution channels and contributes to the development of markets for local products. The internal network also allows movement of people. This influences the pattern of settlement, the time and money cost of commute to work or job site and the ease of delivering public services such as education which by extension affect the medium term quality of the labor force. The principal challenges in the transport sector relate to this third and most costly part of the system. 3.7 The extensive road network of the Indonesian administration was constructed with relatively low standards of design and material. This combined with the harsh climatic and geological conditions created a high maintenance system that deteriorated rapidly during the three years prior to the referendum when the necessary recurrent expenditures were not kept up. The ADB7s Emergency Infrastructure Rehabilitation Project (EIRP) was undertaken to make repairs to allow humanitarian aid to flow, and to provide road drainage, bridge repair, backlog maintenance, administrative improven~entsand training.' 3.8 The ongoing recurrent costs of this system will be high and are budgeted at $16.4 million in FY01102, although almual costs of some $25 million have been estimated as the cost of maintaining in good condition a network of 1200 km of main roads. ' DFID and PKF also were involved in much of the early road rehabilitation. 3.9 Establishment of an asset inanagemeilt systein is critical to ensuring that the road system is efficiently maintained. According to the ADB, current organizational analysis indicates the need for 90 persons even if substantial reliance is placed on local private coiltractors for civil works. However, the roads department faces severe shortages of qualified people and has been able to attract only four local staff. Training programs, the creation of a road administration and adoption of more flexible recruitment and contracting procedures are expected to improve the situatioil somewhat. But a supply constraint necessitating continued use of foreign staff will clearly persist. 3.10 Prior to the referendum, cargo and passenger shipping was an extension of the Indonesian coastal feeder shipping system. There were no direct international exports or imports from Dili as all cargo passed through other Indonesian ports. After the violence, cargo handling at Dili stopped until the amval of the peacekeeping force operations. The port then suffered congestion problems until US$5 million was made available under the EIRP for a third berth, slipway repair and container yard upgrading. Currently the port handles about 2,000 ton equivalent units and 8,000 toils of break bulk cargo. As was the case in Indonesian times, and as is common among smaller islands throughout the Pacific, ships rely entirely on their own equipment for cargo handling. Navigational aids are being restored with Japanese funds. Congestion is no longer a problem: queing time is less than 24 hours, waiting time at the berth is less than six hours and container handling is smooth. There has been a shortage of container storage space but this is being resolved. 3.11 The recurrent cost of port operations for FY01/02 is estimated at US$1.3 million. Within three years, the port budget is expected to be balanced, through the collection of user fees and training of local staff to replace expatriates. 3.12 For the near term, port operations at Dili do not present a serious constraint to business development. However, there is no room for expansion of the port and in the medium term plans will have to be made for further development of the facilities at Hera port 14 krn away from Dili as well as for greater container depot space. Private development of a natural port at Com may be needed to support to service the development of the Timor Sea resources, if warranted by sufficient activity. Power 3.13 As a business environment issue, the availability and cost of power are important considerations. Power outages have been frequent and investors with a need for continuity of supply would not consider locating in East Timor without installing their own capacity. Most commercial businesses of any size have their own generators as do all the hotels and most manufacturing operations. Even for those businesses which are not critically dependent on continuous supply, charges are high. Nevertheless, current policy objectives for this sector are appropriate. These include achievement of financial self sufficiency for the Power Authority through careful rate setting, customer identification efforts, increased metering and lowering the billing threshold. Staff training programs and formulation of a longer-term power sector development strategy under the new Power Authority, including alternatives to diesel generation, should contribute to cost reduction. 3.14 The main issues in the power sector are refurbishment and upgrading of installed capacity, shortages of teclmical and managerial staff and establishment of a viable financial basis for operation of tlle Power Authority. EIRP Generation is entirely diesel based and costs have risen sharply with iiltematioilal oil prices2.Sixty percent of recurrent cost is for fuel. In August 2000, new tariff rates were set at 12.3 cents per kwh. In order to facilitate cost-recovery, UNTAETIETTA has, in Dili in the first instance, worked to install electricity meters on all residential and business premises. This task is perhaps 70 percent complete. Billing became effective on August 1, 2001 with tlle first bills to users issued in October. Cost recovery before then was minimal. The new tariff will be 24.9 cents per kwlhour for businesses and $1 per month for the first 25 kw/hours, plus 24.9 cents for each additional kwlhour for residential users, and will be subject to change. This rate is high by international standards but reflects high unit costs and the nature of supply. The large and temporary international presence and the absence of billing arrangements outside Dili create distortions that will effectively cause several redistributions as follows: Consumers in Dili will be subsidizing consumers in the rest of the country, at least in the near term (until localized cost recovery arrangements are installed outside Dili.) The tariff per kwlhour is considerably higher than those imposed under the Indonesian administration (when fuel prices were heavily subsidized). Thus the new tariff has been set at the lower band of possible pricing options-given that future demand is uncertain and so that the change in electricity prices from the pre- UNTAET period is not too great. 3.15 The imposition of the tariff will bring financial relief to the UNTAETIETTA administration from the expense of providing free electricity, as has been the case since 1999. The tariff will promote a degree of conservation, and it provides electricity at a reasonable cost to low-income consumers. However, since billing began, the rate of payment of invoices has been low. 3.16 Diesel-powered generators, of varying reliability, currently service Dili and the outlying districts. There is no national electricity grid. This means that all power sources need to be localized. A long-term energy strategy is required and efforts, augmented by ADB-funded technical assistance, are underway to determine the best approach in terms of economic and environmental efficiency. 3.17 The violence in September 1999 destroyed most of the telecommunications infrastructure in East Timor. After the arrival of the Interfet, Telstra of Australia was invited by the UN to establish basic public telecoinmunications services on the basis of a revolving short-term contract. During the ongoing transition period, UNTAET had its own separate telecommuilications facilities. Presently, Dili has a reasonable public mobile phone service that also extends coverage on a limited basis to the capitals of Oecussi, Baucau, Kovalima and Bobonaro districts. Telstra has one Inteinet cafe in Dili. Data trailsmission out of Dili via certain mobile phones is possible but slow. Satellite phones are also in common use, especially by busiilesses and organizatioils operating outside Dili. 2See "Summary Sector Reviews in East Tinlor" (ADB, ETTA), for a more thorough coverage of issues in infrastructure. EIRP denotes: Emergency Infrastructure Rehabilitation Project. 24 3.18 East Timor7sDivision of Information, Teclmology, Post and Telecommunications (ITPT) was charged with preparing a telecommunicatioi~spolicy, suppol?ing regulatioils and ail operatioilal plan to establish agreed levels of telecommunications capacity and services in East Timor. A national telecoinrnunications policy was approved in May 2001 laying the basis for passage of the national telecommunications law (UNTAET Regulation 2001115) in July 2001. In turn, this law empowered ITPT to seek, tluough international coillpetitive bid, tenders for a comprehensive and integrated telecon~mui~icationsnetwork. Tllis tender process is now well underway. 3.19 Following considerable external review and internal debate, the administration determined that the best approach would be to bundle together the required system elements into one contract including establishment and maintenance of a network to deliver mobile and terrestrial telephony, an international gateway, and Internet, broadcast and private network access nation-wide. The winner will therefore get the lucrative mobile phone market but must also provide other necessary but less profitable services. Mobile and terrestrial lines will be available in all district capitals. The proposed arrangement requires universal access to the network by all East Timorese and will provide sufficient capacity to support the demands of the future East Timorese government. 3.20 The proposed network will cost an estimated $15 million to establish and the successful contractor will retain ownership of the network, ahead of transfer to the government, for a period to be negotiated (but in the order of 10 to 12 years). A general principle of cost-recovery- oriented basis has been agreed, although the administration has yet to settle on a definitive pricing policy, and this will necessarily be subject to later negotiation with the contracted provider. Expressions of interest from international telecommunications companies have been sought and received. Tenders from the pre-qualified candidates have been requested, with a view to implementation by the end of 2002. Although the current scope of the proposed network does not extend to the sub-district level, the network will be compatible with, for example, possible future radiotelephone connections into the main telecommunications network from East Timor7s482 villages. 3.21 Several Pacific and Caribbean island countries with populations even smaller than East Timor found that use of bundled 10-year exclusive monopoly contracts resulted in relatively high prices-and some are now opting out of the contracts. These circumstances may be somewhat different, however, in that these countries generally had some pre-existing systems for which selective additions and upgrading were needed, whereas East Timor must basically develop a whole system from scratch. Moreover, the proposed exclusive contract in East Timor facilitates the provision of con~munityservice obligations. Nevertheless, the experience of other countries signals the need for caution. 3.22 There is now clear recognition within government of the importance of an effective and efficient telecommunications system to the future economic development of East Timor. It is essential that the coming transition from the current UNTAET-supported telecommunications arrangements to the new privatized structure occur with minimum disruption to private and public business. It is also important that East Timor realize the potential of the proposed system for aiding broader development objectives. 3.23 According to Regulation 199911 the laws that applied in East Tinlor prior to October 25, 1999 will continue to apply. There were certain explicit exceptions to this general provision. For example laws dealing with such issues as capital punislment and national defense were revoked. But as far as the operating e~lvironmentfor business is concerned the legal framework remains that established under Indonesian law. As a stop gap measure this regulation was perhaps reasonable but operational difficulties with the arrangement have quickly emerged. The principal problem is that the Indonesian law contains explicit references to Indonesian administrative bodies that now have no authority or presence in East Timor. For instance BKPM (the Indonesian investment agency) can no longer render decisions regarding foreign investment in East Timor. The law can be amended to replace references to Indonesian institutions with East Timorese ones instead. But this assumes that such local institutions exist and have the capacity to perform the functions previously performed by the Indonesian bodies. There are a number of important gaps here. 3.24 Insome areas there was a need for new legislation, which has now largely been met. A tax code for example has been redrafted, passed and promulgated, and the implementing institutions set up and made operational. Further refinements of the law may be needed and arrangements made to ensure a smooth handoff to the new administration at the end of ETTA mandate. 3.25 In other areas, however, there is a clear need for new laws which has not been met. In part, this may be because of a perception that it was best to wait until after the elections were held and the new Constituent Assembly and the new Council of Ministers were established to give these new representative bodies the responsibility for passing such fundamental legislation. However, it is likely that for the near term the Constituent Assembly will be preoccupied with developing the constitution, and that progress will be slowed by the same problems that were present before the election, namely: New legislation is needed and drafted, but not passed, e.g. Company law, Investment code. New legislation is needed but no policy directions have been given by Cabinet, no drafts have been prepared and no work is in progress e.g. Insurance, Bankruptcy, Commercial (contract) Law. New legislation is needed, drafts are prepared but political problems are thwarting passage e.g. Land Law. Company Law 3.26 In most countries, investors wanting to establish an enterprise can do so using one of several possible legal structures: a sole proprietorship, a partnership, or a public or private incorporated company. The incorporated company structure has several advantages. It can be used to limit the liability of shareholders. It gives shareholders the right to vote and to replace directors if they are not satisfied with the management and profitability of the enterprise. It also provides the framework for creditors to claim collateral when a company fails to make pro~nised payments. These provisions encourage the investors to provide capital and creditors to provide credit. There is no effective Company Law in East Timor at this time so investors cannot establish a corporate vehicle for doing business although this would be the preferred foim for most larger enterprises and for foreign investors. 3.27 New legislation, which contains the following accepted iilternatioilal standards, has been drafted but remains under consideration: A coinpaily has unlimited liability for its debts while the shareholders' liability is limited to the notional value of their shares. A company is a separate legal entity from its shareholders and directors are agents of the company who will incur no liability provided they have not abused their power or exceeded their authority. A company continues to exist until struck off, and survives a change in ownership of shares and directors. A company can own property and assets belong to the company and not to shareholders whose financial interest is represented by the shareholding. A company can enter into contracts. The usual general meeting and reporting standards are applied. Different minimum capital requirements are specified for 100% East Timorese companies ($2,000) and non-100% East Timorese companies ($10,000). 3.28 Ln addition, two of the areas of most concern in the Indonesian law-namely the need to have the Articles of Association prepared by a notary and the 60-day review period-have been amended. There is no reference to the use of a notary and the 60-day period has been reduced to 7 days. These two changes render the regulation much more competitive without diluting its effectiveness. 3.29 There are a few possible difficulties though. The current law does not indicate penalties for violation of the law. This provides little guidance to a judiciary that is stretched by heavy caseloads and a lack of trained judges. Implementation of the law will involve certain recurrent expenditures. In particular, provision must be made to track and monitor companies' compliance with filing requirements including annual income statements and balance sheets. Good laws are of no use where institutioilal support mechanisms are absent. Company law confers benefits (such as limited liability) on firms and their shareholders. The state cannot give substance to the obligation of coinpanies to behave according to the law unless they have the means of checking, for instance, that board members are not bankrupts or that no contracts are signed by bankrupt companies. 3.30 Regulation 200014 requires every person and legal entity intending to operate a business in East Timor to register with the UNTAET Business Registration Unit. But the regulation falls far short of the needs of the Corporate Law. In many countries the Registrar of Companies operates a Register of Claims, which records the use of collateral by borrowers and links the clainls to assets listed in the company balance sheet. Without an actively updated register, collateralized lending is difficult. Claims in case of default are hard to enforce and without the central register there is no check against double pledging of assets by borrowers. The existing Business Registration Unit could be upgraded to undertake these functioils or another entity created for the purpose, but only if budget is made available and staff is trained. Registration is simple and the fees low so it presents no barrier to doing business. But as it is currently set up the Business Registration Unit provides little more than opportunity for an enterprise to register its name. Many names have been registered by visiting business persons who are simply canvassing business opportunities in East Tinlor but are unlikely to actually invest. So the registry is not useful as an indication of the true level of business activity or as a planning instnunent. Foreign Investment Law 3.31 Most developing countries have a foreign investment law (FIL) that sets down the principles according to which foreign companies and individuals can hold property and operate businesses in the country. Their rights and obligations need to be clear and fair in order to minimize the risk and uncertainty they face and to maximize inflows of foreign direct investment. Failure to do so limits growth poteiltial by reducing the country's access to financial capital as well as closing channels of technology transfer. As a result, foreign investment tends to be small and concentrated in lower value added trading and service operations where investors have relatively few assets at risk and payback cycles are short. Box 3.1. The Judiciary East Timor has a two-tier court system. There are four lower level district courts. These courts are not specialized. There is no streaming or classification of cases so each lower court hears all types of cases whether they are criminal or civil and whether they deal with torts, family law or other matters. There is a single higher court, which is the court of appeal. It consists of a three-judge panel including one East Timorese and two foreignjudges. There are 25 judges for the higher and lower court systems combined. The case load is very heavy and backlogs are long. The violence surrounding independence generated a large number of criminal cases. High priority has been attached to resolving these cases first in order to help establish peace and stability. However, it has also meant that civil cases and especially commercial disputes have been left unresolved. Criminal law is based on Indonesian law. UNTAET has made amendments to both the Criminal law and to criminal procedures in order to adapt them to Timorese customs and institutions. The Indonesian-basedcivil law and procedure remain basically un-amended. 3.32 Model legislation for FILS is readily available from a variety of sources and typically embodies some basic principals such as national treatment, and the right of repatriation of dividends and profits. Eligible sector/industries are defined, as are JV (joint venture) arrangements, taxation principles, etc. Emphasis is on making the rules simple, transparent, based on explicit criteria rather than relying on discretionary or interpretive administrative decisions, and very limited in terms of offering special tax or other incentives beyond those necessary to limit anti-export bias. 3.33 Work on a draft Foreign Investment Law for East Timor began in late 2000 and several drafts have been in circulation. These drafts have been commented on by the World Bank Group's Foreign Investment Advisory Service (FIAS) and others. Some reservations have been expressed about the use of reserved and restricted lists and about certain registration processes, but in general terms the drafts appear'to be reasonably close to best practice. The draft law, and underpinning policy, are still subject to ongoing revision and redrafting. 3.34 Given the continuing absence of a FIL, and notwithstanding a general statement of principal that foreign investillent is welcomed and encouraged in East Timor (which has generally been upheld in terms of day-to-day decisions), there is no comprehensive legal framework for FDI. To date the level of foreign investment remains small. However, the importance of this legislative gap should not be overstated. There have been only modest expressions of interest by foreign investors. The biggest investment will likely be in oil and gas, and in East Timor, as in most other countries with resource based industries, foreign investments in this sector are likely to be undertaken through special negotiated arrangen~eiltsrather than a generic foreign investmeilt law. Finally, FILSare a kind of apex legislation that bridges tax law, custoins law, conlpaily law and various other bits of legislation. To work effectively, the other parts must also be in place. Passage of a good FIL is important for the longer-term development of East Timor, but its introduction will not, of itself, trigger a large and irmnediate increase in FDI. Banking and Financial Sector Legislation 3.35 The legal framework governing banks and non-bank financial institutions is further advanced than other areas. Three key regulations were passed in 2000. Regulatioil200015 on ~icensingof Currency Exchange Bureaus Regulation 200016 on Establishment of a Central Payments Office (CPO) Regulation 200018 on Bank Licensing and Supervision 3.36 In addition, the CPO (now transformed into the Banking and Payments Authority (BPA)), has issued a series of instructions detailing acceptable banking practices regarding liquidity, capital adequacy, qualification of administrators, large credit exposures, transactions with related persons and audits. Bank reporting and balance sheet records are required to conform to International Accounting Standards (IAS). 3.37 A further piece of legislation covering non-bank, deposit-taking institutions has been drafted and is expected to become law in the near future. 3.38 This favorable legislative environment does not, however, mean that financial intermediaries are meeting the needs of the business community. A key issue is the inability of financial institutions to collateralize assets. There are several underlying problems. The first is the lack of a functioning Company Law which, as described above, means that creditors cannot make enforceable claims against corporate assets. As in other countries it is sometimes possible to use financial leases to get around this problem. Instead of borrowing to finance an asset purchase, the creditor buys the asset outright and leases it to the operating company. In the event of default there is no issue of enforcing the claim. The creditor can repossess the asset easily because he has clear title to it. Courts can be relied upon to back up the claim if necessary. This can work reasonably for moveable assets but is more cumbersome for fixed asset^.^ 3.39 A second problem is the lack of a bankruptcy law. The blanket provision of Regulation 199911 stating that the pre-1999 Indonesian law is applicable in East Timor, does not provide a workable legal framework for bankruptcy as the local institutions necessary to implement it do not exist. This frustrates loan making since there is no clear mechanism for resolving creditor claims in the case of insolvency. This is the most important block to making loans to persons and unincorporated businesses. At present no cabinet policy has been issued that would initiate work on bankruptcy legislation. 'Financing can sometimes still be arranged even for fixed assets and working capital through purchase and leaseback deals. For instance a firm wanting to buy a fixed asset may finance it by selling different movable assets (typically cars) to the bank. The bank leases the cars back to the fiml and the firm uses the car sale proceeds to buy the machine. 3.40 The third problem steins from difficulties surroundiilg land ownership. This issue is discussed in a broader coiltext below. But as far as finai~ceis concerned, the lack of clarity over owilership of a signification portion of land and real estate assets, especially in the urban areas, cuts off an important source of collateral. 3.41 In nlost countries, systeins of sharing infonnatioil on the credit history of borrowers have evolved between financial iilstitutions including banks, credit unions and leasing conlpanies. These systems compensate for the asyminetric inforn~ationbetween borrows and lenders and make overall lending more efficient by filtering out credit access by borrowers with chroilic credit problems. Credit-information sharing systerns can be run privately, by the financial institutions themselves or by specialist service providers, or by the public sector. No such arrangements exist in East Tiinor presently but one will need to be established in order to support financial sector development. 3.42 With the departure of the Indonesian banks, there are only two banks operating in East Timor. BNU has been present since 1999 but lending on its own risk has been modest, with operations restricted priinaiily to basic bank services such as clearing and transfers. ANZ established more recently and has very limited operations. 3.43 Both banks operate as branches of their overseas parents and as such are subject to the regulatory regime of their parent country authorities. The BPA maintains close correspondence with the Portuguese and Australian supervisory authorities with respect to the operations of the local branches. 3.44 There has been some discussion of establishing locally incorporated banks, but, with respect to commercial banking, caution needs to be exercised given the lack of supervisory capacity and the inability to insure deposits or respond to a banking run. The appropriate response to the dearth of con~merciallending is to address the shortcomings noted above, which if left unaddressed, would equally hinder prudent lending by local banks. As demand grows, attracting additional high quality foreign commercial banks that are supervised and financially backed by their own regulatory authorities should be considered. 3.45 Microfinance programs, by contrast, may play an important role in developing both the private and rural sectors. In recognition of this, several projects have been established including the ADBYsMicrofinance Development Project and a number of independent microfinance operations. The World Bank's Community Empowerment Project (CEP) has issued credits and grants for village level projects. The Small Enterprises Project (SEP) has issued loans to small and medium-sized enterprises throughout all district capitals. SEP 11,continuing on from SEP I, will also seek to extend increased credit for viable businesses to the sub-district level. 3.46 Initially the focus was on the general business sector due to that sector's relatively greater loss of physical capital following the 1999 destruction. However, start up of these programs has been slow. Moreover, loans and credits that are now being extended are being directed almost exclusively at businesses that are downstream from traditional rural activities. Although creating dowilstream denland indirectly benefits farmers, none of the current available credit programs strictly tailors any of its lending to rural con~inunities,taking into account the particular needs of faimers. Since 1999 the agriculture sector has rebounded strongly and returned to seasonal production patterns. There is a need to ensure that farmers and rural-based enterprises have sufficient access to credit, offered in coi1jui1cti011 with technical agriculture initiatives and training of farmers in the responsible use of credit. THELEGAL FRAMEWORK LANDOWNERSI-IIP FOR 3.47 Portuguese colonial rule, followed by the Indoilesian regime and widespread destruction of real estate and land records in 1999, have resulted in a complex set of conflicting land claims. Tbere is an urgent need to settle these claims. Commercial activity and investment in new construction and capital improvements as well as residential construction and restoration are stalled without appropriate mechanisms of establishing ownership and providing security against possible eviction. Unauthorized occupation of land is increasing as is land fraud including false claims and instances of persons renting and selling property to which they do not have clear title. Social discontent is being created as these tensions remain unresolved. 3.48 Accordingly, the Land and Property Unit (LPU) of the Ministry of Infrastructure developed a set of proposals on land rights and unlawful occupation. A Cabinet Sub-committee on Land and Property considered these matters and in October 2000 the ETTA policy on land was set down by Cabinet with the following functions assigned to the LPU. Assign temporary use agreements to confirm use of property Settle disputes related to temporary use Register claims to land, but not land titles Institute a period for lodging claims to land extending well into independence coupled with a public information campaign Develop a single national database of land claims Develop a mechanism to provide some confidence to investors such a regulation allowing longer term (e.g. 20 year) temporary use agreements where the ownership of land is clearly incontestable (e.g. government property) Initiate a national debate on the land question Strengthen capacity for resolution of land disputes at the community level. 3.49 While the Cabinet policy provides a welcome clarification of responsibilities, in fact the LPU has already been performing some of these functions. Assigning of land use rights by the LPU has been an effective means of freeing up assets for productive use. The practice of delegating LPU land use decisions to District Administrators has expedited the process, and more than $1 million has been raised through the sale of use rights on public and abandoned land this fiscal year. However, the average temporary lease is less than one year and investment by leaseholders is limited because of perceived risks to both parties in the current uncertain environment. Provisions for 20-year temporary use leases have not yet been implemented. Moreover, the LPUYsauthority extends only to public and abandoned land, while the large and growing body of land subject to private disputes is not covered. The land use sales are therefore only a temporary strategy and a more permanent solution is needed. 3.50 This will be a substantial undertaking. Even before the destruction of virtually all land records in 1999, less than 25 percent of the land was registered, with the major part of the country being unregistered land comprising forest land, plantations, State land, and land held communally or under traditional ownership rights including land occ~lpiedby squatters and refugees. Establishing a formal land administration system will entail not only confinning the ownership of registered land but also recording details for all other land. The highest priority should be given to urban areas where coilflicting claims and disputed ownership are most acute and pressures for security of tenure are greatest. In most rural areas the traditional follns of tenure will be able to continue satisfactorily for some time, although details of this land should be collected and recorded in a database to assist land admillistration in the districts. 3.51 Land investigations and the action of provisional registration will be dependent on having in place a land registry and a land record system for unregistered lands. The land registry should include the records of rights and ownership of the former registered land under the previous Portuguese and/or Indonesian systems system as well as any new registrations and derivative transactions that may occur. A land record system will be needed to record ownership, occupiers and other details for non-registered lands. In the longer term, it will be desirable to bring all land under the registration system, a process that will require design and implementation of a systematic registration program that makes special provision for dealing with communally held land. The cadastral survey system must also be re-established to define the spatial locations of the registeredJrecorded land. A dispute process is needed to resolve conflicts over ownership, rights and boundaries of registered or unregistered land. 3.52 Accordingly, a Regulation On the Establishment of Panels of the District Court with Exclusive Jurisdiction over Land Disputes (draft November 21, 2000) and a Regulation On the Establishment of a Land Commission (draft October 24, 2000) have been drafted. The Land Disputes Regulation provides for the creation of "panels of judges within the District Court of Dili" to hear and decide land disputes. It also provides for panels within the Court of Appeal for East Timor to hear and decide appeals stemming from the Distiict Court panel decisions. 3.53 The Land Commission Regulation gives the Land Commission authority for: Verification and registration of land lights Cadastral mapping and survey Resolution of land disputes Administration of public and abandoned properties 3.54 These two pieces of legislation have not yet been passed by Cabinet/Council of Ministers. Once this legislation is passed, it will be critically important that the budgetary resources necessary to establish the associated administrative bodies be made available and that recurrent funding for their continued operation be within the scope of anticipated fiscal resources over the medium-term. 3.55 The recently adopted National Constitution prohibits foreign ownership of land. While the Government is exploring alternative means to facilitate foreign investment including through the use of leases, the legal validity of lease arrangements remains constrained by other legal framework deficiencies. Commercial Law 3.56 Coinmercial law provides rules for econon~ictransactioils between parties including the structure of contracts, the transfer of property and mechallisins for resolving disputes. This is fundamental to the creation of a good business envirollrnent. But unlike the Indonesian company law, which provides a fairly modern and folward-looking model for East Tilnor to adapt to its own needs, the Indonesian commercial law is archaic and is based 011 a 1847 Dutch code. Moreover, the I~ldonesiancode is an awkward an~algaillof laws dealing with succession law, family law and property laws as well as what would normally be considered as basic modem commercial or coiltract law. The Ministry of Justice in Dili does not even have a complete set of the Indonesian codes relevant to commercial law issues and it is unlikely that one exists anywhere in East Timor. Somejudges may have a recollection and understanding of the old code but this is a wholly inadequate basis for providing judicial certainty in busiiless affairs. 3.57 At present there are no initiatives to revamp the Indonesian code or introduce anything new. This issue needs attention. Adoption of a modern law would likely be preferable to adaptation of the Indonesian one but this is a judgment the authorities will have to make. Then work should begin as soon as possible on drafting and passing the legislation. Labor Law 3.58 Labor regulations have been the subject of active public discussion in East Timor, beginning with the first UNTAETIILO Tripartite Consultation Workshop on Labor Laws and Employment Rights in April 2000 and continuing with consultations on the proposed Labor Regulations for East Timor held in January 2001 by the Division of Labor and Social Services in the Department of Social Affairs, in collaboration with the ILO. Draft legislation was prepared in July 2001 but has not been passed. 3.59 The draft law confines itself largely to defining the administrative and adjudicative structures, and issues such as minimum wage rates are to be determined by subsequent regulations. There are however a number of explicit quantitative provisions. Several factors need to be taken into consideration in debating and possibly amending the draft as well as in setting the subsequent regulations, to ensure that East Timor is able to maintain adequate worker protection while keeping sufficient labor market flexibility to ensure competitiveness and the creation of employment opportunities. These include: overtime pay rates, duration and frequency severance pay entitlements and notification requirements minimum wage level and frequency of adjustment holiday leave sick/family/maternity leave work week treatment of seasonal workers use of foreign workers 3.60 There is a tradeoff between labor protectioil measures such as the minimum wage and the level of investment and employment generation. Private investment will be attracted on the basis of East Timor's competitiveness, which will be partly deternilled by worker productivity in relation to the basic wage rates and benefits that must be provided. Second, East Timor's unique circumstailces as a new nation suggest that a period of assessment may be necessary so that the economy's competitiveness can be calibrated before these regulatory parameters are set. It will also be important to ensure some consistency in the regulatory environment so that investors do not abruptly face new rules after they have already committed resources. This suggests a go-slow approach: take enough time to set labor standards at levels that are commensurate with East Timor's level of productivity and competitiveness; allow reasoilable transition periods for firms to con~ply;focus initially on the structures and process of standard setting (composition of regulatory boards, deliberative/consultative procedures) without necessarily setting tight deadlines for establislment of quantitative regulatory hurdles. Finally, certain issues in areas such as wrongful dismissal, worker safety and sexual harassment lie more in the realm of worker/human rights and are independent of productivity and competitiveness. Regulation in these areas sl~ouldbe addressed directly. Tax Law 3.61 The tax law has been comprehensively redrafted using the Indonesian law as a base. Tax rates on wage income are as follows: Income band (US$/month) Rate (percent) Less than 100 100-650 650 and up 30 3.62 Tax must be withheld, at rates ranging from 4 or 5 percent for income derived from services such as construction, transportation and oil drilling, up to 15 percent for dividend income interest and income from property. In addition there is a service tax of 10 percent applicable to hotels, restaurants, vehicle rentals and telecommunications companies. The withholding tax rate for non-residents is 20 percent. 3.63 Imports are taxed at an import duty of 5 percent and a sales tax (imports only) of 5 percent. Excise duties are payable on a narrow range of luxuries and "sin tax" goods. This closely parallels the Indonesian system with the exception that there are virtually no exemptions offered in East Timor whereas Indonesian law has a complex exemption regime. 3.64 The tax structure is simple with just three wage tax rate bands and general freedom from exemptions. Consideration was given to using a VAT for indirect taxation. This was rejected however, on the grounds that it would be too complicated and would yield little additional revenue given the thin domestic value added and the high proportion of imports. From the point of view of transparency, minimal compliance costs and ease of administration and reporting, the tax system in East Timor is good. Priority was signaled from the highest levels in getting the draft legislation prepared and passed, in getting the large taxpayer unit in place and constructing the taxpayer rolls. This aspect of taxation has been well handled and contributes to the development of a healthy business environment. Tax incentives, sectoral priorities and the structure of taxation have been the subject of recent debate and are discussed in more detail below. 4. RAISING AGRICULTURAL INCOMES AND PRODUCTIVITY THEIMPORTANCEOF AGRICULTURETHE ECONOMY IN 4.1 East Timor is a small (19,000 km sq), lightly populated mountainous country, with several geographic characteristics that make it less than ideal for agricultural production-rugged, erosion-prone terrain; poor soils, and varying, often unpredictable rainfall. Nonetheless, the ecoiloiny is predominantly agricultural, with about 75 percent of the population living in rural areas where they work as subsistence or small-scale fanners of rice, com, vegetable gardens, tree crops, and livestock. Estimates for 2000 indicate agriculture's share in non-oil GDP at about 26 percent. 4.2 Most of East Timor's land is dominated by subsistence production of staples such as maize, cassava, rice, and sweet potatoes for domestic consumption. The estimated area planted is 121,000 ha for maize, 91,000 ha for cassava, and 39,000 ha for rice. ' Other crops include coffee (89,000 ha). Most of the land is not used intensively and there is potential to increase crops, pastures, and livestock production. Of the estimated 600,000 ha of land suitable for arable production, only 40 percent is cultivated. There may also be potential to diversify crop production for local consumption and possibly export. Much of the south coast is flat lowlands, but land utilization there is limited by the erratic availability of water for irrigation and poor road access to markets. There is abundant rainwater, but rainfall is seasonal and runs freely to the sea. Constructing water capture systems is difficult because rain is torrential and watercourses are unstable and itinerant. Thus many farmers live in highland areas, and most of them utilize diversified farming systems to ensure food security: shifting (swidden-slash and bum) cultivation of rainfed crops, mainly maize; cultivation of rice (in lower, flatter areas); house gardens of vegetables, cassava, beans, and maize; household chickens, pigs, and perhaps a buffalo; and harvesting of forest products such as tamarind, candlenut, sandalwood, and yams. Forest coffee plantations are also found in the highlands, and coconuts are grown in the lowlands. 4.3 The agricultural production systems overlap one another, with a single-family being involved in three or more farming activities. Climatic conditions allow for growing one maize crop and one or two rice crops per year as well as a variety of legumes, and roots and tubers. Reliable monsoon rains fall from December to March when more than half the annual precipitation falls. But even over this period, distribution can be uneven and droughts are common. The mountainous topography gives rise to a diversity of inicroclimates which affect production. Generally rice is grown in lower, warmer areas, maize dominates the medium altitudes, and root crops are found at higher altitudes. 4.4 Livestock, fishery, and forestry industries are not well developed and coiltribute coinparatively little to GDP, though they play important roles in subsistence and in culture. 'MAFF Agriculture Working Group: Draft Analysis of Tinlor Lorosa'e Household Survey, March 2002. 35 4.5 Coffee, and (fomlerly) some live cattle are the only agricultural products that are exported. Agriculture's key positioil in the economy, and its low productivity, has created ail imperative for rapidly rebuilding agricultural capacity for both food security and income-generating reasons. 4.6 Table 4.1 provides a breakdown of the agricultural sector's contribution to East Timor's GDP Table 4.1: Agriculture's Contribution to Rural GDP, 1993 & 1999 1993 1999 Subsector Share (96) Share (%) Food crops 64 6 1 Non-food crops 20 27 Livestock 13 8 Forestry 1 2 Fisheries 2 2 TOTAL 100 100 Source: Central Fiscal Authority, ETTA, 2001 The Major Farmirtg Systenls 4.7 In response to a wide variety of land forms, soils, and climate, farmers have evolved diversified farming systems to ensure food security. The traditional agricultural production systems commonly used by farmers in East Timor are: Swidden cultivation of rainfed crops, mainly maize; Lowland cultivation of rice, either rainfed or irrigated; House gardens with rainfed crops of maize, cassava, and beans, also chickens, goats, and pigs; Production of Bali cattle and buffaloes, also used to puddle the paddy land in preparation for rice planting (tile technique of herding many animals around the field, called rencah); and Harvesting forest products such as tamerind, candle nut, fuel wood and stand-by foods such as yams. 4.8 In addition to these production systems, certain areas have extensive plantings of tree crops-coffee in the highlands and coconuts in the lowlands. Some are large plantations established by the Portuguese but the majority are smallholder plantings. The agricultural systems overlap each other with swidden cultivation being by far the most doininant activity. 111 some parts of the country swidden agriculture is practiced on a sustainable basis while in other areas population densities (more than 50 persons per hectare) are too high for sustainable swidden cultivation. This gives rise for the need to promote conservation and resettlement, addressed later. 4.9 House gardens of typically 0.25 ha, are found around most rural dwellings.2 In some areas they operate well and are very productive. In others they appear more haphazard. The differences are related to climate and etlmicity. House gardens also have h i t trees and small animals. Typically maize is consumed by the household and most root crops are fed to the animals. The short and unreliable wet season limits the development of house gardens and in the drier areas goats replace pigs as the animal of choice. 4.10 Wetland rice culture may be inigated or rainfed. Maize is cultivated on the upland areas at the beginning of the wet season and land preparation of rice occurs after that. After the rice is planted farmers move back to their upland swidden gardens to harvest maize before returning again to the paddy fields to harvest rice. With this system adverse weather conditions for one crop are compensated for by favorable conditions for the other crop. The goal of the farmer is not to intensify production but to ensure food security through diversification. Nonetheless, preliminary results of the 2001-02 poverty survey indicate that nearly 30 percent of households continue to experience food shortages during the traditional 'famine months' (November to February) between the rice and maize harvesting seasons. Post-Conflict Developments 4.11 Following the vote for independence in August 1999, widespread violence by the anti-independence supporters was unleashed on East Timor. Many people were killed and many forcibly removed to West Timor. Houses and property were burned and looted. Agriculture, already suffering from neglect because of political disruption, ' received a further blow as livestock for meat and traction were destroyed or driven into West Timor, farm and food-processing machinery was destroyed or stolen, agricultural inputs (fertilizer, pesticides, hand tools) were removed or burned, food stocks and seeds were looted, and local urban markets were burned. Most farm families store crops in their homes, thus the burning of rural villages and homes destroyed vital food stock and seed supplies. Only seeds stored in trees or buried in the ground remained, and even some of these were consunled during the crisis thus depleting the supply of seeds for the next planting season. Because of the large number of displaced persons, lack of manpower also constrained land preparation and planting. The Department of Agriculture was looted and approximately 80 percent of its staff left for Indonesia and never returned. Rural roads and bridges were not maintained and became impassable thus isolating farmers from markets and input supplies. As a result of these numerous causes from the conflict, rural life was thoroughly disrupted and agricultural production was greatly reduced. At the beginnin3 of the new millennium, East Timor was threatened with malnutrition and starvation. 4.12 After hostilities ended, a major imperative became the rebuilding of the agricultural sector. This rehabilitation period benefited from a large infusion of donor and NGO assistance during the emergency and post-emergency periods. - - ~ The average size of cultivated plots is 0.6-0.9 ha per household, but this includes plots other than house gardens. 4.13 As a result: Starvation and malnutrition were avoided, due to the tradition of solidarity and shaiing in conlmunities, the availability of traditional standby crops (cassava) and wild foods (sago), and the free distribution (often by helicopter into reinote areas) of rice, cooking oil, and other food supplies by national and international aid agencies; There has been considerable reconstruction and replacement of rural infrastructure and ago-processing facilities, such as coffee-bean processing machinery, village rice mills, and maize huskers; Many farmers of Indonesian origin who were settled .in East Timor under new irrigation schemes have left the province, leaving the land to its original owners; The Ministry of Agriculture, Fisheries and Forestry (MAFF) has established a headquarters staff and field officers to manage the rebuilding of the agricultural sector. 4.14 Followii~gsharp drops in output in 1998 reflecting a drought and 1999 due to the disruptions, there is considerable uncertainty about interim production trends (Table 4.2). MAFF estimates for the 2001 harvest indicate a full recovery from pre-conflict (1997) levels for most crops-with both rice and maize production estimated to exceed 1997 levels. However, preliminary estimates froin the household survey indicate considerably lower production and yields during August 2000-August 2001.~While irrigation systems have been rehabilitated, large areas of irrigated land may still require rehabilitation. Recovery of agricultural production has also been limited due to continuing marketing difficulties and in some areas by lack of soil preparation 4.15 Production of green coffee beans was largely unaffected by the conflict, but destruction of processing capacity and lack of maintenance of road infrastructure initially caused severe problems. The cooperatives that process coffee made great efforts to replace destroyed or stolen plant and to repurchase stolen stocks. Production in 2001 appears to be at least at pre-conflict levels. 4.16 Small animal production appears to have recovered. Chickens, pigs, and goats are now seen again in rural and urban areas. Buffalo and Bali cattle are being imported under the ARP, and others have been brought from hiding places in the mountains, but total numbers of large animals remain below pre-1999 levels. 3 Given different ~nethodologies,imputing a trend from the data in Table 4.2 is difficult and should be approached with caution. It is possible that Indonesian data were inflated, and since MAFF estimates for 2001 were based on Indonesian data, they may also be too high. In addition, recent experience suggests that milling rate in East Timor is closer to 50% as opposed to 65% used in the past. This may reflect deteriorating post harvest processing capabilities. Table 4.2: Agricultural Production, 1997 and 2001 Indonesian Data MAFF Data TLHS Data Food Crops Prodrrctiot~ Yield P~.oductiotz Yield Prodrrctiotz Yield (MT) (MTAa) (MT) (MT//Ia) (MT) (MTha) Rice 35,979' Maize 95,000~ Peanut 3,200 Soybean 1,200 Cassava 66,500 Sweet Potato 16,200 Tree Crops Coffee 9,700 Coconut 9,900 Candlenut 690 Cocoa 42 Cloves 12 Livestock Head Local Chickens 585,355 Broilers 503,183 Goats 202,934 Pigs 362,473 Cattle 146,557 Buffaloes 73,818 Horses 32,792 Sources: Indonesia Central Bureau of Statistics,1998,Livestock Service of East Timor Province, 1998, and UNTAET, Ministry of Agriculture Fisheries & Forestry (MAFF) estimates,Timor Lorosa'e Household Survey(2001). Figures for livestock are from JICA (2002) The Study of Integrated Agricultural Development of East Timor. ' Takes into account official Indonesian estimates of paddy production, and 50% milling rate. This is a revised estimateby MAFF experts of 1997maize production. 4.17 Production from house gardens appears to have recovered, but slowly. While fruit trees and root crops remain productive, seeds for annual crops have been unavailable or unaffordable to impoverished households with no cash income. MEDIUM-TERM FRAMEWORK PRINCIPLES AND 4.18 Rebuilding East Timor's agriculture and developing the rest of the nlral sector will take many years and require considerable external assistance. It will also require East Timor's farmers and govenlnlent officials to adopt new habits of thought. 4.19 Prior to the vote for independence, the central Illdonesian government played a considerable role in East Timor's agriculture. The central government fixed the prices paid by consumers for agricultural staples and guaranteed retuills on these staples to fanners. It also coiltrolled the distribution of agricultural inputs, either by producing tl~ein through state-owned entei-prises, or by purchasing and distributing them itself, or through price subsidy mechanisms. Agricultural extension advice was given at no cost to farmers, but internal trade was heavily regulated. 4.20 East Timor has the opportunity to create an enabling environmeilt that will allow farmers to decide for themselves what to produce and how to market what they produce. It will also allow East Tiinor to develop an agricultural policy framework based on global best practices. Ln this new environment, the government would become the "enabler" or "facilitator" of agricultural development rather than a dominant decision maker and provider of agricultural inputs and services. After the transition period, its role may well be limited to determining and carrying out agricultural policy, supervising and financing the govenment's agricultural investments, establishing and enforcing a "light-handed" regulatory framework (particularly on natural resources management and quarantine requirements), and providing such services as vaccination, data collection and analysis, adaptive research on farm products, and agricultural education. 4.21 The adoption of such a framework is based on an assessment of the limited capacity to successfully intervene more aggressively by the government, by the scarcity of resources for doing so on an ongoing basis, and by evidence from other small open economies of the benefits of relying on market signals and international prices for guiding farnler decisions. Moreover, having dismantled the prior system of regulated prices for inputs and outputs, there seems to be little point in returning to a system that is inappropriate to East Timor's circumstances. Whether or not rice should constitute a special case during the transition is discussed in more detail below. 4.22 The following principles for the medium term may serve as a useful guide as East Timor builds its new agricultural system-as long as specific institutional features and the potential for market failures are also borne in mind: Agriculture will continue to be the most important source of livelihood for the majority of the population and the principle economic activity in East Timor. Agricultural production and price decisions should be made by those in the private sector involved in agriculture: farmers, bankers, business owners, suppliers, and traders. Such an approach minimizes demand for government resources as well as opportunities for corruption among civil servants. This approach implies that both internal and external trade should be relatively free of tariffs, quotas, and taxes in order to allow producers to make pricing decisions based on prevailing prices. When public intervention is appropriate, the model should be public financing and private delivery. For example, the government may wish to support mass media or group extension activities. It can do this by contracting with private extension agents or NGOs to deliver the services (as MAFF is currently doing with vaccination campaigns). The govenlineilt should not own revenue-generating enteiprises, such as meat slaugl~terhouses,warehouse facilities, grain storage facilities, tractor pools, or rural service and equipnlent centers. ~ k v e l o ~ m eprojects and constructioll of physical infrastructure facilities at the n t coinmunity level should be guided by participatory processes. 4.23 The following sections use these principles as guidance in addressing the major issues in the rural sector of East Tinlor, and to highlight the priority activities that the donor community might engage in over the transition period. The Role of Government in Agriculture 4.24 Notwithstanding general acceptance of a framework that uses price signals to guide resource allocation, investment decisioils and rewards factors of production, there is considerable debate regarding how rapidly such an economy can be put in place and the role of government in guiding and regulating the market outcomes that result. As is evident from the lack of spoiltaneous markets for farm products in East Timor, markets and institutions (particularly for agriculture) take time to develop. It is therefore useful to think of transitional policies that the government could utilize to assist the development of agriculture consistent with the long-term goal of primary reliance on private decisions based on intemational prices. This will require considerable discipline, a clear framework and timeframe, and a communications plan so that political pressures do not overwhelm the strategy when the transition is deemed complete. 4.25 As the government and the intemational donor group have recognized, a priority for the transition period was the need to boost agricultural productivity, that traditionally was low (Table 4.3). This is a time when unusually large public funds are being directed to the rural community. One-off public expenditures to rehabilitate and improve roads, bridges, irrigated areas, public buildings in rural areas, sanitation and water supply, and communications are appropriate but should be accompailied by viable plans for their continued maintenance either by the communities with or without payment from the government, or by the public sector. But further action may be justified in the short term. It may be necessary for government to assist in the reconstruction of household food storage facilities, to arrange for the availability of farm input supplies in regional centers if spontaneous importers do not arise and to exempt farm inputs from import tariffs, to perhaps rent trucks to traders in the short run while they establish their trading business, and assist in the establishment of rural banking facilities. The idea is to re-establish or build market linkages for farm households as rapidly as possible. The section below on donor assistance and funding requirements and Table 4.4 provide greater specifics on existing donor assistance programs. 4.26 In the longer-term the govelml~entwill want to put in place the building-blocks (investments, laws, regulations, policies, and public good activities) that define the enabling environment for agriculture. The illvestments would include re-establishing adaptive research stations and agricultural training facilities, and providing extension selvices for agriculture, land management, and rural household management/self- reliance. The govelmnent will want to establish rules and regulations for agriculture, livestock, forestry, fisheries, land management, resource management, and conservation. Whenever possible, the governmeilt should adapt existing regulations and policies used by other countries as its ow11 standards, rather than creating new regulations for itself. An example might be conlpetition policy. In addition, it would be advisable to recognize the testing and certification process of other countries, such as Australia, so that East Tiinor need not establish the iilstitutioils that otherwise would be needed to conduct its own certification of agricultural products. 4.27 Foreign grants for agriculture over 2000-05 are likely to be considerable- exceeding US$10 million in FY 2001-02 alone. These foreign grants are likely to be one- off events and should be used to put in place the framework and infrastructure for a vibrant agriculture and for longer-tern1 sustainable development of the sector. 4.28 The chief goals of government policy should include achieving: food security through improvements in the productioil of a diverse set of staple foods and the restoration of household and village crop storage facilities. improvements to the production of niche crops and animals, and on the promotion of internal markets and alternative income generation able to provide cash income to subsistence communities. improvements to swidden farming systems in the uplands that will improve livelihoods and reduce environmental degradation. It would be a mistake, however, to turn swidden farmers into sedentary farmers too soon. Many of those involved in shifting (swidden) agriculture in East Timor use sustainable methods, and there is evidence that extension workers have given such farmers poor advice in the past. Increased Protection for Rice? 4.29 While production of many crops may have largely recovered to pre-1999 levels, the adjustment for rice farmers appears to have been particularly traumatic, or at least better publicized. Though rice production in 2001 exceeds production in 1999 or 2000, farmers have faced difficulties in selling their surpluses, as distribution networks and marketing for imported rice in Dili and upland communities are better developed. Local rice producers face difficulties in accessing markets due in particular to a lack of post harvest processing especially packaging for transport, and unfamiliarity with distribution and marketing given the previous regulated system. Lack of stores to purchase inputs and inadequate access to credit also represent significant hurdles. By contrast, the price of iinported rice at the retail level has not been below local prices-though this inay reflect higher mark-ups and profits for distributors of imported rice-and free donated rice and smuggled rice do reduce the ability of local farmers to sell their s~rpluses.~ 4.30 Annex 2 provides more detail on the characteristics of rice production and the constraints to marketing local rice within the country. The following points summarize the annex: 4 Large quantities of rice are repoi-tedly smuggled tluougll the border in the west and by barge to the east. 42 Most people in East Timor are net coilsumers of rice, whether local or imported. Net rice consun~ers/buyersfall into several categories including:5 (i) highland rural poor; (ii) highland people with higher incomes derived, for example, froin coffee; and (iii) lowland non-rice fanners including people in provincial towns and Dili. Major impediments to the improved sale of local rice at markets throughout East Timor include: (i) dysfunctional post harvest processing, packaging, storage; (ii) distribution problems including networks, roads and transport infrastructure; and (iii) likely higher costs of production for local rice versus imported rice to the extent foreign rice production is subsidized. The decision to eat either rice or corn is not fixed and appears to depend on availability and price; nonetheless, most people prefer rice. The shares of maize and rice consumption thus vary: highland coinrnunities are estimated to consume 70 percent maize to 30 percent rice, while in the lowlands consumption shares are about 50 percent each. Current consumption of rice is estimated at 78,000 tons per annum. Local production is estimated at 29,000 tons (in 2001) and imported rice currently stands at around 50,000 tons. Production, consumption and imports were all disrupted following the 1999 violence but production has picked up and seems to be returning to pre-1999 levels, and consumption is thought to be higher than pre- 1999 levels. Rice imports are undertaken by private traders responding to market forces rather than by a government agency, donors and smugglers. Some 57 percent of traders' imports come from Vietnam, 35 percent from Indonesia and 8 percent from Thailand according to UN Border Control data. Local rice tends to retail at marginally lower prices than foreign rice. Imported rice is slightly more expensive in rural areas including the center, the highlands and some urban areas in the districts. Imported rice is cheaper in urban areas including Dilimaucau and in the west. 4.31 In response to the difficulties experienced by rice farmers, the thrust of policy should be to address the specific problems ailing rice farmers-post harvest processing, packaging, marketing, credit access, and facilitating the availability of inputs coupled with tariff exemption; at the same time, allegations of smuggling need to be investigated. Rehabilitation of irrigation systems, where warranted by long-term sustainability, also needs to be accelerated. But since the solutions to these problems will take time to implement, powerful pressures for increased protection in the form of higher tariffs on rice have arisen. Such a policy could perhaps be warranted if the higher tariff was modest and temporary-to allow a reasonable time frame for the above adjustments to be developed. 4.32 But there are also serious drawbacks to raising the tariff on rice that need to be carefully coilsidered before any decision is reached. First, rice faimers comprise about 26 percent of the population-and the proportion of net sellers is considerably less than Excluding the expatriate community which is expected to diminish in nun~bersand should not be a factor in formulating policy. 43 this; any boost to their incomes from higher protectioil must be weighed against the loss of purchasing power to the rest of the population, in particular the reduced coilsumption of rice by poor coilsumers in urban areas and upland fanners (the latter is where poverty is most acute). Second, as it is a basic consumer good, raising rice prices would exacerbate the problems with competitiveiless discussed in Chapter 2. Third, even if a higher tariff is intended to be temporary, international experience indicates that reversing the higher protection can prove to be very difficult in practice. 4.33 This in turn could promote shifts in cropping patterns not necessarily beneficial or sustainable in the long teim. These factors suggest that a hasty decision on higher protectioil should be avoided, and any such decision should be preceded by developing feasible means to assist the poorest rice consumers, particularly in the uplands, who are currently benefiting from the greater availability of imported rice. Coffee as a Specialty Crop 4.34 Coffee is the dominant export of East Timor, with an estimated 25,000 hectares of productive coffee gardens and an additional 23,000 hectares of unproductive ones. Some 25,000 farm families get a substantial amount of their income from small coffee holdings, while another 15,000 get a minor amount of their income from this source. Yields are around 150 to 200 kilograms of green beans per hectare, less than half their potential if improved cultivation techniques were employed. Most of the large coffee estates established in colonial-timeshave been abandoned, and the coffee bushes on them are picked by nearby farmers. 4.35 There reportedly are many complex cultural and social, as well as economic, factors affecting coffee cultivation in East Timor. Coffee farmers are predominantly the Mumbai-speaking mountain people who are said to be cLnon-mercantile,"viewing themselves as custodians of land rather than farmers. Accordingly, very little work is put into the husbandry of coffee. Bushes are not pruned or mulched, replanting hardly takes place, fertilizer and pesticides are not used to enhance yields, and families are said to haivest coffee only for present cash needs rather than profit maximization. Explanations of a more economic nature can no doubt also be suggested. During most of the time of Indonesian rule, the emphasis reportedly was placed on harvesting a large quantity, regardless of quality, as all Timorese coffee was sent to Java for blending. Only in more recent years did the emerging cooperative organization begin to pay a premium for better quality cherries and to improve the quality of local processing. 4.36 A recent representative survey of coffee growers conducted by the World Bank in the major coffee producing areas concluded that the majority of farmers would be prepared to invest more effort in their activity if that meant increased income. If they were given free saplings of good quality, they would be prepared to replant on a significant scale. Also, most indicated that they would welcome access to communal facilities for processing cllei~iesand thus produce a better quality coffee bean. Almost all were ready to join an association or other cooperative entity if that were to facilitate access to better facilities. 4.37 Because of low inputs into coffee production and the favorable climate, East Timor's coffee can be certified as organic and is in demand, both for this reason and because of its flavor characteristics. The demand for organic produce worldwide is growing, and certified organic coffee from East Timor may become a more valuable source of export revenue. Presently, only about 10percent of East Timor's coffee fetches the organic premium even though all coffee produced is, in fact, organic. Efforts to increase the output and export of coffee should be made carefully, based on improved varieties and cultural practices, not on increased purchased illputs such as fertilizer or pesticides. It would be desirable to accompany these efforts with a campaign to inform farmers of the importance of organic certification, so that East Timor can obtain the organic premium on a larger portion of its exports as this segment of the global market grows. 4.38 For the more immediate future, however, most of the income gains are to be had from improving the quality of processing, a change that requires investment in recuperating and building new communal facilities and the provision of good quality saplings and advice to farmers. With adequate financing arrangements such as micro- finance schemes, part of the cost of such additional infrastructure and services could be borne jointly by the farmers, with the state perhaps providing some form of catastrophe insurance. Fisheries Development potential6 4.39 The fisheries sub-sector is not well developed but may have the potential to improve food security and contribute to small-scale development and exports. East Timor has about 650 km. of coastline but very little continental shelf; depths of 200 meters reached within 2-8 krn of the shore and depths of 2000-3000 meters within 20 km. Coastal resources are limited to a narrow stretch of steeply sloping seabed. The limited statistics available on fishing activity prior to 1999 point to a fleet of some 630 vessels with outboard motors and 1,387 canoes. These data are for 1997, when 9,066 fishers reportedly caught a total of 2,423 tons of marine fish. During the 1999 disturbances, practically all of the larger commercial fishing vessels reportedly left East Timor, although 24 vessels are said to be in Atauro island, part of East Timor and a few miles off from Dili. 4.40 The biggest current problem in the fishing industry is lack of demand due to lack of purchasing power. The poor consume about 10 grams per week per capita, while even the 'iich' consume only about 190 grams according to the latest data from the TLHS 2001. Low local demand for fish will likely increase only as overall incomes rise. 4.41 A recent LrNDP study that focused also on marine needs and priorities in East Timor (Assessing Environmental Needs and Priorities in East Timor; Dili, 2000) concluded that East Timor's marine ecosystem, if used in a planned and non-destructive manner, has considerable potential to support econoinic development and sustain the population. Using figures from other similar areas of marine waters, the study estimated 6This section draws on: "Hera Port Fisheries Facilities Rehabilitation Project," ADB, and "Fish for the Future: A Strategic Plan for the Fisheries of East Tinlor;" Depai-tnlent of Economic Affairs, May 2001. that even after providing East Timor's population with a minimum recolnlnellded annual fish consumption of 13 kg per capita, exports could still be sustained.' JICA has estimated that recent donations of fishing eilgines and gears (China alone donated about 300 outboard engines, 1,500 gill nets, and 3 metric tons of lines) could motorize some 40 percent of the East Timorese fishing fleet.' This could soon strain fragile coastal resources, indicating the need to introduce early systems of management. Given that coral reef fisheries are easily over-fished, restricting it to local indigenous fishers appears warranted. 4.42 In addition to its potential to sustain an important sector of subsistence activity, fishing offers potential job-creation opportunities in related activities such as marketing, processing, provisioning, servicing, etc. Once East Timor delimits its exclusive economic fishing zone, the offshore resources will be available for larger-scale commercial exploitation, either by East Timorese fishers or preferably through licensing arrangements with foreign fishing vessels, as experience in Pacific Island countries indicates. East Timor's pelagic resources include small species such as Indian mackerel, scad, sardines and small tuna, but also high-value species of large tuna. 4.43 MAFF's Fisheries Division has developed a strategic plan for fisheries that aims to sustain East Timor's aquatic resources, establish a profitable private sector-based fishing industry that maximizes sustainable economic returns, and assert jurisdiction and control over the marine resources in the seas surrounding East Timor. The government's role is to be limited to providing the enabling business environment and framework within which the sector's private fishing and associated businesses will operate. 4.44 External assistance at this stage should be channeled towards developing the regulatory and legal framework for inshore fisheries management and investment in offshore fishing. Enforcing fishing regulations through partnerships between government, communities and private operators (co-management arrangements) and recognition of user rights for coastal communities has proven effective in many Asian- Pacific countries. External assistance can also be directed towards rehabilitating the fishing harbor at Hera port, promoting further inland aquaculture (particularly of carp) and ilnproving post-harvest handling and processing. This separation of functions between the private and state sectors seeks to incorporate lessons learned from other countries' failed fisheries development programs. In many cases, direct involvement of government agents in the sector, in activities overlapping those of the private sector, led to overstaffing, inadequate accountability, and deterioration of services and facilities. Thus, in the new Hera facility, for example, it is envisaged that private operators will be responsible for all activities that can be commercially viable, including boat building, workshops, stores, and fuel supply. In this way, each colnmercial enterprise within the port area will have a vested interest in a functional port. Food Security 7 There are, however, a number of constraints to developing significant exports in the short run. 8 JICA (2002) "Study of Integrated Agricultural Development of East Timor." 4.45 East Timor's lack of purchasing power will cause suffering, but depeildeilce on aid will also generate distortions in rural-to-urban migration and agricultural prices. Food security will be achieved when households are productive enough to grow or purchase adequate food for their families. When hunger has been eliminated, absolute poverty has been eliminated. Accordingly, reducing hunger and poverty through food policy is a top priority in East Timor. 4.46 An appropriate food policy is one that has programs or policies in place to cope with short-run food contingeilcies while retaining a consistent long-term vision, policies, and actions for the agricultural sector. The following discussion relates to short-run food contingencies. 4.47 Although there has been discussion within East Timor and its donors of a policy that would ensure self-sufficiency in food production, other countries have found this approach to be expensive and difficult to sustain. It is expensive because to guarantee sufficient local supplies in a low production year, stocks must be purchased and held from above average production years. Holding stocks of grain is expensive in terms of the opportunity cost of capital and physical deterioration of the stock. Further, complex price manipulation and purchase/sell arrangements must be constantly made by the buffer authority. Around the world there is a record of failure of managing public stocks. 4.48 A more recent approach is to establish a buffer fund with which to fund poverty support in times of high food prices (shortages) while creating the conditions for a robust overall economy. Under such conditions, much of the population is able to afford food and farmers receive the right price signals for deciding on production. A buffer fund is a public fund earmarked to provide assistance in years of exceptionally poor harvest to the most needy who would use it to purchase staples in the market. The government does not need to purchase food, store grain, or control prices itself. If the government is uneasy about providing cash, it could provide vouchers that can only be used for the purchase of foodstuffs, or it could purchase staples and provide them to the targeted poor. The last option however requires a larger bureaucracy to manage and administer than other options and hence may be least desirable from the cost perspective. 4.49 At the same time that the government establishes the buffer fund, it would seek productivity gains on existing farms, assist in increasing farm storage capacity to minimize post-harvest losses, and provide farmers with advice on diversifying into other crops, especially export cash crops. Such steps will contribute most to food security. Land Tenure 4.50 There is no cadastre or formal titling of land in the rural areas of East Tiinor, and only a small percentage of land has recogilized status under either Portuguese (colonial) or Indoilesian law. Pennaneiltly cropped land, such as house gardens and rice fields, is recognized by custom as individual property. Ownership of the vast majority of rural land, however, is deteimiiled by clan-inherited usufruct rights under traditional custoinary law. 4.51 All members of a clan (suco) have the right to seek pennissioil to cultivate unused land within the domain of the suco, and this land is not for use by a member of another suco except through marriage. Generally users inay only plant annual crops but inay plant tree crops in their 0.25 ha house garden. 4.52 There are large areas of the country with seemiilgly abandoned terraced gardens and unused pasture. These are referred to as sleeping land. This arises because of disruption of social structures, resettlement, and absentee landlords (the custodial keepers of clan lands who are responsible for distributing land to clan members). 4.53 If agriculture is to reach its development potential, a land registration system that guarantees certainty of ownership will eventually be needed. That will require a study of clan systems and existing tenure, and an adjudication process-a long-term and expensive process. The present arrangements appear to restrict land use and crop selection, but the situation does not require immediate action. Farmer and Rural Community Organizations 4.54 Farmer and other rural organizations have historically been weak in East Timor. The government should support the building of farm and community organizations, such as cooperatives constituted under Rochdale principles9. Other forms of business organization, such as limited liability con~panies,partnerships, associations, and NGOs should also be encouraged. Supporting women's groups is an effective way of empowering women entrepreneurship. The overall objective of all these activities is to foster individual and community self-reliance, and to develop organizations capable of implementing government-financed projects. Agricultural Education 4.55 Agricultural extension, institutional capacity and provisioil of agricultural information are starting from a very low base and warrant a high priority in government expenditures. To leverage resources, innovative partnerships with church groups, overseas donors, and NGOs could be developed. One partner might provide the physical facilities, another the expert instructor, with the government paying building rent and other variable costs. Natural Resource Management 4.56 Because of serious environmental degradation in much of East Timor, new plans for the development and protectioil of forests and watersheds are needed. This is a large task that will require a multiyear commitment by East Timor and its donors. 4.57 Forests. Forests have played a significant role in the development of East Timor. They contribute to family livelihood by providing firewood, nontimber forest products, poles, and timber. Many valuable timber species are indigellous to East Timor, such as 9See Abrahamsen, M. A. (1996), Cooperative Busi~zessEnlerpl-ise, McGraw-Hill, New York. 48 teak, redwood, and mahogany. Expoits of nontimber forest products include beeswax and honey. 4.58 Although the Indonesia11government classified several forest areas in East Timor for protectioil as national parks, deforestation has taken place at a rapid pace, and forest management practices have not been sustainable. In addition, large areas of mountain forests where resistance groups took refuge were destroyed by defoliants and are now degraded scrubland. 4.59 Firewood collection is a major contributor to deforestation. Firewood requirements for cookiilg expanded enormously when subsidies for kerosene, the main cooking fuel, were removed. Acceptable alternatives, such as fuel-efficient stoves, solar energy, and dedicated fuelwood lots need to be introduced to reduce deforestation. Effective forest management has only been achieved where the forest area has been clearly defined and is recognized by the community. 4.60 A forest inventory is planned to generate more complete information on timber species and non-timber forest products. A Forest Division with responsibility for reforestation, forest and watershed management and protected areas has been established at MAFF, and a revised national forest policy is due to be completed in 2002. 4.61 Watersheds. A national assessment of watershed degradation is needed to plan for environmental rehabilitation. Since this degradation is linked to poverty, it should be analyzed together with the results of the poverty assessment. The information from these activities could be coupled with that on land degradation to determine priority areas for intervention. 4.62 Regulatory Framework. East Tiinor will need to develop a regulatory framework for its natural resources. The framework should: Recognize the importance of resource sustainability Enable the participation of all stakeholders Recognize the value of traditional resource management systems Provide security and certainty of land and resource user rights Upland or Lowland Focus 4.63 Much of the early post-conflict assistance to farmers focused on lowland agriculture, especially irrigated rice production. Lowland or upland development choices are not always straightforward-many farmers travel to lowlands to produce cash crops but often choose to live in the upland where many have traditional land and a house garden. However substantial populatioils live solely in upland areas and are most likely to be subsistence farmers with little cash income-with the exception of coffee fanners. These people tend to be the poorest, the most vulnerable to natural disasters, and in some areas inay also be major contributors to natural resource degradation. 4.64 The heavy einpl~asison investment in expensive irrigation sti-uctures for rice, in a country where lice is neither the major nor traditional staple, and where the structures are subject to almost certain destruction by flash floods over time, needs to be examined. This is especially true given that world prices for rice are low and projected to remain low. Once rehabilitation of existing inigatioil systems is complete, plails to expand areas under irrigation need careful analysis. The rate of return on many new irrigation investments in East Timor would appear to be below that of other investment alternatives in the agricultural sector. Given the poverty-alleviation focus of most aid, programs that seek to increase the resilience of upland farming activities, minimize the vulnerability of these coinmunities to natural disasters and climate variability, and plan for a managed transition out of subsistence, are deserving of a higher priority. DONORASSISTANCE FUNDINGREQUIREMENTS AND 4.65 Agriculture rehabilitation efforts by donors, expected to be concluded by mid 2002, have focused on the restoration of lost assets and rapid rehabilitation of services, seed supplies, infrastructure, and equipment. Much of the assistance funding has been committed by TFET, the Chinese, Japanese and Portuguese govemnents, USAID, LTNDP, and the FAO, along with donations from Norway and Macau. Post-emergency food distribution was provided by the World Food Program and other NGO programs. 4.66 Rehabilitation efforts have been followed by a period of transition, which is focusing on helping the East Timor Government put in place its agricultural policies, regulations, and essential functions of MAFF (such as bio-security, forestry, and fisheries regulations), and in smoothing the transition from a heavily subsidized sector to one with limited recurrent funding. 4.67 Donor-supported integrated rural development activities with a long-term focus have started in late 2001 in parallel with trallsitioilal assistance. At its center is a prograin to lift rural incomes through productivity-enhancing activities, participatory rural development, and management of natural resources that seeks to encourage community and household self-reliance, and development of a diversified set of crops to provide cash income. Community-based maintenance of small imgatioil schemes, access roads, water supply systems, and natural resource management are being supported using this approach, through projects funded by TFET (Second Agriculture Rehabilitation Project), AusAID (East Timor Rural Developn~entProgram), UNDP/UNOPs (Ainaro and Manatutu Comn~unityActivation Project), and other projects supported by Portugal, JICA, CIDA, USAID, World Vision and various IVGOs. Several of these programs provide for capital expenditures against a clear commitment by communities to maintain and manage local resources, with the government playing an enabling and regulatory role. The assistance is to be provided in cash rather than in kind, and covers supplemental inputs and specialized assistance. Local labor, materials, and other items would be part of the coinmunities' in-kind contribution (Box 4.1). Additional iilvestinents in this phase may include establishing essential govemnent functions, such as data collection, monitoring, control and surveillance of natural resources, essential applied research, the facilitating of exteilsion and agricultural infoinlation networks, and capacity building. It is anticipated that the Disaster Management Unit, in collaboration with MAFF, will carry out disaster management. Box 4.1. Principles of Donor Assistance for Agricultural Sector During the 2001 Joint Donors Missions, the following priilciples guiding donor assistance to the sector were agreed with MAFF: Donor assistance should move away from rehabilitation to long-term sustainability; Greater focus needs to be given to programs of assistance that seek to promote food security over a wider range of rural communities; For programs focused on marketable agricultural products, it will be critical to undertake a realistic assessment of market viability; Donor assistance should seek to increase the resilience of farming communities and reduce their vulnerability to climatic and economic shocks; Donor-funded programs should seek to strengthen traditional systems of management and self-reliance within rural communities; Future assistance should be demand-based and integrated to ensure that the focus is one of rural development and not simply agriculture; Cash-based assistance (e.g. in the form of direct assistance to communities) should be preferred to input-based assistance; where the latter remains needed, it has to be backed by adequate funding and technical assistance for distribution and follow-up operation and maintenance. Where programs provide benefits to individuals or family groups related to income generation, funds should be provided as credit rather than as grants. Grants should be reserved for activities that offer broad-based community-benefits; Where communities benefit directly from the assistance provided, unskilled labor should to the extent possible not be paid for but provided as a community contribution. This would help improve the sustainability of the interventions; Prograins of assistance should allow for equitable levels of development for both upland and lowland communities. Partnerships between Government, aid agencies, NGOs, the private sector and communities should be fostered to maximize the value of assistance to rural communities. Source: Joint Agriculture Donors Interim Mission to East Timor (November 19-30,2001). Final Mission Report, January 31,2002. 4.68 Several sectoral priorities remain to be funded by donors. These include further technical assistance to the govemnent in policy development, and capacity building for administration and establishment of an efficient public institution at MAFF. Assistance is also needed to provide credit and savings facilities for rural people, to assess the impediments to marketing, and to explore economic opportunities for niche markets. Donor assistance to the forestry sector and watershed management, though started, remains a distinct gap. Further assistance is also needed to help MAFF establish an effective quarantine and pest management system, an early warning system for disasters and climate extremes (such as El NiAo-induced droughts), and to continue its support to GIs and information dissemination. Some assistance may also be needed in developing fui-tiler infrastructure, such as a research laboratory, a holding nursery and seed center, and the constructioil of abattoir facilities in Dili and other major towns, though the sustainability of these facilities would need to be carefully investigated. 4.69 Table 4.3 provides estimated donor assistance to the agricultural sector during 2001-2005.1° This program is expected to be revised periodically during the course of Joint Donor Agricultural Missions. Concluding Comments 4.70 The basic framework for agriculture now being put into place in East Timor should be largely completed within two years, paving the way for the long-term developineilt of the sector. Getting the building blocks of agriculture in place is needed to provide food security and income creation for a large share of the population of East Tiinor. Increasing food security, in turn, will provide a significant source of growth for the country. But to achieve this, all the building blocks noted previously will need to be present. 4.71 In designing its investments, the government will need to ensure that recurrent costs do not exceed long-tenn financial capabilities. Recurrent costs can be managed by maintaining the principle of user-pay-assigning recurrent costs of infrastructure that have identifiable private benefits, such as irrigation, to the beneficiaries. Indeed the government could take the further step of transferring ownership of certain assets to the community which would be responsible to operate and maintain them. The administration should resist patterning public support of agriculture after the top-heavy and unnecessary structures in many industrial countries. Where public action is justified, the first thought should be of public financing but private delivery. There will be strong pressures in favor of government intervention as farmers in East Timor have operated in a culture where the central government intervened heavily in commercial transaction. Changing the farmer's mentality to one of reward based on effort will be a challenge. 4.72 Even though agriculture has recovered very rapidly since 1999 and is presently being provided with a large injection of external capital, it is not realistic to expect the sector to grow rapidly on a sustained basis. With agricultural production for most of the major crops estimated to have recovered to pre-1999 levels, prospective growth will be constrained by shortages of capital, limits on farmer knowledge and expertise (including marketing expertise), the inevitable adjustment period from the previous regulatory regime and the bottlenecks and deficiencies prevailing in the distribution and transport systems. 4.73 Despite the large capital investment program planned by donors, a large subsistence agricultural sector in East Timor is likely to persist in the foreseeable future. Reducing rural poverty, building rural income security, and generating rural development will therefore need to remain high on the policy agenda. 'O This estimate reflects the results of the November 2001 Joint Agriculture Donors Mission. It does not include programs expected to be approved during 2002103, but for which commitments had not yet been made. Table 4.3: Current Estimated Funding for Agriculture aild Rural Development, 2001-2005 (US$ millions) Japarl TFET Aust~.nlin" FA0 USAID" Porf~rgnl UADP Clzir~~ 0then1' -E s r i ~ n n f e d - - - - - - - - - Reltnbilitntiotr (2001) F a ~ mto market rural roads 0.5 some Irrigation rehabilitationlwater 8.0 some Vaccination 0.4 SeedIStorage Some 0.5" some Livestock 1.5 Tractorsh4achinery Hand tools 0.15 some some Fishing Equipment some Post-Emergency Food Distrib. 0.09 Distribution of Farm Equipment Major Market Rehabilitation some Trntrsition (2001-2002) Vulnerable people some Rapid community pilots Sustainable vaccination 1.3 lnformation to farmers 0.8 Facilitating inputs (fert., pest.) 2.0 Organising farmers 1.I 0.15 some 0.3 Reducing post-harvest losses 0.09 0.4 some Micro-finance 0.2 Agriculture PolicyIStrategy 3.74 0.3 0.05 some Regulatory Framework: ForestryProtected Areas Fisheries 0.21 1.3 Quarantinelcertification Some some some Capacity building 0.23 0.9 0.03 some some S~rstaitrableDevelopmetit (2001-2005) Participatory Development and 0.9 1.5 Natural Res. Management Regulatory review Monitoring, Control and Surveillance Applied research 0.33 0.2 0.15 0.92 Rural Water Supply and Sanitation 7.5 Facilitating extension 0.15 some Monitoring/statistics/inventories 0.6 Land tenure 1.9 Sector coordination and partnerships some Major im'gation maintenance Disasterfrisk nlanagement Management ofprotected areas some Capacity building and administrative 1 .O 0.15 some 0.19 Source: Joint Agricultural Donors 111terimMission foEast Ti~ttor(November 19-30, 2001), Final Mission Report. January 31, 2002. I 'Australia donor assista~iceincludes activitiesfunded by AuAlD and the Auhalian Centre for International Agricultural Research (ACIAR) "USAID is providing a revolving credit of $12 million that will be divided in a yet to be specified way between the noted activities in the coffeeand rural wholesale sectors. " Others include donors not represented on this nussioci but known to be pla~llli~igacLivities in East Timor. These include New Zealand, Canada, Sweden,Norway, Macau,the European Union, ADB, the World Food Programalcdother UN Ageacies. l4 53.36m funded by Japan " fundedby Japan '"unded by Japan 5. ADDRESSING HUMAN RESOURCE DEVELOPMENT NEEDS Introduction 5.1 Human development is at the 11eal-tof development and the well being of people. h~ternationalevidence indicates and there is a positive relationship between human development-by which is meant the level and quality of education embodied in the workforce and the health status of the populatioll-and growth. Growth permits improvements in human development by creating resources which can be devoted to education and health services (by government and by individual families), while improvelnents in education and health status of the labor force contribute directly to the productivity of labor. Differences in human capital endowments can therefore imply different prospective growth rates. 5.2 The linkages between education and productivity are complex, but international evidence has shown, for example, that improved levels of education are associated with increased capacity to: (a) improve the input mix and enhance agricultural productivity; (b) alter the mix of commodities produced in response to changing market signals; and (c) switch occupations and sectors in response to economic opportunities. Improved health and nutritional status are, particularly for the poor, associated with improved capacity to work and lead productive lives. Higher education levels, particularly for girls are associated with lower population growth rates and healthier mothers and children due to a range of factors--delays in marriage, increased birth spacing, reduced child mortality and improved knowledge about nutrition and disease. In short, human development helps people build productive and healthy lives and a more cohesive society. 5.3 East Timor's specific needs provide the starting point for a discussion of its strategic human development priorities. As discussed, the country was devastated by the violeilce of September 1999. Not only was the capacity to improve its human development severely impeded, East Timor started from a low human resource base reflecting its colonial history. Always one of the poorest regions in Indonesia, East Timor continues to face the need for extensive rebuilding and rehabilitation of its economic and social infrastructure (including a sizable proportion of its schools, hospitals and health clinics-80 percent of which were destroyed after the September violence). Education 5.4 Prior to the Indonesian administration of East Timor, the Portuguese colonial adininistratioil had done little to promote mass education-at least in rural areas. Consequently, illiteracy was still about 90 percent in 1975. Under Indonesian rule, the public education system was significantly expanded. Indonesia set out to implemeilt its national education policy objective of 9 years of coinpulsory basic education for children aged 7-15 years. The Indonesian curriculum replaced the Portuguese system and Bahasa was introduced as the language of instruction. Structurally the education system comprised 2 years of pre-school, 6 years of priinary (from age 7 years), 3 years of junior secoildary and 3 years of academic or technical and vocational education, 2 years polytechnic education and 3-4 years of university education. 5.5 Over the period 1976 to 1998, enrolmeilt in primary schools increased more than 11 fold (an impressive rate of 16 percent per annum); junior secondary enrolment increased more than 100 fold, and senior secondary some 228 fold. Nevertheless, this was fiom a very small base in 1976 of 13,500 primary children, 315 junior secondary children and 64 senior secondary children. In 1999, East Timor still lagged significantly behind educational achievements in the rest of Indonesia. Primary gross enrolment rates had oilly reached 95 percent, junior secondary 64 percent and senior secondary 39 percent. Perhaps more telling, net enrolment rates (i.e., the proportion of the age cohort attending each level of schooling) were only 74 percent for primary school aged children, 36 percent for junior secondary school aged children and 20 percent for senior secondary school children. 5.6 There were a wide range of inefficiencies in the structure and financing of education and the system did not adequately cater to the needs of the poor. In 1999, Indonesia planned a $11 million budget for education, equivalent to some 2.9 percent of estimated provincial GDP with 35 percent allocated to primary education, 27 percent to junior secondary education; 17 percent senior secondary education; 7 percent for tertiary and a relatively high 9 percent for administration. Student-teacher ratios were relatively low with ratios of 25:1 in primary schools, 16:1 in junior secondary schools and 14:l in senior secondary schools. The comparable average for low-income countries is currently 41:1 in primary schools and 24:1 for all secondary schools. Overall public expenditures on education were contained in Indonesia by paying teachers a relatively low salary. 1 One consequence of this strategy was a wide range of informal fees for school attendance designed to supplement teacher incomes. This may well have impacted on the capacity of the poor to afford to pay for their children's education. It is also arguable that there were too many small and thus inefficient schools-particularly primary schools-although sparse populations, as in many parts of East Timor, do pose particular problems in this regard and suggest the need to consider options that involve multi-grade teaching in small schools. 5.7 Patteins of rural-urban inequality were strongly evident in school enrolments. Primary net enrolment rates in rural areas in 1999 were 73 percent, some 16 percentage points lower than in urban areas. Similarly, junior secondary school rural net enrolment rates were 33 percent, 34 percentage points lower than for urban centers while the comparable senior secondary school rate was 17 percent in rural areas, 23 percentage points lower than in urban areas. Thus, even universal primary education was far from being reached, and a very sizable proportioil of primary school age children in rural areas-more than 25 percent-were not in school. At higher levels of education the 'Average teacher salaries were equivalent to $91 per month prior to the financial crisis but only $20 per month when the rupiah collapsed in 1988-99. The average teacher salary was estimated to be only 58 percent of GDP per capita. differences between urban and rural areas were higher, and in absolute terms the propoi-tions not in school were very significant. For example, tlu-ee quarters of rural junior secondary school age children in rural areas were not injunior secoildary schools. 5.8 School coinpletion rates were very low, pai-ticularly for the poor and females. Indonesian Susenas survey data for East Timor for 1999 indicate that less than 60 percent of the poorest quilltile of 16-18 year old males and 47 percent of females had competed six years of primary schooling with comparable numbers for nine years of schooling being 22 percent for males and 27 percent for females. On the other hand, 74 percent of the richest quintile of males and 79 percent of females had completed six years of schooling. Fifty three percent of males and 48 percent of females had competed 9 years of schooling. For older cohorts of the population, education levels fell off dramatically, reflecting the fact that any attempt at mass education in East Timor was a post 1975 phenomena. 5.9 As a consequence of this history of educatioilal development and the low completion rates of even recent cohorts of children, illiteracy remains high. In 1997,just under one half of the population was reported as having no schooling. The 1995 Inter- censal survey also showed illiteracy varied significantly by district with some reporting over 60 percent as illiterate and that 90 percent of the population over 50 years reported as illiterate but with almost 25 percent of the 10-19 year olds also reporting as illiterate. 5.10 These data underscore the fact that East Timor starts with a low stock of well educated population. Illiteracy remains high even among today's youth. When combined with the fact that a large proportion of the teaching force, health staff and civil service more generally-particularly middle and upper level management-were Indonesian and have now departed the country, it is clear that teacher supply and the capacity to supply qualified staff for the health system, the civil service and the private enterprise are severely constrained. For example, in 1999 East Timorese accounted for 78 percent of primary school teachers, but only 3 percent of junior secondary school teachers and 8 percent of senior secondary school teachers. Only about 30 of the original 160 doctors remained in East Timor after the disturbances. Healtb 5.11 A significant public health system outside of a few urban areas had not been developed until the Indonesian administration expanded health expenditures. An extensive array of government hospitals, clinics and health posts were established across all districts staffed by about 160 doctors and about 2,000 nurses and midwives. The Catholic Church, other religious charities, and the coffee cooperatives also established a small but significant number of clinics and small hospitals. 5.12 Notwithstanding the development of this health infrastiucture, inany limitations of its effectiveness and relevance to the population's health needs have been identified. Specifically, many public health services were inadequate and limited-even in the provision of clean water and sanitation. Health outconles and the health status of East Tinlor's population during the 1990s remained unacceptable, with the disease burden falling disproportionately on the poor. 5.13 Life expectancy was low and the infant mortality rate was high. UN figures estimated that life expectancy was below 50 years in 1995 and that the under 5 mortality rate for East Timor was a very high 201 per 1000 live births. The 1995 Intercensal Survey estimated that the infant mortality rate was 78 per 1000 live births. While other estimates suggest the under 5 mortality rate was about 124 per 1000 and average life expectancy at birth was 55-58 years, there is little doubt that infant and child mortality rates were unacceptably high. 5.14 The reproductive health status of the population remained a concern. Only 22 percent of births in 1998 were attended by a doctor or midwife with most of the rest perfonned by family members. Thus women of child bearing age face considerable dangers as a result of inadequate pre-natal and post-natal care andlor a lack of willingness to access the care if available. While hard to document, it is clear that mateinal mortality rates are very high and that many mothers die unnecessarily as a consequence of child birth. Indonesian Demographic Household Survey data indicate the contraceptive prevalence rate was 27 percent (compared to 57 percent for all Indonesia) and that the total fertility rate (the number of births each female is expected to have) was 4.43 (compared to 2.78 for all Indonesia. The natural rate of population increase is estimated at a relatively high 2.2 percent per annum-although net out-migration and conflict contained the overall population growth rate. 5.15 The disease burden of preventable and communicablelvector borne diseases is significant and adversely affects productivity and the quality of life. While data on disease remains limited, it is evident that the prevalence of preventable diseases remains significant-although the immunization program had reduced the burden of disease associated with the six childhood diseases associated with the program. Notwithstanding this, disease outbreaks from this group of diseases were not uncommon, and WHO targets of 80 percent immunization were not achieved nor sustained. Malaria and TB constitute the country's two most serious specific disease problems with malaria being the likely cause of 65 percent of all new outpatient visits to government health facilities in 1998 for school age children (5-14 years) and 50 percent for the working age population (15-44 years). TB accounted for a staggering 25 percent of new outpatient visits among the working age population and 46 percent of the 45 years and older population. 5.16 Nevertheless, the health system was underutilized. The number of outpatient visits per capita per year was just over one-low by international standards-and indicating very low numbers (less than five) of outpatient visits per staff member per day employed in the health system. As expected, the most important users of the health system are the very young and the aged. 5.17 The low aggregate demand for health sei-vices, indicates a health system not working well. A working group, associated with the Joint Assessment Mission in January 2000 made the following observatioils on the effects of the Indonesian system: "The facilities or services provided in East Tinlor were based on a standard that was not relevant to local populatioil needs, situation and/or capacity to maintain ... This legacy is evidenced in the high ratio of facilities to the population served as well as the presence of facilities that are unnecessary or too costly to maintain." The report continued: "this highly-centralized system employed too few East Timorese at managerial and decision- making levels (provincial), resulting in a severe dearth of managerial capacity in health. Conversely, the very low field presence of doctors meant that more junior East Timorese health professionals (especially nurses and midwives) were put in charge of health facilities, which then fueled the need to quickly produce paramedical level personnel resulting in the current over-capacity." The departure of Indonesian senior management and of at least 130 of 160 doctors also meant that a different health system model would be required in an independent East Timor. 5.18 The transitional administration and donor community have accorded very high priority to human resource development. It was recognized that the low educational and health coverage and attainment were due to previously low levels of public expenditure on education and health, inefficiencies in its allocation, and an inability of households to bear the private costs of education. The need to make unscheduled fee payments for health care services and a lack of trust of the population in the Indonesian dominated health care system probably also account for the low utilization of government health facilities. As indicated in Table 5.1, expenditure on education and health now compares favorably to countries with similar income levels to that of East Timor. Education 5.19 The core of the transitional strategy for education was to: (a) restart schooling as fast as possible; (b) restore educational infrastructure which would be needed for a newly revamped post independence education system; and (c) begin dealing with a wide range of strategic policy and delivery issues confronting the education system including the need to deal with three competing objectives: (i) expand primary school enrolments to embrace those children not attending school (one quarter of the total who were disproportionately rural, female and urban poor) while judicially restoring capacity at higher levels of education; (ii) improve the quality of education (including a reform of the overcrowded curriculum while also introducing a new language of instruction); and (iii) establish a sustainable public financing framework for education. 5.20 Wllile data and management information remains problematic in the confusion of the transition, the restoration of scllooling was spectacularly successful as a consequence of the rapid initial response of international agencies (particularly UNICEF), and the international community more generally in the subsequent support of UNTAET efforts to reestablish scl~ooling.Estimates of key education statistics are presented in Table 5.2 for 1998 and 2000 to coinpare the situatioil before and after the referendum. Table 5.1: Cross-Country Comparisons of Per Capita Spending in Health and Education GNP per capita. health spending per education spending per $us capita $US capita $US East Timor (CFET only) 400 9.5 21.5 Higher income countries Philippines 1050 Georgia 970 Sri Lanka 810 ' China 750 CBte dlIvoire 700 Indonesia 640 Cameroon 610 Similar income countries Guinea Azerbaijan Pakistan .. Armenia India Vietnam Uganda Nigeria Lower income countries Cambodia 260 1 8 Rwanda Nepal Burundi 140 1 6 Ethiopia 100 2 4 Source: UNTAET/World Bank background paper for Canberra Donors' Meeting on East Timor Table 5.2: Key Education Statistics Before and After the Referendum (Public and Private) Teachers Of which A11 ET Student/ Student/ Student1 Schools Students Total Timorese Teachers Teacher ET Tea. Teacher 1998 2000 1998 2000 1998 1998 2000 1998 1998 2000 Kindergarten 64 na 4,502 na 330 na na 14 Primary 788 674 167,181 188,900 6,672 5,172 3,000 25 32 63 Junior Secondary 114 103 32,197 29,586 1,963 65 884 16 495 33 Senior Secondary 37 34 14,626 10,007 1,059 87 583 14 168 17 Vocational Second 17 na 4,347 na 478 55 na 9 79 na Other (incl. Univ.) 7 1 5,166 na 338 128 na 15 40 na Total 1027 812 228,019 228,493 10,840 5,507 4,579 2 1 41 50 Sources: The figures in this table are drawn from a variety of sources, including the report of the Joint Assessment Mission November, 1999 and East Timor in Figures 1997 (1998) for pre referendum statistics, and the estimates from District Education Office in East Timor for data after October, 2000. 5.21 A number of points arising froin the transitional strategy are woi-thy of mention. First, aggregate enrolments for the systein as a whole were restored quickly with priinary school enrolments being considerably above pre-disturbance levels. Nevertheless enrolments in junior secondary and senior secoildary schools have yet to reach pre- disturbance levels. This was to be expected because of both qualified teacher supply constraints at the secondary level and the fact that secondary school buildings require significant investments to rebuild. 5.22 Second, disaggregated primary school enrolment data suggest the increase has been experienced in all grades. There was an increase in grade 1 enrollment-of about 50 percent-due in large part to the policy of making all first graders of 1999 to repeat grade 1 in 2000. While field surveys suggest this policy was not always observed, there does appear to have been a real increase in grade 1 enrolments, and net enrolments in grade 1 have dramatically increased. It is also hypothesized that the abolition of fees and the requirement to have a school uniform has reduced the private costs of education, eliciting a positive demand response. Further, the transitional administration has also tried to improve the teaching environment by ensuring quality enhancing inputs are supplied including textbooks and other learning materials. This contrasts strongly with the past. Under the Indonesian system there was a rapid decline in enrolments in the upper grades-in part attributable to constraints in school places at the post primary level -but also due to internal inefficiencies. It is possible that net enrolments rates may now have reached into the 90 percent range-a remarkable achievement if it has been achieved and can be sustained. 5.23 Third, it is conceivable, as a result of these achievements by the transitional administration, that there will not be a need to significantly expand aggregate primary school capacity (and thus the number of teachers, unless decisions are taken to reduce the student-teacher ratio - see below) beyond the rate of student population growth. Two caveats to this are noted. School mapping exercises currently nearing completion may identify areas still in need of services. And aggregate capacity will need to increase if there continue to be significant numbers of student repeaters in each grade. Management infoimation will need to be collected, carefully monitored and policies developed to improve the internal efficiency of schooling. 5.24 Fourth, a major effort was made to eliminate one of the most glaring inefficiencies of the old Indonesian system. There was a dramatic increase in the student to teacher ratios while the remuneration of teachers (and other civil servants) was increased compared to Indonesia11 levels. This enabled the number of teachers to be reduced and their pay increased while still reducing the aggregate payroll. Initially, primary student-teacher ratios were increased from 25:l to 63:l; junior secondary ratios froin 16:1 to 33:l; and senior secondary froin 14:l to 17:1. For the system as a whole, the number of teachers was decreased from 10,840 to 4,579. Primary school teacher numbers were reduced from 6,672 to 3,000. The rapid enrolment response was a dramatic success for the transitional administration. But it was recognized that the student-teacher ratios-particularly in piiinary schools-had become too high to sustain a quality education and that more teachers would be required at significant budgetary cost. Thus the 2000-01 budget provided for an additional 1,000 teachers designed to reduce the student teacher ratios particularly at the priillary level, and the 2001-02 budget also provided a modest increase in teacher numbers. 5.25 Fifth, East Timor faces a shortage of trained teachers-particularly at the junior and senior seconda-y level where only about 5 percent of teachers were East Timorese. A serious effort was made to select teachers from the pool of East Timorese primary teachers on the basis of a national test of knowledge of curriculum content and pedagogy. At the high school level, skill shortages have dictated that teachers be selected from interested university graduates, individuals who have completed a substantial proportion of a university program (5-6 semesters) or primary school teachers who did well in the primary school test. Not all vacant posts are filled yet due to supply constraints. Teachers are the single most important factor to determine the quality of education so it is important that the best possible teachers are hired. The skill constraints strongly suggest that in-service training programs and additional new teacher capacity-particularly for secondary schools will be required. It is probable that additional resources will be needed for this: teachers account for more than 50 percent of public sector employees and their wages claim 75 percent of public recurrent expenditure on education. Thus, their productivity and the efficiency in deployment will have a significant impact on the quality of public expenditures overall. Health 5.26 The core of the transitional strategy for health was: (a) the immediate re- establishment of essential health care provision-particularly at the sub-district level where most people could access services; (b) to re-establish essential inpatient services (hospital care); (c) re-establishment of a central health authority that would become the country's Ministry of Health; and (d) health human resource capacity building. Most importantly, it was recognized from the outset by both East Timorese authorities and donors that there was a need: (i) to develop a new health policy quickly to provide the framework for the sustainable delivery of health services (and to ensure that reconstruction of a revamped health system would not preempt necessary health reforms); (ii) for managerial and technical inputs (doctors in particular) to be provided from external sources during the transition; and (iii) for donors to work with government within the framework of a sector-wide approach to health programming and financing. The sector-wide approach was designed to work at two inter-related levels: (a) to permit the definition of needs across-the-board-service delivery, rehabilitation of the health infrastructure, health systems and health policy development; and (b) to coordinate donor financing of the sector and support from UN technical agencies in support of the agreed program. 5.27 Consistent with the transitional strategy, initial effort focused on: (i) improving access to preventive services and to curative care at the first level-including support for rehabilitation and reconstruction of damaged infi-astructure which was located in appropriate populatioil centers; (ii) re-establishing only five of the approximately twenty hospitals under the Indoilesian administration strategically placed across the country- albeit on a temporary basis and with significantly reduced bed strength while a hospital reconstruction strategy was developed; (iii) establishing an autonomous medical store to manage the purchasing and logistics of divg and key supplies for the health system; and (iv) establishing a health policy developinent process within the embryonic Division of Health Services. 5.28 From a programinatic perspective, emphasis was place on preventive health programs-particularly activities designed to increase the knowledge of the population about personal health care and services that would be provided by the health system. The strategy provided for acceleration of selected high priority activities, including immunization, TB control, basic clinical care (including treatment of malaria), and development of a referral mechanism for emergency care including for difficult births. It was to be delivered through strategically placed and rehabilitated health facilities and mobile clinics and integrated into a basic package of services4rawing heavily on the technical expertise of the UN technical agencies (particularly WHO, UNICEF and LWPA) and a few specialized NGOs. 5.29 Implementation of this transitional strategy was, at least initially, based on a strategic agreement with a number of international NGOs which entered East Timor during the emergency period (most with financing from ECHO of the European Commission) to stay and take the initial lead for service delivery at the district level and for management of the hospitals. Staffing of the health facilities was through the payment of stipends from the CFET budget to a maximum of 1087 East Timorese staff pending formal recruitment of staff by the DHS. As with education, the staff ceiling set dramatically reduced numbers of staff within the revamped health system and average salary levels were also increased. Adequate provision was also made for essential drugs and equipment within the CFET operating budget from the outset. 5.30 The FY02 combined sources budget and the fonvard projections for the operating budget (CFET) for the period to FY05 has made a number of important steps towards placing the budget within a medium-tern public expenditure framework despite the uncertainty facing both expenditures and revenues. For the social sectors in particular, the projections of the CFET operating budget tries to take into account the fact that: (i) the population is expected to be growing at around 2 percent per annum; (ii) some allowances have to be made for capital equipment and maintenance programs for the infrastructure being put in place as part of the reconstruction effort; (iii) there are a range of essential and continuing expenditures currently being fi~ndedfrom other donor financed budgets (including TFET, bilateral projects and technical assistance programs) and (iv) some adjustments are needed to current staffing levels to improve the effectiveness of service delivery and, in the case of education, an expanding post primary system. 5.31 There remains some uncertainty, however, as to whether all the medium-tern pressures on the education and health budgets-which constitute 38 percent of the total CFET budget of $65 million-have been captured as discussed below. This raises an important point about the future development of the budget planning process. It is fundamental to a sound public expenditure planning process that there is a mechanism which eilcourages priority setting and budget sciutiny-both by the line agency (e.g., educatioil or health) and by the central agencies (those agencies responsible for finance, plailniilg and deciding persoilnel policies). It is also iinportailt that there is a close professioilal working relationship between the central agencies and the line agencies in this process, and that the budgeting process does not become incremental or an arbitrary bargaining process. 5.32 There needs to be an agreement that resources-both capital and recurrent-will be allocated to meet the objectives of specific agreed policies. The starting point is that the costs of implemeilting a specific policy in the most cost effective manner needs to be determined as a basis for a decision on whether or not to include the policy or activity within the budget. The budget process should thus focus on the implications (outcomes) of doing more or less in support of the costed policy and not on arbitrary "across-the- board" cuts (or additions) to specific inputs (e.g., staff, text books, pharmaceuticals etc.). A number of conditions need to be met to enable such a process to work: Line agencies together with central agencies need to have an agreed sector strategy which sets out program priorities (e.g., commitments to universal basic education, and the priority for secondary and tertiary education; or public health programs and the priority of curative hospital care) within a resource envelop that is feasible and sustainable. The broad resource envelop needs to be determined by the central agencies in consultation with the line agency and agreed and understood by all parties as a mechanism to encourage priority setting-initially by the line agency. In the case of East Timor, with multiple sources of funds from a range of donors, it is important that donors support the government's sector strategy. While this is a negotiated process, it is crucial that the full costs of the policy or program- capital which may be financed by the donor, and recurrent which will eventually become the responsibility of government-are reflected fully in current and forward budgets ,as part of the approval process. A failure to do this can significantly distort sector priorities. 5.33 As discussed, the health program, with agreement of all donors and the transitional administration, set out from the outset to adopt this "programmatic approach" in the belief that a program which integrates all sources of financing and has a single set of service targets across all sources would improve the effectiveness of expenditures. Certainly, health sector management has found this a very useful management and planning tool. Donors have worked cooperatively with government to develop and implement a transitional strategy which required a major reduction (about $6-7 per capita) in operational health expenditures while also focusing on the strategic health policies and priorities being developed as part of a national health policy development process. The education program, through TFET, now supports a policy development process which can fonn the basis for the establishment of a sector wide or programmatic approach. The recent establishment of the Planning Commission can also facilitate such an approach under government leadership. 5.34 There is also a need to focus on the whole resource inanageinent process-not just the budgeting process. Agencies need to make and iinpleinent regular activity plans to implement agreed policies, ensure the timely resourcing of these activities, inanage fund flows and cash, and ensure that accounting and auditing mechanisms and procurement processes are in place. This has received less attention in the past than the budgeting process. However, sound public sector management requires clear and transparent accountability mechanisms. This includes consolidated and audited expenditures which should be presented to Parliament annually and be subject to open discussion and review. This would typically require departmental or Ministry public accounts committees to review financial management procedures and outcomes and to be responsible to Parliament for the use of expenditures. 5.35 This will require skilled staff in both the central and line agencies and the cooperation of donors. At a time when inappropriate decisions by individual line agencies or donors have the potential to seriously distort domestic resource allocations (e.g., establishing a new hospital or expanding a university program outside of the agreed policy framework), planning, budgeting and financial management skills are in short supply. Significant attention needs to be given to developing this capacity; much of this will need to be learned through experience. Even in the Ministry of Health, which has a reputation of having handled these problems better than most, strategic management of the budget process was largely in the hands of expatriates, despite significant efforts to include East Timorese in the process. 5.36 East Timorese health program managers, as with the majority of managers in the civil service, are new to management. They have been focused, appropriately, on developing and managing the day-to-day implementation of rolling three month work programs, and, in the context of rapidly changing circumstances, constantly need to adjust activity plans. In the short term, this will be the reality of management so it is also important that the system does not plan itself to death. One option is to focus planning efforts initially on those areas of government which claim rapid increases in domestic resources or are in receipt of significant donor-financed project assistance. Medium-Term Issues Facing Education 5.37 The transitional administration, with the support of donors, has accorded very high priority to the public finance of education. The 2001-02 combined sources budget for education anticipates expenditure of $54.6 million which represents about 19 percent of total public expenditure. This represents an increase of 18 percent over the 2000-01 all sources budget for education. The expenditure is to be financed from: (i) CFET funds for the operating budget of $16.9 million (31 percent of the total) up 26 percent from $13.4 million in 2000-01; (ii) TFET funds of $13.8 inillion (24 percent of the total) essentially for the reconstiuction and development of the primary and junior secondary education system; (iii) bilateral funding of $23.7 million (43 percent of the total) three quarters of which is for tertiary level training including overseas scholarships and university development ($17.5 million or 32 percent of the total expenditure on education); and (iv) $1 million from the UN Assessed Contribution (2 percent of the total). The overall pattern of bilateral donor expenditure in support of education-with a strong bias in favor of the tertiary sector for 2001-02-was similar in 2000-01, and prospective programs show no signs of change. 5.38 One of the ci-ucial policy questions is how fast to expand each level of education (primary, secondary, tertiary), keeping in mind the overall resource constraints facing the country, the severe qualified teacher supply constraints (particularly at the post primary level) and the relative benefits of expanded education to the economy, society and individuals. International evidence shows that the social benefits of a universal basic education of 9 years (which in East Timor rneans primary and junior secondary education) are very high-typically higher .than any other investment by government. A literate and numerate society and workforce is better able to adjust to changing economic circumstances and increase productivity-even in traditional agricultural occupations. More generally, a wide range of other social benefits accrue to society from universal basic education through improved health and healthier behaviors which result in lower infant and child mortality, increased birth spacing, smaller families and lower maternal mortality rates. Individual "voice" in society is also increased. A literate society is better able to participate in its own governance and the development of democratic institutions. A commitment to ensuring that all children complete at least primary school should therefore be the number one education priority, and is achievable in the first years of independence. 5.39 It will be more difficult for East Timor to achieve universal junior secondary education in the medium term. This would imply a capital investment of approximately $30 million over and above currently financed programs and an additional $2.5 million (assuming a student teacher ratio of 40:l) in operating budgets above the current $2.7 million. Moreover, the binding constraint is likely to be the supply of trained teachers. When combined with the demand for health human resources and other civil servant positions, there may be a requirement for a modest expansion of both junior secondary and senior secondary high school capacity to supply the inputs to the training programs required. 5.40 Subject to quantification of the likely annual demand for senior secondary school graduates (for direct employment and for future studies) it is probable that maintaining the current continuation rates between junior secondary and senior secondary schooling in the medium term would be appropriate. This would entail modest expansion of senior secondary capacity and cost about $5-8 million over the next five years and an additional recurrent cost by the end of this period of $0.6 million per annum. Expansion of senior secondary education in the short run is, however, not a priority. More important will be efforts to increase the skills and competencies of the teaching force. 5.41 Tertiary education, in all its forms, provides East Timor with its biggest education challenge. The private returns to higher levels of educatioil in East Timor (as in most countries) are very high. One World Balk estimate for the private return to tertiary education in East Timor compared to "no education" was over 100 percent immediately prior to the disturbance^.^ Subsequently, skilled wages have substantially increased and Project Appraisal Document of the Fundamental School Quality Project, September, 2001 Annex 2. 65 East Tinlor can expect to face large skill deficits for some time as a consequence of the departure of skilled Indonesians. This will place continued pressure on skilled wages and mean that private retuins to education will have increased significantly and will be sustained for some time, aided by the abolishment of school fees. This will create a strong private (and political) demand for access to higher and higher levels of education. The reiiltroduction of fees-at least for families who can afford them-is therefore warranted. 5.42 The social returns to tertiary education (i.e., the benefits to the economy as a whole) in the medium to longer term will, however, be much lower than the returns to primary and junior secondary education. The social returns to education in East Timor prior to the disturbances were estimated to be 7 percent for primary, 10 percent for secondary (including senior secondary) and oilly 2 percent for tertiary education. These accord with the relative patterns of social returns to education observed internationally. 5.43 In the short run, however, East Timor faces severe skill shortages. As a consequence, short-run social returns to tertiary education are likely to be very high (as one of the major benefits accrues from "savings" on expatriate salaries compared to the domestic wage structure). However, these short-run social returns to tertiary education should not be used to rapidly expand tertiary education. There is a need to take a longer- term perspective-at least 10 to 15 years-to justify the level of expansion of tertiary education. Onejustification for aid, in the form of technical assistance, is to enable short- term peak demands for skills to be met while sustainable and demand-driven capacity is developed. 5.44 Tertiary education however covers a wide range of skills and levels of education (degrees, diplomas, certificates, professional and vocational skills etc.). East Timor needs to make decisions on how best it can meet its economy's skill requirements. This will require an overall human resource assessment to examine the demand side and an analysis of cost-effective options to ensure sustainable supplies for each group of skills. It will not be possible, at least in the short to medium term, for East Timor to develop all the required training capacity domestically for a range of reasons including supply constraints for trainers and teachers, and the cost of providing some categories of graduates. These two constraints will be further exacerbated if the language of instruction is changed from Bahasa. 5.45 The 2001-02 and subsequent CFET budget now includes a significant domestic budget for tertiary training. It provides for a capacity of 5,000 students at a unit cost of $440 per student. While this is an average of a wide variety of programs there will be a wide variation of costs between programs. It is very unlikely that a quality university education can be delivered for anywhere near this cost. For example, the average unit costs at the University of the South Pacific are at least 4-5 times this average; the costs at the University of Papa New Guinea are even higher. Given its small population and labor market, East Tiinor should thus be wary of developing a university offering a full range of programs. 5.46 Donors are major financiers for: (a) the reconstructioil of physical facilities for tertiary institutions; (b) crash in-service training programs; (c) technical assistance; and (d) scholarships for overseas training. This amounts to a sizable proportion of total public expenditures on all levels of education, represents a very large public subsidy to relatively few students, and hence warrants careful analysis of its cost effectiveness. Further, there is a need to ensure the specific investments being made are: (a) supporting the optimal mix of skills required by the economy; and (b) that "all the costs of policy" as discussed above, including the long-run costs of technical assistance are clearly identified before new programs are established. Only if this is done can there be any confidence that future budgets for tertiary education are sustainable. In a situation of severe skill constraints, it is not possible to judge whether individual training programs are sustainable from the perspective of trainer supply constraints without looking at the aggregate demand and supply for each set of skills. This reinforces the need for an overall human resource assessment. 5.47 This discussion has highlighted the fact that East Timor faces some difficult choices on both the levels and amounts of education to provide-and to allow donors to finance. First priority should go to primary and junior secondary education while higher levels of education are judicially expanded in response to clearly identified skill needs in the economy. It is likely that teacher training and in-service skill upgrading will yield h g h returns in the short to medium term. Decisions on significantly increasing unit costs within the education system to improve quality-particularly efforts to reduce student- teacher ratios need to be taken with great care as to the benefits compared to the costs. Once again, "whole of policy costing," and comparing these costs to benefits will assist in this regard, and should become standard practice prior to budget approvals for policy changes. 5.48 Clearly, generous donor support for the sector is required but it is also important: (a) that it is optimized to meet the immediate skill and educational development needs of the economy; and (b) domestic education and training capacity is developed consistent with the long-term needs of the economy and in a manner which is sustainable. Considerable donor resources are also being devoted to overseas scholarships. Every effort should be made to ensure this is done as effectively as possible. Quality "third country" programs in many cases, in part because students are less likely to migrate permanently to such countries, may be more appropriate and less costly. Finally, if overseas migration of internationally trained East Timorese staff becomes a significant problem-as it has for many parts of the Pacific Islands-this will have to be reflected in cost-effectiveness calculations of such training. Medium-Term Issues Facing Health 5.49 The transitional government and donors have accorded the reestablishrneilt of health services high priority. The combiiled sources budget for 2001-02 allocates $27.0 million to health equivalent to $32 per capita or 11 percent of pla~lnedtotal public expenditures. This represents an increase of $7 million or about $8 per capita over planned expenditures for 2000-01. The expenditure is to be financed from: (i) CFET f ~ ~ n dfor the operating budget of $7.6 million, 29 percent of the total (equivalent to s almost $9 per capita); (ii) TFET funds of $10.0 million ( 37 percent of the total) for recoilstructioll and development of health facilities, policy and progranl develop~nentand support to district health services; (iii) bilateral funding of $8.6 million for health related projects and programs (33 percent of the total) ;and (iv) a UN Assessed Contribution of $0.8 illillion for technical support of the administration (1 percent of the total). 5.50 The Divisioil of Health Seilrices (now the Ministry of Health) made significant advances in the development of a program budget for 2001-02 consistent with the principles of public expenditure planning outlined above. However, more still needs to be done to further integrate bilateral projects into the framework and to more clearly identify the sustaining costs of these projects which will need to be financed by the operating budget in future years. The Ministry is responsible for providing core health services including the support of ongoing service delivery, improving the range and quality of services available, and developing health sector policy and supportive legislation. The Ministry has developed a three "program" budget to: Support ongoing service delivery: to ensure reliable access by the population to a basic package of services to be implemented through District Health Plans (13) and supporting established hospitals (4) and laboratory services; Improve range and quality of services: to ensure adequate functioning of central pharmacy and blood bank, develop range and improve quality of basic services; take over responsibility (from WHO) for communicable disease surveillance and develop communicable disease control and health promotion activities; Develop and implement policy and management systems: to ensure a technically sound health policy with supporting legislation, develops and plans the human resource capacities, administrative and monitoring systems, and an adequate asset replacement (fleet and equipment) program. 5.51 The Ministry, consistent with its efforts to focus on outcomes, and to direct resources-human and financial-in support of agreed programmatic objectives, has established monitorable indicators for each sub-program and project. Each district, initially with the support of International NGOs has developed district plans which are designed to focus the health effort on the core health programs. The second TFET project structured its project components and sub-projects within this framework and a matrix was created for the first project-which was established prior to the budget structure. 5.52 Notwithstanding all the attempts to develop a program budget for health and a health program focused on outcomes, a range of issues continue to confront the health system which will need to be the focus of attention over the medium term. These include: A need to finalize the medium-term hospital development strategy consistent with the agreed strategy to contain hospital costs to below 35-40 percent of the total health budget. Thei-e are thee significant pressures on the hospital budgets stenlining from a desire to: (i) establish more hospitals to deal with problems associated with travel time to hospitals, rather than focusing on strategies to get referral systems working; (ii) establish more bedded facilities without reference to the problenls of adequately staffing them; and (iii) increase staff per hospital bed. Attention needs to be given to arrangements for hospital management to ensure they are managed efficiently. Optioils include the establishment of hospital inanagement boards and professionalization of management. a The need to finalize the transition of responsibilities for district health programs and the inlplementation of district health plans to district health management teams. The TFET program has absorbed the major costs of international IVGO support to the program-including the supply of 30 doctors to be recruited from relevant developing countries overseas. This is within a framework which is reducing the costs of delivering district health services by $6 per capita over two years to levels which are both sustainable $10-12 per capita and are considerable above average low income country levels of $3-4 per capita (Table 5.1) . The ongoing costs of this for the operating budget will perhaps be in the range of $1-2 per capita by the end of the TFET program ($1-1.6 million annually)--depending on how many East Timorese doctors can successfully complete their studies and return to East Timor. The need for an urgent effort to finalize a human resource development strategy including the development and implementation of a nurse practitioner cadre to partially replace some of the activities previously done by doctors. This would reduce the number of doctors that are absolutely required to sustain adequate medical services in hospitals. It will also contain sustaining costs for overseas doctors. A need to continue to refocus resources and efforts on those elements of the core health care package which will have the greatest impact of health outcomes (immunization, communicable and vector-borne diseases including malaria, TB and HIV/AIDs, child health, reproductive health and preventive health programs). As part of this effort, it is important to develop an improved understanding of health seeking behavior. Available data suggest average per capita visits to health facilities has almost reached pre-disturbance levels. But health outcomes suggest that needs should be higher than utilization data imply, and the current capacity of the system is certainly well above current demands. A need to finalize the health policy framework including the development of a pricing policy which takes account of the needs of the poor. 6. TIMOR SEA REVENUE: ISSUES OF MANAGEMENT AND GOVERNANCE 6.1 The prospect of significant oil revenue is a blessing for the nascent economy of East Timor. The new government has few resources to fund its development needs and provide crucial public goods and services. At present, it is almost totally dependent on external donors. Oil revenue from the Ti~norSea can help it to contaiil the tax burden 011 a poor population and reduce-or even eliminate-its dependence on aid or external borrowing. However, if experience from other countries is anything to go by, this happy prospect could easily sour. Resource-poor countries have consisteiltly outperformed resource-rich countries in terms of growth and poverty reduction. This is especially true for oil exporting countries.' The challenge for East Timor is to avoid the errors of so many oil exporters and ensure that its oil wealth becomes a blessing for all its people and indeed for future generations aS.wel1. This chapter presents an overview of lessons from managing expenditure, revenue and savings from oil and gas flows that stem from the experience of other countries but may also be relevant for East ~imor.' The Problems with Oil 6.2 There are many explanations of why oil (and more generally mineral) economies suffer. Most of these explanations are in some way related to the volatility of oil revenue and to fiscal mismanagement. Mismanaging wilzclfaalls.There is tendency to over-spend during booms caused by rising oil prices, often leaving the country even worse off than it was before the boom, when the boom tunls to bust. One reason for this is that as spending rises during the boom (and especially government spending since the government is usually the main direct beneficiary of the boom), non-traded goods prices rise. This draws resources from non-oil traded goods sectors (including agriculture) or slows the increase in resource flows to these sectors. When the boom is over it is extremely difficult to resuscitate or restore growth to the affected sectors. (This is frequently called the Dutch disease.) Absorptive capacity. Many governments believe that increasing public investment will help avoid the Dutch disease by increasing productivity. However, they frequently do not have the capacity to manage the increased investment efficiently I See Auty and Miksell, "Sustainable Development in Mineral Economies", 1999; Sachs and Wanler, "Natural Resource Abundance and Economic Growth," 1995; Gelb et al, "Oil Windfalls: Blessing or Curse," (1988). Reference to oil and gas in this chapter is frequently shortened to oil alone. and therefore much of the investment is simply ~ a s t e d .Also, the investments ~ are often not chosen to ineet the countries' developnlent needs. Volatility. The volatility of oil revenues often creates an unstable, stop-go, patte~n of government investment and expenditure. These cyclical patterns will affect the entire economy and inay make it inore prone to business cycles. Goverrzarzce. Corruption and poor governance seem to be more corninon in oil exporting c~untries.~This too is the result of the concentration of the rents in the fiscal authority. In a nutshell, the presence of oil rents tends to cause an over- expansion of government leading to rent seeking and corruption and, consequently, to the misallocation of resources. 6.3 These problems are all relevant for East Timor; a key challenge for the government will therefore be to ensure that the oil windfall is not squandered and indeed is wisely invested or saved.5 6.4 The key to avoiding or at least minimizing some of the effects of the above resource or oil-associated problems is to "smooth" expenditure. This will insulate (or sterilize) the economy from the deleterious effects of the "boom", and also help to contain the shift in resources towards non-traded sectors. The benefits to the economy from its natural resources would derive from the income generated by the saving. These would be permanent, stable benefits and could be used to benefit the population as a whole-and, importantly, provide a means of sharing the proceeds of the offshore revenue between current and future generations. A related issue concerns the nature of the investments. Frequently, governments use the oil revenues to establish inefficient state enterprises which can be nourished so long as the boom continues but become unmanageable burdens afterwards. Similarly, oil revenues are often used to promote so-called downstream petrochemical industries. However, the presence of oil and gas does not necessarily mean that a country has a comparative advantage in petrochemicals, which tend to be highly capital intensive. Hence, this kind of industrial policy often turns out to be a huge mistake and an unsustainable fiscal burden. See Kaufmann et al, "Governa~zceMatters ", Finance and Development, June 2000. Also, oil exporting countries tend to be clustered in the bottom one-third of the countries listed in Transparency International's ailnual Corruption Perceptions Index. East Timor's leaders have recognized both the promise and the problems of the oil revenue: "...(oil revenue) can mean the difference between one teacher for every 100 students, and one for every 40. It can allow the country to extend basic health care, such as malaria prophylaxis and immunization of children, to its people. It will not make East Timor rich. However, if the money is well spent, it will give the people of East Timor the opportunity to escape the grinding poverty that is the legacy of occupation and war." (Mari Alkatiri and Peter Galbraith in Tlze Age, July 7, 2001. Jose Ramos-Horta strikes a cautious note: "The danger of wealth, the danger of the dollar, the petro-dollar - it is obvious. It is clear that we must have solid and democratic institutions. We must work with great transparency and with a great sense of responsibility, otherwise we could have all the oil in the world but the country would still not develop. There are so many examples of this." BBC Monitoring Service, July 5, 2001. In the same interview Alkatiri added: "Until a system is set up to ensure that the money, these resources, will indeed benefit the whole population and not just some bank accounts, it will be a good idea to take great care. And above all we must develop a transparent, solid, effective system to absorb this money and spend it positively to develop the whole country." 6.5 One can thi~lkof the oil as "capital". The consuinption stream (i.e. the benefits) that is derived from the oil call oilly be realized by its extraction and sale. When that happens the stock of "capital" is gradually depleted and so over time it will generate less and less consumption. To ensure a smooth consumption stream it is necessary to replace the depleted "capital" with physical or financial capital. This has been called replacing the assets below the ground with assets above the ground.6 This means calculating the maximum level of expenditure that can be sustained in perpetuity. 6.6 A fixed expenditure, based on pennanent rather than transitory income, will prevent any precipitous rise in domestic prices and an unabated decline of the non-oil traded goods sectors and industries. Because the East Timor economy has been so ravaged by conflict, even more of a concern than the "decline" of the traded goods sector is the slowing of the flow of resources to the traded goods sector, which is crucial to the development of the non-oil economy. Stable government expenditure would also contribute to a more stable and predictable private sector investment climate. Of course, pennanent production and price shocks or new resource discoveries could necessitate adjustments to the permanent level of expenditure. However, these adjustments will tend to be more moderate if the savings strategy is maintained.' 6.7 An expenditure smoothing strategy could also help match the level of public investment to the government's absorptive capacity. One could argue that investment in physical assets such as infrastructure also fulfills the requirement of replacing the assets below the ground with assets above the ground. The constraints, however, are the absorptive capacity of the economy, the recurrent costs of maintaining the physical assets, and the need to maintain or develop good governance.8 6.8 Avoiding a rapid build-up in government expenditures can also contribute to maintaining good government. Although this is neither necessary nor sufficient, rent seeking and corruption are far less likely if "expenditure smoothing7'rules are invoked. Government would remain more limited and investment would be built-up gradually thus lowering the benefits to rent seeking and corruption. At the same time, independent efforts at maintaining fiscal transparency and accountability will be required. Indeed, there is a "chicken and egg" element to this: without good government it is unlikely that the government will be able to sustain a prudent expenditure policy; but, without a prudent expenditure policy good governance is very unlikely. 6.9 These factors provide strong reasons for utilizing an expenditure-smoothing rule once oil revenue exceeds the sustainable level of spending. They do not, however, - See Corden, W. Max, "Booming Sector and Dutch Disease Econonlics: Survey and Consolidation," Oxford Eco~io~~iicPapels, 36 (Nov. 1984). 'A stable nominal expenditure level would mean a decline in per capita ternx and in terms of income, as the economy grows. This is appropriate as the development process, by definition, involves the accumulation of productive assets for the benefit of future generations. Thus, the assets accumulated from this windfall will become increasingly insignificant as time goes by. Collier (2001) suggests that aid flows to East Tinlor may already be above levels that call be efficiently absorbed. Excessive spending from oil revenue could also be inefficient and bequeath unsustainable subsequent spending commitments. See "East Timor: Developmeilt Challenges for the World's Newest Nation", ISEAS. preclude the need for judgment. Importantly, the prospective revenue flow will continually be subject to revision due to fluctuations in prices and production capability (production may also be positively correlated with price, adding to volatility), and the resulting adjustments in the evaluation of "permanent" income. Moreover, certain public investments may be lumpy, with rates of return that are time sensitive-which could warrant an exception to the agreed upon spending rule. 6.10 Decisions on smoothing rules could also range from: conservative such as maintain a constant expenditure in perpetuity in real per capita terms; to the simplest such as adhering to constant nominal expenditure in perpetuity; to those which discount future spending. Selecting an expenditure-smoothing rule therefore involves thorough debate of development objectives. There is, however, a simple argument for favoring a fixed dollar amount of expenditure: because that is the "permanent" income. If there is dollar inflation of x percent, then real income from the oil has gone down by x percent if the price of oil or proven reserves has not changed. If the price of oil changes, then, as noted, the estimate of "permanent" income would change commensurately, warranting an adjustment to prospective expenditure. MANAGEMENTREVENUE OF AND SAVINGS Revenue Uncertainties 6.11 On July 5, 2001, representatives of LJNTAET and Australia agreed to an arrangement on the division of oil and gas resources from the Joint Petroleum Development Authority (JPDA) of the Tilnor Sea. Under the so-called Timor Sea Arrangement (TSA), which is subject to ratification by the East Timor and Australian parliaments, 90 percent of oil and gas revenue in the JPDA would accrue to East Timor, with Australia receiving 10percent (excluding revenue from related onshore investments in ~ustralia).~The Bayu-Undan liquids project was approved in 2000 and is due to begin production in 2004. Full gas development at Bayu-Undan depends upon Australian regulatory approval of parts of understandings on fiscal terms reached between East Timor and Bayu-Undan contractors in December 2001. Greater Sunrise is another large discovered gas-condensate field, which lies partly in the JPDA as defined in the TSA, but a development concept for this field has yet to be agreed among the joint venture partners. The only fields that have been developed to date (Elang Kakatua) may close in 2002 or 2003. 6.12 In light of the above, there is still much uncertainty about revenue prospects from the Timor Sea, which extends beyond the usual volatility surrounding prices and production. Revenue flow estimates have been subject to significant changes-both in aggregate and in their distribution over time-reflecting the evolutioll of negotiations Oil and gas resources in the Tiinor Sea had been governed by the 1989 "Zone of Cooperation" Treaty between Australia and Indonesia, which partitioned the "Tinlor Gap" into thee areas, of which the JPDA was one. Oil production in the Timor Gap is subject to production-sharing contracts between the commercial operators and the Joint Authority, which was established as an executive organ representing the two countries. In October 1999, UNTAET assumed Indonesia's rights and obligations relating to the ZOC. The Timor Sea Ai-rangelnent is due to apply following East Tinlor's independence. with the coinn~ercialoperators. As a result, it is not yet possible to derive firm revenue projections of the oil and gas resources from the Timor Sea. 6.13 While there may be considerable uncertainty about Timor Sea revenue flows, it is clear that for at least the next two years such revenue under any scenario will be insufficient to finance government expenditure needs. Initial estimates indicate that revenue from the Timor Sea will rise sufficiently by FY06 to cover anticipated public expenditure needs but that large fiscal surpluses are unlikely before the end of the decade. Issues in Allocating Offshore Revenue to Public Expenditure 6.14 Based on currently available information, East Timor will experience a revenue windfall for some twenty years beginning in the middle of this decade. But there is less clarity on the viability of significant new fields, and hence offshore revenue at the end of this period could diminish sharply or cease. 6.15 Under these circumstances, and for the reasons discussed above, a strategy for allocating the revenue stream between public expenditures and saving needs to be developed. There will be ample time to develop such a strategy-given that fiscal surpluses are not expected to emerge until FY06. Nevertheless, once there is greater assurance regarding revenue flows, it will be important to begin assessing alternative options-to infonn the public debate, and to develop the parameters with which to guide public investment decisions, fiscal policy and assessment of financing needs.'' 6.16 A starting point for this analysis is to estimate the constant nominal dollar allocation in pei-petuity to public expenditure that would be implied by a particular revenue stream (given an assumed rate of return on accumulated savings). Additional adjustments can then be made to accornnlodate alternative strategies: for example, a more conservative strategy would be to maintain spending levels constant in I-ealor real per capita terns. Alternatively, a more aggressive spending strategy could also be developed, predicated on the need for greater public investment in the short term. 6.17 Selecting an expenditure strategy should involve a number of criteria, including evaluation of absorptive capacity for implementing new projects and the prospects for donor financing. More fundamentally, the allocation of revenue between savings and expenditure should be linked to an assessment of development potential in the non-oil sectors: if non-oil growth prospects in the long term are assessed as poor or uncertain, a conservative spending strategy may be warranted; if strong growth opportunities were to emerge in the non-oil sector, a case could be made for a more aggressive spending strategy on the assumption that the non-oil economy would provide a stronger base for future tax revenue. 6.18 Any smoothing scenario will need to be reviewed and updated on a regular basis to reflect changes in production, prices and returns on investment, particularly in 10 Based on information received from the CFA following the initialing of the Timor Sea Arrangement, the Bank had outlined a number of illustrative expenditure smoothing options. These are not reported here as the underlying revenue estimates are being revised. response to inajor developillents sucl~as energy price shocks or the discovery of new exploitable resewes. It will also be ilnportant to make available to the public accurate illformation about revenue flows to develop a national consensus behind the savings strategy, and to minimize the scope for diverting resources from public priorities. Institutional Aspects of Revenue and Savings Management 6.19 Good revenue management will not guarantee a stable expenditure pattern but it can help. Oil revenue is typically volatile owing mainly to volatile international prices. If volatility affects expenditure then, as noted, the instability of oil markets will be transmitted to the rest of the economy. These considerations have led many natural resource exporting countries to establish natural resource funds to manage the revenues. There are essentially two main categories of funds (though one could think of combinations and variations of these): stabilization funds and savings funds (or funds for the future.) The former focuses on the problem of volatility and the latter on the questions of exhaustibility1'and "windfalls." 6.20 Stabilizatiorz funds. These funds try to stabilize the flow of revenue to the government by "borrowing" from the fund when the price is low and accumulating funds when the price is high. Thus stabilization funds may have a role where the output stream is a steady-state feature, with income depending mainly on price volatility. But since volatility in East Timor's case has as much to do with volume as price, and the windfall is expected to last some 20 years followed by zero income, at least on current assumptions, the use of a stabilization fund would appear to be less relevant to East Timor's current situation.] 6.21 Savings frirtds. These funds focus on the inter-temporal questions discussed above. Examples include Norway, Kuwait and the State of Alaska. Typically, such a fund would be managed by professionals and invest part of the current revenue offshore in order to sterilize the effect of windfalls on the economy and smooth the revenue to the budget over time. For this type of fund to be effective in smoothing expenditure it is necessary for the fiscal authority to feel bound by the "rules." Thus, if expenditures were to rise with a windfall, even as revenue from the fund to the budget was constant, then the government may have to borrow to finance a resulting deficit. The net asset position of the government would be unchanged, as the increased borrowing would offset the accumulation of assets by the fund (particularly if the borrowing is external, which is the relevant case for East Timor). But the inter-temporal objectives of the fund would be undermined. As in the case of stabilization funds, fiscal discipline is essential to prevent the govemmeilt from "raiding" the fund. Sometimes, the funds are "implicitly raided". This happens when the existence of the fund gives legislators or fiscal authorities a false sense of security and so they tend to run higher deficits than they otl~erwisewould have. " This is somewhat of a simplification. In particular, the savings type fund is also motivated by the problem of limited government investment capacity. It aims therefore to "diversify" the government's assets so as not to over-extend the public sector investment program at the time of the windfall. This is likely to be very relevant for East Tinlor. l 2 A thorough assessment of both types of funds with a discussion of country cases is given in: IMF (2001), Stabilization and Savings Funds For Nonrenewable Resources, Occasional Paper No. 205. In this way they undermine the rules of the hnd without actually dipping into its capital.13 6.22 The savings fund is therefore not a panacea. It can, however, help mitigate some of the problems discussed above in the following ways: It could serve as an institutional mechanism to reinforce fiscal prudence. Although the government can always break the rules, the existence of a find created by it may make it somewhat more difficult to do so. A properly established fund can be an institutional mechanism to enhance transparency. Since the fund can be required to adhere to pre-arranged rules and accountability standards it can serve as a good example to the country and the fiscal authority, as well as ensuring that the resource revenues are well accounted for. An improperly constituted fund could have exactly the opposite effect, however. 6.23 I~zstitutiolzalfrallzework for tlze fund. The establishment of either category of fund also requires an institutional framework to deal with three central aspects of fund management: integration within the fiscal framework, an asset management strategy and insurance mechanisms for transparency and acc~untabilit~.'~Once the decision has been taken to implement a fund, it should be integrated within the budget process to maintain unified control of fiscal policy and avoid the emergence of two budgets, namely the traditional budget and a separate expenditure program financed by the fund. It is important to ensure that spending decisions are taken within the context of the budget and that the expenditure is included in the budget in a comprehensive way so that fund resources are spent through the nonnal budgetary approval process. 6.24 Irztegratiolz into the budgetprocess. Ideally, the fund's operations should be incorporated into the fiscal accounts with transfers to and from the fund requiring parliamentary approval. Proper integration of the fund and the budget helps to maintain a unified control of fiscal policy and avoid problems in expenditure coordination. The find should have no independent spending authority and there should be no targeting of fund resource for certain items of expenditure, ensuring that all revenues and expenditures from the fund are on-budget. In the case of Norway's State Petroleum Fund, for example, transfers to and from the fund need parliamentary approval and the fund's operations are incorporated into the fiscal accounts. In the case of the Alaska Permanent Fund, the principal is the permanent part of the fund which cannot be spent without a majority vote of the state's residents while the income or money earned from the investment of the principal is designated towards inflation proofing the principal, increasing the size of the principal and paying dividends to each Alaskan citizen. Any leftover surplus from the income is put into the Earnings Reserve Account until the state legislature decides what to do with it. " Cordenhasremarkedthatmostcountriestend totreatpositiveshocksaspermanent andnegativeshocks as transitory. 14 This section is based on IMF (2001), op cit. 6.25 Asset Mnrzagenzerzt. Given weak capacity within East Timor, inailaging of the accuinulated savings could be contracted out to two or t h e e top-rated international fund managers, with each assigned to invest half or a third of the available funds with the objective of maximizing the yield subject to carefully chosen investment criteria including prudential investineilt rules, published guidelines of the desired risk-return combination, the proportions to be invested in different types of assets, etc. Performance could be monitored frequently and followiilg periodic reviews, the worst perfonner could be replaced. Box 6.1 provides an example of successful management and investment outcomes for a country with little investment expertise. 6.26 Ideally the asset management strategy for the fund should reflect a consolidated portfolio of the government, with short-term asset operations conducted in a manner coordinated and consistent with the debt management operations of the ministry of finance, the treasury's management of government cash flow and the financial assets already held as part of the government's balance sheet. Box 6.1. The Tuvalu Trust Fund The Tuvalu Trust fund was set up in 1987, under an International Agreement signed by Tuvalu and development partners namely New Zealand, Australia and the UK, who initially contributed to the Fund. Tuvalu Trust funds are invested internationally in a combination of shares, listed property, fixed interest deposits and bonds. There are two accounts: the A account is the growth capital which is managed with the objective of maintaining the real purchasing power of the Fund, with distributions made from the A account when fund growth exceeds the rate of inflation. The B account is a buffer fund of unspent distributions. Since 1996, the B account has been referred to in the Govenlment's Budget as the Consolidated Investment Fund and plays a key role given that the value of the A fund can fluctuate depending on international financial markets. The Tuvalu Trust Fund is managed by professional fund managers in Australia (currently 50 percent with Country Nat West and 50 percent with JP Morgan Chase) with oversight by four Trust Directors, representing the governments of Tuvalu (Chair), New Zealand, Australia and the UK and an advisory Committee;both bodies meet twice a year. Britain for example contributesto the work of the fund by providing two independent consultants to serve on the Board and Advisory Committee to provide advice on and assistance with the operation of the Fund. Monitoring is carried out by consulting actuaries and the Fund is audited internationally. Fund information is readily available to the public. The average rates of return from the A account from 1987-97 was 10.5 percent per annum nominal and 6.5 percent real (though the favorable investment environment during this period should be noted). The Government of Tuvalu has decided to maintain the value of the hnds in real per capita terms, with the implicationthat it will need to continue reinvesting from the funds that are potentially available for distribution from the A account, with implications for the amount of funds available for budgetary support. 6.27 Countries with pressing infrastructure needs may be tenlpted to undertake investment in domestic physical assets as well as to promote the growth of the non- resource sectors. However, there are few successful examples of expenditures of oil money in this regard. Spending of this type tends to rise to unsustainable levels, with rapid increases leading to poor quality projects. Investment should ideally be guided by medium-term recui-rent implications rather than the availability of resources in the fund. Furthermore, this strategy tends to worsen rent seeking activities and make the fund prone to abuse, while making it difficult to achieve and assess real returns to such activities. 6.28 Trarzsparency. Clearly defined rules, purposes and objectives for the Fund are important for improving transparency and establishing the legitimacy of fund operations. Regular and frequent disclosure and reporting on principles governing the fund, its inflows and outflows and the allocation of assets is also recommended, as is regular notification to the legislature and the general public. A detailed annual report can also serve to provide important information such as asset allocation, summary statistics on the performance of the portfolio and retrospective of activities of the fund during the year. The Norwegian State Petroleum Fund and the Alaska Permanent Fund are best practice examples in this regard. 6.29 Transparent management of both funds is accomplished by a number of requirements. For example, the Norges Bank, which manages Norway's State Petroleum Fund, is required by law to provide information concerning the fund's management to the public, and comprehensive accounts and data on the SPF's operations are readily available on the Internet on a timely basis. In the case of the Alaska Fund an annual dividend is paid to each permanent citizen from the earnings of the APF and public oversight of the Fund balances with respect to dividend payments helps to improve transparency. 6.30 Finally, there has been some discussion regarding a proposal to establish a government owned oil company. Before any decision is taken on such a move, its rationale should be carefully considered, given the mixed record of equivalent entities in other countries-including on governance issues. The government clearly needs expertise in keeping track of oil technology to develop an independent view and to be able to negotiate on equal terms with commercial firms, but the creation of a new firm for this purpose may not be warranted. Conclusions 6.31 The key to ensuring that the oil revenue is a blessing will be a prudent fiscal policy-the subject of the next chapter-which can integrate in a sustainable manner all revenue prospects, expenditure needs and fiilancing availability. In this context, the illvolvement of IF1 and donor finance in the medium term can reinforce fiscal discipline. Donors and IFIs should not be put off by the reluctance of the government to spend its oil revenue in the medium teim. On the contrary: the objective of external involvement in the medium tenn sl~ouldbe to ensure that fiscal policy meets the development objectives of the country; and, to allow the government to save a large part of the oil revenue thereby creating a permanent stream of stable income. 6.32 A savings fund may be useful if it can enhance transparency and fiscal prudence. If managed by professionals it could also relieve the pressure on the institutional capacity in East Timor. 7. MAINTAINING PRUDENT FISCAL POLICY 7.1 In the medium-tenll, East Timor's public finances will be heavily shaped by the extent and duration of Timor Sea revenues, which, while subject to coiIsiderable uncertainty, will be by far the largest revenue source. Indeed, once oil revenues begin to exceed public expenditure requirements, the principal public finance issue could become how much of offshore revenue should be spent versus saved. But until more significant offshore revenues do come on stream, the nation will remain heavily dependent on donor financing: in the current 2001-02 fiscal year, for example, projected revenue of $31 million (including offshore revenue) comprises less than 60 percent of Consolidated Fund expenditures and only 14 percent of overall expenditure including spending financed by multilateral and bilateral agencies. 7.2 In such an environment, proper analysis and prioritization of all expenditures, within a sustainable medium-term framework, is particularly important. This will require the development of realistic and costed strategies for each sector as well as budget processes that explicitly address priorities and sustainability-including the recurrent cost implications of past and prospective spending choices. The significant work already done in establishing budget processes and institutions will need to be consolidated. The in-house fiscal management capacity already developed needs to be reinforced and coordinated with the nascent legislative authority to ensure that spending and revenue legislation and policies are compatible with rigorous, transparent and effective economic management and in harmony with the country's economic development objectives. 7.3 Developing robust, non-corrupt administrative systems for collecting taxes is an important discipline to be instilled from the outset. At the same time, the temptation to raise taxes to reduce a massive budget deficit in the short term will have to be tempered by the poverty and lack of purchasing power of the bulk of the population, the paucity of private enterprise, the shortage of skilled administration officials, and the high cost (relative to collection proceeds) of maintaining an effective tax administration. In fact, the imperative to generate jobs and the solid prospect of offshore revenues that, if well managed, can cover the bulk of efficient public spending in the medium term, suggest a tax regime that enhances con~petitivenessand is simple to administer. In practice this inay amount to a regime of minimal taxation of productive activity, at least during an initial phase when the impediments to private investment are significant. 7.4 Following an overview of fiscal developments, this chapter comments on a range of issues with expenditure, tax and financing policy repercussions, and illustrates these issues through the development of a sustainable long-term fiscal scenario. 7.5 East Timor has benefited enormously from the generosity of the donor community. But the multiplicity of channels tl.11ough which aid has flowed has created a major challenge in tenns of consolidating the fiscal accounts and, more importantly, coordiilatiilg donor initiatives to ensure they remain compatible with strategic objectives and sustainability. 7.6 In this regard, the Ministry of Finance and its predecessor the Cenh-a1 Fiscal Authority's commendable efforts to coilsolidate the various accoullts have helped to facilitate a compreheilsive view. Table 7.1 presents a synopsis of the various sources of expenditure and ETTA revenue and Table 7.2 provides a breakdown of the "Combined Sources Budget" which brings together the details of planned expenditure under CFET, TFET, by bilateral donors and for internatioilal staff funded under the United Nations Assessed Contributions budget in ~ ~ 2 0 0 1 - 0 2 .All financing has been in the form of ' grants. Figure 7.1 indicates the projected allocation of total resources by sector, and Figure 7.2 provides the coilsolidated capital spending allocations. These expenditure levels are a modest 14 percent of GDP if only CFET expenditures are included in the total, but jump to 56 percent if expenditures financed by multilateral and bilateral grants are also included. Table 7.1: Expenditure and ETTA Revenue ($m) Total Expenditure CFET TFET Total CFET +TFET Bilateral UNTAETa ETTA ~ e v e n u e ~ 28 31 Non-oil and gas revenueC 15 19 Tinlor Sea revenued 13 12 Source: Ministry of Finance a. Estimate of expenditure in East Timor excluding offshore payments. b. Revenue estimates rounded to nearest million c. Excluding revenue retention by autonomous agencies. d. Includes first tranche petroleum payments and revenue from the existing ElangIKakatua oil field and the Bayu Undan field which is being developed. ' Chapter 8 outlines the budget formulation process. Acronyms: East Timor Transitional Administration (ETTA); Consolidated Fund for East Timor (CFET)--for recurrent-type spending, administered by the UN, with contributions from donors; Trust Fund for East Tinlor (TFET)--for capital projects, administered by the World Bank and ADB with contributions by bilateral donors and the World Bank; Central Fiscal Authority (CFA)--precursor of Ministry of Finance. Table 7.2: Combined Sources Budget 2001-02 ($'000) 21 CFET TFET Bilateral Assessed Total Deputy Transitional Administrator Constituent Assembly Office of the Inspector General Central Administrative Services Central Payments Office Cabinet Secretariat National Planning and Development Agency Transitional Cabinet East Timor Defence Force Police and Emergency Services East Timor Police Service ETTA Security Fire and Rescue National Disaster Management Office Internal Administration Civil Service and Public Employment District Affairs Border Service Civil Registrar National Archives Justice Judicial Affairs Prisons Management Serious Crimes and Prosecution Economic Affairs Agriculture Indushy, Mineral Resources and Tourism Trade and Investment Social Affairs Education, Youth and Cultural Services Health Services Social and Labour Services Political, Constitutional, Electoral Foreign Affairs Central Fiscal Authority Infrastructure Communications District Public Works Land and Property Unit Power Authority Transport Water and Sanitation Authority Aviation Authority Maritime Port Authority Whole of Government and Other TFET CEP SEP Micro-finance Other Bilateral Transitional Administrator Contingency Reserve Unappropriated TOTAL a/ Ex antc FY02 presentation. Figure 7.1: Expenditure allocations, 2001-02 H e a l t h E d u c a t i o n . Y o u t h 8 1 1 % lnfrastructure Cultural 21 % 20% Agriculture 1 8% Internal Defense, Police 8\ O t h e r Justice E m e r g e n c y 1 5 % Administration 4 % 7 % 6% Figure 7.2 -- Capital s p e n d i n g allocations - 2001-02 Defense, General Other Police & government Emergency 11% 4% lnfrastructure 29O h Agriculture 11O h Private sector Education, \and community Youth & development 9% Cultural 16% 14% Source: CFA (ex-ante FY02 presentation) 7.7 Fiscal developments in FY02 indicate that revenue performance has been on target in aggregate, with direct taxes (income tax collections) slightly higher than initial estimates, offset by a small shortfall in indirect taxes (import and service taxes). Of continuing concern, however, are shortfalls in the collection of electricity tariffs, which amouilted to only 36 percent of pro-rata targets through March 2002. CFET expenditure has been running well below target reflecting: slow budget execution on goods and services; slower than anticipated recruitment of ETPA staff positions; and a fieeze on capital spending in anticipation of UNTAET asset transfers. As a result of these factors, a $5 illillion carryforward at the end of the fiscal year has been estimated (Table 7.3). 7.8 Table 7.3 provides estimates and projectioils of revenue, expenditure and financing requirements as developed in the context of preparing the Government's FY2003 budget. "Consolidated Fund" revenues in Table 7.3 comprise of non-oil revenue and non-FTP oil revenue, whereas Timor Sea Fund revenue consists of first tranche petroleunl (FTP) royalties, which the authorities intend to save tluougll FY05 given prevailing uncertainties in other budgetary items (expenditure, revenue and financing). Non-oil revenue is projected to grow very gradually in the medium teinl reflecting current uncertainties with respect to the evolution of the domestic economy and the pace of developnlent of productive enteiprise-Table 7.4 provides more detail on non-oil revenue composition. Oil revenue projections grow more significantly even though they have been discounted by 25 percent to incorporate possible production delays and downside risks to prices. Table 7.3: Consolidated Fund Accounts and Estimates, FYOl to FY06 ($ million) Projections Component Actual Est. Core Total FYOl FY02 FY03 FY03 FY04 FY05 FY06 Revenues 28.5 31.3 42.1 42.1 52.8 97.7 99.1 ConsolidatedFund 25.5 25.3 41.3 41.3 46.2 73.5 76.8 Timor Sea Fund 3.O 6.0 0.8 0.8 6.6 24.2 22.3 Expenditure 51.3 53.9 . 67.6 77.2 82.9 96.3 101.1 Salaries and Wages 13.9 19.8 25.8 25.9 27.4 28.3 28.9 Goods and Services 15.7 23.5 33.9 35.0 40.6 42.7 42.6 Capital 21.7 10.6 4.9 9.3 4.9 5.3 4.6 Development 3.0 7.0 10.0 20.0 25.0 Savings (FTP) 3.0 6.0 0.8 0.8 6.6 24.2 22.3 - DeficitISurplus (25.8) (28.6) (26.3) (35.9) (36.7) (22.8) (24.3) Net Financing 25.8 28.6 26.3 35.9 36.7 22.8 24.3 CFET Reserves (3.4) 15.0 5.1 5.1 - Timor Sea Fund 24.3 External Financing 29.2 13.6 21.2 30.8 36.7 22.8 - ExpenditureIGDP (%) 14.5 14.2 18.9 22.0 23.6 26.0 25.3 Source: Ministry of Finance Note: For reserves: (-)designates illcrease 7.9 Two expenditure scenarios have been developed in the context of the FY03 budget reflecting prevailing uncertainties in budgetary execution and financing availability; the smaller core budget is nearly $10 million below the alternative expenditure scenario, with the difference primarily reflecting reduced capital and development expenditures. But even if the larger expenditure scenario is assumed for FY03, the three year cumulative financing requirement for FY03-05 is $78 million less than presented by the authorities at the Oslo donors' meeting in December 2001. (Deficit estimates still exclude First Tranche Petroleum (FTP) royalties, given the decision to save all such revenue through FY05.) ~ e h c ewhile significant donor support will continue to , be needed during the next three fiscal years, the projections of financing requirements have been substailtially reduced, and the prospect that budgetary support from the donor community can be forgone after FY05 appears favorable. 7.10 The two years of UNTAET adilliilistration have served to reestablish a nunlber of spending decisions in the various sectors of the economy, thus also shaping the main fiscal trends illto the future. Any changes the independent government inay wish to introduce to the established revenue and spending patterns will need to be analyzed as alteratioils to the current forecasts. Assuiniilg that current sources of revenue and levels of services are maintained and that services currently provided by the UN agencies off- budget are gradually phased into the budget, a sizeable financing gap will remain over the next tluee fiscal years, highlighting the need for vigilance and careful budgetary oversight. 7.11 The new administration may wish to introduce alterations impacting revenues andlor expenditures. Careful consideration also needs to be given to other possible tax sources (a land tax and a national lottery have been mentioned) and to the possibility of tax exernptioils to promote job creation. In each of these cases, the tax amount directly gained or lost is estimated at no more than $2 million. 7.12 Changes on the expenditure front could involve larger amounts. Widespread subsidies to agriculture or a new level of regional administration, for example, could add more significantly to overall expenditure and to the financing gap over the next few years. Anlong their first priorities, the new administration and parliament, in the terms set out by the new constitution, need to focus on the existing framework of revenues and expenditures and decide on possible changes to bring public structures and policies more in line with national priorities. Each of these potential changes should then be carefully costed over a multi-year perspective before a decision is taken on which changes are possible to introduce. 7.13 A potentially important decision, in terms of its impact on overall administration costs, has to do with the model of local government that will be established or, in other words, the chosen model of administrative decentralisation. The 13-district model that the Ii~donesianadministration inherited from the Portuguese colonial period was developed early last century and there is no reason why it should continue to be the most appropriate, democratic or cost-effective model to satisfy present aspirations and circumstances. Communications improvements have shortened distances and the principle of subsidiarity, implying that the lowest capable level of a hierarchy should take each decision, is now a more entrenched tenet of democracy. 7.14 There is no off-the-shelf model of local government that is readily available and can be recommended. The solution most appropriate to East Timor should emerge from broad consultation with all stakeholders and a realistic assessment of experience thus far with local decision-making structures as, for example, those supported by the Community Empowerment Projects. The issue of enhancing local community participation is addressed in more detail in Chapter 9. 7.15 The UNTAET adininistration saw as one of its priority functions the creation of the standard organs of a modern bureaucracy, even though the underlying activities barely existed. Thus, East Timor has a bureaucracy to regulate financial services but no indigenous financial institutions and a tax administration but a very small tax base, other than for import tariffs. The allocation of scarce local huinan resources to such "pre- emptive capacity building" may be questioned, but the fact is that the tax service has been put in place and some local staff has received training. Table 7.4: Actual and Projected Revenue, FYOl to FY06 ($ million) Source Actual Proiections FYOl FY02 FY03 FY04 FY05 FY06 Non-Oil Revenue 15.5 19.3 20.8 20.1 21.3 22.9 Taxes 12.2 17.0 18.6 18.0 19.1 20.3 Direct Taxes 0.8 4.6 4.3 4.3 4.6 5.O Indirect Taxes 11.4 12.4 14.3 13.7 14.5 15.3 Non-Tax Revenue 3.3 2.3 2.2 2.1 2.2 2.6 Oil Revenue 13.0 12.0 21.3 32.7 76.4 76.4 Non-FTP 10.0 6.0 20.5 26.1 52.2 54.1 FTP 3.O 6.0 0.8 6.6 24.2 22.3 Consolidated Fund Revenue 28.5 31.3 42.1 52.8 97.7 99.3 Assigned ~evenue" 1.4 3.6 6.5 7.5 9.5 13.5 Non-Oil RevenueIGDP(%) 4.4 5.1 5.8 5.7 5.8 5.7 Oil-RevenueIGDP (%) 3.7 3.2 5.9 9.3 20.6 19.1 Total RevenueIGDP (%) 8.0 8.2 11.8 15.O 26.4 24.8 Source: Ministry of Finance dAssigned to Power Authority,Dili port, civil aviation. 7.16 Table 7.4 presents the sources of tax revenue, past and projected, which suggests that the bulk of onshore tax revenue will continue to originate from import taxes- consisting of a flat 5% import duty plus a 5% surcharge with relatively few items exempted. 10% services tax applies to hotels, restaurants, telecoinmunications suppliers and car rentals.' Income is taxed as follows: resident monthly wages are taxed at 10% for wages between $100 and $650 and at 30% for wages above $650; dividends, interest and royalties are taxed at 15%; non-resident income is taxed at 20%; and most other income sources are subject to three tax rates of lo%, 15% and 30% with the top 30% rate applyiilg to annual income above $6,737.3 Revenue from most of these sources is expected to remain modest and does not factor in the administrative cost of collections. 7.17 For most economies, taxation is a necessary evil-a dampener on economic initiative and activity, but a generator of revenue to cover the cost of providing needed services to the community. In East Timor, the bulk of the revenue provision function in the medium tenn can, in all likelihood, be met by revenue from offshore oil and gas. At the same time, other taxable productive activity in the econonly is minimal, and the prospect of higher oil and gas resources raises a different potential problem-that the competitiveiless of prospective activities in other sectors will be undernlined, perpetuating widespread underemployment. 2 Small increases in selected indirect taxes are envisaged at the start of FY03. 3 Further details may be found in UNTAET Regulation No. 2000118. 7.18 These factors suggest the need to view tax policy in a broad and dynainic frainework that does not exacerbate already serious problems of competitiveness. Thus, consideration might be given to a reduction or even elimination of the corporate illcome tax on private entei-prise. In the case of companies from countries that have agreements with East Timor to avoid double taxation, voluntary tax payment in East Tiinor could be encouraged up to the level of the company's home country liability. In this way, East Tinlor could benefit froin some revenue without placing any marginal tax burden on the conlpaily conceined and without giving a ''free ride" to the conlpany's home country tax authority. Such a pro-job tax policy could be advertised as an incentive to productive investment, both local and foreign. Adjustments to personal income tax rates could also be considered, for example, to raise exemption threshold levels in order to limit income taxes to relatively high-income individuals, thereby eliminating the tax component from the wages of most workers. 7.19 By contrast, additional effort is needed to address the problem of poor collection of user fees and charges-to avoid waste and promote economic use of scarce resources as well as to limit budgetary allocations to cover losses. The prime example is that of the Power Authority, which has only recently begun sending electricity bills to its largest custonlers; through early 2002, collections were running well below target, with about a $2 million shortfall in cost recovery. And it is important to maintain administrative vigilance over the collection of import tariffs, which account for more than half of onshore revenue. 7.20 A traditional concenl of tax officials is to search for new sources of revenue. In this context, suggestions have included a national lottery and taxes on property and on casino activities. Such proposals should be very carefully considered from a social perspective before decisions are taken. Several countries have come to seriously question whether the revenues generated from taxes on gambling warrant the social costs of addiction and the problems caused by the presence of organized crime frequently associated with casino activities. As for a property tax, it is doubtful whether, in a country with very few large property owners, the revenue collected will cover the cost of applying the tax, especially if property used injob-creating activities is exempted. 7.21 Irrespective of tax policy on onshore assets and even under the most favorable assumptions about Timor Sea commercial agreements, East Timor will continue to be heavily dependent on donors funding, at least until 2004, to pay for the existing policy framework. In large measure, the funding gap reflects the recurrent cost repercussions of investment projects financed by the donors themselves plus the cost of a substantially pared-down civil service. Contiilued donor financing would permit East Timor to inaxiillize the flexibility with which it can devise a long-term saving strategy as oil revenues begin to exceed current needs within 2-3 years-i.e., without the burden of accumulatiilg expensive comn~ercialdebt in the interim period. 7.22 The Timorese authorities will need to be particularly vigilant with regard to new spending initiatives in the next few years to constrail1the fulilndinggap to levels that do not require resort to commercial financing. Given the high political risk of lending to a new nation and the uncertainty associated with offshore revenues, ally offer of commercial financing would likely come at a high cost, most probably iilvolving securitization of future offshore revenue. The prudent means of avoiding this would be to contain government expenditures to levels that can be financed by donor grants or concessional loans; terms for the latter include long maturity and grace periods and service charges amounting to less than one percent per annum. 7.23 If in fact donor funding were not available to cover the entire funding gap, there are two further optioils that could be considered. First, the decision to save all first tranche petroleum royalties until FYOG could be reconsidered, which would free up additional, though limited, resources for the budget. Particularly if higher Timor Sea revenues were to materialize, there will be ample opportunity to build up savings; given the many urgent investment needs, it would not seem appropriate to forego vital expenditure in the interim period. As a last resort, it may be possible to place a limited quantity of Timorese bonds at reasonable interest rates in countries with large Diaspora communities, such as Australia and Portugal. Such bond placements would not collateralize future oil revenues nor require the high risk premium common to regular emerging market debt-if they can tap into patriotic support fi-omDiaspora communities. 7.24 Most importantly, maximum effort is needed to conclude agreements with the Timor Sea commercial counterparts and to formally ratify these, which should minimize the need for financial assistance fi-om the donor commu~lityafter 2004 or 2005 (though technical assistance will continue to be needed for an extended period). 7.25 Any attempt to assess the sustainability of a fiscal fi-amework for East Timor is dependent primarily on what is assumed about the largest and most volatile variable-the size and timing of revenue flows from offshore oil and gas exploration. On the basis of Timor Sea reserves already assessed, there are different projects that could proceed, but the Bayu Undan project is the only one on which development commitments have been made. Until agreement is reached on other projects, it is prudent to assess sustainability on the basis of only the Bayu Undan project. 7.26 In preparing for the Oslo donors' conference in December 2001, the sustainability of econonlic management in East Timor was tested on the basis of projected revenues from Bayu Undan liquids only, as that was the only part of the project already approved at the time. A scenario was constructed through 2030, in which significant donor hnding was needed only for the period 2002-05, until large oil revenues became available. Beginning with fiscal year 2005-06, oil revenues provided the residual funding for expenditures and excess revenues were saved, in amounts well in excess of the FTP portion. Bilateral funding to cover teclu~icalassistance and IF1 concessional loans, amounting to about $30 million and $15 million, respectively, in the initial years, were the only external funding assumed. Interest earned on saved oil revenues gradually increased as a source of budget financing and, by the time oil flows were expected to cease, the oil fund would have accunlulated sufficient capital to provide a constant revenue flow for the budget, in peiyetuity. 7.27 For the oil-only sceilario to be self-sustaining after the 20-year oil boom was over, the non-oil econoiny would need to grow sufficiently during that period to be able to replace oil as a tax base large enough to meet the growing expenditure needs. Two important preconditions would need to be met for such a vibrant non-oil economy to be a reality. One is that iilstitutions and procedures, legal and other, be in place, and policies implemented, so as to encourage continued rapid growth of output and incomes in the non-oil sectors. The other is that governnleilt spending be restrained at the outset and then grow at a slower pace than the non-oil economy. The private sector, not the state, would be the economy's engine of growth under this scenario. 7.28 In March 2002, the oil company responsible for developing the Bayu-Undan project announced that it had agreed with Japanese utilities to buy practically all of the field's proveil reserves under a 17-year contract due to begin in 2006. On this basis, it had decided to go ahead with the gas project, with plans to pipe its gas ashore to Darwin. Tax arrangements between the company and the Timorese authorities have been agreed and are being negotiated between the company and the Australian authorities. Adding the gas to the Bayu-Undan project raises the estimated total cash flow over the project life by some 50 percent, but the additional investment in the early years means that the large revenue flows are delayed by about four years, until FY 09/10. Between FY 05/06 and FY 08/09, however, expected flows from oil and gas still exceed the financing needs of the Tiinorese administration, but it will not be possible to maintain the policy of saving all FTP revenues. If agreement is reached between the oil company and the Australian tax authorities, the sustaiilability conclusioilspresented at Oslo are reinforced by the new revenue perspectives, and it is anticipated that no additional budgetary financing by donors will be required after FY 04/05, though technical assistance will continue to be needed. 7.29 I11 summary, even if no new reserves are found and if agreement is not reached on the exploitation of confirmed reserves, such as Sunrise-Troubador, fiscal sustainability could be assured as long as tax and other policies succeed in establishing and maintaining a vibrant non-oil economy that can serve as an effective source for fiscal revenue as oil reserves taper off. In the event that agreement will be reached with oil companies to exploit all known reserves and that new reserves may be discovered and brought to market, the situation becomes less constrained, permitting an acceleration of investment in measures to reduce poverty. In that case, government expenditure could grow.more rapidly, reducing the reliance on private sector growth for tax revenue, but increasing the need for secure governance mechanisms to ensure that higher public spending is in fact effective in accelerating poverty reduction. 8. ENSURING EFFECTIVE PUBLIC EXPENDITURE MANAGEMENT 8.1 Effective and efficient expenditure management systems, capable of delivering resources to service delivery units on a timely and predictable basis, whilst ensuring coinpliance with policy directives and expenditure limits, are critical for ensuring macroeconomic stability, achieving value-for-money in the application of scarce resources and attaining developn~entobjectives. Such a system requires a sound legislative and regulatory framework, operated by competent, well-motivated and well supervised staff, within a governance framework that holds individuals and institutions to account for their performance. 8.2 ETTA has put in place the key elements of a transparent, performance oriented public expenditure management system applying international best practice. A Central Fiscal Authority was established in January 2000, with responsibility for budget formulation, budget execution, the rendering of accounts and treasury management. Subsequent framework legislation, complemented by internal Directives, lays out the principles and procedures for public expenditure management across Government. A Central Payments Office was also established in January 2000, as ETTA's banker, holding many of the hnctions of a Central Bank. 8.3 Six challenges are identified for the future: development of a medium-tern1 framework for expenditure management, consistent with macroeconomic stability, that can allocate resources in line with priorities and deliver a predictable flow of resources to agencies; strengthening the performance and service delivery orientation of management systems; maintenance of expenditure controls, to ensure that resources are applied in line with policy directives and comply with expenditure limits; creation of governance structures that guarantee transparency and hold the executive to account for its performance; formulation of realistic, resource constrained plans to guide decision-making throughout government; and development of an external assistance nlanagement system that will align external resources with Government's developmental priorities and facilitate sound expenditure management. 8.4 In each of these areas the issue of capacity and capacity building is central to the effectiveness of the systems in the short and medium-ternl. BUDGETING, EXPENDITURE PROGRAMMING AND PERFORMANCE 8.5 CFA launched the 2001/02 budget process, for the July to June fiscal year, with the dissemination of a Budget Circular to Departments in mid-March 2001. This provided guidance on budget coverage, expenditure limits for the forthcoming year and the preseiltatioil of proposals. Departineilts were required to present forward estimates of expeilditures by program and source of funds up to 2004105, together with a staffing profile for national and iilternatioilal personnel, a breakdown of expenditures following a suinmary GFS economic classification and service delivery targets for the budget year. Programs coincide with Departments' agency structure: in most departments, budgets were foilllulated by agency managers and then consolidated at departmental level. Department proposals were submitted in mid-April, negotiated through a series of bilateral budget meetings with CFA and finalized by Cabinet in May 2001. The Combined Sources Budget was presented to the Canberra Donors Conference in June 2001 and then approved by the National Council (now Constituent Assembly). The budget is subject to a mid-year review, in November, when appropriations are revised in line with execution and proposals for additional spending appraised. 8.6 Important features of the present budget process are: firstly, the medium-term perspective for aggregate, departmental and agency expenditure programming, supported by a review process in order to reconcile agency proposals with hard aggregate resource constraints; secondly, its performance orientation, whereby program level allocations are related by quantitative and qualitative service delivery targets. However, these features are still at an early stage of development. 8.7 The 2001102 exercise was hindered by the absence of a clear policy framework for the prioritization of expenditures and indicative resource allocations at the Departmental level. Although the Budget Circular recommended that Departments should limit increases in total spending to 10% above the revised 2000101 Consolidated Budget-in order to ensure consistency with the medium-term expenditure aggregates presented at the November 2000 Brussels Donor Conference-few Departments respected these limits. Departmental proposals for CFET totaled $112.6 million, double the agreed forward estimates, including a 34% increase in the number of civil servants to ilearly 16,000 posts. Proposals were subsequently trimmed to $64.9 millioil by CFA and Cabinet. The mid-year budget revision has again generated proposals for increases in spending amounting to approximately $15 million, though CFA has subsequently cut these back to $2 million, and only $1 million is likely to be approved by Cabinet. 8.8 As yet, no fornlal procedures for performance review have been put in place. While most agencies managed to present key performance indicators and targets, performance monitoring in some sectors is hampered by the lack of baseline data and systems for routine data collection. In the absence of adequate service costing infonnation, particularly unit costs for deliverables and assessments of implementation capacity, the relatioilship between expenditure and performance indicators is poorly understood, making it difficult to set realistic targets. Furthermore, the structure of agency budgets does not encourage a service delivery orientation. Resources are allocated between vertical programs, managed by central agencies, with no iildication of the distributioil of expenditures between levels of administration. Consequently, it is difficult to detenniile what proportion of spending on, for example, primary education, will actually finance services in schools, how resources will be distributed geographically, or track expenditures to the field level during execution. Inevitably this increases the risk that resources will be captured by the central administratioil at the expense of field level service delivery. This risk inay be con~poundedby the presence of large nunlbers of expatriate advisors at central level, creating uilrealistic expectations regarding the working environment and promoting a proliferation of management informatioil and decision-making systems, thereby increasing adininistrative overheads. 8.9 The Ministry of Finance is well aware of these weaknesses, which it intends to address through a medium-term strategy for the developnlent of the budget system, currently in draft form. This strategy focuses on core-budgeting systems and institutions. However, it is equally important to coilsider how resources are managed by line Ministries down to the field level. Some of the key issues to be addressed include: Regulating the budget process. Regulation 2001113, which currently provides the basis for budget fornlulation, only addresses the presentation of the Consolidated Budget. A specific Budget Regulation, or Guidelines issued as a Directive, is needed to provide a statutory basis for the budget formulation process, establishing a budget timetable, the form of presentation of departmental proposals, the basis for costing these proposals, and the appraisal process. Ideally, the regulation would also outline agency budget management systems, including, for example, the constitution of Budget Management Committees to oversee the preparation of budget proposals and in-year prioritization of expenditures, thereby improving transparency and ensuring adequate participation of agency and district managers. Moving to a MTFF based on the National Development Plan. In principle, the forward estimates presented in the 2001102 Co~lsolidatedBudget should provide the basis for the 2002103 budget exercise. In practice, the results of the first forward budget exercise will probably have to be adjusted to reflect execution in 2001102 and the priorities identified in the National Development Plan. Thereafter, rolling- forward estimates within a Medium-Tern Fiscal Framework should provide the indicative allocations that will serve as the baseline for agency budget estimates. If agencies are to prepare realistic forward estimates, however, coilsiderable work will be needed at agency level in dete~miningservice delivery unit costings and cost drivers. This, in turn, will have to reflect policy decisions regarding service standards and the appropriate trade-offs between service coverage and quality. Establishing a formal policy and budget review process. The Ministry of Finance's strategy envisages a multi-stage budget process. This would entail the presentation of policy proposals and their budgetary in~plicationsby Ministers shortly after the approval of the revised budget. These would be reviewed by a Council of Ministers sub-committee in the light of projected resource constraints and the trade-offs between competing applications of funds, prior to the agencies' formulation of detailed budget estimates. Estimates justifying the retention of forward estimates and additional spending would then be subject to a secondary review before submission of the budget for approval by the Council of Ministers. Ideally, these annual routines would be compleinented by a rolling-program of periodic expenditure reviews, either across government or focusing on priority sectors, perhaps on a three yearly cycle (such as the United Kingdom's Fundamental Expenditure Reviews or the Public Expenditure Reviews undertaken in some African countries). These would focus attention on the budget base- ongoing expenditures-providiilg an opportunity for a critical analysis of public policy, its impact and expenditure implications, iilcludiilg an assessment of the efficiency and effectiveiless of alternative expenditure policies and alteinative service providers. Performance monitoring. If targets are to act as a spur to improved performance, mechanisms will have to be put in place for routine reporting on key performance indicators. Reporting should address the requirements of both agency management and the legislature. However, the use of administrative data for performalce assessment may taint these sources. Triangulation of data is critical. An independent perfoilnance audit provides one triangulation point. Periodic service delivery surveys and participatory assessments provide another, allowing decision- makers to assess the respoilsiveness and adequacy of services fiom the clients' perspective. In the longer-term, some consideration should be given to the integration of agency level performance assessments with the personnel performance assessments for senior civil servants. Improving the service delivery orientation of sector spending. Expenditure tracking studies in Uganda have demonstrated that centralized agency budgets and funds release mechanisms facilitate administrative capture of resources intended for service delivery. Uganda has resolved these problems by adopting a system of block grants to individual service delivery units, supported by participatory management committees and publication of expenditure information at the village level. Deconcentration of financial management responsibility to field level service delivery units may be appropriate in East Timor, given limited resources and management capacity. Though clearly, this must be placed within a governance framework at local level. As an intermediate step, however, District allocations within agency budgets would improve transparency. Moreover, where these funds are decentralized to District Finance Offices for the purposes of budget execution, it would be possible to track execution against allocations, reducing the risk that funds will be retained by agencies' headquarters. Using public spending to respond to and empower communities. The experience of East Timor's Community Empowerment Project, financed through TFET, demonstrates that there are significant advantages to be gained from the decentralization of decision-making and resources to the community level. Comn~unitiesare better placed to identify their development needs and priorities, and respond quickly when these needs are identified; communities can mobilize substantial resources, in labor and materials, significantly reducing the unit cost of infrastructure; cornnlunities have a vested interest in the maintenance of infrastructure and ensuring the service delivery standards are maintained; and community structures are able to oversee the use of funds, reducing the risk of comption. Regulation 2000113 provides a framework for community development initiatives, establishing 418 village councils, comprising elected representatives from hamlets. Since May 2000 the councils have implemented about 1,400 CEP project activities, valued at nearly $5 million. Clearly, a strong case may be made for institutionalizing the CEP approach, by providing support through a Departlnental structure within the Ministry of Internal Administration, rather than a project managenlent unit, and placing a system of community grants on budget. However, if the community driven approach is to realize its full potential a number of issues will have to be addressed. Firstly, the functional responsibilities of levels of government should be clarified, so that it is clear, for instance, that feeder roads will be the responsibility of beneficiary communities rather than central government. Secondly, government fuilctional departments should be encouraged to respond to community initiatives by providing technical support, particularly in areas such as public works and agriculture. Thirdly, appropriate mechanisins should be put in place, backed up by incentives, to ensure transparency and accountability in the use of funds by communities and by sectoral agencies. This may entail empowerment of village councils to oversee the management of service delivery units such as schools and health posts'. Lastly, a system of territorial planning should be developed to ensure a rational distribution of service delivery units and adequate coordination between coinmunity and government initiatives. Budget Execution and Control 8.10 ETTA has adopted a centralized system of budget executioil and control for CFET, whereby Treasury authorizes commitments and transactions and executes payments on behalf of agencies from its Consolidated Fund account (Box 8.1). Within Treasury each stage of the payment process is registered on a computerized Financial Management Information System. This system is currently being rolled out to agencies, allowing them read-only access to the system. It is not intended to introduce agency-level electronic payments processing in the near future. Although the system is centralized, some flexibility is built in. Agencies may requisition petty cash accounts for direct procurements up to $1,000, though these funds are usually limited to $3,000. Agencies are also allowed to vire (reallocate) expenditures between recurrent and capital appropriations, up to 10% of the total appropriation, though there are restrictions on the virement of funds to personnel. 8.11 Recruitment within the civil service is also centralized. Established posts are agreed through the budget process, though proposals for new posts must be endorsed by the Civil Service and Public Einployment Service (CSPES), which verifies job descriptions to ensure that there is not excessive duplication of functions. CSPES then oversees selection and recruitment of candidates. Payroll is centrally managed, with agencies verifying that staff in place before final payroll lists and cheques are drawn. Payroll for district staff is drawn in cash in Dili and disbursed by the District Finance Officer. Again some flexibility is built into the system by allowiilg agencies to recruit temporary staff directly, mainly for support staff, drawing funds against their goods and services allocations. 8.12 In line with World Bank and bilateral donors' requirements, management of TFET and bilaterally financed projects is delegated to agency project iillplementatioil units. Treasury authorizes projects to open accounts in commercial banks. Payments are then executed following donor specific procedures. Quarterly statements of receipts and payments are submitted to Treasury. These are registered against ETTA's Capital Fund ' Discussed at greater length in Chapter 9. for the purposes of budgetary repoi-ting. Inevitably, the lack of prior-expenditure controls on project accounts increases pressure for these to acconlmodate expenditures that cannot be covered from CFET. This is likely to increase the burden of recurrent expenditures on nonlinally capital projects subverting centralized expenditure controls. Treasury intends to adapt IFMIS to allow registration of project expenditure through the Governnlent's financial system, while at the same time generating financial reports and statements required by the World Bank and-potentially-other donors. This should facilitate the application of CFET expenditure controls to projects executed by Government agencies. 8.13 Regulation 2001113 has also iiltroduced specific, decentralized procedures for autonomous agencies-the Power Authority, Dili Airport Authority, and Dili Harbour Authority. This provides for the capitalization of agencies fiom budgetary appropriations and financing of service delivery from earmarked revenues, thereby creating incentives for enhanced revenue collection. Government may purchase goods and services from the autonomous agencies at market piices. No provision is made for transfers to finance s~~bsidiesor insolvency. In order to address these issues it may be appropriate to issue specific legislation, together with articles of incorporation for each of the agencies specifying the scope of operation and their internal governance structure. Box 8.1. The Budget Execution Process Following the approval of appropriations,Treasury authorizes the release of funds through warrants (Expenditure AuthorizationNotice), usually on a quarterly basis. Agencies may apportion these funds to subordinate spendingunits by issuing departmentalauthorizations. Spending units initiate purchases by issuing a Commitment and Payment Voucher (CPV) which is endorsed by heads of Department and Agencies (Authorizing Officers) and then approved by a Treasury CertifyingOfficer, after verifying budget coverage. Commitments are regstered on both the agency and Consolidated Fund ledgers. Spending units' Authorizing Officers issue purchase orders against approved commitments. After deliveiy, Agency Certifying Officers confirm receipt of goods and services and endorse the CPV for payment. The Treasury Authorizing Officer verifies the CPV against the approved commitment and issues Treasury Payment Order for liquidation by the Central Payments Office (ERA'S banker), either by cheque or bank transfer, and registers the payment approval against the Consolidate Fund ledger. CPO submits daily statements to Treasury for reconciliation. Once the payment has been liquidated the Treasury registers the process as closed. At district level, payment of salaries and goods and services is usually made in cash, withdrawn from CPO in Dili and held in safes, owing to the lack of banlung facilities. The Ministry of Finance remits funds to Sub-Treasuries, constituting an imprest account, on the basis of departmental expenditure authorizations for goods and service in favor of District Offices. Departmental expenditure authorizations are approximately $2,000 per month for health, rising to $15,000 for public works. Agencies' guidelines indicate eligible expenditures against these allocations. Payments are executed by the District Finance Officer, as approved by the agency's District Certifying Officer and the District Administrator as Authorizing Officer. All payments are registered against agency ledgers to ensure that expenditurelimits are respected: most payments are pre-authorisedby the District Finance officer to verify eligibility and budget coverage. 8.14 It should be stressed that the budget executioil procedures applied by ETTA provide the basis for rigorous expenditure control and timely reporting execution on core government expenditures. Monthly agency execution reports and quarterly financial reports are generated within days of the end-of-period, the former distributed within Government and the latter available to donors as well. The Ministry of Finance submitted its report and accounts on the 2000101 Consolidated Budget within three months of the end of the fiscal year. The report has now been audited by an international firm and certified without qualification. 8.15 Systems for procurement, asset management and inventory control are less robust. A central procurement agency has been established, under CFA, which is supposed to process procurements initiated by line agencies and ensure that goods received are certified and recorded on a central assets register. In practice, many projects undertake their own procurement, often following donor-specific procedures; there is no systematic register linking procurements to payments; and assets registers and inventories are incomplete. The elimination of tax exemptions for ETTA procurements provides an opportunity to tighten up on these procedures. However, considerable investments are also required in systems development-particularly as regards linkages between procurement and asset management and Treasury's IFMIS-and capacity building. 8.16 Difficulties have also arisen in liquidity management, owing to delays in the transfer of LJNTAET funds by the UN at the beginning of the 2001102 fiscal year. Since the Ministry of Finance currently has no access to a ways-and-means facility with the CPO, and is not authorized to contract short-term debt through commercial channels, the liquidity constraint had to be accommodated by restrictions on the release of fi~nds+nly monthly warrants were issued-and delays in the authorization of agency commitments. For the purposes of efficient budget execution these restrictions should be relaxed following Independence. The Budget and Financial Regulation (2001113) already provides for government borrowing, requiring the legislature's authorization of borrowing limits through a Government Finances Regulation. It may be prudent to introduce restrictions on the level of long and shoi-t-term debt within framework legislation. Often restrictions on short-term debt are set as a proportion of Consolidated Fund revenues in preceding years. Some consideration should also be given to the alteration of the CPO Regulation (200016) so as to provide the Government with a ways- and-means facility for unforeseen, short-term coverage of liquidity constraints. Again, prudential limits should be established. Before these facilities are put in place, Treasury will have to put in place operational procedures and capacity for debt management. 8.17 The present, centralized expenditure control system is cuirently seen as international best-practice in public expenditure management, since it is the best guarantee of compliance in budget execution. However, system efficiency and effectiveness-in terms of timely authorization of commitinents and compliance with established procedures and appropriations-depends on a Treasury is staffed by con~petent,well motivated and properly supervised personnel. If delays in commitment autl~orizationbecome routine, impeding prograin implementation, the temptation to delegate responsibility for expenditure coiltrol to agency level may arise. Several countries have dismantled centralized payments systems-for instance, Mozan~biq~~e, Malawi and Jamaica-replacing them with delegated systems where funds are released to agency bank accounts, with the agencies themselves enforcing commitnlent and expenditure controls. The poor record of these systems in terms of excess expenditures, build up of arrears and unprogrammed budget deficits is well documented. Clearly, the present, centralized system of expenditure control is better suited to East Timor's needs. Maintenance of the present system will, however, require a strong political commitment to expenditure control and the presence of suitably qualified staff within Treasury, even if this signifies that key line positions will continue to be occupied by expatriate staff in the medium tenn. Oversight Institutions and Transparency 8.18 East Timor's Constitution-which is still in preparation-should define the iilstitutional framework that will guarantee oversight and accountability of the executive though the legislature. The key institution is an independent auditor, reporting to the legislature, following the principles established by the International Organization of Supreine Audit Institutions (INTOSAI) and the IMF's Fiscal Transparency Code. The auditor's independence should be established in the constitution, providing for appointment by the legislature or by the executive and legislature jointly, fixed terms of appointment with clearly specified criteria and procedures for dismissal in case of coilflict of interest or incompetence, together with a statutory budget allocation, determined by the legislature. Given the lack of skilled auditors in East Timor, the auditor's functions would have to be developed over the medium-term. At the very least, the audit institution should present a compliance statement. In the longer term, these functions may be extended to include a performance audit of voted appropriations. 8.19 The Government has indicated that, given the lack of national capacity, it inay be appropriate to undertake audits using a commercial firm, as well as submitting accounts to a national audit institution. There is a danger, however, that this may result in conflicting audit opinions, possibly undermining the authority of the national audit institution. To avoid this problem any commercial auditor should be contracted by and report to the national audit institution rather than the executive. 8.20 Alongside the independent audit, the Government requires internal oversight functions. These include the existing Internal Audit unit, reporting to the Minister of Finance, primarily respoilsible for the identification of weaknesses in financial nlanagenlent systems and proposal of systemic improvements. It should be stressed that internal audit is primarily a management-rather than a compliance-function, which may be extended to include assessments of economy, efficiency and effectiveness in the application of public funds. ETTA also has an Inspector General, reporting directly to the Transitional Administration, and presumably the Chief Minister after Independence. The Office of the Inspector General is broadly mandated to investigate cases of con-uption and provide public infonnation on corruption issues. OIG has undertaken twenty investigations in 2001, most relating to recruitment and procurement, leading to changes in procedures and the strengthening of internal supervision in the respective institutions. 8.21 Regulation 2001113 requires the preparation of detailed budget and reporting documents, conlplying with international transparency standards. The budget and reports are to be presented on both a cash basis, indicating revenues received and expenditures incurred during the fiscal year, including tax expenditures (income forgone froin tax concessions), and an accrual basis, indicating changes in assets and liabilities, including contingent liabilities (potential liabilities arising, for example, from loan guarantees). Compliance with these requirements requires significant analytical capability that will take time to develop in East Timor. While it would not be appropriate for East Tinlor to adopt the lower standards of other developiilg countries, some coilsideration should be given to the phased introduction of reporting requirements. This will be particularly important if capacity is significantly reduced in coming years owing to the departure of intemational personnel. 8.22 Regulation 2001113 does not require the Government to present performance targets in the budget and actual implementation in interim and end-of-year reports. Although, perfomlance targets were presented in the 2001102 Consolidated Budget, there is no statutory requirement for the Government to do so. Since performance targets and actual implementation data are likely to be a more useful guide to the relative merits of programs for decision-makers and legislators, there is a strong case for including this requirement in budget framework legislation. Role of the Legislature 8.23 Regulation 2001113 requires the Ministry of Finance to present to the National Council (predecessor of the present Constituent Assernbly): the budget; quarterly expenditure reports; final accounts; and the report of the independent auditor. The 2001102 Collsolidated Budget was submitted to the National Council in May 2001. h m a l appropriations-covering only CFET expenditures, broken down to the program level, distinguishing appropriations for personnel, goods and services and capital-were analyzed and voted by the Council over a period of six weeks. Substantive alterations were introduced to the allocations of some sectors. The Constituent Assembly's current standing orders require the budget be voted in plenary, though other legislation is delegated to a Specialized Legislative Committee. 8.24 While plenary approval of the budget is the norm, most Commonwealth, European and Congressional systems also provide for review of the budget in sub- committee, such as a Budget and Finance Committee, prior to plenary debate. Often, the Budget Committee is chaired by a member of the majority party. The Co~lstitutionor legislature's standing orders may specify the maximum duration of such deliberations, requiring the Committee to present reconlmendations to the plenary. It is also common practice for the legislature to establish a separate Public Accounts Committee, which deliberates on the government and agencies consolidated accounts, with authority to call public officials to account and make recommendations intended to improve compliance. Often the Public Accounts Committee is chaired by a member of the opposition, so as to ensure critical analysis of Government reports. 8.25 While the configuration of legislature's voting procedures and committee structures will differ depending on each country's requirements, it should be noted that there is a substailtial body of theory and empirical evidence demonstrating that these arrangements can have a considerable impact on budget outcomes. These arrangements merit coilsideration by technical specialists. Development Planning 8.26 In October 2001, the Government started preparation of a National Development Plan, which it illtends to finalize before Independence in May 2002. For this purpose, the Government established a Planning Commission, comprising five ministers, chaired by the Chief Minister, a Secretariat to provide technical and managerial support, and a Consultative Con~missionof Civil Society, including five representatives from the NGO sector. As the first step in the planning process, each Ministry prepared a Status Report during November 2001, coinprising an overview of the sector under Indonesian administration, a comprehensive listing of reconstruction and developmental activities, identification of gaps in the current program, a sector policy framework and development vision. Subsequent steps identified in preliminary documentation include broad consultations with NGOs, communities and focus groups, in order to generate a popular developinental vision, and the preparation of thematic analyses based on the status report, focusiilg on poverty reduction issues. 8.27 It is still unclear what form the National Development Plan will finally take. The most realistic option would be to limit the plan to a medium to long-term visioning exercise, setting out the broad priorities for development with a preliminary identification of key development strategies and developmental targets for the medium to long-term. In order to operationalize the vision, the new, post-Independence Governmeilt would then have to prepare a medium-term strategy, covering its electoral mandate, and annual operational plans and budgets. The alternative would be to prepare an operationally oriented development plan with specific resource allocations and targets for the short to medium-term. This would be a time-consuming process and is unlikely to be concluded before May. 8.28 The purpose of the National Development Plan must be clarified before it is possible to identify an appropriate planning process. Equally important is the design of follow-up operational planning and monitoring mechanisms. The critical choice lies in the institutioilal division of responsibilities for planning and expenditure management functions. Outside South Asia, there has been a marked trend for the consolidation of these fuilctions in a single core institution, such as a Ministry of Planning and Finance, and in corresponding departments within line Ministries. This facilitates the preparation of medium-tern1 and annual plans on the basis of realistic resource constraints, linking of resources to developineilt targets, and in-year adjustment of resource allocations and targets in line with implementation. The Medium-Term Fiscal Framework and Budget process developed by the Ministry of Finance already has these features. Should the Goveimnent opt for an institutioilal separation of planning and expenditure management fiinctions-for instance, attaching planning to the Ministry of Eco~lomicDevelopment- care will have to be taken to ensure that plans are based on a realistic resource envelope and that in-year adjustments are consistent between instruments. This would require specific regulations coveling the planning process and its linkages with the annual budget and MTEF exercise. 8.29 Biannual Donors' Conferences have provided a forum for dissemination of infoi~nationregarding finailcing requirements and implementation, and coilsultation between Government and donors as regards priorities and policy measures. At the sectoral level, TFET has established a system of biannual Joint Donor Assessnlents for each project. These allow representatives of financing agencies, collaborating agencies (such as NGOs) and government departments to assess progress and participate in the fonnulation of work plans. 8.30 Despite these mechanisms, both the Government and the donor community have complained that aid coordination is weak, that registers of projects and financial flows are incomplete and stand-alone projects are proliferating. This is partly a reflection of the diversity of aid delivery mechanisms, including UN administered, Trust Fund and bilateral programs, and the substantial proportion of in-kind assistance. However, management is also hindered by the lack of a coherent institutional, policy and legal framework for aid management within Government. 8.31 At present, responsibility for aid coordination rests with the Plailning Commission's Donor Coordination Unit. The unit consolidates information on external financing collected from the donor agencies. It is also supposed to match donors with Government financing priorities by drawing on an "unfunded priorities list", comprising a pipeline of projects at various stages of formulation proposed by line Ministers and approved by the Council of Ministers. In principle, Ministlies may only seek external financing for projects on this list and the final approval of financing agreements is also subject to the Council of Ministers approval. 111practice, the list lacks a coherent policy framework and is viewed as a "shopping list" by most donors. Bilateral negotiations between sectors and donors continue and it is likely that most financing agreements are initiated at this level, either through specific Ministry requests or donor project identification missions. 8.32 Regulation 2001113 provides for presentatioil of a special funds budget, comprising in-kind aid and external financing that does not pass through the Consolidated Fund. Agencies are supposed to provide information on programmed external financing by project through budget forms. Coverage froin this source is more comprehensive than in most aid-dependent countries, though it is still incomplete and the information available is not consistent with Goveinrnent budget classifications, preventing consolidated analysis. Moreover, since external financing agreements are signed by the Transitional Administrator and retained by UNTAET, the Budget Unit is unable to verify aid flows reported by agencies against the original agreements. This prevents the Ministry tracking disbursements against commitments. 8.33 Given the importance of extemal financing for both recurrent and capital expenditures in the period to 2004105, and the contiilued importance of exteinally financed teclmical assistance thereafter, the establishment of an effective external assistance management system should be a priority. Ideally, this system should be managed by the Ministry of Finance, so as to ensure that external resources are progranuned aloilgside domestic financing. Key elements of the system would include: an exteinal assistance strategy; a centralized external assistance tracking system; and the development of integrated programs and working groups at the sectoral level. 8.34 Agreement between Government and Donors on the broad principles of external assistance management and prioritization is a much more useful tool for donor coordination than a project pipeline. Tanzania's External Assistance Strategy provides a good example. This lays out the nlechanisms for donor coordination at an inter- governmental and sectoral level, the criteria for prioritization and broad principles regarding the aid delivery mechanisms to be employed, including a commitment by donors to untie aid, increase the proportion of programmatic financing, harmonize procedures in procurement and financial management, and provide information on commitments and disbursements in a budget-friendly format. On the Government's side, the strategy includes commitments regarding disclosure of information, mechanisms for policy consultatio~land participation in internal planning and monitoring processes. In this way, doilors are encouraged to use the Government's planning system as the basis for their programming. 8.35 The budget process is not an effective mechanism for gathering information on aid flows because line Ministries may be ill-informed regarding project disbursements and have an incentive to under-report external financing and in-kind contributions. To ensure comprehensive coverage, the Ministry of Finance should also track disbursements against financing agreements, using illformatioil collected directly from the external agency. For such a system to work effectively, the present Regulation 2001113 would have to be amended so as to require the Minister of Finance's approval of all external financing agreements that entail Government counterpart financing during implementation or operation. To ensure transparency, disbursements against all outstailding financing agreements would be included as a memorandum item in the budget, distinguishing funding channeled though the Treasury system and the valuation of in-kind contributions. 8.36 A centralized aid management and reporting system is only likely to be successful if it offers govenlnleilt and donor agencies incentives for the registration of agreements and disbursement plans with the Ministry of Finance. Four instruments are important in this respect: firstly, elimiilation of import duty exemptions for imports financed by donors on behalf of government, so that these charges have to be covered by agencies' appropriations, thereby ensuring that programmed expenditures are registered on budget; secondly, restriction of exenlptions for NGOs to imports identified in programs registered with the Ministry of Finance; thirdly, restriction on work permits for technical assistance personnel to projects and programs registered with the Ministry of Finance; and lastly, requirement for Treasury approval for all bank accounts for projects of which the Govenunent is a beneficiary. It is encouraging to note that Government imports and local purchases under CFET and TFET are now taxable. However, imports by bilateral agencies continue to enjoy blanket exemptions. These issues should be addressed in the Government's External Assistance Strategy and appropriate procedures laid out in the Finance Regulation and supplementary directives. 8.37 Coordination between donors and government is best achieved at the sectoral level, since donor agencies may be expected to concentrate their assistance in those areas in which they have an agency mandate and particular competence. Effective sectoral coordination has to be based on clear sector policies and strategies, reflecting medium- term resource constraints and integrating both external and internal finance. The TFET approach, with its consolidated financing and joint donor assessments, provides an excellent model. However, it is important to accominodate a wide range of aid delivery and financing mechanisms within an integrated sector program (sector wide approach). The health sector appears to be closest to this ideal, though even here there are stand- alone projects that lie outside the sector program. Again, the mechanisms for donor coordination are best defined in an External Assistance Strategy, which lays out the structures for consultation andjoint-programming. 8.38 Throughout the civil service there is an acute shortage of experienced and qualified East Timorese staff: 25% of L5 and L6 posts were unfilled as of November 2001, and many of the staff selected at these levels do not have the desired qualifications. Qualified accountants and economists are in particularly short supply. Although the CFA has its full complement of approved posts, other than two L7 Directors, most of the senior management posts continue to be occupied by expatriates. Much the same is true at the sectoral level, where many agency Authorizing and Certifying Officers are expatriates. Training of East Timorese staff has mainly been geared to routine, procedural tasks, which continue to be overseen by expatriate personnel, rather than broader managerial and supervisory functions. Consequently, few of the posts currently occupied by expatriates, particularly those in CFA, can be filled through intellla1proillotions in the near future. 8.39 In order to avoid deterioration of core Governinellt services, ETTA is seeking funding for expatriate staff in priority positions following Independence. UNDP has identified a total of 96 priority positions2, 49 of which are in financial and personnel management hnctions, including: 12 positions in Treasury; 7 in the Revenue Service; 2 in the Budget Unit; 9 in the Central Paynlents Office; 3 in Customs; and 4 in IT; 4 in Procurement; 2 in the Inspectorate General; 2 in Public Service Management; and 4 in Finance/Administrative Officer posts within sectoral Ministries. This list does not include staffing needs for institutions to be created at Independence, such as a national audit office. 2U N D P , Prioriq) Hlannn Reso~ircesNeerlsfor /lie East TinloreseGovernvlent nt I~irlepel~rlence, October 2001. 8.40 Although the UN has indicated its intention to finance some of the priority posts, it is still uilknowil how much funding will be available. Seine funding has been made available froin other sources, including the IMF part-funding of CPO posts and possible Australian f~~ndingof Budget posts, but a considerable financing gap is likely to remain. In order to initiate recruitment, this gap should be closed early in the first quarter of 2002. Ideally, fuildiilg for these posts should be provided through a coilsolidated financing channel, since piece-meal recruitment through bilateral project agreements would significantly increase costs to Government and hinder participation by East Timorese in the selection process. Particular care should be taken to ensure that the expatriate staff selected are qualified and experienced, even though this is likely to increase the cost of recruitment. Box 8.2. Botswana: Lessons in Managing Mineral-Led Development How Botswana, a landlocked country in Southern Africa with a population today of 1.6 million, managed its expanding flows of nlineral rents, converting them into broad-based economic and social developnlent, offers lessons for East Timor. When it gained independence from Britain in 1966, Botswana had an economy based on cattle and subsistence arable farming. Household income was vulnerable to periodic droughts and the country was rated one of the poorest 25 in the world, in terms of per capita GDP. Domestic revenues were limited and, much like East Timor today, it required program aid to cover a recurrent budget gap, and the whole of the modest development budget was donor funded. Local skilled manpower was limited, and the government was critically dependent on expatriates to fill professional, managerial and technical positions. Thirty five years later, Botswana is a middle income country, with an increasingly diversified economy, foreign exchange reserves exceeding that of its much larger neighbor, South Africa, and insignificant external debt. Aid has been phased out (in the mid 1970s Botswana was one of the most successful in the region in negotiating and using aid while retaining control over policy), and the government progressively (but not abruptly) reduced its dependence on expatriate skills. Its social and economic indicators, although currently under threat from HIV/AIDS, are amongst the most favorable in Africa. Since independence, per capita GDP growth has averaged over 7% per annum, amongst the highest in the world. Mineral revenues, principally from diamonds, fuelled economic growth, but it was how the rents were managed, through good policies and robust institutions that provides the key to Botswana's success. On gaining independence, and well before the onset of diamond revenues, Botswana developed an effective planning and budgeting system. Still maintained today, it was based on the regular production of national development plans (currently six year plans, reviewed mid-term, as in Malaysia), feeding into a traditional dual budget structure, comprising a recurrent and development budget. National plans have always contained three key elements: (i) a MTEF based on realistic macro projections, to guide annual budget-making, (ii) clear sector policies and programs, and (iii) a public investment program, supported by sound routines for project preparation. Plans were put together in a highly consultative way, which ensure both ownership by ministers and credibility to civil society and the private sector. The planning and budgeting system was originally developed to attract aid, and use it and local resources effectively on physical and social infrastructure. When mineral revenues began to flow, this meant that there were already plans and programs in place to absorb the resources. When revenues exceeded forecasts, instead of expanding budgets, governments adhered to fiscal targets and sterilized the balances through the central bank. When mineral revenues dipped (as they did in the early 1990s when diamonds had to be stockpiled), the govenunent was able to cushion the shock by drawing down reserves. White elephant projects were avoided, as was a large parastatal sector. Early in the mineral boom special funds were created to absorb excess revenues, but these proved less important than adhering to a well designed fiscal framework and sector expenditure programs. Prudent monetary policy was maintained, the exchange rate was kept realistic, and the economy open to trade within the Southern Africa Customs Union Area. The success in managing the economy and attracting aid donors in the years before the mineral boom entrenched confidence in the plailnilig and budget system, and meant that the political leadership was able to delegate the design of sound policies and their implementation to technocrats. 8.41 Some consideration should be given to the exit strategy for expatriate in-line staff. UNDP's proposal provides for only 18 moilths of expatriate assistance in all posts. If the intention is to ensure the maintenance of perfomlance standards, this may be unrealistic. Ideally, the appointment of nationals to in-line posts should proceed as qualified and experienced staff become available. This will require considerable investments in training at University level. The alternative, advisor-model is not particularly well suited to financial management functions, where systems are in place and critical decisions are of an operational nature. Naturally, Timorese officials may be uilcomfortable with prolonged posting of expatriates in in-line functions. However, a number of countries have successfully accommodated expatriate staff within the management structure over many years (Box 8.2 on Botswana). POTENTIALTHREATSAND OPPORTUNITIES 8.42 One of the fundamental choices in the design of public expenditure management is in the appropriate balance between flexibility and control. Flexibility allows managers to manage, allocating resources as necessary to achieve targets. Coiltrol mechanisms ensure that managers apply resources in line with policy priorities and reduce the risk of corruption. The current system is control oriented, particularly as regards systems for budget execution, ensuring that all major agency transactions are verified independently by Treasury. 8.43 During the 1980s and 1990s, many OECD countries sought to relax central controls, allowing nlailagers greater flexibility through the delegation of recruitment, procurement and expenditure management functions to agency level. It should be stressed, however, that flexibility has been conceded within an environment that ensures rigorous adherence to aggregate expenditure constraints at the agency level, strong systems of internal control and accountability for performance, in terms of service delivery targets and compliance with procedures. 8.44 While the delegation of expenditure control to agency level might a p pear an attractive option to a new Government, it is important to consider whether this approach would be appropriate in East Timor. East Timor lacks strong embedded governance structures-such as notions of compliance and contract, its institutional enviroiunent is weak, particularly as regards national capacity at a technical and managerial level, and the institutional framework for accountability has yet to be put in place. In this context, attempts to dismantle centralized systems of expenditure control, recruitment and procurement, nln the risk of loss of discipline. The expeiience of Mozambique, Malawnd Jamaica following the introduction of delegated expenditure control systems illustrate the potential dangers. 8.45 111 the immediate post-Independence period, therefore, the maintenance and efficient operation of existing, centralized public expenditure control mechai~ismsshould be considered a priority. This will require both political commitlnellt and considerable investment in experienced expatriate personnel and capacity building. 8.46 It should be stressed that centralized expenditure control is not incon~patiblewith deconceilti-ation and decentralization within the public sector. Close coordination between territorial and sectoral planning can ensure that locally identified needs and priorities are taken into account when drawing up sectoral spending plans. Simple measures, such as the involvement of district managers in the preparation of plans and budgets and the identification of district allocations within agency budgets, can significantly strengthen the role of local administration in budgetary decision making. Ideally, an increasing proportion of funding would be executed through district offices, though in the absence of a district banking network this will prove difficult to implement. 8.47 A systein of block grants direct to communities, such as those implemented by the Community Einpowennent Project, can provide important channels for financing local initiatives, at relatively low cost. Equally important, this would strengthen communities' hand when negotiating with central and local administrations regarding the location and relative priority of government interventions. The key feature of these systems is that the distant coiltrols of Treasury and District Finance officers are complemented by the immediate controls instituted by the community, to ensure that funds are correctly applied to address the community's priority concerns. 9. PROMOTING EFFECTIVE GOVERNANCE 9.1 An honest, efficient and service-oriented public administration will be essential to maintain confidence in the new government and the goodwill of the donor con~munity. This chapter addresses three dimensions of this objective. First, it considers a set of issues relating to the size, composition, co~npensationand training needs of the civil service that will support the objective of sound administration while remaining coilsiste~lt with fiscal constraints. Second, it turns to the priorities for developing transparent and accouiltable structures that can minimize the growth of cormpt practices withill the bureaucracy. Finally, it explores the potential benefits that an enhanced role for local communities can provide in terms of quality and cost-effectiveness of service delivery. 9.2 It should be stressed that each of these topics contains difficult issues for a new State to confront-and each will need further discussion and debate. The intention of this chapter is to contribute to that debate rather than to recommend blueprints for East Timor to adopt. BUILDING THE CIVILSERVICE Background and Current Situation 9.3 Following the establishment of a United Nations administration (LNTAET) in East Tiinor in October 1999, it was decided to develop a locally-recruited administration, (ETTA) as a basis for a new government. The circumstances in which ETTA was established were extremely difficult. Alongside the violence, disruptions and damage, there was a collapse of the state and key institutions and the permanent departure of the upper echelon of civil servants. 9.4 Much progress has been made in restoring public services. At the beginning of the administration, essential services, including the provision of education and health, were largely non-functional. Gradually, these services were restored, initially with the help of humanitarian agencies, NGOs and unpaid volunteers with previous experience in key occupations such as teaching and nursing, who provided a starting point for the staffing of ETTA. Terms of einployment and pay scales for ETTA staff, as defined in the budget process, became effective on July 1, 2000'. New key institutions were established including a Cabinet structure with 50 percent Tiinorese membership, a National Couilcil to advise on policy, legal instnlmeilts and planning, a judiciary, a Central Fiscal Authoiity and a Central Payments Office. In September 2001, following the electioil in August of a Constituent Assembly charged with producing a constitution, the Cabinet is now wholly Timorese. 9.5 ETTA was founded under the principle that there should be an efficient and effective civil senlice much smaller than its predecessor under Indonesian rule. Given limited budgetary resources and extensive unemployment, there were strong initial Directive No 200014: Public Seivice Commission, Terms of Employment. 106 pressures to set civil service salaries at low levels and to start off with a very simple salary and grading structure with only seven grades. Employnlent of permanent civilian staff by ETTA was 1,579 in July 2000, and had risen to about 9,600 by November 2001 (Table 9.1)-less than the amount budgeted for as of June 2001, and far below the some 28,000 civil servants in office during Indonesian rule. Table 9.1: Civil Service Employment by Level (2000-2001) 1February June 30 November Minilnun1 July 2000 2001 2001 2001 Education (budget) (actual) Level 1(General Service) 1,664 1,402 Level 2 (Basic Technician) Junior School 765 562 Level 3 (Interned. Technician) High School 4,986 4,815 Level 4 (Senior Technician) 3-year psc* 2,299 2,413 Level 5 (Professional) 5-year psc 603 278 Level 6 (Senior Professional) 7-year psc 201 107 Level 7 (Manager) Specific to job 36 28 Total Civil Service Of which permanent: Armed forces Total government employrneilt -- 8,997 11,154 * Psc refers to post secondary certification. 9.6 Although recruitment in the aggregate kept broadly on track, two main problems became evident: significant shortfalls remain at higher grades, particularly Levels 5 & 6; apart from teachers and nurses, middle-to-higher level staff have usually been recruited with lower qualificatioi~sand experience than originally envisaged. 9.7 With these and other problems in mind, the structure of pay and grades was re- exanlined in March 2001'. A number of recommendations were made including: the decompression of salaries by increasing the salaries of employees on the higher levels (levels 5 to 7); the replacement of a single payment for each level with a more conventional multi-step salary scale; the mapping of staff at the decompressed salary levels onto appropriate salary scales. 9.8 These recominendations remain under consideration. Related work is also underway concerning the design and implementation of an establishment database and an "ivil Service Pay in East Timor: Current Status and Reconlmendations, World Bank, March 30, 2001. 107 establishment approval and control system linked to the budget process and the payroll system. Scope of Government 9.9 The modem concept of goveilunent is limited in scope, with a high degree of delegation and decentralization. The role of govenlrnent is seen as essentially enabliilg and, to a degree, regulatory, rather than executive. Together with this liinitation of scope comes a strong emphasis on value for money and coininitment to service delivery: recognition that the government should be efficient and effective in its use of national resources. 9.10 According to this concept, paying for most goods and services should be the responsibility of the private-sector market economy. There are, however, certain services that the private market econoiny cannot or will not pay for: the government should be responsible for paying for those services on behalf of the nation. That does not necessarily mean it should itself produce the services, still less deliver them. However, the State should retain central responsibility for: Q the maintenance of a framework of legislation and economic policy conducive to law and order, and sustained social and economic development; Q the provision of essential social and physical infrastructure, of the kind that can oilly be provided collectively, such as border controls, external relations and national defence, law enforcement, legal and corrective services, social security, education, health care, roads and public utilities; Q certain social and economic support services, but only where the benefits can be demonstrated tojustify the cost. Agricultural extension services are an example. The Size of the East Timor Civil Service 9.11 In comparison with other low-income countries, the targeted East Timor civil service staffing level is much as it should be. As a comparator, a sample of low-income countries in the 1999 GDP per capita range of $200-550 is used, excluding very large countries (Bangladesh, India, Nigeria and Pakistan) and former Soviet Union countries. The civil service in East Timor constitutes a noticeably lower percentage of the population than the average for the con~paratorcountries (1.3 vs. 1.9percent-Table 9.4), altl~oughin a number of comparator countries in the period of comparison, government bureaucracies were still over-staffed. 9.12 Considerable progress was made towards achieving staffing levels, though staffing levels remain below the levels budgeted for the previous fiscal year (Table 9.1). At the beginning of the previous fiscal year there were only 3,586 employees of whom over half were are on temporary stipend3 contracts. There are 28 permanent positions 'Temporary or stipend workers are divided according to 5 grades corresponding to Levels 1, 2, 3, 4 and 5 of their permanent cotlnterparts and their rates of pay are slightly lower (1-5 percent). The hiring of temporaries does not require the formal procedures needed for permanent staff. filled at Level 7 of which alillost all are judges. The defense forces are currently above target and their ilunlber are due to grow to 1,500by 2005. 9.13 Significant shortfalls remain at higher grades, particularly Levels 5 and 6 where conversioil of existing tenlporaries to penllailent status would be quite insufficient to meet target einploynent levels. School teachers and nurses apart, recruitment experience thus far has been that middle-to-higher level staff have usually been hired with lower qualificatioils and experience than originally envisaged. The aim in such cases is to upgrade their skills subsequently. 9.14 As in all developing countries, there is a concern that some payments may be to non-existent 'ghost' workers. Currently, however, there is a close match between payroll and einploynent information and it seems unlikely that this problem has had sufficient time to develop. The authorities are aware of the possibility of this problem arising and intend to monitor the problem through regular use of attendance records. 9.15 East Timor therefore has the potential to avoid one of the most crippling problems faced by many other developing and transitional countries, an overstaffed civil service, provided that it maintains a strict ceiling on total staff numbers and the wage bill. Overstaffing would reduce the ability to pay adequate salaries to attract qualified and experienced staff, and cause loss of perfonnance standards and management problems. Once it had taken hold it would lead to "crowding out" of non-staff items (such as goods and services) and further loss of efficiency. In this connection, it is important to look at the total wage bill, including teachers, other civilian staff, the police and military persoimel, since ring-fencing any particular category would close off the option of transfeiring resources between categories. Civil Service Pay 9.16 There are serious concerns that the existing civil service salary structure will, if left unchanged, prove unsatisfactory for the needs of the new administration. In particular: the differentials between the highest and lowest paid are narrow by international standards, creating a concern that it will become difficult to recruit andlor retain well-qualified employees. the limited number of pay levels in the present structure may also allow insufficient flexibility to differentiate remuneration levels between individuals possessing different levels of experience and specialist qualifications. 9.17 ETTA pays less than UNTAET, International Agencies and the upper end of salaries paid by the NGOs. ETTA salaries are at least 30 percent less than those currently paid in UNTAET and the relative gap is larger at higher levels (Table 9.2). Comparisons of salaries between the two institutions understate the differences as UNTAET pays an additional family allowance to married enlployees of $11 per month, offers free medical care and inakes an employer's contribution of 12.6 percent of salary to a pension scheme for einployees with inore than 5 years of service. In contrast, ETTA pays no additional benefits on a regular basis at the present time, altl~oughstaff attending approved training courses are paid an additional $5 per day if they reside in Dili and $15 per day otheiwise. At present, ETTA has no increnlents in its pay structure, but only a single fixed pay amount for each Level. LNTAET~has 4 steps according to years of experience on the scale for each Level, but most employees are currently on the lowest step. Salaries of other UN agencies are comparable to UNTAET though with longer scales. One agency currently pays somewhat more than the UN norm but has frozen its pay levels until comparability is achieved. Table 9.2: A Comparison of Montllly Salaries (USD) ETTA UNTAET' 1 NGO ~ ~ r e e ~ i i e n tNGOs Actual I ~ Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 Level 7 Source: Salary scales supplied by ETTA, UNTAET and NGOs operating in East Timor. 1. UNTAET salaries shown include a recruitment allowance applicable from the second step of the scale for each level The allowance varies from USD 8 for step B of Level 1 to USD 124 at step D of Level 7. 2. Guidelines for minimum standards and conditions for NGOs employing local staff in East Tirnor (October 2000). 9.18 ETTA salaries correspond reasonably closely with the midpoints of the salary guidelines agreed by the NGOs and it seems that the NGOs used ETTA scales as a reference point. Salaries actually paid by the NGOs are similar to the agreed rates at the bottom end of each Level, but there is a noticeable tendency for the agreed guidelines to be breached at the upper end. At Level 6, NGOs pay consistently more than the guidelines or the ETTA rate, suggesting that market pressures warrant higher pay levels. 9.19 The private wage sector is extremely underdeveloped in East Timor at present and only limited comparisons can be made with government salaries. Local hotels/lodges in Dili pay about $120 per month for literate staff with limited secondary education - somewhat higher than Level 2 at ETTA. Inexperienced staff with a post-school diploma earn about $170 a month-a little higher than ETTA Level 4. Post-school diploma staff with 10 years experience earn $400 per montll-a higher rate of pay than currently available in the civil service at any level. Professionally-trained local university graduates at managerial level cornilland salaries fioin $15,000 to $25,000 per annum (including a $5,000 housiilg subsidy) depending on experience. Even after subtracting a 10-20 percent discount for the greater job security5 typically found in government UNTAET scales were adopted from those origiilally devised by the UNDP. The UNDP scales are based on a cost-of-living survey carried out in MarchlApril2000 and UNDP norms. This is a norm based on studies of international comparisons of public and private sector wage rates. en~ploynent,the indications are that the private sector pays higher wages to more qualified and experienced employees. 9.20 TA pay scales are conlpressed relative to East Timorese comparators. In general ETTA pay scales are inore conlpressed than those of their counterparts as pay differentials between the highest- anETd lowest-paid are relatively low among ETTA employees when compared with the UNTAET pay structure and actual salaries paid by the NGOs. The highest pay in ETTA is about 4.4 times that of the lowest paid compared with a corresponding ratio in UNTAET of 7.4. Even if it is assumed that all UNTAET employees are on the lowest salary step of their Level, the UNTAET ratio is still higher at 5.3. The same conclusion holds if Level 6 is taken as the highest point in the pay structure6. In ETTA, the compressioil ratio is 3.1 compared to 4.9 and 8.0 in UNTAET and the NGOs respectively. 9.21 ETTA civil service pay is also relatively compressed by international standards. Currently OECD countries have a higher compression ratio than ETTA despite the fact that wage differentials by skill are lower in developed countries. Historical comparisons suggest that coinpression ratios were much higher than the current ETTA value in many developiilg countries in the 1980s (Table 9.3). Lower compression ratios are usually only found in former Soviet Union countries such as Ukraine and Russia where regulations mandated that the ratio of highest-to-lowest pay across the civil service and state-owned enterprises could not be greater than 2:l. The pay structure in China (not shown) is similarly coinpressed in the urban state sector. In the 1980s, there was a tendency for pay structures to become inore compressed in many developing countries as the real value of civil service pay fell in the face of economic contraction and associated fiscal pressures and govenlments felt obliged to protect the real incomes of the lowest paid. Compression ratios nevertheless remained above the current ETTA value. Ghana and Laos were noticeable exceptions where salaries were decompressed under programs supported by the INIF and the World Bank. 6There are currently no UNTAET employees at Level 7. Table 9.3: Civil Service Pay Compressio~lRatios for Various Countries Earliest Period Latest Period Reference Years OECD COWTRJES Belgium -- 5.5 Late-1990s Canada -. 5.2 Late-I 990s France .. 5.7 Late-1990s Germany -- Late-1990s Netherlands -- Late-1990s UK .- Late-1990s USA -. 8.9 Late-1990s Develooina Countries Argentina Pre-1990 Burundi 1984 Cameroon 1989 C.A.R. 1985-88 Gambia 1985-88 Ghana 1984-92 Guinea 1985-89 Guinea-Bissau 1988-89 Laos 1987-88 Malawi 1975-83 Mali 1985 Mauritania 1975-85 Mozambique 1985-90 Niger 1975-85 Nigeria 1975-83 Russia 1980-96 Senegal 1980-85 Sudan 1975-84 Togo 1985 Ukraine 1980-96 Uganda 1983184 Zaire 1985 Zambia 1975-84 Mean East Timor -- 4.4 2001 Sources: World Bank data files and B. Nunberg, "Experience with civil service pay and employment reform: an overview". D. Lindauer and B. Nunberg (eds) Rehnbilitntirig Goverrtntent. World Bank. 1994. Compression ratio equals ratio of highest- to lowest-paid civil servant. Allowances included ~vlvherepossible. 9.22 The case of Indonesia provides a particularly interesting comparison as the ETTA salary structure is to some extent derived from that of the previous administration. In structural (management-based) positions the compression ratio is 6.6:1, but in functional (profession-based) positions it is almost 14:1.The reason for this sharp discrepancy in the degree of compression between the two types of positioils is not that the base salary structures are different, but rather that the incidence of allowances and Izo~zovaviadiffers markedly. Non-transparent payments of this kind are uildesirable as they distort incentives and can create substailtial horizontal inequities in pay between employees holding identical rank. However, even anlong staff holding stn~cturalpositions the compression ratio is higher than in ETTA despite a widely-held perception that pay is too coinpressed anlong this group of Iildonesian civil servants to give adequate work incentives. Table 9.4: Iilternatioiial Conlparisolls of Civil ServiceWages and Employment East Timor Comparator (200012001) ~ountries'(late 1990s) Average annual civil service wage (US$) 1,560 1,415 Civil Service wage relative to GDP per capita (per cent) 4.75 4.25 Government wage 'billrelative to GDP (per cent) I 6.3 I 6.42 Civilian governnlent employees relative to population (per cent) / 1.31 1.90 GDP per capita (US$) I about 329 II 333 Source: IMF and World Bank estimates. 1 . The comparatorcountries are Benin,Burkina Faso, Cambodia,Chad, Gambia, Ghana,Kenya, Lao PDR,Lesotho, Madagascar, Mali, Mauritania,Mozambique, Senegal,Tanzania,Togo, Uganda, Vietnam, Yemen, Zambia and Zimbabwe. 9.23 Taking salary compression and the desire for a 'lean and mean' civil service into account, East Timor's governrnent wage and employment levels are broadly comparable to those of other low-income countries. In comparison with other low-income countries, East Timor's average govenlmeilt wage and its civil service staffing level are much as they should be. As a comparator, a sample of low-income coulltries in the 1999 GDP per capita range of $200-550 is used excluding very large countries (Bangladesh, India, Nigeria and Pakistan) and former Soviet Union countiies. East Timor's average wage is a little higher than that of the average for the comparators (Table 9.4), and relative to GDP per capita. The laudable desire for a lean and mean civil service is reflected in the fact that civil servants constitute a noticeably lower percentage of the population than the average for the comparator countiies. In a number of the comparator countries, governrnent bureaucracies are still over-staffed despite attempts at rationalizatioil and should be reduced further. 9.24 The analysis above suggests two main recommendations, which, however, will need to be reconciled with the emerging fiscal situation; to the extent that the financing gap cannot be closed, the proposals below would have to be postponed.7 Decompress salaries of Levels 5 through 7 and Level 4 excluding education and medical personnel relative to a shadow wage for lowest-skilled employees. Map staff at the decompressed salary Levels into appropriate salary scales. 9.25 There may be a resistance to the proposed decompression of salaries, both on pl~losophicalegalitarian grounds and on the more pragmatic argument that it inight lead to strong protests on the part of the lower-paid, even though their pay would remain significantly above the shadow wage. The fact remains that those employed in the civil service, even at the lower levels, are well off relative to the majority of the population. If it is necessary to raise salaries at the higher levels in order to facilitate recruitment and retain a small group of talented individuals, this would indeed be a worthwhile investinent for the nascent civil service-whereas a corresponding rise at the lower levels cannot be justified on either egalitarian or fiscal grounds. A practical concern about decompression, however, relates to the availability of qualified staff-without which ' Annex 3 sunlrnarizes the decompression proposal and provides cost implications. decoinpression may not have a significant impact on the supply of higher level staff needed by the administration. In certain instances, it may take a generation to raise local skills to needed levels-in which case, decompressioil alone will not suffice, and other means to supplement the local workforce may be needed (as discussed in the previous chapter). 9.26 It is evident that the mapping of staff onto salaiy scales would have to await a thorough job evaluation exercise, since it would otherwise lack transparency and credibility. Job evaluation is not yet a critical issue within ETTA, but it will become so. When differential pay scales are introduced, individual staff throughout the organization will raise justifiable concerns about the duties and respoilsibilities of their own jobs in coinparison to ostensibly identical (and identically-paid) jobs elsewhere. In this context, a case can be made for the introduction of a personnel management cadre-made more effective through training and the provisioil of tools such as coinputerized systems for HR management, Establishment Control and Training Needs 9.27 It certainly appears that there will be a continuing and substantial need for training over the next few years and the success or otherwise of the training program may have a critical impact on the future development or otherwise of civil service staff. Training can also be an effective response to the concerns posed by low qualifications of middle-to-higher level staff. 9.28 The UNTAET Civil Service Academy is a good basis on which to develop training programs for ETTA staff at all stages in their careers. The effectiveness of training programs can be enhanced through a con~prehensivetraining needs analysis of the civil service and, if warranted, the restructuring of training prograins in the light of this analysis. It is likely that many of the elements of training will be drawn from the existing Civil Service Academy program. Other elements could be entirely new, however. Equally, the course materials and content and types of trainers used may be different. Civil Service Law and Regulations 9.29 Legal traditions and systems differ, but there is a consensus among democratic nations that the rule of law, technical and managerial conlpetence, reliability, predictability, accountability, transparency, and citizen's participation are all principles that should guide public administration. In addition to transmitting general principles and standards for performance through civil service legislation, the state or state organizations must also fulfill the traditional responsibilities of an employer. This requires specific regulations covering staffing and career management policies-including systems for selection, recruitment, promotion, and remuneration. 8 Care is needed to ensure that the such a system is tailored to East Tirnor's specific needs. 9.30 Employneilt in ETTA is curreiltly goveined by UNTAET Directive No 200014, "Public Sellrice Commission Terms of Employment", which became effective as of July 1, 2000. However, it covers oilly the bare essentials, and a comprehensive set of regulations to govern the employment of public servants therefore needs to be drafted. 9.31 According to the OECD Civil Service Legislatioil Contents checklistg, a law defining the civil service should contain provisions protectiilg the civil service from political iilterfereilce or other kinds of interference; equally importantly, it should contain provisioils aiming to raise the quality and the performance of the staff subject to the law. In order to benefit from more secure appointment conditions, civil servants should be required to meet certain quality and perfonnailce standards. 9.32 Civil Service Law should thus strike a balance between the duties and accountability implied in a public office and the rights that ensure professional integrity in carrying out the office. It must meet a number of different objectives. According to the OECD Checklist on Secondary Legislation (and Other Regulatory ~nstrurnents)'~, regulatory instnunents are invariably required to implement primary legislation relating to the civil service. They are typically made by governmental bodies that, under the terms of the Civil Service Act, either administer the civil-service legislation, or some part of it, or are given management responsibilities in relation to the civil service. One example of a code of conduct is the New Zealand Public Service Code of z on duct". PROMOTINGEFFECTIVEGOVERNANCE REDUCINGCORRUPTIONOPPORTUNITIES AND 9.33 In 1997, a World Bank paper on identified two important areas in which policies may be directed to reduce opportunities for corruption: nzacroeconomic and sector policy refornzs that contribute to the expansion of markets and reduction of rents; and strengthening institutions to control corruption through: building traditional systems of well-performing government: a professional civil service, souild financial management, disciplined policymaking, and a balance of responsibilities among central, state, and local governments; strengthening the legal framework, including the judicial system; and increasing transpareilcy and introducing other measures that strengthen the role of civil society in demanding better government. 9.34 Timorese participants in the Bank-facilitated anti-corruption workshops in January and Febi-uary 2001 identified a range of priority actions that they believed should form the basis for near- and medium-term action to build transparency and accountability in their new nation's government, focusing primarily on strengthening institutions to control coiruption and strengthening the role of civil society in fostering transparency and accountability. The findings from the workshops in early 2001 were elaborated upoil in 9 SIGMA Papers No. 5, Organisation for Economic Co-operation and Development, Paris 1996. 10 SIGMA Papers No. 14, Organisation For Economic Cooperation and Development, Paris 1996. I IPublished pursuant to section 57 of the State Sector Act 1988. I2 Helping Countries Combat Co~luption:The Role of the World Bank. World Bank, September 1997. the Irnplenzentatiolz Plarz for Elenzerzts in tlze Franzework Anti-Corniption Strategy for East Tinzor, and are utilized here. Individuals and groups in civil society, political parties, and the transitional government moved forward to lay the groundwork for some of the identified priority actions during mid-2001, especially in civil society, but most of the institutional initiatives were placed on hold to await the results of the Constituent Assembly and the installation of a new govenmlent. On September 20, 2001, the Second Transitional Government of East Timor was announced, with Ministers and Vice Ministers selected from among the winlli~lgparties from the August 30 election. Institutional Development and Phasing 9.35 Competing priorities and severe capacity constraints make it iinperative to define a sequence for governance reforms. The discussion here attempts to do so, setting forth short and medium-term recommendations-refemng to within six months and three years, respectively. These suggestions do not include a number of "integrity plan" items that arise in most ailalyses of corruption and governance reforn~.For example, strengthening criminal laws against corruption and establishing an anti-comption commission are common features of integrity plans; work on these will need to proceed based on experience developed by the newly independent nation, but are not included here. Basic law and order are of course fundamental to good governance-and will need to be addressed in parallel with anti-comiption efforts-though the post-99 ballot experience to date on law and order is clearly encouraging. 9.36 The steps that are suggested follow a sequence from (i) the easiest and most pressing measures for improving transparency and the rule of law, to (ii) core systems for administration and oversight of public resources, along with elaboration of basic administrative and economic laws-and their implementation. Seeking to do more than this within the first few years of independent government in East Timor would risk a loss of focus and ~nomentum.There is also a wealth of comparative experience and research to suggest that getting the core functions right is necessary to effective governance, and thus would ideally precede efforts to establish anti-comption bodies and codes. 9.37 The analysis that follows focuses on the future, but at the same time must take account of structure and practice of governance during this transition period. Especially in ally discussion of legal change, one must pay close attention to the two existing sources of Tiinorese law: Indonesia and UNTAET. The new administration, like successor regimes in other contexts, will need to build on those elements-improving some, discarding others, and using the remainder as residual law until reforms can be enacted. In addition, Indonesian and transition-period governance practices have already laid down a pattern of path-dependency that will shape the post-transition context. To the Indonesian pattern of cronyisn~,UNTAET brought the complicatio~lsof an international enclave-including dramatic salary differentials and massive commodity imports. The other factors likely to constrain post-transition governance include fear and mistrust brought about by the violence of 1999, and a co~ltinui~lgpattenl of goveining by fiat in a situation of legal and procedural flux. Administrative ~riorities'~in the Short-term - Citizen Illoltitorilignlid ColnplnilitMeclinnis~ns 9.38 Civic nzolzitoring: A high priority has already been assigned to increasing the transpareilcy and accountability of both donors and government in the allocation of resources (including trust fund monies) and implementation of reconstnlction and developillent projects. Between the February workshops and the August election, actioils in this area have been taken by civil society groups: NGOs and youth groups have iilcluded anti-corruption and good governance elements in their civic education campaigns, and a sub-grant fiilanced by the Danish Trust Fund for Governailce was given to Yayasan Hak, a highly-regarded Timorese human and civil rights NGO, to develop staff skills and capacity for inonitoriilg government and advocating for transpareilcy and accountability. The Yayasan Hak activities include cross-visits and training with Iildonesian NGOs involved in similar activities, public information campaigns, and outreach. 9.39 These initial civil society initiatives could be broadened to include the ongoing moilitoring of and reporting on government performance in areas that directly affect the general public, such as the delivery of services-e.g. utilities, transport, infrastructure, health, education, sailitation, police. Some activities along these lines are underway as NGO initiatives or under such decentralized donor projects as the Community Empowerment Project. Such efforts could have a more powerhl result if they were combined illto a progranlme of focused research and reporting along the lines of the "service delivery survey" (examples of these have been supported by WBI in several countries, including Tanzania and El Salvador) or the "report card on public services" (these focus on urban areas in India, and are produced by the Public Affairs Centre in Bangalore) or "social monitoring" studies (such as those conducted by SMERU in Indonesia). 9.40 Citizen conzplaint nzechanisnzs:Acting upon suggestions to establish confidential coinplaint mechanisms, the sitting Inspector General in the previous ETTA cabinet secured conlplaint boxes in two locations in the ETTA headquarters in Dili, and discussions were held with representatives from local NGOs on other possible confidential and secure channels for complaints-boxes in each district office and/or in local churches, boxes in Yayasan Hak's district offices, a post-office box number (when mail systems are re-established). Complaints received through any of these channels could be logged by the group managing that channel and then forwarded directly to the Ombudsman or equivalent (depending on what institutions are adopted), which would log their receipt and keep complete records of steps taken as well as the date of each event in the case. Credible reports of corrupt acts should be required to be referred automatically to criminal prosecution. This type of information should be the subject of regular public reporting by such a watchdog agency. -- 13 Excludes items covered in previous chapters. 9.41 Anti-corrtption in civic edzrcntio~i: Many NGOs, religious and social organizations, and educatioilal iilstitutioils have included ailti-corruption principles and concepts in their outreach during the election campaign period. This outreach could be broadened to iilclude standards of govenlmental integrity and performance, and information on relevant government agencies and complaint channels. Such a campaign might also inform communities about how to set up public monitoring and audit procedures focused on local governmeilt operations, infiastnlcture projects and public service provision. Review and ClarilS,Civil Service Riiles, Biiild Citizen I~ip~itilito Piiblic Service Conznrission 9.42 Civil-service recruitment has been viewed as problematic. Strong criticism has been leveled at the perceived lack of transparency in recruitment to the public service during the transition. A teacher recruitment scandal uncovered by the Office of the Inspector General reinforced this negative perception. A short-term priority should be the establislment of an independent commission to oversee this area, with substantial participation fiom civil society. UNTAET has in fact acted in this area, establishing a Public Service Commission under regulation no. 200013. This cormnission sets personnel and recruitment policies, arbitrates employment-related disputes and cames out oversight. The Transitional Administrator appoints and has the power to remove all members according to (loosely) stated criteria. 9.43 Improving trailsparency and integrity in public service recruitment inay indeed be a potential "quick win" and therefore an important early focus. While it should help improve the efficiency and effectiveness of the public sector, its actual importance is likely to be mostly symbolic. A noticeable improvement in this area can send the signals, first, that it is possible for East Tinlor to effectively address governance problems, and second, that relatively petty forms of corruption are intolerable-and by implication more serious abuses are even less so. Given the "hot button" reaction of many Tiinorese to this set of problems, this initiative might attract the most willing participation, and so could be the best one to start with. But there is a risk that momentum could stop here; hence, action in this area should not be allowed to substitute for equally or more important goals. Clarification of the hiring practices in autonomous agencies such as the Power Authority is also needed. More broadly, oversight capacity from both the government and civil society needs to develop into an effective means of ensuring strong standards of governance in autonomous agencies, particularly when the latter may be major contributors to, or claimants of, the government budget. Enforceable Official Code of Condiict Applicable 9.44 UNTAET regulation No. 2000/23 established the cabinet and went some distance toward adopting a code of conduct for ministers. Cabinet members are required under the regulation to declare their assets and business interests. This infornlatiou is kept in a register maintained by the Transitional Administrator-there are no provisions for disclosiilg it to the legislature or the p~~blic.The regulation also applies conflict of interest i111es to the cabinet and provides for the Transitional Administrator to elaborate a code of conduct for cabinet members. 9.45 The Second Transitional Government could send an important signal of its commitment to govenmental transparency and integrity by strengthening existing regulations on official conduct. This could begin with a debate on "what kind of governnleilt East Timor wants." However, if this drags on, the utility of such an activity as a signal of seriousness would be lost. It would be a useful objective to prepare a legislative bill on official conduct as one of the first acts of the new Timorese legislature. This will require both consensus-building and technical drafting work. 9.46 The Constituent Assembly, working with the Second Transitional Government, could develop a public sector "transparency plan" to provide for the publication of policies and budgets, as well as information that would have to be disclosed under other refornls discussed in this chapter (proposed regulations, personnel procedures, asset declarations, bidding procedures and awards, etc). Sirrzple Cllanges to Adr~zirristrative Procedures 9.47 UNTAET has taken some additional steps towards administrative accountability, but much remains to be done. Regulation No. 1999/4 establishes the Official Gazette as the organ for publishing regulations, but unfortunately it does not require publication as a pre-condition to a rule's effectiveness. Indonesian law applicable to East Timor does not include legislation on administrative procedures, although in principle government actions are subject to judicial review on grounds of legality and rationality. Indonesia has a system of administrative courts that have evolved a modest administrative case law, although these courts' lack of enforcement powers and sanctions allows their decisions to be largely ignored. 9.48 This is unsatisfactory for East Timor, and both short and long-term improvements in administrative transparency and legal accountability will be needed. In the short term, it should be possible to strengthen the UNTAET provisions in force, mentioned above, ensuring that basic transparency standards and procedures are in place. This could take the form of a fairly simple act anlending regulation No. 1999/4 to require publication in the Official Gazette (in all or most of the relevant languages) as a preco~zditionof a regulation's or directive's coming into legal effect. One could add to this some other simple procedural requirements such as public opportunity for notice and comment in advance of promulgation. Such a regulation could easily be drafted and proposed within a couple of months. Administrative Procedures - in the Medium-Term 9.49 East Tiinor's doininant source of revenue in the inedium tern1 is expected to accrue from offshore oil and gas revenue and royalties. Governance arrangements relating to these prospective flows were discussed in Chapter 6. In the short tenn, however, revenue from onshore taxes will be higher. 9.50 As customs revenue is likely to remain the major source of oilshore revenue in the medium-term, it is important to ensure integrity in this area. Concerns have already been expressed about under-valuation and manipulation of customs exemptions. A strengthened Supreme Audit function (see Chapter 8) would serve to allay such concerns, but it would also be worth reviewing procedures and practices with a view to simplifying them, increasing transparency and aligning incentives. Minimizing exemptions from the existing tariff rate would also reduce the opportunity for fraud. In the longer term, particularly if offshore revenue materializes as anticipated, consideration could be given to reducing the tariff rate while maintaining its uniform structure. Developrrzerzt arzd I~rrplerrzerrtatio~zof Standardsfor Freedorrr of I~~forr~rationarzd Adrrrirzistrative Procedure arzd Review 9.51 F~-eedonzof Ilzformation: A medium-term objective would be to institutionalize transparency requirements and procedures by means of a freedom of information act (FOIA). This kind of legislation creates a legal presumption in favor of disclosure of official information of all kinds, with carefully crafted, narrow exceptions. It also spells out procedures for requesting information from government agencies, third party review and adjudication (in different models, by a parliameiltary committee or a court), and appeals. 9.52 Ad~nirzistrativeProcedure: This is defined in administrative law, which refers to the standards of legality, rationality, and conformity to rule-making procedures that are used to restrain ministerial and bureaucratic actions. Thus, an agency creating or implementing a regulation can be sued on the grounds that the rule violates its enabling statute, general principles of legality or rationality, standards of proper procedure, general limits on the agency's or the bureaucrat's authority, etc. Soine countries (e.g. Indonesia) develop these rules case-by-case and agency-by-agency, bringing in some general principles of accountability across the board. Other countries (e.g. the US) have an Administrative Procedure Act that provides uniform standards applicable to all, or most, agencies, in addition to agency-specific rules and case law. A functioning systein of administrative law can be a vital anti-corruption bulwark, but creating such a system in a context of weak governance presents a difficult challenge. 9.53 Developing an Administrative Procedures Code is a longer-term undei-taking that would require some comparative study, a review of applicable law, and careful attention to its confonnity with the constitution to be drafted for East Timor. The first step here inight in fact be a constitutional provision setting forth some simple objectives and standards of administrative accountability, and requiring the post-independence legislature to enact an Administrative Procedures Code within a year or two. The US APA provides one possible model, although there are clearly others that would be relevant. In addition to drafting and enacting such a law, the Timorese will need to plan how it is to be implemented in the fonn of agency-specific rules, and design the mechanism of administrative and/or judicial review. A new act will need to designate where judicial review is to take place-regular or administrative courts-preferably the fornler, since this approach would bring the full range of judicial authority and sanctions to bear on the iillplementation of administrative law. Review rrrtd Strengthen Proc~irerrzerztRules and Practices 9.54 UNTAET regulation No.2000110 sets forth a reasonably strong public procurement law. The regulation covers the threshold amounts required for different tendering methods, pre-qualification and bidding procedures, content of bid documentation, publication and record-keeping requirements, and protest and appeal mechanisms. It also prohibits the offering of "inducements" to governmeilt procurement personnel, though it does not set any penalties. 9.55 This area would benefit from a few modest steps. One would be a review of the rules with a view toward possibly simplifyng them (they may be too complex in the East Timorese context). A second step would be to ensure that these procedures are not relaxed in future. Broad procurement standards have been written into some recent constitutions. For example, the South African constitution of 1994 requires parliament to make legislative provision for "independent and impartial tender boards" operating under procedures that are "fair, public and con~petitive," and protected from political interference. Last, and most importantly, there will need to be a review of implementation and compliance, with a view toward ensuring strong enforcement in the post-transition era. Legal prioritiest4- in the Short-Term 9.56 The rule of law requires some consensus on the sources of law. It will therefore be important to clarify the basis and content of East Timorese law. W T A E T Regulation No. 199911 recognizes "the laws applied in East Timor prior to 25 October 1999" as the law of the land, except to the extent to which they may be inconsistent with applicable huinan rights conventions, the WTAET mandate, or UlVTAET regulations. The regulation also specifically repealed some Indonesian legislation. Where not contradicted by regulations or otbenvise invalid, Indonesian law applies exclusively (a result of annexation, which replaced previous Portuguese law with Indonesian). This makes I4This section collsolidates legal priorities discussed elsewhere in the report and links them to integrity and goven~anceconcerns. Indonesian law the residual source of law for public administration, judicial procedure, and commercial transactions. 9.57 A number of laws that could be quite important in East Ti~norwere enacted in Indonesia in mid-1999, prior to the cut-off date fixed by regulation 199911. These include: Law 22 of 1999 on Local Governance; Law 25 of 1999 on the Financial Distribution between Central and Local Governments; Law 28 of 1999Regarding State Administrators Clean and Free of Corruption, Colh~sion,and Nepotism; and Law 31of 1999 Regarding Eradication of Criminal Acts of Corniption. 9.58 These and other laws apply in the absence of later (UNTAET) regulations or findings to the contrary. Such laws (in principle) provide the starting point for East Timorese legal development. The decentralization and local government finance laws would require substantial devolution of power to elected councils at the district level (in Indonesia, kotn/kclbupnterz), as well as a 25 percent share of domestic revenues to be transferred (using a formula not specified in the law) to the districts. The anti-corruption laws provide for asset declarations by senior officials and an investigating commission to verify them, as well as criminal penalties and an anti-corruption commission to be specified in a later enactment. 9.59 In practice, it appears these laws have been disregarded during the transition period. Apart from the repealed legislation listed in Regulation No. 199911, no attempt appears to have been made to clarify the general principle that Indonesian law applies where not inconsistent with UNTAET regulations. In East Timor, this situation is highly problematic. A capable common law court could sort this out through case-by-case interpretation, but East Timorese courts are not trained in common law teclmiques. One response would be for the drafters of the new constitution to write provisions clarifying the relationship of received Indonesian law to both UNTAET and post-transition legislation-e.g. what provisions are repealed, replaced, modified, etc. Property Rigltts 9.60 Investment growth requires clear definition, recognition, and disciplined enforcement of property rights in such areas as: land and other real estate; movables, intangibles, security interests; corporate shares and control rights; transfer, i.e. freedom of contract and contract enforcement; and protection of usage rights, hence investment expectations, against detrimental regulatory, political, or criminal action. 9.61 In govenlance ternls, inlprovemeilts in the protectioil of property rights places the initiative for their ellforcement and for their profitable exploitatioil (leaving externalities aside) in the hands of owners. In doing so, the system restricts the potential role of government and of middle-men as holders of coiltrol rights (monopolists), aiid subjects official action in this area (e.g. adjudication and regulation) to stringent standards of procedural regularity and transparency. 9.62 The clarification of one set of property rights-interim land rights-is a relatively accessible short-term target. During the UNTAET administration, a number of early steps were taken to alleviate the chaos in the land tenure system brought about by the violence, migration, and destructioil of records in 1999. The Land and Property Unit, along with District Administrators, have been resolving land-claim disputes under guidelines developed in early 2000. 9.63 It will be important to review the laws, procedures, and forums in which land rights are recognized. There are several integrity concerns potentially arising here. Conlpetiilg priorities-equity, harmony, security, investment-often lead to conlplex rules and procedures that open up large avenues of discretion and potential abuse. One steptaken in 2000 that is of questionable legality and probably sent a bad signal was the retroactive invalidatioil of land sales and leases involving non-resident Indonesians and Indoilesian firnls (Reg. No.2000127). Concerning the involvement of District Administrators in adjudicating land disputes, there are countless cases in other countries where placing this function in the hands of local executives lends itself to favoritism and corruption of all kinds. There should be a review of these processes and their results. Such a review could suggest alternative approaches, such as a temporary land tribunal including a few trained and credible people, which could "ride circuit" in order to hear either first-instance cases or appeals from the District Administrators. 9.64 Other points deserving intervention are these: (a) looking to best international practice and local conditioils to design interim property rights vehicles for potential investors; (b) if liinits on land transfers to foreigners cannot be avoided, working to define them clearly and appropriately; and (c) assessing property registries (land, collateral, companies, etc.) to increase effectiveness and squeeze discretion out of the registration process. 9.65 As discussed in Chapter 3, there is a need to review commercial, banking, bankruptcy and contract legislation in order to foster an improved business environment climate. UNTAET Regulation No. 200018 sets out a simple but reasonably effective banking law, with a uniform $2 million miilin~urncapital requirement, stringent rules governiilg the process of granting licenses, and a liberal regime for foreign bank branches. 110 other exanlples of legislatioil on these matters were available, and so Indonesian law applies to those areas. Revisiolis to tile Draft I~tvestrrterttCode 9.66 Irzvestrnerzt code: An important governance benefit of a well-crafted i~lvestment regime in East Timor would be to encourage the entry of "respectable" investors and owners. East Timor will clearly need to draw on international capital markets, business and production know-how, and regulatory capability. 9.67 The Investment Institute has drafted a law creating an investment "gatekeeper" agency. In other countries, these types of agencies have frequently been the scene of substantial corruption, since they often command the most important control point in the chain of approvals. Since the legal structure-into whch the investment rules will be inserted-is in flux, the prospects for such legislation to be well-implemented are not strong. Moreover, other laws and regulations are not explicitly pre-empted, and other regulatory functions are not brought under the investment authorities. The draft law therefore does not provide a "one-stop shop," nor does it attempt to assess or to rationalize the body of rules and procedures potentially applicable to foreign investors. A combination of liberal rules, oversight, appeal rights, transparency, and internal competition could drastically reduce the risk. The draft should be carefully reviewed and strengthened with these concerns in mind. Legal Priorities - inthe Medium-Term 9.68 Regulatory review leading to a strategy for regulatory sinzplzfication and/or- deregulation under the new government. Here again, LINTAET has taken a positive step, setting out a simple regime of business licensing in Regulation No. 2000/4. Fees and penalties are spelled out in the regulation, and street hawkers are exempt. Unfoi-tunately, the rule contains no standard of timeliness in the granting of licenses, and it states explicitly that other regulatory requirements remain in force. Again, one would need to look to Indonesian law and current local practice to see what these are, and the design of regulatory reforms can begin from that point. Plarz for Stre~tgtlzenirzgand Settirzg zrp Altenzntives to Con21ne1-cia1Adjiidicatiorz 9.69 Here, the first step would be to review commercial adjudication institutions and practices, in order to design improvements or alternatives to be implemented post- election. UNTAET has enacted regulations on the Judicial Service Commission (No. 1999/3) and the organization of the courts (No. 2000111). These rules set up a commission under the authority of the Transitional Administrator, and a system of eight district courts and an appeals court. The accountability of judges to the Commission is defined fairly clearly. As the judicial system is strengthened over the long term, it will likely be worthwl~ileto look into the alternatives of a conlmercial chanlber at the district and appeals level, or a separate commercial court, andlor a set of alternative dispute resolution inechanisms that could include an arbitration and mediation bodies. The involvement of foreigners should be encouraged in this process-foreign judges and arbitrators, as well as international appeal forums. AILElectoral Law, wit11 Political Iiltegrity Rilles 9.70 East Timor entered its first elections with only rudimentary UNTAET-issued regulations guiding the August 30, 2001 election of the Constituent Assembly; an early priority for the Second Transitional Govenllnent will be adopting its own law defining standards of campaign behavior, financing practices, and party governance. The Pact of National Unity, signed in the run-up to the August 30 election, did include a code of conduct for political parties, and this could be the basis for developing an electoral law. Reform in this area presents serious challenges in even the most stable industrial democracies. At an appropriate time between election cycles, Timor could convene a working group coillprised of party, NGO, government, and other stakeholder representatives to develop a specific plan for this area-and to extract commitments (both domestic and intenlational) to implement such a plan. International technical assistance (featuring assistance by experts from leading Asian democracies) will probably play a key role. 9.71 As a complement to the traditional approaches to improving governance, detailed above, the authorities might wish to consider the benefits of less conventional approaches that give more decision-making power, and resources, to local communities. Numerous countries have moved away from the traditional model, where the central ministry is responsible for both policy formulation and service provision and all personnel in its area of activity are civil servants, to outsourcing of specific non-core activities or more innovative forms of public-private partnerships. The experiences of other countries can be useful in helping East Timor avoid some of the shortcomings of the traditional model. 9.72 Policy direction and project implementation in East Tirnor have followed traditional sectoral paths--each ministry has received policy advice and resources to spend in building structures and providing services in its "area of competence". Similarly, a civil service has been established along traditional lines of each central ministry providing and controlling the personnel required to deliver that ministry's services. The results already include comnlon problems of civil service absenteeism, lack of discipline, private charging for publicly-provided "free" services and corporative rent- seeking on the part of the scarcer human resources. At the receiving end, local communities do not have the information or the mechanisms (nor the perception of rights, given their past experience) to ensure that public service provision complies with requisite standards. 9.73 The preceding sections of this chapter provide practical suggestions on mechanisms to increase transparency and promote complaints, but these are unlikely to be fully effective solutions. An alternative would be to rethink the structures of public service policy fonnulation and provision in ways that enhance the power of the people directly affected-the users of the services concerned. This would entail a different system of service delivery, with enhanced decision-making and monitoring by local communities. The central principle applied would be that of subsidiarity, namely that each decision should be taken as close as possible to the citizens directly affected. The effect would be an eillpowe~nlentof local communities, and the initial experience with the CEP projects would suggest that significant benefits in tenlls of goveillance and cost- efficiency of service delivery could be expected. 9.74 In recent years, many countries have made advances in the direction of decentralization, particularly in education, health and roads, each a major expenditure item in national budgets. The experience has been that major improveineilts can be expected where managers are made accountable to the clients and have the authority and resources to make key management decisions (including who gets hired and how the central ministry's regulations are modified to suit local conditions or preferences). 9.75 Governance is improved by making the service-providing personnel, such as teachers and nurses, respond to their immediate employer, the local community council, for example, rather than to a distant ministry. The local community could decide how many teachers or nurses to employ and how much to pay them, with the central ministry providing quality control in the form of training and accreditation of personnel, the setting of core syllabus requirements and the provision of text-books, for example. Cost- efficiency is enhanced by local communities having a direct stake in getting as much value as possible out of their resource allocation, selecting cheaper inputs and complementing purchased inputs with their own materials or labor inputs. 9.76 This option would imply transferring general grants to local communities, together with the responsibility for deciding on the level and composition of basic services they would purchase in each budget year. Communities would decide on what primary health care, basic education, road maintenance, agricultural extension services, etc, they would purchase, and from whom. The service personnel would be centrally- accredited but locally hired (and fired if they don't perform); the compensatioil packages could also be a matter for local negotiation. By bringing the transaction to the local level, the efficiency of resource allocation and the responsiveness of service provision can potentially be improved. 9.77 Deciding on the amount of the annual grant for each community could be done on the basis of a variety of indicators of need-population size (perhaps broken down illto age cohorts to cater for special needs of more vulnerable sectors), area covered and its nature, length of existing road network, poverty indicators, etc. The grant would also take into account the amount of revenues local governments themselves can mobilize. The central agencies would be available to provide advice or technical assistance, preferably for a fee (to prevent wastage of a "free" good and to facilitate monitoring of the usefulness of the central agencies' services; in financial terms, it washes out). Communities could purchase services from private suppliers if they preferred (e.g. church schools, NGO clinics) and could be allowed to supplement their block grant with local tax resources or goods and services, by collective decision. 9.78 This system could also have significant benefits in tenns of contaiiliilg the overall cost of administration, both national and sub-national. The 13-district inodel that the I~ldoilesianadministration inherited froin the Portuguese coloilial period was developed early last century and there is no reason why it should coiltinue to be the most appropriate, deinocratic or cost-effective model to satisfy present aspirations and cii-c~~n~stances.Coinmunications iinproveinents have shortened distances and the principle that the lowest capable level of a hierarchy should take each decision, is now a more entrenched tenet of denlocracy. 9.79 In a small couiltry like East Timor, there would appear to be little justification for more than two levels of government, a central administration (with some generic supervision and coordination roles decoilceiltrated at the current district level) and a strong local government, perhaps consisting of cominunity councils at the cunent sz~co level. I11 any event, this is a decision with significant budgetary repercussions and should not be rushed into. It is preferable to start with a simple structure and expand it gradually if needed, as it is much easier to create new civil service positions later than to reduce them if an initial structure proves to be uimecessarily heavy. 9.80 In ternls of policy-making, some areas, such as national security, economic and financial policies and overall policies withrespect to health, education, social security, and national infrastructures, are clearly the responsibility of central government. But most of these areas have important local impacts and the decentralization model should be able to accommodate local preferences. Given East Timor's rugged geography, many areas are sparsely populated and difficult to access, making it difficult to ensure the provision of reasonable services of, for example, health and education to all communities. Local decision-making should perniit communities to opt, for example, between spending a given local budget eilvelope on improving the access roads and buying a school bus or covering the cost of miming an additional small school or clinic. Also, flexible devolution, with block grants as outlined above, should encourage con~munitiesto complenlent their centrally funded budget with local resources such as, for example, voluntary labor for road maintenance. 9.81 There is no off-the-shelf model of local government that is readily available and can be recommended; rather, the solution most appropriate to East Timor should emerge from a sequence of activities within a decision making process. The issue of local govemnent is expected to be one of the themes for discussion in the context of a wide- ranging national consultative process that will provide input for the new nation's first National Development Plan, to be ready by independence. Studies are already being undertaken, and these should provide guidance to the East Timorese in deciding on their government structure. There are several options for the future structure and role of elected councils, and the way that they should interact with the sectoral agencies. A legal basis for the village and sub-district developnlent councils already exists (UNTAET Regulation 13/2000), but this will need to be rethought now that the country is getting over the reconstruction phase. ANNEX 1: CHARACTERISTICS OF POVERTY 1. This annex provides an overview of poverty in East Tinlor based on h~donesian data through 1999. 2. During 1999, 56 percent of East Timor's population were classified as poor,' inlplying some 500,000 people lived in poverty. Povei-ty in rural areas was significantly higher than in urban areas and, given the distributioil of the population, the majority of the poor lived in rural areas. These poverty estiinates rely on Indonesia11 data sources, in particular the SUSENAS, the national socio-economic household survey.3 East Tirnor's poverty rate was twice the Iildoilesia~lpoverty rate of 27.1 percent. The 1999 poverty rates reflect the effect of the financial crisis. Most important for the poor was the rise in inflation, in particular the dramatic rise in the price of rice. The relative price of food rose 160 percent from February 1996 to February 1999, in contrast to the price of non-food components of the CPI which rose 81percent over this period. Education of Household Head 3. Poverty is multidimensional-besides expenditure poverty it also encompasses access to education and health, which are both important in their own right and in ensuring that individuals can take advantage of opportunities to improve their welfare. In East Timor as elsewhere, there is a strong correlation between expenditure poverty and low levels of education. Anlong the poorest quintile of the population, based on consumption expenditures, almost tluee quarters of household heads have no schooling (Table 1). This drops to 28 percent among the top quintile of the population based on expenditures. - I A4eastlrenlents of Poverty in hzdoilesin, 1996, 1999aiid Beyoiid, (2000), Menno Pradhan, Asep Suryahadi, Sudaino Sumarto, and Lant Pritchett, SMERU Working Paper, Jakarta. A household is deemed poor if it has insufficient expenditures to afford a basic bundle of foods that gives 2,100 kcals per person per day, and an essential bundle of non-food goods (see Pradhan, Suryahadi, Sumarto and Pritchett (2000) for tlie construction of the poverty lines). ; Tlie SUSENAS was administered by the Iiidonesian statistics agency and, until recently, was the oilly comprehensive data source on poverty available. Tlie SUSENAS has a core-rotating module design. The best data for poverty estimates is based on the expenditure module data conducted every three years, with 1996 and 1999 being the last two such surveys. This survey is representative for provinces and its urbantrural breakdown. I11 addition, every year a larger survey, called tlie core SUSENAS, is implemented annually. Estimates of poverty can also be derived from this survey as a short consumption module is also included as part of the survey. Tlie 1999 numbers presented are based on the 1999 expenditure module. ANNEX 1 Table 1: Poverty and Educatiol~of Household Head, East Timor 1998 I Percent I Quintile1 Quintile 2 Quintile 3 Quintile 4 Quilltile 5 Total No School 67.0 x + 7 + K Less than primary 10.7 Primary 11.3 Junior high school 1 3.1 4.8 More thanjunior high school 2.4 6.3 Total 100.0 100.0 Source:Pedersen and Arnberg (1999) Occupational Status of Household Head 4. Almost 70 percent of household heads report agriculture as their occupation, whereas among the poorest quintile of the population based on expenditure, 92 percent live in households with the household head employed in agriculture. This fraction declines to 32 percent in the top quintile of expenditure. Another 20 percent of the population lives in households headed by someone employed in services and this sector is largely where the non-poor work. Of the bottom quintile, only 5 percent of households belong to this sector and almost half of the top quintile are headed by households employed in services. Demographic Structure 5. There is no discernible relationship between poverty and age of the household head. The presence of a widow correlates with poverty level. 18 percent of the poorest quintile have a widow but this declines to 11 percent in the richest quintile. Poverty is also strongly correlated with household size-with poorer families having larger families. However, this result should be viewed with caution as the analysis does not account for economies of scale. Figure 1: Poverty by District, East Timor, 1998 Povertv Incidence bv District. East Timor. 1998 I I Share of Poor bv District. 1998 ,- OECUSI KOVALIMA L I So~uce:National Socio-economic survey (SUSENAS) core, 1998 Regional Dimension of Poverty 6. The SUSENAS core has the advantage that it is representative at the district level and therefore allows poverty estimates to be generated at the district level, though sample sizes are small and the results should be interpreted with caution. The estimates indicate that Dili, the only relatively urbanized district, had the lowest poverty incidence (around 10 percent) in 1998, while Baucau had the highest poverty incidence, with over half the population living in poverty. But the distribution of the poor across districts depends on ANNEX 1 the population in the district. Baucau had the highest poverty incidence, and the largest share of the poor living there (19 percent), but Aileu with the next highest incidence of poverty had only 6 percent of the poor resident there. Table 2: Education Indicators by District Gross junior secondary Gross Primary Enrollment Rate enrollment rate District I Males I Females 1Total I Males 1 Females / Total Kovalima 91.6 104.8 97.3 45.0 58.5 52.0 Ainaro 66.0 66.5 66.2 40.7 34.0 37.5 Manufahi 99.4 77.7 88.7 33.8 36.6 35.1 Viqueque 79.4 97.9 87.7 71.7 41.8 55.0 Lautem 106.7 104.1 105.2 74.9 64.8 71.7 Baucau 80.5 80.0 80.3 64.6 62.1 63.4 Manatuto 102.1 125.2 112.9 61.4 45.3 52.9 Dili 107.9 112.6 110.0 91.3 77.3 84.4 Aileu 91.6 80.2 86.5 49.1 58.1 52.4 Liquica 91.4 86.3 88.7 37.3 28.3 33.3 Ermera 66.9 66.8 66.9 27.7 38.5 32.5 Bobonaro 96.1 92.9 94.4 32.9 33.7 33.3 Oecussi 90.8 76.9 83.7 50.0 50.9 50.4 East Timor 90.1 90.2 90.1 57.0 52.7 55.0 Sollrce: National Socio-Econornic Survey (SUSENAS), 1998 Table 3: Health Indicators by District Percent of births attended % of children District by modern medical personnel immunized Kovalima 14.8 79.5 Ainaro 8.3 58.6 ~ a n i f a h i 13.6 70.8 Viqueque 10.8 51.5 Lautem 19.0 81.0 Baucau 12.3 78.6 Manatuto 42.8 90.3 Dili 54.5 90.4 Aileu 14.0 95.9 Liquica 17.3 87.8 Ermera 7.3 63.0 Bobonaro 12.7 67.1 Oecussi 15.4 86.0 East Tinlor 22.2 86.4 7. District level indicators for education and health status are also available. These iildicators are correlated with the poverty incidence rates, though the correlation is by no means perfect. The gross primary enrollment rate in East Timor for boys and girls was 90 percent in 1998. Ainaro and Ermera had among the lowest gross primary enrollment rates (66 percent each). Dili and Manauto had the highest gross priinary enrolltnent rates. The overall gross junior secondary enrollment rate was 55 percent, with some difference across genders, with the rate for boys being 57 percent and for girls 53 percent. Dili still had the highest gross junior secondary enrollmeilt rate (87 percent), with Ennera, Bobonaro and Liquica indicating the lowest enrollment rates (33 percent). 8. Some health related data are also available from the 1998 socio-economic survey. As for the education indicators, Dili had the best indicators for both the proportion of births attended by modem medical personnel, and for the proportioil of children immunized. Ainaro and Ermera fared badly on the health indicators too. ANNEX 2 ANNEX 2: RICE PRODUCTION, IMPORTS, CONSCTMPTIONAND MARKETS: AN OVERVIEW 1. Prodzictiorl of paddy and rice. Figure 1 and Table 1 below indicate that production of paddy in East Timor between 1991 to 1997 was relatively stable, peaking in 1994 and 1995 but generally fluctuating between 55,000 tons to 65,000 tons. Several sets of production data are available from different departments of the former Indonesian administration and reconciling some of them is difficult. 2. Prior to the civil disruptions, paddy production was guided by Indonesian government policy and implemented by BULOG. The policy provided floor prices for rice and markets for the products either locally or to other provinces. Production decisions by farmers was based on this regulated environment and there was a sense of security that whatever was being produced would be purchased. In addition, there was little competitioil for rice from other rice producing countries. 3. Table 1 indicates that paddy production dropped by 21,741 tons (from 67,144 to 45,403 or 32%) between 1997 and 1999 as a result of civil disruptions and the drought of 1998. Production began to recover in 2000 and by 2001 production is estimated1to have recovered to 80% of its level in 1997Figure 1below shows these trends. Figure 1: Estimated Annual Paddy Production, (Tons) for East Timor for 1991-2001 1 -Paddy roduction MAFF, 2000 - - - - -Paddy Production TLHS, 2001) P a d d y Froduction tKanror Statr Lalam Angka) -a- Paddy Productron [FAO, 2000) Source: Various irlchldirzg (i) Tin1 Tim Dalarn Artgka, Statistik Pertatlian, MAFfigures (ii) Estinlatesfor 1995 and 2001(iii) Suser~asfor popttlation data used irt cor~sttnlptiortfi~tres(assttrrtes65kg/cap/anrlltrr1rice consttmption based on MAFfigures) (iv) Ir~rportfi@tresbased UN Border Control Data 2000 artd 2001 TLHS 2001 Table 1: Estimated Paddy Procluction, Rice Consnmption, ant1 Rice Import (tonslannum) for East Timor for 1991-2001 Paddy production (MAFF, 2000) 56,214 56,359 65,060 67,072 74,744 67,091 67,144 64,812 50,920 Paddy Production (Kantor Stats Dalam Angka) 56,213 56,358 64,060 66,985 66,697 69,465 71,958 Paddy Production (TLHS, 200 1) 57,854 Paddy Production (FAO, 2000) 45,403 Rice Production (50% milling efficiency) 28,107 28,179 32,030 33,493 33,348 34,733 35,979 32,406 22,702 25,460 28,927 Population 772,027 796,317 811,637 827,727 842,696 857,018 881,600 887,686 695,325 753,189 737,811 Population (TLHS 2001) 825,205 Potential Rice Consumption (ARP - 84kg/amudcap) 64,850 66,891 68,178 69,529 70,786 71,990 74,054 74,566 58,407 63,268 Potential Rice Consumption (TLHS- 95kg/annudcap) 78,394 Estimated Imports (UN Border Control) 10,600 20,422 Estimated Imports (TLHS, 2001) 49,705 a. Population data used in consumption figuresbased on 1.11991& 1992:The East Tirnor Project -Volume 1; 2.11993- 7: East Tirnor in Figures 1997; 1998-2000: Special Report FAOlWFP 2000 Crop and Food Supply Assessment Mission to East Timor, 2000; 4.12001: Civil Registry in East Tirnor - Results July 2001; 5.12001: TLHS 2001. (assumes84kg/caplannurnrice based on ARP IIMAFF) b. Data on pre-ballot use of "imported" (Indonesian)rice are not available. Sotirce: Asfor Fieire 3 4. I~trportedrice has been accessing markets throughout East Timor especially where local rice has difficulty penetrating. According to LJN border control data, rice imports have increased from an estimated 10,600 tons in 2000 to 20,500 tons in 2001, 57% from Vietnam, 35% from Indonesia and 8% fiom Thailand (via Singapore). (Figure 2). ANNEX 2 Figure 2: Official Channels: Rice Imports into East Tinlor between July 2000 and July 2001 as recorded by UN Border Control, 2001 (Total 17,232Tons valued at USS2.66 million) Austraia 64.6Tons (03 \ 6.055Tons (354 Vieinam 9.781Tom (5/% 1.331Tons ( 8 4 Source: UN Border Control,2001 5. Prior to 1999, rice was 'imported' into East Timor by BULOG. Annual imports are estimated to have been between 36-38,000 tons per annum. Following the 1999 disruptions, Indonesian imports suddenly ceased and were replaced by other sources. These included (i) the donor community, (ii) traders and (iii) smugglers. Quantities imported by the donor community and traders are registered at UN Borcer Control and figures obtained from them suggest that in 2000 and 2001, about 10,600 and 20,500 tons of rice were imported respectively. However, recent data from the TLHS (2001) suggests that some 50,000 tons of imported rice was consumed in East Timor in 2001. Based on this, it may be that smuggling would account for about 30,000 tons in 2001. The distributioil of imported rice between donorsltraders and smugglers (which has huge inlplications on the effectiveness of any tariff) is however not clear, and will require further investigation. 6. Consu~tlptior~: Time series data on the per capita consumption of rice in East Timor is difficult to obtain. Based on investigations camed out by the Pilot Agricultural Service Center design team and advice from MAFF, a consumption of 84.kg/capita/annumis estimated. Per capita consumption is about 65% that for Indonesia, reflecting cultural differences and the use of maize as a substit~te.~Based on this figure and population estimates in Table 1, annual consulrnption of rice rose to 74,566 tons in 1998, declined in 1999 to some 58,400 tons and rose again to 78,394 tons in 2001. Analysis of TLHS data suggests that consumption of rice per capita in late 2001 was around 95 kglannum. (Population is estimated by the TLHS to be 825,000 people compared to earlier estimates of just under 740,000). Thus, latest estimates of national annual consumption of rice suggest consumption to be around 78,000 tons. Note that consumption patteins will change over time depending on availability of rice, prices and ability to pay. The latest data fiom the TLHS suggests that consumption has climbed to about 90-95 kg/capita/person due to the availability of cheap foreign rice and increased cash from donor activity. Figure 3: Rice Procluction, Consumption and Imports in East Timor (1991-2001) Potential Rice Cons~~mption(High (84kglannumlcap)) +Rice Production (50% milling efficiency) Potential Rice Consl~mption(TLHS(95kgIann11mlcap)) - - -X- - -Estimated Imports (LJN Border Control) +Local Production - - -X- -Estimated Imports (TLHS 2001) +Local Production - - -EstimatedImports(IndonesianGovernment)+LocalProcluction C Source: Si~senas;UN Border Coritrol; TLHS (2001); ARP II 7. Assuming 50% milling efficiency and accounting for losses, the production of rice is estimated to have been around 28,000 tonslannuin in 1991 rising to almost 36,000 tonslannum in 1997. The difference between local production and consumption would have been met by BULOG prior to 1997. Since UN Border Control records only around 21,000 tons for that period, smuggling of rice into East Timor across the porous borders in the west and by barge in the east probably make up the difference. Rice in the rural center and Dili may be mostly imported rice via the port. (This scenario is supported by prices of rice as shown in Figure 6.) While production has managed to recover to 80% of 1997 levels, data from the TLHS (2001) suggest that consumption has increased beyond pre-1999 levels, facilitated by higher imports 8. Figure 3 demonstrates that East Timor is not self sufficient in rice. While rice imports were controlled by the Indonesian government prior to 1999, this function is now taken up by private traders who respond to commercial interests and market signals to determine prices and quantities of import. A priori, transferring these import functions to traders responding to market forces could benefit East Timor in the sense that these operators will attempt to maximize efficiency, will assume the risk and provide the cheapest possible product on the market. However, smuggling is likely to affect the efficiency of these functions. 9. Markets: As reflected in the design of both ARP I and ARP 11, major impediments to the improved sale of rice at markets througllout East Tinlor include (i) dysfunctional post harvest processing, packaging, storage which affect quality; (ii) distribution problems including networks, roads and transport infrastl-ucture; and (iii) uncompetitive costs of production for local rice. ANNEX 2 10. Essential links from producer to consumer were not developed by the market during the period of the Indonesian administration. The Indonesian system relied on a centralized collection and distribution system for rice through Dili where the market was created by a guaranteed floor price, collection was organized and re-distributed to markets outside East Timor, and rice froin other provinces was sold in local markets. Some local rice may have also been distributed to local markets. Following civil disruptions at the time of the referendum and the collapse of the Indonesian government system, this artificial rice market and distribution system suddenly stopped operating. Quality and Price Factors 11. Retail patterns for rice are impacted by: (i) the nature of the markets as discussed above; (ii) consumer choices, as affected by quality of imported versus local rice availability; (iii) personal stocks kept by rice farmers; and (iv) the price difference for bulk (50kg or more) and small buyers. The latter are likely to be lower income people not involved in rice growing activities. 12. Market access problems for locally produced rice include: lack of transport, logistical issues, infancy of networks, lack of post harvest processing, especially packaging for transport, the Indonesian legacy of regulating markets, poor education of a majority of rural people, and the need to maintain a subsistence mode of life to ensure self sufficiency and survival. Inlported rice by contrast has better access to most markets. Consumer choices are, inter nlia, affected by quality. Reports from field discussions in different parts of East Timor with consumers and local sellers (not traders who typically sell imported rice) confirm that local rice can be of better quality than imported rice. East Timorese prefer to eat locally produced rice when it matches the quality of the imported rice. However, quality control is not good and many local mills that produced good. quality grade rice when initially set up quickly lapsed into mediocre production with up to 80% broken rice. 13. Many farmers reported keeping much of the rice produced for their own consumption between harvests. This is due to their inability to purchase sufficient rice with cash during off-seasons to meet their needs. By comparison, many upland rural people, especially those involved in coffee, have more cash to buy rice. Typically, these consumers buy imported rice because it is available at the upland markets in sufficient quantities and in convenient package. In times of large harvest, rice is abundant in rice growing regions and many farmers try to sell their surplus produce at markets. However, as described previously, the sale mechanisms are not developed and the quality of local rice is often lower and so some rice remains unsold. 14. Prices. Figure 4 indicates that the cost of imported rice at the port of Dili rose from about Rp.1000lkg to Rp.1500fkg between June 2000 and in July 2001. These values were coilverted from US$ at current exchange rates for each month. The cost of rice per unit weight is roughly similar for the different countries exporting to East Timor. Figure 4: Import Prices of Rice at Dili Port from Various Countries 1 I I--c~ i e t n a r n-D- Thailand -A- Indonesia X Aastralia 1 Source: Census and Statistics Unit (2001). S~rweyof Sucos ofTimor Lorosae 15. The recent suco census confirms that on average, local rice retailed at markets and stores is marginally cheaper or at least on par with imported rice (Figure 5). The average imported price for rice is Rp. 3,982kg, while local rice sells for an average Rp. 3,592kg. The latest data from the TLHS confirms that prices have been falling and that on average local rice is still marginally cheaper than imported rice (imported rice: Rp. 3512kg vesus Rp. 3455kg). 16. Figure 5 indicates that both local and imported prices have adjusted and declined since early 2001 at the retail level. Figure 5 also shows that while imported rice is more expensive than local rice if purchased by the kilo, it is cheaper (on a per kg basis) if purchased by the 50kg bag. Thus, a 50kg bag could typically be sold at Rp.150,000 or Rp. 3,00Okg, but if only one kg is sold, the price increases to around Rp. 4000. 17. Imported retail prices have dropped because distribution channels may have opened up and larger quantities may be making their way into inland markets. Imported rice may also follow price trends of local rice, which are dependent on harvest seasons. This is confirmed by data collected from the Border Control Unit that shows that the moving average of imported quantities since Jun 2000 has increased from 17,000 to 20,500 tons for 2000 and 2001 respectively. In addition, a good harvest in 2001 means that there may currently be a surplus of rice in East Timor especially in rice growing areas such as Viqueque and Covalima. 18. Figure 5 also shows that average prices of rice at the border (i.e. at port of Dili) have been steadily increasing from about 1000 Rpldkg to just over 1500 Rphlkg. Thus the profit margin of importers or wl~olesalershas declined froin about 1840 R p n g in January 2001 to 1550 Rphkg in May 2001 and 1050 Rphkg in August 2001. At current estimates of import levels (20,500 tons per annuin), profit margins are valued at some US$2.3 million. The profit margin for the retailers who sell imported rice by the kg at markets has also decreased in the same time period from 780 to 400 Rpllfkg. ANNEX 2 Figure 5: Prices of rice in East Tinior (rplllkg) +Import(Diliport) +Import (sold by the Kg) -A- Local (sold by Kg) U l r n p o r t (sold by 50kg Bag) Sources: (1) UN Border Control (all "Import, Dili Port" values) ;(2) Suco Survey (January 2001 values for "Import sold by Kg" and "Local sold by kg"); (3) World Bank field observations values collected during ARP pre-appraisal mission in various districts (all May 2001 values); (4) JICA field observation values collected during field visits in various districts (all August 2001), Values for "Inlport sold by 50kg bag" in January 2001 and "import (Dili Port)"in August 2001 are estimated (5) TLHS, 2001 for September 2001 values. 19. The latest data from the TLHS is shown for "Nov 01" in Figure 5. It shows that prices of both local and imported rice increased again late last year reflecting lower rice supply post harvest. In addition, the data shows that imported and local rice prices compete for market share. Since the price of imported rice follows the seasonality of local rice, it must be the local rice availability that is driving the imported rice price fluctuations. Thus, traders' profit margins also fluctuates with the seasons. 20. Prices of rice per Kg are summarized from the TLHS in Figure 6 for the different strata. The figure indicates that imported rice is slightly more expensive in rural areas including the center, the highlands and some urban areas in the districts. This may well be related to transport costs and the fact that local rice may not be able to reach these markets effectively giving imported rice more of a monopoly. Imported rice is cheaper in urban areas including DilifBaucau and in the west. These areas are rice-producing areas and one would have expected to see local rice being cheaper especially since imported rice would have high transport costs to reach these areas. However, it may be that traders sell at reduced prices to compete with the local rice and can still make a profit. If rice is smuggled from Indonesia to the rural west, transport costs would be minimized and the current 10%tariff avoided allowing imported rice to compete effectively. Figure 6: Prices of rice in different regions in East Timor, September 2001 Local Rice Imported Rice - Source: TLHS 2001 Characteristics of Rice Growers 21. Figure 7 shows the number of households who grow listed crops. According to the data (from the TLHS), some 26% of households grow rice. Figure 7: Number of Households in East Timor Who Grow a Particular Crop I Potato Soy bean Gogo Rlca coconuls Peanus Mungb a n Othr huR VC~dabhS Kldnsyb a n Coffse cherries Coffee dry b a n Squash Rlca Bananas Taro Sweet potato Cauava ANNEX 2 Figure 8: Estimate of the Number of Rice Farmers at District Level in East Timor (2000) -Number of people living in Sucos where rice is the main crop (aggregatedat the District level) +Percent of people living in Sucos where rice is main crop to total number of people in all Sucos 22. The results of the Suco Survey indicated that some 30% of East Timorese are involved in rice growing activities (Figure 8). This was an approximate figure as the Suco Survey did not permit accurate determination of the actual number of people who were rice farmers.3 Table 2: Rice Farmers by District (2000) District A B Viqueque k,.,>60,967 66% Oecussi ?::41-,247 30% Bobonaro 28,511 60% ~ 0 ~ a 1 i m g $j.:5,c25I462, - as% .F.r;.i~-. ....... Manatuto 19,082 29% Lautem 18.337 35% Baucau 13,736 36% Manufahi 12,925 17% Aileu 10,869 27% Ermera 7,327 15% Ainaro 4,125 12% Dili 2,178 5% Liquica 1,864 2% ET 246,630 29% A: Number of people living in Sucos where rice is the main crop. B: Share of people in Sucos where rice is main crop. The relevant questions asked (i) Are parts of the Suco irrigated? (yes, no) (ii) Please list the 3 main crops in this Suco in order of importance (rice, corn, coffee), (iii) How many people are living in this Suco at the moment? 23. For example, in Aileu, some 60,000 people live in Sucos where rice is the major crop being grown. It is assumed that most people in these sucos would be rice farmers. This population represents about 65% of the total population in all the Sucos in the District of Aileu. Likewise, in Ainaro, just over 40,000 people live in those Sucos where rice is the major crop being grown and that population represents about 30% of the total population in all the Sucos in the District of Ainaro. In total, nearly 250,000 people, or about 30% of the population are estimated to be reliant on rice farming for at least part of their livelihood. 24. Much of the production of rice is consumed by the farmer and his family. Table 3 shows prod~~ctionof milled rice (estimated at 50% of paddy production) for different income quintile groups in society. The Q1 group is the poorest 20% of the population with 42, Q3, 4 4 and Q5 being progressively richer. Thus, Table 3 shows that 64% of production in the Q1 group is consumed by the farmer and his family. This proportion is relatively stable for all quintile groups, being 69% for the richer Q5 group. The quantity of rice sold however rises from 500 t.p.a. to 1,279 t.p.a. indicating that the richer rice farmers can dispose of more of their home-grown crop. They have more quantity to spare and also have better access to markets. This is shown in Table 4, which indicates richer farmers have about three times the expenditure on transport. Table 3: Protluction & Consumption of Milled Rice in East Timor Sorted by Income Quintiles Local Rice Purchased 443 6% 982 8% 833 6% 1,456 8% 1,2136% 4,928 7% Local Rice Home Grown 2,254 32% 3,164 27% 3,000 23% 3,951 23% 3,552 16% 15,921 22% Local Rice Received 121 2% 339 3% 169 '1% 306 2% 730 3% 1,665 2% Imported Rice Purchased 4,208 60% 7,208 62% 9,328 70% 11,544 67% 16,462 75% 48,751 68% Total Consumption 7,027 100% 11,693 100% 13,331 100% 17,257 100% 21,957 100% 71,265 100% . . Lost Production 447 12% 223 5% 641 10% 425 5% 286 5% 2,022 7% Consumed Production 2,371 64% 3,295 69% 4,181 63% 5,206 67% 4,118 69% 19,169 66% Sold 500 13% 1,035 22% 1,189 18% 1,551 20% 1,262 21% 5,537 19% Barter +Payment 366 10% 188 4% 453 7% 418 5% 159 3% 1,585 5% Surplus Production 20 1% 31 1% 151 2% 218 3% 164 3% 585 2% 100 Total Production 3,704 % 4,772 100% 6,615 100% 7,817 100% 5,990 100% 28,898 100% Source: TLHS 2001 25. Overall, Table 3 shows that about 25% of the production enters the market place by sale, payment or bartering. Surplus production seems to increase with income and losses seem to decrease with income. A slight decrease in production in the Q5 compared to the 44 group may be due to shifts in income earning activities away from rice to higher paying crops like coffee, fmit, and vegetables at the Q5 level. Similarly, consumption of local rice also tapers off at the Q5 level, perhaps reflecting consumer choice towards imported rice that may be of more consistent quality, more convenient packaging and relatively more easily available. Given the localized surpluses of unsold rice and the extent of imports, members of the farming community with some wider ' support have called for the imposition of a higher tariff. , ANNEX 2 Table 4. Expenditure on Transportation for the Poorest ancl Richest Farmers DiliIBaucau - 102,093 Other Urban 77,302 57,261 Rural Highland 22,598 150,545 Rural Lowla~ld 26,234 87,384 Average 31,534 99,321 Soitrce. TLHS2001 26. Based on the results of the TLHS, total national annual production of rice was about 29,000 tons between the third quarter 2000 to the same period in 2001. This compares to total national annual consumption of just under 72,000 tons.4 The consumption data is divided into local rice and imported rice. Total annual consumption of imported rice is reported at about 49,000 tons or 68% of total while consumption from domestic production accounts for about 31% (22,514 tons/annum5) including purchased, col~sumptionof home grown and received rice. Based on these data, it is evident that East Timor is far from being self-sufficient in rice production. Figure 9: Annual Consumption of Local Rice and Imported Rice (tons) for different income quintiles in East Timor. (Ql=poorest to QS=richest) 4 2 4 3 4 4 Q 5 mTotal Local rice W In~portedRice Source: TLHS 2001 The TLHS is "snapshot" view of the nation between August 2001 and December 2001 and as such does not give information across time. The recall period for the production data was 12 months while that for the consumption data was one week. This explains the differences in the categories, e.g. "local rice home grown" should be equivalent to "consumed production;" "local rice purchased" should be equal to "sold" production. The person in the household answering the questions on production was different to the person answering questions on the consumption. This also helps explains differences. Production calculated from consumption - imported rice = 22,514 t.p.a. When calculated from the production table, it is 26,291 t.p.a. (net of surplus and losses). The difference is probably due to the different recall periods and the different respondents as explained in footnote 4 above. Expenditure Impact of Higher Rice Prices 27. The effect of a rise in import tariffs would be a correspoilding rise in the prices paid by the consumer of imported rice. In terms of expenditure, the purchase of imported rice occupies roughly the same importance in relative terms, ranging from 63% in Q1 to 75% in Q5 groups. In nominal terms, however, the richer Q5 groups spend nearly four times as much on imported rice (6,444 rphlweek-Q5: 1,741 rph/week-Ql). (Table 5 and Figure 10). Table 5. Average Per Capita Expenditure on Rice for Poorest (Ql) to Richest (Q5) Income Groups (Rphlweek) Local Rice- Purchased 179 6% 417 9% 337 6% 535 8% 481 6% Imported Rice -purchased 1,741 63% 2,889 62% 3,771 71% 4,604 68% 6,444 75% Local Rice - Home Grown 801 29% 1,184 26% 1,129 21% 1,517 22% 1,352 16% Local Rice-Received 42 2% 127 3% 57 1% 128 2% 264 3% Imported Rice -Received 11 0% 16 0% 9 0% 0 0% 99 1% Total 2,774 100% 4,632 100% 5,303 100% 6,783 100% 8,640 100% Figure 10 I Expenditure on Rice (Rphhveek) I I Quintile I Quintile2 Quintile 3 Quintile4 Quintile 5 1 Local Rice- Purchased 8Imported Rice -Purchased · LocalRice Home Grown - 8LocalRice - Received 8Imorted Rice -Received 28. This analysis does not take into account behavioral changes that might be prompted as a result of the imposition ofa tariff and the resulting changed circumstances. The changes in consumption of a good in response to a change in price is called the price elasticity of demand and it is a useful way of predicting, inter nlia, the impact of a tariff (assuming that the price response to a tariff would be increased prices at the market). Initial calculations of income elasticity of demand across quintiles were undertaken using the TLHS data. The results indicate that the richer quintiles modify their consumption patterns less than the poor as income fluctuates. Figure 1I shows that when faced with a 1% change in income, richer farmers in quintile 5 would modify their consumption of rice by 0.49%, while for the same change, poorer (Q1 fanners) would modify their ANNEX 2 consumptioil of rice by 1.12%. This shows that for the poor, the decision to buy imported rice is more dependent on their revenue; if during a season, there is less revenue, the purchase of rice would be sacrificed more readily than for the richer farmers. Purchase of rice is relatively more dependent on income for the poor than for the rich and therefore an increase in the price of rice due to a tariff would affect the poor more than the rich. Figure 11:IncomeElasticity of Demaiid for Rice in East Timor Across Different Quintile ~ r o u ~ s ~ - - - Sotrrce: TLHS 2001 29. It is useful also to look at expenditure7on rice in comparison to other staple crops. Figure 12 shows that the purchase of imported rice constitutes a major proportion of the expenses incurred by the poorest people in East Timor (57%). Other major groups include vegetables (20%), corn (7%) and local rice (6%). By comparison, households in the richer quintile (Q5) on average spend only 26% of the total expenditure on imported rice, with meat, total vegetables and fish (26%, 22%, 10%) being other important foods. Note that the nominal value of the expenditure for the rich isjust under 24,000 Rph while the for the poor it is only 3,059 Rph. In terms of percentage of total expenditure, therefore, the poor spend more than the rich on purchasing rice. The TLHS is not the best source of data to calculate elasticities because we are using averaged values of consumption and income across quintiles to show how consumption changes based on income. Ideally, larger samples should be used. A small specific survey based on the TLHS methodology should be undertaken to more accurately determine both income and price elasticities of rice and cross price elasticities. Expenditure is calculated based on a group of agricultural products derived from the Consumption and Expenditure section of the TLHS questionnaire. It is not the same group of general commodities used for the calculation of the poverty line. The purpose here is to show relative expenditure on major agricultural products for two different income groups in East Timor. Figure 12: National Average Per Capita Percentage Weekly Expenditure for tlie Very Poor (Ql) ant1 the Rich (Q5) Percentage weekly expenditure for very poor (Quintilel) for purchased food (Total:3,059 Rph). Total Local Rice 1 LegumeslNut Total Fruits 6% Unhusked /,- Total 2% 1 % Rice /coy