Sri Lanka - Public Expenditure Review

Document of The World Bank FOR OFFICIAL USE ONLY CONFIDENTIAL REPORT NO. 12337-CE SRI LANKA: PUBLIC EXPENDITURE REVIEW December 22, 1993 FILE COPY Country Operations, Industry & Finance Division Country Department "' CONFIDENTIAL South Asia Region Report No: 12337 CE DOCUMENT OF THE WORLD BANK Type: ECO This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS (Annual Averages) Rs per US$1.00 US$ per Rs 1.00 1981 19.25 0.052 1982 20.81 0.048 1983 23.53 0.042 1984 25.44 0.039 1985 27.16 0.037 1986 28.02 0.036 1987 29.44 0.034 1988 31.81 0.031 1989 36.04 0.028 1990 40.06 0.025 1991 41.37 0.024 1992 43.83 0.023 Rs = Sri Lanka Rupee FISCAL YEAR (FY) January 1 to December 31 FOR OFFICIAL USE ONLY GLOSSARY OF ACRONYMS ADB Asian Development Bank BOI Board of Investment CCPI Colombo Consumer's Price Index CPC Ceylon Petroleum Corporation CEB Ceylon Electricity Board CWE Cooperative Wholesale Establishment EDB Export Development Board ERC Economic Restructuring Credit FDI Foreign Direct Investment FIAS Foreign Investment Advisory Services FSP Poor Relief Food Stamp Program GATT General Agreement on Tariffs and Trade GCEC Greater Colombo Economic Commission GDP Gross Domestic Product GNP Gross National Product IDA International Development Association IFC International Finance Corporation JSP Janasaviya Program MDM Mid-Day Meal Program MFA Multifiber Arrangement NTBs Non-Tariff Barriers PFDC Private Finance Development Credit PIP Public Investment Program PMEAC Public Manufacturing Enterprise Adjustment Credit PRDA Provincial Road Development Authority REER Real Effective Exchange Rate SLR Sri Lanka Railways TIN Tax Identification Number VAT Value-added Tax SRI LANKA PUBLIC EXPENDITURE REVIEW CONTENTS Page No. EXECUTIVE SUMMARY ................................................ i I. PUBLIC EXPENDITURES AND THE PUBLIC INVESTMENT PROGRAM ........ II. ECONOMICINFRASTRUCTURE..................................... 23 III. AGRICULTURE ................................................. 33 IV. THE SOCIAL SECTOR ............................................ 43 V. PUBLIC ADMINISTRATION AND THE CIVIL SERVICE .................... 63 ANNEXES STATISTICAL ANNEX OF TABLES (Tables 1 to12) ............................ 78 M A P ... .... ....................................... ........ .... ... 94 This report is based on the findings of a World Bank mission which visited Sri Lanka in February, 1993. The mission was composed of Mohammed Allak (ASTPH); Maurice Gervais and Jonathan Stevens (ASTPS); A.G. Karunasena, Terrence Abeysekera and Dilesh Jayantha (RMSL); Thomas Selden, Neville Edirisinghe and Ruben Suarez (Consultants); and Menahem Prywes (SA3CI and task manager). The mission was supported by and shared consultants with a concurrent Poverty Assessment mission led by Hugo Diaz. Nearly all members of the Sri Lanka Country Team contributed, in some way, to this report. In particular, contributions to the report were made by David Steedman (ASTPS); Abderraouf Benbrahim (SA3CI); Douglas Lister and Ivan Serejski (SA3AG); Inai Bradfield, Mohinder Manrai, Lea Donaldson, Claus Schroeder, Alex McPhail (SA3EI); and Paul Blay (SA3PH). Pedro Alba (SA3CI) also contributed to the drafting of the report. L. F. Goh provided valuable secretarial assistance. LIST OF TEXT TABLES Page No. Chapter I Table 1.1: Key Economic Indicators, 1985-1992 ................................. 4 Table 1.2: Key Economic Targets, 1991-1997 .................................. 6 Table 1.3: Savings and Investment Balance Targets, 1994-1997 ...................... 7 Table 1.4: Total and Capital Expenditure, 1958-1993 ............................. 8 Table 1.5: Economic Classification of Central Government Expenditures ................ 9 Table 1.6: Functional Classification of Central Government Expenditures .............. 11 Table 1.7: Public Investment:1993-97 ....................................... 12 Table 1.8: Identified Public Investment - Shares in GDP .......................... 13 Table 1.9: Investment Programs of Selected State-Owned Enterprises ................. 15 Table 1.10: The BudgetCycle ............................................. 17 Table 1.11: Planning Cycle for Public Investment Projects .......................... 19 Table 1.12: Proposed Cuts in the Public Investment Program 1994-1997 ................ 21 Table 1.13: Proposed Reductions in Central Government Expenditures: 1994-1997 ......... 22 Chapter II Table H. 1: Public Investment Plan in Transportation, 1981-1996 ..................... 25 Chapter III Table 111.1: Sectoral Composition of Public Expenditure in Agriculture, 1980-93 .......... 34 Table 111.2: Crop and Agriculture Research Budget by Ministry in Sri Lanka, 1989 ........ 41 Table 111.3: Comparison of Expenditures on Research and Extension in Sri Lanka with OtherCountries................................ 41 Chapter IV Table IV. 1: Eligibility Criteria for the Food Stamps Program ........................ 46 Table IV.2: Budgetary Projections for Poor Relief Food Stamp Program ................ 46 Table IV.3: Budgetary Projections for Janasaviya ................................ 50 Table IV.4: Cost Projections for Reform of the Social Transfers ...................... 57 Table IV.5: Composition of the Health Public Investment Program, 1993-97 ............. 62 Chapter V Table V.1: Employment in the PublicSector................................... 64 Table V.2: Trends in the Costs of Wages and Pensions ............................ 66 Table V.3: Central Government Approved Cadre by Grade, 1993 .................... 68 Table V.4: Projections of Central Government Wages and Pensions ................... 71 Table V.5: Financial Operations of the Provincial Councils ......................... 75 Table V.6: Recurrent Expenditures of the Central Provincial Council: M inistry of Education .......................................... 76 Table V.7: Recurrent Expenditures of the Central Provincial Council: M inistry of Health ............................................. 76 LIST OF TEXT FIGURES AND BOXES Page No. Figure 1.1: Central GovernmentDeficit ....................................... 3 Figure 1.2: DefenseExpenditures ........................................... 10 Figure 1.3: Public Investment: Actual 1985-91; Planned 1992-96 & 1993-97 ............. 11I Figure I.4: Major Components of Identified Planned Public Investment ................ 14 Figure 1.5: Actual/Planned Public Investment .................................. 16 Box 1. : Alternative Concepts of Government Capital Expenditure ................... 8 Box 1.2: Criteria for Evaluating Projects and Programs .......................... 20 Box 11.1: Rehabilitation and Maintenance in Water Supply ........................ 27 Box 11.2: Road Rehabilitation and Maintenance ................................ 31 Box III. 1: Operations and Maintenance of Mahaweli ............................. 36 Box 111.2: Performance of the State-Owned Plantations ........................... 38 Box IV. 1: Social TransferPrograms ........................................ 44 Box IV.2: School M aintenance ............................................ 58 Box IV.3: Repairs and Maintenance of Health Facilities ........................... 62 EXECUTIVE SUMMARY 1. Introduction. The primary objective of the Public Expenditure Review (PER) is to assist the Government of Sri Lanka in reforming public expenditures as part of its broader program of adjustment and structural reform. To this end, the PER analyzes the overall level, composition and efficiency of the public expenditure program, with particular emphasis on the Public Investment Program (PIP) for 1993-97. In this regard, the PER reviews broad issues such as expenditure programming and budgeting as well as public expenditure issues in selected sectors (economic infrastructure, agriculture, education and health). The PER then focuses on: (i) the cost, mix and targeting of expenditures on the social transfer programs; and (ii) the cost and structure of the public administration. 2. Overview of Expenditure Policies. The Government has made impressive progress in reducing the central government budget deficit from a peak of 23.1% in 1980 to 7.4% in 1992. Since 1988, this intensive adjustment effort brought a significant decline of 7 points of GDP in the central government's deficit. This was achieved mostly through a sharp cutback in expenditures from 34.5% of GDP in 1988 to 28.6% in 1992. The adjustment effort has been accompanied by several broad improvements in the composition of public capital expenditures. First, there was a sharp decline in the funding of massive infrastructure projects, in particular the Mahaweli program. Second, the Government's emphasis on privatization and promotion of private sector activity led to a reduction in transfers to inefficient state-owned enterprises. 3. Despite these efforts, the quality of the fiscal adjustment raises several concerns. First, practically all the fiscal adjustment was achieved at the expense of lower capital expenditure. For example, by 1992, capital expenditures, adjusted for financial payments such as privatization proceeds, had declined to 7.9% of GDP from 12.7% in 1988. In contrast, the Government's efforts to contain current expenditure had limited success. Little progress was achieved in limiting wage and pension payments, new social transfer programs were introduced, and defence expenditures (above 4% of GDP) remained a significant fiscal burden. Such a one-sided adjustment is not sustainable in view of Sri Lanka's many development needs. The Government therefore, needs to intensify its efforts to reduce current expenditures as a share of GDP, in particular, wages and pensions and certain poorly targeted social transfers. Initial efforts to control these expenditures should reduce interest rates and induce a further decline in current expenditure by bringing down interest payments on the domestic component of the public debt. Second, the PIP includes certain projects with a low development impact that suffer from inadequate preparation or are not of immediate priority and they should be eliminated or postponed. 4. Weaknesses in expenditure programming and budgeting have contributed to these quality and compositional issues. For instance, the coverage of the public investment program needs major improvement so that it will provide a basis for choice among alternative projects. The coverage is particularly inadequate for projects of the state-owned corporations that are funded from non- budgetary sources. Some of these projects are very large, e.g., the proposed purchase of five A340s by Air Lanka, which would have major adverse implications for macroeconomic management. Furthermore, there is a pressing need to subject all projects, both on and off-budget, to rate of return analysis before admitting them into a comprehensive public investment program. Monitoring and follow-up of expenditures and projects is also weak, especially those of the provincial councils. 5. Another major expenditure planning issue is the need to provide a medium-term perspective to expenditure decisions. The Government has often undertaken sharp contractions in expenditures in a partially ad-hoc manner because of resource pressures, and not based on systematic reviews of expenditure programs and sectoral strategies. This points to the need for a 'core' investment program that would protect priority projects and to a medium-term plan for current - ii - expenditures. Such a plan would help avoid the under-funding of operations and maintenance in certain sectors. Proper funding for operations results in improved efficiency in the delivery of public goods and services, while proper maintenance will save major expenditures for rehabilitation in the future. The sectors that suffer from under-funding include education, health, irrigation and transportation. 6. The Public Investment Proaram. The PIP for 1993-97 calls for a approximately steady ratio of public investment/GDP of about 9%-10%. Its broad sectoral priorities remain mostly unchanged from the recent past: progressive disengagement from massive infrastructure projects and from areas, such as industrial enterprises, that are best left to the private sector, and increasing commitment to social and economic infrastructure. Although the overall level and broad composition of the proposed PIP seem consistent with the Government's macroeconomic framework and broad development strategy, there are a number of serious issues that need to be addressed. 7. First, the Government needs to increase its investments in sub-sectors where inadequate infrastructure is constraining development of the private sector and where private financing is difficult to mobilize--such as rehabilitation. The need for rehabilitation is especially urgent in the road, irrigation, power, rail, health and education sub-sectors. More intense efforts to rehabilitate roads, in particular, would support the development of export-oriented agriculture, small manufacturing, and tourism in rural areas. Moreover, rehabilitation of irrigation often yields high returns, especially when compared to new projects. In the power sub-sector, the rehabilitation of existing generating plants is the most cost-efficient way assure adequate power supply over the short- term. To assure the reliability of supply, the transmission of distribution system needs to be further strengthened. There is also a need for continued public investment in water supply and sanitation, and to mitigate the environmental impacts of rapid industrialization. More work would be necessary to recommend specific new public investments in new infrastructure. It may be possible, however, to make a case for construction of certain new roads, especially in the Colombo area. 8. The second issue is that, as mentioned above, the PIP includes certain large projects that are inadequately prepared or of low priority. These investments, with doubtful economic rate of return, are mostly in the transportation sector, but also include three irrigation projects and a thermal power plant. Some of these projects are financed from budget sources and amount to some 0.3% of GDP p.a. over the period 1994-97. However, the largest questionable investments are projects of state-owned enterprises financed outside the budget (and excluded from the official PIP) amounting to about 0.8% of GDP p.a. over the same time period. The latter threaten to burden the central government with debt service in the future, since the Government bears the ultimate liability. Thus, the PER recommends that the following ten projects be canceled: Kalu Ganga irrigation project (Rs. 12 billion), Mahaweli System L (Rs. 2 billion), Udawalawe - Right Bank (Rs. I billion), procurement of Airbus 340s (Rs. 30 billion), Thermal Electricity Plant (Rs. 3 billion), the Matara/Katargama and Batticaloa/Pottuvil rail extensions (Rs. I billion), Rural Electrification (Rs. I billion), Second Airport (Rs. I billion) and Kelani Valley rail line (Rs. 0.1 billion). 9. Operations and Maintenance. Inadequate expenditure on maintenance is a problem in Health and Education, as well as in Roads, Irrigation, and in the Railway; other recurrent expenditures are adequate in most sectors outside of Health and Education. In Health, expenditures on operations and maintenance are insufficient to operate part of the system of clinics and hospitals--let alone maintain the structures. The increase in labor costs displaced expenditures which enhance the quality of education: school supplies and equipment, and maintenance. Most maintenance in Health and Education, and some in roads, is in the domain of the Provincial Councils; the effectiveness of those expenditures has been hampered by slow progress in advancing the capacities of the Councils. In Roads, the management of maintenance needs improvement so that any increase in these expenditures - iii - on maintenance will be used effectively. While greater budgetary allocation of resources to operations and maintenance is needed, especially for roads, education and health, an important reason for the inadequacy of maintenance is that the beneficiaries of infrastructure often do not contribute to its maintenance. Therefore, in some sectors, the solution is to be found in charging more appropriate user fees to finance better operations and maintenance practices (e.g., power, water supply) and for encouraging user participation in maintenance (e.g., irrigation canals). 10. Pricing Policy. Sri Lanka's development needs are large and meeting them will require the Government to take action on several fronts. The expenditure policy reforms described above, which would enhance the efficiency of public resource use, need to be complemented by improved cost recovery and by measures to increase the role of the private sector. With respect to cost recovery, one fundamental issue is the low level of tariffs in Sri Lanka, especially in the economic infrastructure sectors. For example, even after a 30.5% electricity tariff increase in July 1993, the average power rate (at about 6 US cents per kwh) is around two-thirds of the long-run marginal cost of producing electricity so that a further 30% increase in the electricity tariff will be necessary by the end of 1993. Moreover, there is cross-subsidization of household consumers, some of whom pay only 15% of long-run marginal cost. Similarly, the internal water tariff is below the marginal cost of production and the collection system is relatively ineffective. In addition rail tariffs are well below long run marginal costs so that the Govemment must make huge annual transfers to continue operation of the railway. 11. In addition to improving efficiency of use, raising tariffs to their economic levels will help generate funds for operations, maintenance and investment while reducing dependence on budgetary transfers. In the irrigation sector, transferring certain maintenance responsibilities to farmers organizations will also enhance cost recovery. Prices of infrastructure services need to be increased towards long-run marginal costs to reduce and eventually eliminate implicit subsidies, unless the subsidies are clearly necessary to correct a market failure or because of equity considerations. Adequate tariffs are also important to attract private sector investment. 12. Role of the Private Sector. In Sri Lanka, as in many other countries, it appears that the public sector has become overextended, and public resources should be redirected and concentrated in those areas in which public sector intervention is required because of market failures or social objectives. Many state-owned enterprises, in particular, are absorbing scarce resources, are operating inefficiently, and not investing adequately. The rationale for public participation in these areas is also not convincing. In recognition of these concerns, the Govemment of Sri Lanka has already embarked upon a substantial program of privatization and other reforms to encourage a greater role for the private sector. The Govemment should continue to pursue this effort, within an appropriate regulatory framework. Private participation is already leading to significant efficiency gains in the provision of services and stimulating additional investment. For instance, the private sector is now participating in a significant expansion of Sri Lanka's telecommunication lines and is providing cellular telephone service. The Government should move forward with particular urgency with the privatization of Air Lanka and fuller privatization of the bus companies. 13. Also, build, own and operate (BOO) and build, operate and transfer (BOT) arrangements can often provide a framework for private sector financing of large infrastructure projects. The most important example is in the power sector, where the Govemment is negotiating BOO/BOT arrangements to build a large volume of thermal power capacity. Where privatization is not possible in the short-term, the private sector can manage certain state assets, such as the state plantations, with appropriate provisions for risk sharing and movement toward full privatization over the long-term. Furthermore, the private sector can improve the efficiency of operation of many government agencies and state-owned enterprises through the contracting of certain operations, such as - iv - maintenance of roads and of railway track and equipment. 14. The Social Sector. Sri Lanka has established an excellent record in social development, as evidenced by indicators that compare favorably with much higher income economies. However, there are some areas of insufficient progress, for instance, the nutritional status of children is poor. Weak income growth, especially in agriculture, raise concerns about the sustainability of the poverty alleviation effort. Moreover, the limited availability of resources and policy constraints threaten to undermine the achievements to date. For example, the social transfer programs, which accounted for some 1.8% of GDP in 1992, generally serve a very useful role in reducing poverty. Nevertheless, they face significant targeting and efficiency problems. A careful reassessment of these programs would not only help assure that the truly poor benefit from the programs, but could also free financial resources for alternative uses. Similarly, the cost effectiveness of expenditures on education and health also needs to be analyzed. 15. The major social transfer programs are Janasaviya, which is a two-year program of intensive transfers accompanied by some poverty-alleviation efforts, the Mid-day Meal and the Food Stamp program. The Government has made progress in targeting through screening methods developed as part of Janasaviya. The screening is conducted through inspections and community meetings and focuses on observable indicators such as possession of consumer durables. While Janasaviya screening is relatively successful, there is little evidence that a substantial fraction of Janasaviya graduates are sustainably exiting poverty. While the Janasaviya, the Mid-Day Meal and the Food-Stamp program absorb large budgetary resources, programs that support the poorest of the poor are severely under-funded. True safety-net programs such as Public Assistance to the indigent, and aid to the elderly and orphans, in particular, are very low. 16. Expenditures on the social transfer programs can be made more effective through better targeting and restructuring, as these programs benefit many middle and upper income households. The Food Stamp program, in particular, benefits nearly 40% of the population, including many non-poor but excluding some of the truly poor. Inflation has eroded the real value of the income transfers, reducing the effectiveness of the programs as safety-nets. Nevertheless, the large number of recipients of social transfers makes it fiscally difficult to raise transfers to levels that would be necessary to adequately support the truly poor. 17. The PER presents a proposal to improve the effectiveness of the social transfer programs in supporting the poor. The proposal is to consolidate the transfer components of the existing programs into a single cash transfer program with two levels of benefit: a higher level for the ultra-poor (defined as households that would qualify under the current rules for eligibility for Public Assistance and aid to the orphans and elderly), and another level for other recipients. The Households would be admitted into the consolidated program after a screening using community-participation methods developed as part of the Janasaviya program. 'Interest' payments on Janasaviya Capital Certificates would be canceled and recipient households that passed screening would be admitted into the consolidated program. The fiscal savings generated by the consolidation could be re-directed into increased benefits for the truly poor and into well-designed poverty alleviation programs. Until screening is completed and new programs are implemented, the savings should reduce the fiscal deficit. 18. Government efforts in Education and Health have been relatively successful. Sri Lanka has a comprehensive public education system which provides primary and secondary school education in most regions and which has achieved literacy rates near industrial country levels. In recent years the recurrent cost of education increased as the Government recruited massive numbers of teachers despite the already favorable teacher/student ratio. The demographic structure of the population implies that the number of students receiving general education will decline over the next decade. For these reasons, the PER recommends a freeze on hiring of teachers until completion of a review to determine the necessary cadre and a pause in the construction of school buildings. 19. Sri Lanka also has a fine reputation for delivering health care. Life expectancy at birth rose from about 58 years in 1950 to 71 in 1991. One of the major challenges for the future is defining and increasing the role of the private sector--which absorbs around 60 percent of national spending on health. At present, private practice by public sector medical personnel diminishes the poor's access to public services; it also lowers the level of care, increases queues, and builds incentives for rent-seeking. The private sector, however, should play a greater role in providing tertiary services. The PER recommends reducing costs and increasing the efficiency of expenditure by shifting the composition of public expenditures away from tertiary facilities (often hospitals) and toward primary and secondary facilities and preventive care and by starting a system of user charges for tertiary facilities. 20. Public Administration. Despite a long-standing tradition of technical excellence and integrity, excessive political interference in hiring and promotion, the impact on morale of ethnic and political conflict, and misguided remuneration policies have reduced the effectiveness of the civil service. At the same time, as a result of over-staffing and an overly generous pension scheme, the fiscal cost of the civil service (including wages and salaries and pension benefits) is high (6.4% of GDP). Since 1990, the Government has been implementing, with mixed results, a program to increase the effectiveness of the civil service and to control the fiscal cost of wages and pensions. In 1990, the Government offered generous severance packages for voluntary retirement. While many accepted the package, the reduction in staff numbers has been short-lived because of the lack of effective controls on civil service hiring. For example, vacancies created by departures have not been eliminated; the number of vacancies and authorized positions (cadre) has actually grown since 1990. In addition, the restructuring scheme was not selective enough, and many key personnel left the civil service attracted by higher private sector wages. Moreover, the scheme was overly generous and resulted in a doubling of the pension bill. 21. Building on the lessons learned from the 1990 experience, the PER recommends that the Government should quickly: (i) abolish the approximately 60,000 vacant positions; (ii) impose strict limitations on hiring of new staff including in Health, Education, the Provincial Councils and the statutory boards, with no replacement of departing low-skilled staff, replacement of one out of every two departing technical and professional staff, and full replacement of senior managers; and (iii) withdraw Circular 44/90, which authorizes the special pension benefits introduced in 1990. In the near-term, the Government should identify a short-list of units which have lost their function or have over-lapping responsibilities and eliminate or merge the units. Over the medium-term, the Government should conduct administrative and functions reviews to establish the cadre necessary to operate each department. Redundant staff should be transferred or receive severance packages. Preliminary work done under the umbrella of the PER suggests elimination of the Department of Food, consolidation of the certain ministries handling agriculture, irrigation and forestry, and consolidation of the ministries which carry out poverty alleviation programs. Another medium-term recommendation is to establish a capitalized pension fund, which would reduce the fiscal burden on the Government by linking pension benefits to the return on capital and by requiring employee contributions. The Government should take the first step by making payments into a capitalized fund that would cover some of its future pension liabilities. It would also be necessary to strengthen the authority of the Civil Service Commission and the Salary and Cadre Committee--which play a crucial role in auditing requests for new positions and in making salary adjustments. - Vi - SUMMARY OF RECOMMENDATIONS Sub-sector Short-term Medium- to Long-term Reason Reporting of the (i) Complete the coverage of the central government (i) Include all investments of (i) Establish a proper basis Public PIP by including all on-budget investments state-owned enterprises, for allocation of scarce Investment (ni) Develop and publish a consolidated public sector boards and authorities in the resources between Program investment program that records all public consolidated PIP alternative projects investments of the state-owned enterprises, (ii) Evaluate and thus better boards and authorities with important manage the investments JEDB, SLSPC, CEB, LECo, CWE, Government's financial CPC, the NWSDB, Telecom, Port Authority, and obligations resulting Air Lanka from its capital expenditures Expenditure (i) Avoid across-the-board cuts in expenditures (i) Carry out rate-of-return (i) Improve the Programming (ii) Develop a core investment program for 1994 studies before admitting effectiveness of public projects to the PIP expenditures in (n) Publish a core investment providing public program and protect core services, and in projects from cuts in particular, improve the expenditures rate of return on public (iii) Estimate operations & capital expenditure maintenance requirements of the PIP Roads (i) Re-examine the Katunayake-Colombo (i) Continue to budget adequate (i) Increase the return on Expressway, in cooperation with major donors, funds for maintenance for public investment in to assure that it is economically feasible and roads roads does not displace an excessive number of people (ii) Encourage the (ii) Increase funding of road maintenance and development of rehabilitation while improving management agriculture, mining and capacity tourism Railway (i) Raise tariffs by 25 percent (i) Raise tariffs to a level that (i) Enhance the efficiency (ii) Cancel or postpone the Matara to Kataragama covers variable costs and of operation of the and the Batticcaloa to Pottuvil rail extensions contributes to overhead railway and the broad-gaging of the Kelani Valley line (ii) Divest commercial property (n) Reduce and eventually (iii) Continue current maintenance and rehabilitation owned by the Rail Authority eliminate the fiscal efforts (iii) Contract out maintenance burden on the (iv) Create a Sri Lanka Railway Authority operations to the private Government ansing (v) Adequately capitalize the Authority taking into sector from the railway account liabilities to the Government (iv) Phase-out budget transfers (vi) Budgetize subsidies in part by putting public and then treat all borrowing service activities of the Railway Authority on a on commercial terms public service obligation contract basis Airports (i) Delay any plan to invest in a second (i) No additional large (i) Improve the rate of international airport at Hingurakgoda until it is investment appears necessary return on public supported by a cost-benefit study as the existing international investment airport can serve anticipated (ii) Limit the budget deficit demand over the medium- by avoiding unnecessary term capital expenditures Airline (i) Negotiate cancellation of Air Lanka's purchase (i) Partly or fully privatize Air (i) Reduce the contingent of 5 Airbus A340s Lanka financial liability of the (ii) Make a commitment to privatize Air Lanka (ni) Place credits to Air Lanka on Govemrnment commercial terms (ii) Increase the efficiency of operation of Air Lanka Bus Lines (i) Discontinue government-financed import and (i) Fully privatize the bus (i) Reduce the contingent assembly of buses once the current contracts companies financial liability of the have expired Government (n) Place the cost of these operations in the budget (n) Increase the efficiency rather than through off-budget financing from of operation of the bus the state-owned banks companies - Vii - SUMMARY OF RECOMMENDATIONS (Continued) Sub-sector Short-term Medium- to Long-term Reasons Power (i) , Increase tanffs by 30%, prepare a program to (i) Move tariffs up to the long- (i) Increase the efficiency of eliminate distortions in the tariff structure and run marginal cost of operation of the CEB improve collections supplying electricity (in) Use price incentives to (iu) Make central government transfers to the CEB (it) Limit annual investment in conserve power and for the capital and operating costs of rural rural electrification to the allocate power efficiently electnfication schemes 1993 rate of Rs 650 million among alternative uses (in) Prepare a revised public investment program (in) Contract out maintenance to (nm) Increase the rate of based on a revised power demand forecast, in the private sector return on investment in agreement with IDA (iv) Encourage private investment power through (iv) Reduce the growth of public investment in in the power sector through cancellation or thermal generation BOO/BOT arrangements postponement of low (v) Assure that financing of BOO/BOT arrangements return projects (such as is on commercial terms and not guaranteed by the certain rural government electrification projects) (iv) Reduce the contingent liability of the Government ansing from investments of the CEB Water Supply (i) Tanffs should be raised toward the long-run (i) Increase the role of the (i) Improve the quality of marginal cost of supplying water and the price private sector through water supply structure should be reformed to eliminate cross- contracting out of tasks and (in) Increase the efficiency of subsidies and provide an incentive for through BOO/BOT operation of the conservation arrangements for capital NWSDB construction (iii) Use price incentives to (ii) Emphasize maintenance and conserve water and rehabilitation in the allocate water efficiently expenditure program for among alternative uses water supply Mahaweli (i) Continue to withhold funding of the Kalu Ganga (i) Avoid investment in new (i) Increase the rate of extension of the Mahaweli system massive infrastructure return on public (i) Cancel or postpone investment in massive new (ii) Enhance maintenance by investment infrastructure such as System L and Udawalawe, shifting responsibility for (ii) Reduce the fiscal deficit Right Bank tertiary canals to farmer by avoiding low priority (in) Privatize the Mahaweli Authority's commercial organizations capital investment operations Plantations (i) Shift from existing contracts with private (i) Increase the flexibility and (i) Restore the economic management companies to long-term leases of 25 productivity of labor by viability of the years or more modifying the guaranteed six- plantations by creating (ii) The management companies should not be day work week and the incentives for investment financed by government-guaranteed debt but restriction on transfernng first through private should provide their own equity or borrow from labor from deficit to surplus management on long- the banking system at competitive rates estates term lease and eventually through private ownership (it) Limit the contingent liability of the Government by reducing borrowing from the state- owned banks on non- market terms Agricultural (i) Increase funding of recurrent expenditures on (i) Continue to shift focus of (i) Improve the growth of Research and agricultural research research and extension away the agricultural sector Extension (i) Improve the qualification of extension agents and from rice toward higher value support them through a mass-media campaign crops contracted out to the private sector - Viii - SUMMARY OF RECOMMENDATIONS (Continued) Sub-sector Short-term Medium- to Long-term Reasons Social (i) Consolidate the transfer components (i) Once screening of (i) Enhance the Transfers of the Janasaviya, Janasaviya Capital beneficiaries has been effectiveness of Certificate, Mid-Day Meal, Food successfully completed, the social transfers Stamp, and Public Assistance raise the amount of in supporting the programs into a single program with transfers. poor through two levels of transfers: one for (ii) Re-direct some or all of better targeting. households who would qualify under the savings from the (ii) Free resources for the current Food Stamp Program and reform to well designed use in poverty another, higher, level for the 'very poverty alleviation alleviation rather poor', defined as households who programs. than transfer would qualify under the current programs. Public Assistance program. (ii) Screen all beneficiaries. (iii) Cancel interest payments on Janasaviya Capital Certificates and move households that pass the screening to the consolidated program. Education (i) Raise expenditures for repair and (i) Raise expenditures on (i) Improve the maintenance of schools. repair and maintenance quality of (i) Freeze recruitment of teachers until a of schools to about Rs. education by review is completed that establishes 100 million in 1992 improving the priorities for current and emerging prices. quality of the shortages and establishes the (ii) Consider encouraging a teaching staff, necessary cadre. role for the private through better (iii) Increase cost recovery in higher sector in financing and maintenance, and education. managing higher by seeking a role education. for the private sector. Health (i) Increase expenditures on repairs and (i) Increase expenditures on (i) Improve public maintenance of health facilities. repairs and maintenance health by (ii) Increase funding of recurrent costs of of health facilities. improving the under-utilized facilities. (ii) Increase the role of the return to (iii) Focus public expenditures on private sector in government prevention of infectious and parasitic providing curative expenditures on diseases. services and establish an health. adequate regulatory system. - ix - SUMMARY OF RECOMMENDATIONS (Continued) Sub-sector Short-term Medium- to Long-term Reasons Civil Service (i) To reduce the size of the civil service, (i) As the size of the civil (i) Improve the and Pensions partially freeze replacement of service decreases, quality of departing staff, including in health, improve incentives to government education and the Provincial Councils. enhance the efficiency of services by Specifically, there should be no Government A reform improving the replacement of departing lower level of salaries should quality of the civil staff, replacement of I out of every 2 integrate allowances in service. departing professional and technical the civilian (ii) Reduce the fiscal staff; and full replacement of senior administration into deficit by reducing managers. The freeze should remain personal emoluments. civil service in effect until reviews to determine (ii) Establish a capitalized emoluments. the necessary cadre are completed. pension fund with (iii) Limit the (ii) Reduce the authorized cadre by about government and civil contingent liability 60,000 budgeted but vacant positions. servant contributions. of the Government (iii) Identify a short-list of Central by establishing a Government departments which have capitalized pension lost their function or have over- fund. lapping responsibilities. Reduce or eliminate unnecessary departments, merge certain line ministries, and retrench their lower-level staff. (iv) Withdraw Circular 44/90, which establishes high pension benefits. (v) Introduce the pre-1985 pension scheme for new entrants to the civil service. (vi) Establish a capitalized fund to meet future pension obligations into which the Government pays a percentage of wages and salaries. (vii) Strengthen the Civil Service Commission and the Salary and Cadre committee. Provincial (i) Establish central control over (i) For reasons of (i) Improve the Councils recruitment in the Provincial Councils administrative and fiscal quality of the as part of a national policy of efficiency, Sri Lanka health, education, management of the civil service. should choose a single public assistance, (ii) Improve monitoring of expenditures system of local and road of the Provincial Councils. government, that is, maintenance choose between the services run by the Provincial Councils and Provincial the Divisional Councils. Secretariats. (ii) Reduce waste of public resources by establishing better accountability. I. PUBLIC EXPENDITURES AND THE PUBLIC INVESTMENT PROGRAM I.A. Introduction and Overview 1. Introduction. The primary objective of the Public Expenditure Review (PER) is to assist the Government of Sri Lanka in reforming public expenditures as part of its broader program of adjustment and structural reforms. To this end, the PER will serve as the major background piece for the Banks's dialogue with the Government on expenditure policies. The report, for example, proposes measures to be supported by the proposed Second Economic Restructuring Credit (ERC-II). Another important objective of the PER is to inform the donor community about the effectiveness of public expenditures. 2. The PER seeks to support the ongoing adjustment process by addressing some of the most important issues in public expenditures. To this end, after this introduction, Chapter I analyzes trends in central government expenditures since 1980 and evaluates the overall level and broad mix of expenditures. The PER then focuses on the cost and efficiency of the Central Government's Public Investment Program (PIP) for 1993-97, and of selected investments of the state-owned enterprises. While broader issues such as expenditure programming, budgeting and coverage are reviewed in Chapter I, sector and project specific issues are discussed later in the report. In particular, Chapters II and III review public expenditure issues in the infrastructure and agriculture sectors, including the adequacy of expenditures on operations and maintenance. In Chapter IV, the PER analyzes the cost, mix and targeting of expenditures in the social sector: the social transfer programs, health and education. Finally, the cost and structure of the public administration and civil service are the subject of Chapter V. Chapter V also briefly examines the role of the Provincial Councils because they are now responsible for most expenditures in the Health and Education sectors and absorb substantial resources. A summary of the Report's recommendations can be found at the end of the Executive Summary. 3. Overview. The last in-depth review of public expenditures took place in the context of the Country Economic Memorandum (CEM) of 1988.' Despite progress in addressing some of the problems identified in the CEM, many of the issues remain the same. For example, the CEM of 1988 recommended that the Government should improve the rate of return on public infrastructure by ending large new investments in Mahaweli and by reducing transfers to public corporations with low- yielding capital investments. The CEM pointed to opportunities to improve the return on public investment by concentrating on rehabilitation and maintenance, for example, of existing irrigation canals in the Mahaweli project, roads and the railway. The CEM examined the problems of leakages and targeting in the Food Stamp program and recommended the consolidation of the social transfer programs into a nutrition program and an income transfer program targeted to the poorest segment of the population. The Government adopted some of the recommendations of the CEM in the agreement that led to the Economic Restructuring Credit of 1990 (ERC-I), and has successfully carried out some Sri Lanka, A Break with the Past: The 1987-90 Program of Economic Reforms and Adiustment (Report No. 7220-CE), May 27, 1988. There was an update in Sri Lanka, Strengthened Adiustment for Growth and Poverty Reduction (Report No. 10079-CE), January 8, 1992. -2- reforms.2 The ERC-I emphasized administrative reform and, in particular, the restructuring of staff and adjustment of civil service wages to levels comparable with the private sector. 4. Since 1988, progress in adjusting public expenditures has been mixed: strong in adjusting the level of expenditures, but weaker in improving programming and budgeting and addressing some of the issues regarding quality and composition of the public expenditure program. The major achievement was the sharp cutback in expenditure as a ratio to GDP from 34.5% in 1988 to 27.8% in 1992. This intense adjustment effort, together with a relatively stable revenue to GDP ratio, led to a very significant decline of some 7.2 percentage points of GDP in the central government deficit. The adjustment effort was also accompanied by broad improvements in the composition of public expenditures. In particular, there was a sharp decline in the funding of massive infrastructure projects, especially in the Mahaweli program. Similarly, the Government's emphasis on privatization and promotion of private sector activity led to a reduction in transfers to inefficient state-owned enterprises. 5. Notwithstanding the improvements mentioned above, the composition and quality of the expenditure program remain of concern. First, the Government's efforts to contain current expenditures have had limited success. In particular, little progress was achieved in containing wage and pension payments, new social transfer programs were introduced, and defence expenditures (above 4% of GDP) remained a significant fiscal burden. As a result, practically all the fiscal adjustment has been achieved at the expense of capital expenditures. By 1992, capital expenditures, excluding financial payments such as privatization proceeds and debt repayments by the state-owned enterprises, had declined to 7.9% of GDP from 13.8% in 1988. In view of Sri Lanka's many development needs, and that further fiscal adjustment is essential to meet macroeconomic targets, such a one-sided adjustment is not sustainable: the Government needs to emphasize the reduction of current rather than capital expenditures. Reform of the civil service would permanently reduce current expenditures (Chap. V). Reform of the social transfer programs would save on current expenditures, at least in the short-term (Chap. IV); much or all of these savings could eventually be channelled into well designed poverty-alleviation programs. Taking these first steps would induce further savings in current expenditures as a lower government domestic borrowing would drive down interest rates and lower interest payments. 6. Revisiting the relative roles of the public and private sector should also provide good opportunities to contain expenditures. A second issue regarding quality is the low development impact of certain investment projects. As analyzed in more detail in Chapters II-IV, certain projects suffer from inadequate preparation while others are not of immediate priority. 7. Weaknesses in expenditure programming and budgeting have contributed to these quality and composition issues. In this regard, the sharp correction in expenditures was partially the result of ad hoc decisions taken because of resource constraints, and not based on systematic reviews of expenditure programs and sectoral strategies. For example, the Government has resorted repeatedly to withholding funds on a across-the-board basis in order to meet fiscal targets. A second area that needs major improvement is the coverage of the investment program. This is particularly true for projects implemented by the state-owned companies funded from non-budgetary sources, in particular Report and Recommendation of the President of the International Development Association to the Executive Directors on a Proposed Development Credit of SDR 68.3 to the Democratic Socialist Republic of Sri Lanka for An Economic Restructuring Credit (Report No. P-5271- CE), April 6, 1990. -3- the state-owned banks. The budgetary process could also be improved by correcting systematic over- budgeting, in particular of some social programs and, on occasion, public investment. Finally, the beneficial impact of improved budgetary and programming procedures would be compromised if monitoring and follow-up of expenditures are not reinforced in parallel. I.B. The Macro-economic Framework 8. Between Independence and 1977, the State gradually expanded its role in the economy. It imposed price controls and restrictions on trade, for example, between rice producing and consuming areas. Public investment was oriented toward capital-intensive projects to increase food production at the expense of plantation-based export crops and toward the establishment of import- competing industries. Trade policy was highly protectionist, with quotas against both agricultural and industrial imports. The liberalization of 1977 brought improved growth by enhancing the efficiency of the economy. Many price controls were eliminated, many quotas were replaced by tariffs, and the foreign exchange regime was liberalized. Rice production and private manufacturing output grew rapidly. Exports of manufactures (especially garments) and tourism increased along with workers remittances from Sri Lankan workers abroad. The period of intense civil conflict in 1988-89 set back the reform process and the growth of the economy. After the restoration of relative calm, the Government recommitted itself to reform. It has since made considerable progress in areas such as reduction of the budget deficit and privatization of state-owned enterprises. 9. The past decade has been a period of broad macro-economic adjustment, despite the setback caused by the intensification of the civil conflict in the late 1980s. The growth of real GDP recovered to the 4 to 5 percent range while inflation decelerated and the current account deficit of the balance of payments narrowed (Table 1.1). The steep downward trend in central government budget deficit played a key role in the adjustment (Figure 1.1 and Annex Table 1). CENTRAL GOVERNMENT DEFICIT CPERCENTAGE POINT SHARE OF GDP) 24 23 - 22 - 21 - 20 - 19 14 - 13 12 - 11 10 - Oudget 1990 1992 199 1995 1998 1990 1992 1991 2 199 19 9 19 9 19'99 199 1 19 3 Figure I.1 -4- Table 1.1: KEY ECONOMIC INDICATORS, 1985-92 (in percentage points) 1985 1986 1987 1988 1989 1990 1991 1992 Growth of real GDP 5.0 4.3 1.5 2.7 2.3 6.2 4.6 4.3 GNP per capita (US$) 337 354 360 375 367 417 460 494 Change in the GNP deflator 0.7 5.5 6.8 11.5 9.9 20.0 10.5 9.9 Real interest rate on 3-month Treasury bills 10.7 3.1 3.3 5.1 5.9 -3.6 4.8 5.8 Current hccount/GDP -9.9 -9.6 -8.0 -8.5 -7.1 -5.5 -7.6 -5.5 Exchange rate (Rs./US$) 27.2 28.0 29.4 31.8 36.0 40.1 41.4 43.8 Memo. Population growth rate 1.5 1.8 1.5 1.4 1.3 1.1 1.5 0.9 Source: Central Bank and World Bank A decline in public expenditure as a percentage of GDP was the source of the decline of the budget deficit since 1980, a period when the share of revenues and grants in GDP held broadly steady. Expenditures declined from an extra-ordinary peak in the late 1970s because of a contraction in public investment, and in particular, in the massive Mahaweli irrigation and power system. Expenditures also declined because the withdrawal of the government from industry and the privatization program reducing budget transfers to the state-owned enterprises. 10. Despite the substantial progress to date, there are several reasons to believe that further fiscal adjustment is required to achieve the Government's objectives of reducing inflation and promoting private investment and higher growth. First, the figures on the central government deficit appear to understate the broader consolidated government deficit because they do not include the share of the deficit of the state-owned enterprises which is financed off-budget by the state-owned banks, supplier's credits or by arrears. There are no official measures of the consolidated deficit, though a rough estimate places the consolidated deficit at I to 2 percentage points of GDP above the central government deficit in the late 1980s. Data on credit to the state-owned enterprises suggests that this off-budget deficit narrowed in the early 1990s. So the consolidated deficit would probably show the same broad pattern of adjustment as the published deficit. 11. Second, while most (but not all) of external debt comes from concessional sources (OECF, ADB, IDA and others), domestic debt is high cost, accounts for more than 80% of interest payments and has generated high interest rates, a high cost of doing business in the private sector, and foreign capital inflows that add to the inflation pressure. External public debt is now about 55% of GDP and domestic debt another 40% of GDP. 12. The most revealing indication of the importance of this problem is that, since 1992, the cost (to the Treasury) of the domestic debt has been higher than the cost of the war. In 1993, total interest payments on domestic debt are Rs. 26.3 billion or about 5.3% of GDP, compared to Rs. 20.7 billion spent on defence. In other words, interest payments and defense account for about half of the total revenues of the public sector. If the deficit is not reduced, the government will have to go on borrowing and paying increasing amounts of interest, the pressure on interest rates will continue and the negative impact on the private sector will become greater. -5- 13. One way to reverse this trend rapidly, would be to limit the fiscal deficit to the amount that can be financed externally with concessional financing. This would eliminate new net domestic borrowing, would maintain the domestic debt constant in nominal terms and would reduce it gradually in real terms and as a share of GDP. The initial reduction in the deficit would pay for itself very quickly in terms of real spending power by the Treasury: if interest rates fall by 5 percentage points, the savings to the Treasury would be about Rs. 5 billion or nearly 1 percent of GDP. This means that the initial reduction in the deficit would be in reality smaller as this savings become available to finance expenditures other than interest payments. 14. It would be difficult to further narrow the deficit principally by increasing revenues, as the share of revenues is GDP is already relatively high for a developing country with Sri Lanka's per- capita income. This suggests that much of the reduction in the deficit will have to be achieved by cutting expenditures. However, care needs to be taken to ensure that expenditure adjustments do not compromise the growth objectives. In particular, capital spending in priority areas of infrastructure, agriculture and social sector should be protected, along with more adequate provision for operations and maintenance. This suggests that much of the adjustment will need to come from restraining the civil service wage bill, cutting subsidies and dropping low priority projects from the public investment program. Continued privatization and restructuring of state-owned enterprises can be expected to reduce current transfers and to reduce the government's role in the investments of the state-owned enterprises. 15. The presence of concessional aid is not a reason for running a higher than programmed deficit. First, any public expenditures beyond the program would probably be financed by high cost domestic debt, rather than concessional aid. Furthermore, projects financed by concessional aid have domestic resource requirement, which must be domestically financed. The financing of these domestic resource requirements is subject to aggregate limits on taxation, credit, and money supply which are necessary to assure macro-economic stability. 16. The PER makes recommendations that would help continue the process of adjustment of the budget deficit by achieving the expenditure targets set in the Policy Framework Paper agreed with the Bank and the Fund. The desired fiscal adjustment is consistent with realization of the macro- economic targets shown on Table 1.2. 17. The scenario postulates an improvement in growth performance to about 6 percent per year by 1995 compared to the historic range of 4 to 5 percent. Trade is a major source of growth: the share of exports of goods and non-factor services rises from 31.9 percent of GDP in 1992 to 34.0 percent in 1997; over the same period, the share of imports in GDP increases more modestly because macro-economic policy sets limits on the expansion of demand. Agricultural exports recover as the sector diversifies into higher value-added export crops and as private sector management of the plantations increases the supply of exportable tea and rubber. Industry makes the greatest contribution to export growth as reduced tariffs bring invigorating competition from imports and as foreign investors in export industries respond to the more liberal trade regime and diminished domestic inflation and interest rates. Achievement of this growth scenario requires an increase in gross domestic investment, with public investment focused on priority areas, that would need to be financed by an increase in savings. -6- Table 1.2: KEY ECONOMIC TARGETS, 1991-1997 (in percentage points) Act Est Targets 1992 1993 1994 1995 1996 1997 Real Growth Rates GDP 4.3 5.3 5.7 6.0 6.0 6.0 GDP Per Capita 3.6 3.9 4.3 4.7 4.7 4.7 Private Consumption Per Capita 1.6 3.1 3.2 3.7 3.6 3.7 National Accounts Gross Domestic Investment/GDP 23.4 23.8 24.2 24.7 25.2 25.2 ICOR 5.3 4.3 4.0 3.9 4.0 4.0 Central Government Revenues 20.3 20.6 20.9 21.0 21.2 21.4 Current Expenditures 21.0 19.7 18.9 18.4 18.4 18.4 Capital Expenditures (incl. net Id) 6.8 9.0 9.0 9.0 9.2 9.4 Deficit (excl. grants) -7.4 -8.1 -7.0 -6.4 -6.4 -6.4 Balance of Payments Exports GNFS/GDP 31.9 33.1 32.6 33.1 33.4 34.0 Imports GNFS/GDP 40.2 40.8 40.3 40.8 41.1 41.4 Resource Balance/GDP -8.3 -7.7 -7.7 -7.7 -7.7 -7.4 Current Balance/GDP -5.4 -4.8 -4.6 -4.5 -4.4 -4.3 GDP Deflator 9.9 12.0 7.0 6.0 5.0 5.0 Source. The World Bank 18. The targets for savings that are consistent with the scenario reflect a continuation of the historical shift away from foreign financing of domestic investment (Table 1.3). Foreign savings (the current account deficit of the balance of payments) fall and national savings rise. The Government savings component of national savings turns positive as revenues and grants exceed public current expenditures, which are restrained and made more efficient through reforms suggested in this report. Improved investment opportunities and income growth resulting from further progress with the Government's reform program, stabilize the private savings component at their current high levels. The rise of government national savings is sufficient to fund an increase in both government and private gross domestic investment--despite the fall in foreign savings. 19. This is an optimistic but achievable scenario in the sense that it assumes fully realized adjustment. Without rapid progress in liberalizing trade and in privatizing or restructuring state-owned enterprises, Sri Lanka would not be able to sustainably increase its GDP growth rate to 6%, which would lead to weaker fiscal revenues. Moreover, the scenario assumes that the Government will meet its expenditure targets by sticking to the programs for rationalization of the civil service and the social transfers. Slippage from the program for adjustment would thus produce a higher fiscal deficit than projected. -7- Table 1.3: SAVINGS AND INVESTMENT BALANCE TARGETS: 1994-1997 (percentage point shares in GDP) Actual Targets 1991 1992 1993 1994 1995 1996 1997 Foreign Savings 7.6 5.4 4.8 4.6 4.5 4.4 4.3 National Savings 15.1 18.0 19.0 19.6 20.2 20.8 21.2 Private 17.1 18.6 18.5 18.4 18.3 18.2 18.3 Government -2.0 -0.6 0.5 1.2 1.9 2.6 2.9 Gross Domestic Investment 22.7 23.4 23.8 24.2 24.7 25.2 25.5 Private 13.2 16.8 14.8 15.2 15.7 16.2 16.1 Government 9.5 6.8 9.0 9.0 9.0 9.2 9.4 Source: The World Bank I.C. Past Trends in Central Government Expenditures 20. The success of macroeconomic adjustment thus depends critically on improving government savings and reducing domestic financing of the budget. In turn, this implies that public current expenditures need to decline significantly. A review of past adjustment in public expenditures should offer lessons for the future. 21. There has been a substantial cutback in the level of public expenditures since 1980, reversing the increasing trend of the 1970's. Total central government expenditures leaped by nearly 11 points of GDP from 1973-77 to 1978-82 as the Government accelerated its on-budget capital expenditure on large infrastructure projects such as Mahaweli (Table 1.4). In contrast, total outlays declined from some 43% of GDP in 1980 to 28% in 1992. Much of this downward adjustment has been focussed on capital expenditures, which have fallen from 24% of GDP to 7% during this same period (Box I.1).' Capital expenditures declined in the 1980s as major projects matured and the Government undertook fewer massive new projects. Recently, the renewed commitment to fiscal adjustment during 1988-93 also contributed to the decline in the ratio of on-budget investment to GDP. In particular, there was a reduction in capital transfers to state-owned enterprises, partly because of the privatization program. 2' On-budget capital expenditures slumped to an extraordinary low of 6.8% of GDP in 1992. The decline, however, is overstated because of a statistical distortion. Capital expenditures includes net on-lending of foreign borrowing, which includes privatization proceeds (recorded as a financing item--i.e., as a negative entry), restructuring costs, and, of course, repayment of on-lending. All three items represent financial entries rather than tangible real investment, so they should be excluded from an economic concept of real investment. Adjusting for these financing items raises on-budget capital expenditures to 7.9% in 1992. -8- Table 1.4: TOTAL AND CAPITAL EXPENDITURE: 1958-93 (percentage point share in GDP) YEAR TOTAL CAPITAL 1958-62 25.5 7.3 1963-67 27.7 7.3 1968-72 28.4 7.1 1973-77 24.3 6.7 1978-82 35.1 15.2 1983-87 32.6 14.0 1988-93 31.4 10.1 Source: Central Bank and CEM, 1988 Alternative Concepts of Government Capital Expenditures There are several different concepts of public capital expenditure commonly used in Sri Lanka. These are: On-budget Capital Expenditures: Expenditures on investments of: (i) central government ministries; (ii) provincial and local governments that are financed through transfers; (iii) state-owned enterprises, boards, and authorities that are financed through transfers from the central government and through net on-lending of foreign borrowing; and (iv) other net on-lending. (Sources: Ministry of Finance and the Central Bank). Off-budget Capital Expenditures: Expenditures on investments of state-owned enterprises, boards and authorities that are financed from sources other than the government budget. The financing is from own-funds, the state-owned banks, or from suppliers and sometimes bears a government guarantee. (Sources: no official or complete record of off-budget capital expenditures). Public Investment: The concept of public investment in the PIP seeks to approximate the national income account concept of public gross domestic investment. It is calculated as the sum of identified public investment, a provision for investment in projects which have not yet been identified, an estimate of extra-budgetary investment, and some smaller adjustments. The estimate of extra-budgetary investment does not appear to include all off- budget investments. (Source: Ministry of Policy Planning and Implementation). Identified Public Investment. Identified public investment from the PIP is the sum of planned investments in specific projects. This does not include some on-budget projects and some large off-budget investments of the state-owned enterprises. (Source: Ministry of Policy, Planning and Implementation). Box 1.1 -9- 22. Efforts to reduce current expenditures have not been successful (Table I.5)." On the contrary, current outlays of the Central Government have increased from 18.5% of GDP in 1980 to 21%-22% of GDP in the late 1980s-early 1990s, despite a reduction in current transfers/subsidies, in particular the elimination of the fertilizer subsidy. With the outbreak of civil strife in the late 1980's, defense expenditures surged to 4.8% of GDP in 1987; they have since stabilized at slightly over 4% of GDP (Figure 1.2). Interest payments have also risen significantly, from 3.4% to 6% of GDP between 1980 and 1992, reflecting the growing public debt burden. In addition, the introduction of new social transfer, such as Janasaviya and the Mid-Day Meal, more than offset the decline in the cost of the Food-Stamp Program. Finally, despite adjustment efforts, personnel costs, including both wage and pension payments, remain high. In 1990 and 1991, the Government reduced the wage bill by cutting the size of the civil service by some 40,000 through attrition and a voluntary retirement program. However, pensions jumped in 1991 as a result of the generous incentives introduced as part of the voluntary retirement scheme. Furthermore, the recent large scale hiring of teachers and health officials is significantly eroding these efforts to reduce staff. Compared to 1988, by 1992 the reduction in wages was more than offset by the increase in pension payments and, while data are uncertain, it appears that the number of civil servants had increased. Table 1.5: ECONOMIC CLASSIFICATION OF CENTRAL GOVERNMENT EXPENDITURES (percentage point shares of GDP) Prov. Bud. 1980 1988 1989 1990 1991 1992 1993 Total Expenditures" 42.7 34.5 32.6 31.1 31.9 27.8 28.7 Current Expenditures 18.5 20.8 22.6 22.4 22.4 21.0 19.7 Salaries & Wages 5.0 4.5 5.7 4.9 4.8 4.9 4.6 Goods & Services 2.6 4.8 4.4 4.5 4.8 4.3 4.0 Interest 3.4 5.7 5.7 6.4 5.9 6.0 6.2 Transfers 7.6 5.8 6.8 6.5 6.9 5.7 5.2 of which, to Households 6.1 4.0 4.8 4.7 5.4 4.0 4.2 of which, Pensions 1.1 1.9 1.9 1.4 2.4 1.8 1.8 Capital Expendituresb/ 24.1 13.7 10.0 8.8 9.5 6.8 9.0 memo item: Defence' 1.4 3.7 3.4 4.3 4.1 4.2 4.2 " Figures for 1992 and 1993 include budget for under-expenditure. b/ Includes net lending. ' Defence is a component of Salaries & Wages plus Goods & Services. Source: Central Bank The detailed economic classification appears in Annex Tables 2 and 3. - 10 - DEFENSE EXPENDITURES CPERCENTAGE POINT SHARE IN GDP) 5 45 Est 3 5 3 25 2 15 1 1980 1 19632 1 9184 19186 1 19188 1 9490 1 19'92 - - 1961 1963 1985 1967 19899 1991 1993 Figure 1.2 23. A review of the functional classification of central government expenditures reveals major shifts in public expenditure priorities (Tables 1.6 and Annex Tables 4-7).' The largest use of public expenditures in the early 1980s was economic services, in particular agriculture. However, outlays on these services have declined sharply since 1987 as major components of the Mahaweli system were completed and as the Government disengaged from manufacturing and mining. On the other hand, there has been some efforts to increase spending on energy and water supply. Expenditures on social services have remained a high priority of the Government, claiming 8.6% of GDP in the 1993 budget. I.D. Overview of the Public Investment Program for 1993-1997 24. Broad Composition. The PIP for 1993-97 that was published by the Ministry of Policy Planning and Implementation in May, 1993 calls for a broadly steady ratio of public investment to GDP of 9-10 percent over 1993-97, about the same as the average public investment from the PIP for 1992-96 that was published in October, 1992 (Figure 1.3). Given the need to balance further fiscal adjustment with the many outstanding development needs, the proposed level of public investment appears reasonable, though there is room for a small reduction in capital expenditures through cancellation or postponement of low priority investments. Annex Tables 4 and 5 show the detailed breakdown of public expenditures by function, and Annex Tables 6 and 7 show current and capital expenditures by function. - 11 - Table 1.6: FUNCTIONAL CLASSIFICATION OF CENTRAL GOVERNMENT EXPENDITURES (percentage point shares of function of GDP) Prov. Bud. 1987 1988 1989 1990 1991 1992 1993 Total" 38.0 33.4 33.2 30.6 31.2 28.6 28.9 General Servicesb/ 4.1 7.3 6.6 6.7 6.0 6.2 5.6 Social Services 11.2 8.5 9.4 8.5 9.3 8.4 8.6 Economic Services 16.5 11.6 9.6 7.6 8.0 5.6 7.7 Other" 6.1 6.0 7.6 7.8 8.0 8.0 7.4 a The figures are approximate because they include gross rather than net on-lending and do not include advance accounts. As a result the total in the functional classification differs from the more accurate total in the economic classification. b/ Includes civil administration, defence, and police. Cl Mostly interest payments. Source Central Bank and World Bank Staff Estimates PUBLIC INVESTMENT ACTUAL 1985-91; PLANNED 1992-96 & 1993-97 (PERCENTAGE POINT SHARE IN GDP) 15 14 13 12 - 11 PIP 93-97 (6ky 93) ID - 1P 92-96 COCt 92) 9 1995 1996 1997 1989 199 1990 1991 1992 1993 1994 1995 1996 1997 Figure 1.3 25. The PIP quantifies real investment of the broad public sector (Table 1.7). The largest and most meaningful component of public investment is "identified public investment," the sum of specific projects. The second most important component is the 'Supplementary Provision,' which is calculated to keep public investment at the target level of about 9-10 percent of GDP. The 'Supplementary Provision' becomes large in 1996 and 1997 as the Government has not yet identified - 12 - projects to absorb those funds for relatively distant years. Adding the 'Supplementary Provision' to identified public investment projected gives budgetary capital expenditure.' The third component of public investment is "extra-budgetary investment," an estimate of investments of the state-owned enterprises that are financed off-budget, although these numbers understate off-budget public investment, as explained below. A detailed sectoral and project breakdown is only available for identified public investment (Table 1.8 and Annex Table 8). Table 1.7: PUBLIC INVESTMENT: 1993-97 (percentage point share in GDP) 1993 1994 1995 1996 1997 Identified Public Investment' 9.5 10.0 9.9 7.8 6.2 + Supplementary Provisionsb/ 0.0 0.2 0.2 2.3 3.9 = Budgetary Capital Expenditure' 9.5 10.2 10.1 10.1 10.2 + Extra-budgetary Investment 0.7 0.7 0.6 0.5 0.5 - Land, etc.' 0.1 0.1 0.1 0.1 0.1 - Other adjustments' 0.2 0.1 0.1 0.1 0.1 = Public Investment' 9.9 10.5 10.4 10.5 10.4 Investment in specific projects. b/ Investment planned for projects which have not yet been identified. Excludes repayment of on-lending. These are capital expenditures that do not contribute to formation of new capital assets such as purchases of land, compensation, payments of guarantees, and capital transfers to foreign institutions. Largely capital expenditures in defence related activities. e Intended to approximate the national accounts concept of public gross domestic investment. Source Ministry of Policy Planning and Implementation 26. The broad sectoral priorities of identified public investment for 1993-97 remain mostly unchanged from the recent past: progressive disengagement from massive new investment in Agriculture, reduced allocations to industrial enterprises, and a modestly increasing commitment to investment in Human Settlements and Social Infrastructure (Figure 1.4). As explained before, the decline in public investment in Agriculture is made possible by the completion of major headworks and canals of the Mahaweli system. However, there is a stronger emphasis on economic infrastructure than in the past. These priorities are broadly consistent with the overall development objectives of Sri These figures do not correspond to capital expenditures in the economic classification of the budget--mostly because they exclude repayment of on-lending. Table 1.8: IDENTIFIED PUBLIC INVESTMENT -- SHARES IN GDP Actual Planned* 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1 Agriculture 4.9 4.3 4.1 3 1 29 2.0 1.8 1.3 1.5 1.4 1.5 1.3 1.1 (a) Mahaweli 33 2.7 2.3 1 5 1.8 09 0.7 0.6 05 0.5 0.7 0.6 0.6 (b) Other Irrigation 05 0.6 08 05 0.3 03 02 0.2 0.2 02 0.2 0.2 0.1 (c) Forestry 0.3 0.3 04 02 0.1 0.2 02 02 0.1 0.1 0.1 0.1 0.1 (d) Land 00 00 00 0.0 00 00 0.0 0.0 0.2 0.2 0.1 0.1 0.1 (e) Field,Export Crops 0.4 02 04 04 03 0.2 02 02 0.2 0.2 0.1 0.1 0.0 (f) Livestock 00 00 0.0 00 00 0.0 00 00 00 0.0 0.0 0.0 0.0 (g) Fisheries 01 0.1 0.1 01 0.1 00 00 00 0.1 01 0.1 0.1 0.1 (h) Plantations 03 03 0.2 03 02 04 04 02 01 0.1 0.2 0.2 0.2 2 Industries, Tourism & Trade 0.0 00 03 03 05 05 07 08 02 0.2 0.2 0.2 0.2 3 Human Settlements 1 0 08 1.0 08 07 1.1 1 2 1 5 1.7 1.6 1.4 1.2 0.9 (a) Housing 0.3 03 02 02 01 01 01 0.1 0.1 0.1 0.1 0.1 0.1 (b) Urban Infrastructure 0.0 0.0 0 1 0 1 0 0 02 0 0 0.2 0.3 0 3 0.3 0.2 0.1 (c) Other Construction 0.2 0.1 0.1 0.1 01 00 0.1 0.0 00 0.0 00 00 0.0 (d) Environmental Management 00 0.0 0.0 0.0 00 0.0 0.0 00 0.0 0.0 0.0 0.0 0.0 (e) Water Supply & Sanitation 0.5 0.4 0.6 04 0.3 02 04 0.7 0.6 0.6 0.4 0.2 0.2 0 (f) Provincial/Regional Development 0.0 0.0 0.0 0.0 0 1 0.4 0 4 0.4 0.4 0.4 0.4 0.4 0.3 (g) IRDPP 0.0 0.0 0.0 0.0 0.1 0.2 0.2 0.1 0.2 0.1 0.2 0.2 0.1 4 Economic Infrastructure 1.1 1.7 2.2 36 2.6 2.9 3.1 1.9 3.4 4.1 3.9 2.5 1.5 (a) Transport 0.3 0.6 0.5 07 0.6 1 1 1.5 1.0 1.3 1.6 1.8 1.5 1.1 (b) Power & Energy 0.3 06 1.0 2.0 1.5 1 6 1.1 0.6 1.2 1.0 0.5 0.5 0.3 (c) Posts & Telecommunications 0.3 0.4 0.5 0.8 0.5 0.2 02 0.1 0.4 0.8 0.5 0.3 0.1 (d) Other Economic Overheads 0.2 0.2 0.2 0.2 0.0 0.1 0.3 0.2 0.4 0.7 1.1 0.3 0.0 5 Social Infrastructure 0.8 1.2 1.5 1.1 0.9 09 1.2 1.1 1.1 1.2 1.3 1.1 1.0 (a) Education 0.6 0.7 0.6 0.5 0.4 0.3 0.6 0.6 0.6 0.6 0.6 0.6 0.5 (b) Health 0.2 0.2 0 7 0.5 0.5 0.5 0.5 0.4 0.4 0.5 0.6 0.4 0.4 (c) Others 0.1 0.3 0.3 0.1 0.0 0.0 01 0.1 0.1 0.1 0.1 0.1 0.1 6 Administrative Overhead 4.8 6.3 4.8 4.0 3.0 2.8 2.5 26 1.7 1.6 1.6 1.6 1.6 (a) Administrative Overheads 2.5 3.7 3.7 3.4 1.2 1.9 1.5 1.4 1.1 1.2 1.2 1.2 1.2 (b) Head Miscellaneous 2.3 2.6 1.1 0.7 1.7 0.9 0.9 1.2 0.6 04 0.4 0.4 0.4 TOTAL IDENTIFIED 12.5 14.2 14.0 12.9 10.6 10.2 10.5 9.3 9.5 10.0 9.9 7.8 6.2 'From Public Investment Plan, 1993-97 (May 93). Source: Ministry of Policy Planning and Implementation - 14 - Lanka. However, as explained within later chapters of the Report, there are major issues surrounding the composition of the proposed program and its individual components. In addition, there are two important process-related issues that need careful review and reform: (i) coverage of the PIP; and (ii) the Budget and Public Investment Planning Process. MAJOR COMPONENTS OF IDENTIFIED PLANNED PUBLIC INVESTMENT (ACTUAL 1985-92, PLANNED 1993-97) 55 5 45- AQflculture Economic Infrastructure 35 3 Soc Infr & H 25 2 15 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 Figure 1.4 27. Coverage of the Public Investment Program. The usefulness of the identified PIP as a guide to the intentions of the broad public sector is compromised by its incomplete coverage of public investment. The most significant recording problem is that the record of identified planned public investment omits substantial investments of the state-owned enterprises. In practice, the National Planning Department includes an investment project in the PIP if it satisfies all the following criteria: i. approved by the cabinet ii. financing is channelled through the budget iii. funding has been identified. These criteria exclude investments of the state enterprises from the PIP which are financed through: i. retained earnings ii. supplier's credit iii. other credit not channeled through the budget (with or without a government guarantee) iv. BOT arrangements. - 15 - 28. For instance, the identified PIPs for 1992-96 and 1993-97 do not include the purchase of airbuses by Air Lanka, of ground buses by the bus companies, many of the power plants planned by the Ceylon Electricity Board and the entire capital investment program of the Lanka Electricity Company (Annex Tables 9 and 10). The off-PIP investment of just six large state-owned enterprises amounts to 2.2% of GDP (Table 1.9). Public investment published in the PIP includes an adjustment for 'extra-budgetary' expenditures that is meant to capture these investments. This adjustment, however, appears under-stated. Table 1.9 shows that the off-PIP investments of the six state-owned enterprises sum to much more than the 'extra-budgetary' component of public investment. Table 1.9: OFF-PIP INVESTMENT OF SELECTED STATE-OWNED ENTERPRISES (Billions of Current Rupees) 1993 Air Lanka 2.58 Ceylon Electricity Board 2.19 Ceylon Petroleum Corp. 0.20 Lanka Electricity Co. 2.35 Port Authority 1.02 Telecom 2.56 Total off-identified PIP 10.90 Note: 'Extra-budgetary' investment from PIP 3.6 Source Business Plans of State-owned Enterprises and Mission Estimates. 29. The investment of state-owned enterprises should be treated the same as other public investments. Many state-owned enterprises are under the control of line ministries, which can choose their investments. Some of the state-owned enterprises and authorities do, however, have a degree of independence. Still, the investments of the state-owned enterprises often subject the Government to a contingent liability because they bear an explicit Government guarantee. Whatever the exact form of financing, the Government bears the ultimate responsibility for servicing the debt incurred by the state- owned enterprises. 30. Another aspect of incomplete coverage is that the identified PIP omits some on-budget projects. Several on-budget projects mentioned in the text of the previous PIP, Public Investment: 1992-1996, are not recorded in its tables: for example, the construction of the second international airport at Hingurakoda and the Katunayke (main airport) expressway." There are also several other projects which would be financed through the budget but are not in the PIP for 1992-96. These are the Colombo-Trincomalee Freeway, the Colombo-Galle coastal Expressway, and the Colombo- Ratnapura Motorway. The PIP for 1993-97 provides a better record than the earlier PIP in several respects. It records projects such as the Colombo-Katunayake Expressway and several large power projects, such as the Kukule, that were omitted from the PIP for 1992-96. The omissions of projects, Z/ The tables list the small design cost but not the large construction cost. - 16 - probably explains why actual expenditures on public investment are usually higher than identified planned expenditures from the PIP of the previous year (Figure 1.5). ACTUAL / PLANNED PUBLIC INVESTMENT (IN PERCENTAGE POINTS) 170 150 - 140 - 130 - 120 - 110 - 100 - 90 - 70 - 60 - 50 - 40 - 30 20 - 10 0 1995 1995 1997 1988 ie 1990 1991 Figure 1.5 31. A more complete measure of public investment would contribute to the quality of the final PIP and the budget. Investment and budget policy-makers would have a clearer view of the over-all physical investment and the corresponding financial obligations of the public sector. When confronted with this information, policy-makers might reallocate funds during the budget and public investment planning process. 32. The Budget and Public Investment Planning Process. The Government of Sri Lanka has a well established process for preparing the budget for the oncoming year. The PIP supports the budget process by providing a framework for setting capital expenditures in the budget, although the weaknesses in coverage discussed earlier significantly reduce its usefulness. The key institutions in managing the budget and PIP development processes are the Treasury Department of the Ministry of Finance and the National Planning Department of the Ministry of Policy Planning and Implementation. In Sri Lanka, the fiscal year of the government budget corresponds to the calendar year. 33. The budget and PIP processes run in parallel. They both start in March when the budget call circular is sent out to the finance divisions of line ministries (Table I.10). Next, individual departments are asked to submit their proposed estimates of expenditures for programs and projects during April to their ministry's finance division, which examines the proposals and then submits them to the Budget Division (in the Treasury Department, Ministry of Finance).' The External Resources Division of the Ministry of Policy Planning estimates foreign funding of projects. In June, the Budget Division discusses the proposals with the ministries. The emphasis is often on cutting back excessive demands; sometimes the resulting cuts are perceived as across-the-board. Over July and August, the Treasury prepares the budget based on an estimate of revenues. The budget is submitted to the Y With copies to the National Planning Department. - 17 - Cabinet in September; the Finance Minister presents the budget to Parliament in October for a budget debate that is limited to 26 days so that the Parliament approves the budget in November. Table 1.10: THE BUDGET CYCLE (for the next Jan. I fiscal year) Quarter Steps Responsible institutions Q I and Q2 1. Overall policy Cabinet 2. Formulation of programs Departments and other state-controlled entities. and projects. 3a. Discuss programs and projects in National Planning Department sectoral committee and formulate Public Investment Program. 3b. Negotiate foreign aid. External Resources Division. 4 Call for annual budget proposals. Treasury. Q3 5. Formulate revenue estimate. Treasury Fiscal Policy Division in consultation with Ministries and the Central Bank. 6. Formulate expenditure estimate. For current expenditures, Treasury Budget Division in consultation with Ministries. For capital expenditures, Treasury and National Planning Department in consultation with Ministries. 7. Examine and approve budget Cabinet. Q4 8. Presentation to parliament. Finance Minister. 9 Examination and approval of budget. Parliament 10. Authorization of expenditure Finance Minister 34. Sri Lanka also has a well established process for developing the PIP, which has weakened recently. The initiating stages (project identification, preparation, and preliminary appraisal) are the responsibility of the departments, state enterprises and possible donors who propose projects to the sponsoring ministry. The line ministries have their planning units, which use the proposal to generate project and sectoral plans in accordance with guidelines set out by the National Planning Departments and then submit the plans to the National Planning Department. In the past, the key institution, that granted preliminary approval, was the Committee of Secretaries, which included all ministry Secretaries and provincial Chief Secretaries, and the National Planning Department of the Ministry of Policy, Planning and Implementation. The Committee vetted projects to assure their economic quality and their consistency with sector strategies and resource constraints, meeting weekly on the day before the cabinet meeting. Major projects and other projects with significant implications for external funding were subject to a two-tier approval process: first, they were submitted to the Committee of Secretaries for approval in principal and later for approval in detail. The External Resources Division of the Treasury functions as a coordinator and forms the link between external - 18 - agencies and executing agencies and also monitors disbursement. The Cabinet must give final approval to projects, which only then would be included in the PIP. 35. However the Committee of Secretaries was recently abolished. The External Resources Division and the National Planning Department now comment on project proposals that come before the Cabinet. But line ministers sometimes submit project proposals directly to the Cabinet without subjecting them to a centralized vetting process. 36. Despite these elaborate arrangements, as analyzed below, there are still a number of large projects which suffer from inadequate preparation or are not of immediate priority. 37. Improving the Budgetary Process and Investment Programming. One of the strengths of the public investment planning and budget process is its linkage to senior leadership. In recent years, the President has held the portfolios for the Ministry of Policy, Planning and Implementation and the Ministry of Finance, which oversee the process. The budget process is also successful in that the budget is approved by parliament before the beginning of the year; thus, there is no need for a temporary funding arrangement during the first part of the year. One of the difficulties in the budget process is that departments and ministries generally calculate their expenditure proposals for the forthcoming year as an estimate of current year expenditures plus an increment. The length of the budget-preparation cycle means that the departments and ministries estimate current year expenditures early in the year, when data on actual expenditures are only available for the first quarter. The long duration of the budget preparation cycle is partly caused by translation of the budget into three languages and a lack of computerization of the process. 38. In recent years, the final budget has over-funded major items such as the social transfer programs, and, on occasion, public investment. On occasion, budgeted expenditures on the Janasaviya Program have been higher than necessary because effective screening has reduced the number of recipients. The Government sometimes under-spends on projects by withholding disbursement during the year. The budget for 1993 shows an explicit entry for under-expenditure of over Rs. 2 billion (Annex Table 2). No breakdown is given across programs, but the sum appears arises mostly from under-expenditure on the major social transfer programs and on capital expenditures rather than on cadre. 39. The practices used in the budgeting of the personal emoluments of civil servants should be improved. The budget for personal emoluments in the forthcoming year is calculated from actual expenditures in the first quarter of the present year, adjusted for increases in wages and benefits. Net departures from the civil service during the final three quarters of the present year can thus lead to over-budgeting. The more serious problem is that the budget lists, but does not budget for, the approved cadre, rather than actual employment. At end-March 1993 there were some 60,000 authorized unfilled positions. This leaves open the possibility the vacant positions will be filled for which funds were not budgeted. 40. To improve the early stages of the process, the budget call circular sent to the finance divisions of the line ministries could be improved to provide specific instructions on the factors to be used in estimating the costs of imported goods, fuel and other goods and services. More importantly, the budget process could also be improved by scheduling reviews of several major recurrent expenditure programs each year to determine whether the programs are effectively serving a justifiable purpose. This would help eliminate programs which have lost their usefulness and are now crowding out more justifiable expenditures. This practice might be reinforced by establishing a 'sunset' clause that cancels each recurrent expenditure program after, for instance, five years unless as specific - 19 - decision is made to retain the program. Another way to improve the process for forming the recurrent budget would be to establish a multi-year framework for the recurrent budget that would guide the Departments and Ministries in preparation of proposals. For example, such a framework would help assure budgeting of sufficient funds for maintenance. To avoid over-running the budget for personal emoluments, the vacant positions should be cancelled, pending reviews to determine the necessary cadre (Chap. V). Table 1.11: PLANNING CYCLE FOR PUBLIC INVESTMENT PROJECTS Steps Responsible institutions Identification, preliminary Departments, state enterprises, and possible preparation and prehminary appraisal aid donors submit projects through the sponsoring ministry to the Committee of Secretaries. Preliminary approval. Committee of Secretaries, supported by the National Planning Department. Project preparation and feasibility. Sponsoring ministry and aid donor. Appraisal. National Planning Department. Approval. Committee of Secretaries. Revisions, if necessary, and Sponsoring ministry. preparation of Cabinet memorandum. Final approval. Cabinet. Record of investment plans Department of Planning records up to five-year costs of project in the Public Investment Program and Treasury records the following year's costs in the national budget. Implementation. Sponsoring ministry. 41. The Government should also conduct adequate rate of return analyses before admitting projects into the PIP. Chapters II and III below mention several large capital investment projects that passed through the budget process without adequate studies. Once the PIP is established, it should be linked more directly to the budget. Projects should not be allowed into the budget without passing through the PIP vetting process and project cost in the budget should not be allowed to greatly exceed those presented in the PIP. 42. In the end successful control of expenditures requires following-up the budget process. The funding of the recent decision to move up the civil service wage increase that had been deferred to January 1, 1993 provides an example. The wage increase is being funded by across the board cuts in capital expenditures and in non-wage and non-interest recurrent expenditures. This is likely to sacrifice priority items in the budget--such as necessary maintenance. In order to strengthen budgetary implementation and control, the authorities will also need to reinforce monitoring of public expenditures. Without adequate and timely information on actual expenditures, it is difficult to ensure - 20 - that budgetary allocations are being respected. Monitoring of expenditures of Provincial Councils is especially weak. 43. The budget process, in particular the preparation of the PIP, is also seriously compromised by the weaknesses in coverage discussed earlier. The omission of large off-budget investments of the state-owned enterprises, through the state-owned banks and other financial savings under the control of the Government, is of particular concern. Not only do these omissions distort the measurement of the financial operations of the public sector, but they may also compromise the quality of the investment program if omitted projects are subject to a weaker approval process. As discussed below, many of the more questionable public investments (e.g., the five Airbus 340s, the bus-kit scheme) are off-budget. I.E. Methodology for Evaluating Expenditures 44. Against this background reviewing generic issues regarding the overall level and broad thrust of public expenditures, as well as investment selection procedures and budgeting, the remainder of the report evaluates public expenditure policies and programs in specific sectors. Below we briefly review the methodology followed within the Report in evaluating expenditure projects and programs, and provide a summary of the estimated fiscal impact of the PER recommendations. 45. The expenditure programs that issue from the budget process and especially the on-- and off--budget public investments are the subject of the PER. The conclusions of the PER with regard to these programs emerge by subjecting the main projects and programs to generally accepted development criteria (Box 1.2). The central criteria are economic efficiency and sustainability. For Criteria for Evaluating Proiects and Programs a Economic Efficiency (i) Is there a well thought-out sector strategy? (ii) Is the project or program consistent with agreements on the relative roles of the public and private sectors? (iii) To the extent that economic rates of return are available, how does the project or program perform in this light? (iv) Does it have spill-over effects that detract from or contribute to investment and growth? (v) Are expenditures on operations & maintenance for each sector adequate? * Technical Efficiency (i) Was the least-cost technology adopted? * Fiscal Sustainability (i) Is the contribution of the project or program to the budget consistent with the desired macroeconomic adjustment? (ii) Are cost recovery policies underlying the operation of the project or program appropriate? 0 Administrative Sustainability (i) Is it well coordinated across levels of government and across agencies? (ii) Is it well coordinated across government agencies? (iii) Are there major implementation constraints? (iv) Is it transparent and are there opportunities for rent seeking? Box 1.2 - 21 - projects, this refers to whether the program or project will probably produce a low return or loss and the extent to which the expenditure is consistent with macroeconomic adjustment. The criteria also set standards regarding the technical and administrative quality of the project or program and its distributional effect on incomes. Annex Tables 11 and 12 summarize the evaluation of the current expenditure programs reviewed by the PER and the marginal investments in the broad PIP, while Table 1.12 summarizes on and off-budget projects which should be cancelled or delayed until they can be better justified. Table 1.12: PROPOSED CUTS IN THE PUBLIC INVESTMENT PROGRAM: 1994-97 (Billions of Current Rupees) 1994 1995 1996 1997 Railway extensions Matara/Katargama 0.2 0.3 0.5 0.5 Batticaloa/Pottuvil 0.2 0.6 0.6 0.7 Broad-gaging of Kelani Valley Line 0.1 -- -- -- Mahaweli System L 0.2 0.7 0.5 0.3 Udawalawe, Right Bank 0.3 0.2 -- -- Thermal Plant (Diesel 40 MW)/ 0.3 1.8 1.1 -- Total On-budget 1.3 3.6 2.7 1.5 % of GDP 0.2 0.6 0.4 0.2 Air Lanka (5 A340 Airbuses). 11.4 6.1 -- -- Rural Electrification 1.1 -- -- -- Total Off-budget 12.5 6.1 0.0 0.0 % of GDP 2.2 1.0 0.0 0.0 J Includes associated transmission. Source. World Bank Staff Estimates 46. Fiscal Impact of the Recommendations. Applying these criteria in evaluation of the expenditure programs leads to recommendations to reduce total public expenditures, both on and off- budget (Table 1.13). The proposed on-budget savings would be greater than the target decline in the total government expenditure of 0.9 percent of GDP by 1997 (see Table 1.2) if they are not offset by expenditures on new programs. A greater than projected cut in expenditures from 1993 to 1997 might be necessary to offset any over-shooting that might develop in expenditures in the 1993 base year. However, the savings achieved through the rationalization of the main social transfer programs could be used on well designed poverty alleviation programs. Until such programs can be implemented, any savings should contribute to the reduction of the fiscal deficit. Reform would bring a net savings in total expenditures in 1994 because it would take at until 1995 to identify and implement new poverty alleviation programs. - 22 - 47. There are several likely declines in public expenditures that are not quantified on Table 1.12. For instance, the savings shown for reform of the civil service is only the part attributable to reform (0.3 percent of GDP by 1996), computed by subtracting wage and pension costs in the reform scenario from those in the no-reform scenario. However, the share of wages and pensions in GDP already falls by 1.6 percent of GDP in the no-reform scenario because of the assumption that the Government maintains its current restrictions on new hiring. Moreover, realization of the proposed initial cuts, both on and off-budget, would set in motion a virtuous cycle of declining deficits and real interest rates, as described above. Table 1.13: PROPOSED REDUCTIONS IN CENTRAL GOVERNMENT EXPENDITURES: 1994-1997 (percentage point shares in GDP) 1994 1995 1996 1997 Capital Expenditures On-budget 0.2 0.6 0.4 0.2 Off-budget 2.2 1.0 0.0 0.0 Current Expenditures Civil Service reform" 0.1 0.0 0.3 0.3 Maintenanceb/ -0.2 -0.2 -0.2 -0.2 Social transfer programs" 1.1 0.7 0.8 0.8 Total Reduction, if savings on social transfers Reduce total expenditures 1.2 1.1 1.3 1.1 Fund poverty alleviation pgms after '94 1.2 0.4 0.5 0.3 a/ The reform projection compared to the no reform projection. b/ Maintenance in roads, schools and health structures. Negative numbers signify and increase in spending. Savings that should be partly or entirely re-directed into well designed poverty alleviation programs. Source- The World Bank Staff Estimates. - 23 - II. ECONOMIC INFRASTRUCTURE II.A. Introduction 48. Sri Lanka's development needs in the economic infrastructure area are large. In a recent informal survey carried out by the July 1993 Private Sector Assessment mission, the private sector ranked deficiencies in transport, telecommunications and power among the most important constraints to private sector activity in Sri Lanka. Similar results are reported in other papers analyzing impediments to private sector growth. Recognizing these deficiencies, the Government, in recent years, has been gradually increasing the share of public resources allocated to developing economic infrastructure. 49. However, meeting these development needs will require the Government to take action on several fronts. Tight public sector resource constraints and deficit reduction targets make these actions all the more critical. In particular, the Government should move rapidly on three key issues that are common to practically all the sector: (i) Tariff Policy. One fundamental issue is the low level of tariffs in Sri Lanka. Unless there is a clear justification (correction of market failure or equity considerations), tariffs need to be increased towards long-run marginal costs to reduce and eventually eliminate implicit subsidies. Equally important, raising tariffs will help generate sufficient funds for operations, maintenance and investment while reducing dependence on budgetary transfers. Adequate tariffs are also important to attract private sector investment. The record in this area is weak, as will be seen in the sector review below. (ii) Role of the Private Sector. A second theme is to put in place policies to generate private sector interest in operating infrastructure and in providing new infrastructure. In Sri Lanka, as in many other countries, it appears that the public sector has become overextended, and public resources should be redirected and concentrated in those areas in which public sector intervention is required because of market failures or social objectives. Private activity in infrastructure would not only increase the amount of investment in the sector without overburdening the budget, but should also lead to significant efficiency gains in the provision of these services. In some subsectors, such as telecoms and road and air transport services, deregulation and outright privatization of state-owned enterprises should be pursued, while adequate regulatory frameworks are developed. In other subsectors, impediments to private investment through BOO and BOT schemes should be eliminated. Once again, developing an appropriate regulatory framework is crucial. The Government should ensure that risks are shared adequately between the public and private sectors and public guarantees to debt contracted by private operators should only be granted when appropriate. Also, the authorities should not provide subsidies to private operators, including subsidized credit from state-owned banks and excessive tax incentives. While progress has been achieved in tackling these issues, these efforts should be pursued with increased vigor. (iii) Composition and Quality of the Public Expenditures. Weaknesses in the composition and quality of the public expenditure program in infrastructure also need to be addressed. The Government should establish clear priorities among activities to - 24 - remain within the public realm. This is a difficult task, but as explained in the previous Chapter, in principle priorities should be assigned to activities with the highest economic returns, taking into account social considerations that may require modifications to the approach. For infrastructure in Sri Lanka, two general recommendations seem relevant. First, allocations to maintenance activities would in most subsectors have high rates of return. Funding of these activities has been insufficient, with the result that the operating efficiency of facilities is low and that the existing public capital stock is deteriorating. Proper maintenance does not only yield high economic benefits, but can also save major expenditures for rehabilitation in the future. Recent increases in maintenance funding should be protected from cuts and in some cases even expanded further. Second, investment projects need to be based on careful technical and economic analysis and only those with high rates of return should be implemented. The review of the public investment program reveals that a number of proposed relatively large projects have been insufficiently analyzed or whose phasing is overambitious. These projects need to be canceled, restructured and/or delayed until the deficiencies are corrected. 50. This chapter reviews government expenditure in three principal sub-sectors of infrastructure--water supply, power, and transportation--emphasizing the three issues discussed above. The other major sub-sector, telecoms, is not reviewed in this report. The large state-owned telecoms utility is being restructured and its privatization is under preparation by the Government. Major investments in the sector have been delayed awaiting privatization and much of the new investment is expected to come from the private sector. II.B. Water Supply and Sanitation 51. Background. While the Ministry of Housing and Construction is responsible for providing water, it has delegated this responsibility primarily to the National Water Supply and Drainage Board (NWSDB). Local authorities are, in principal, responsible for the provision of services within their boundaries, although in practice the NWSDB often also discharges this responsibility. In some cases, the NWSDB supplies water in bulk to local authorities who then distribute the water. The NWSDB concentrates its activities in urban areas and installs larger piped water schemes in rural ares. Many other organizations also provide service to rural areas, including, the Community Water Supply and Sanitation Project Unit of the Ministry of Housing, the Ministry of Health, the state-owned plantations, and the Mahaweli Authority. The state-owned plantations, the Sri Lanka Plantation Corporation and the Janatha Estates Development Board, provide water supply for their workers and their families on the plantations. The Mahaweli Authority provides water supply for people resident in its area. 52. Service Delivery and Coverage. About 88% of the urban sector and 65% of the rural sector have access to water. Coverage varies considerably by district with the Colombo district at 93% and the Ratnapura district at a low of 40%. Coverage for sanitation is about 81% in the urban sector and 70% in the rural sector. Coverage is not the only meaningful indicator of water service, as quantity and hours of service appear to have been emphasized over quality. In Greater Colombo, water leaving the main treatment plant meets government and WHO standards but falls below those standards at the point of distribution. The main technical reasons for the sometimes poor quality of water at the points of distribution are intermittent supply and an inadequate distribution system. There have even been instances of raw water being distributed to consumers. The start of a replacement program for distribution pipes and a limited training program of operations staff in improved process control mark a beginning in addressing this problem. - 25 - 53. One of the most significant issues in water supply is the large volume of un-accounted for water. This contributes to the sub-sector's financial and technical problems by reducing cost recovery and thus the funds available for maintenance and rehabilitation. The level of un-accounted for water in Colombo, for instance, is about 39%, of which half is caused by leaks in the system and the other half by meter related difficulties, illegal connections and by-passes. It is estimated, for instance, that about 12% of meters need repair. Illegal connections and bypasses have been discovered in domestic premises and in industrial and commercial establishments. Water quality can fall because the illegal connections reduce pressure, which permits foreign materials to enter. 54. Sector Finances. The major financial issue in the sub-sector is the relatively low level of tariffs, which is a constraint on the financial health of the NWSDB. Tariffs are low by regional standards; moreover the national water tariff is below the marginal costs of production and the collection system is relatively ineffective. The tariff structure discourages consumption in low-cost regions and encourages it in higher cost regions through a cross-subsidy. Tariffs should be raised toward long-run marginal cost and the structure should be reformed to provide more incentives for conservation. Table 11.1: PUBLIC INVESTMENT PLAN IN TRANSPORTATION, 1981-96 / (Percentage Point Shares in Total) Mode of Transport 1981-85 1984-88 1986-90 1988-92 1989-93 1990-94 1992-96 Highways 19.9 26.3 34.2 40.2 39.1 51.0 36.1 Railways 31.3 26.5 42.5 42.2 34.6 37.5 36.0 SLCTB/Public Bus 26.1 0.0 0.1 2.6 7.3 0.0 0.5b Ports & Shipping 12.9 11.2 5.7 3.3 14.1 8.6 25.4 Air Transport 9.7 36.0 17.6 11.7 4.9 2.9 2.0c/ Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Note: Total in Rs. millions 5,369 10,423 12,549 22,912 25,911 26,230 46,517 as a % of PIP 6.0 9.8 9.3 13.9 17.7 13.1 17.8d' a/ Allocations do not correspond to actual expenditures. b/ Includes expenditures on the Transport Studies and Planning Center, National Transport Commission and the Dept. of Motor Vehicles. The SLCTB is being privatized. Includes the estimated cost of the second international airport but not the purchase of air buses by Air Lanka. d Excluding expenditure on the second airport, the percentage is reduced to 17.5 percent. Source: Ministry of Policy Planning and Implementation. 55. Partly because of the low tariffs, the water supply sector remains dependent on the Government and donors for the development of infrastructure. However, the situation appears to be improving. In 1992 the NWSDB recovered operations and maintenance costs plus two-thirds of annual debt service from collection. In 1993 it is anticipated that the NWSDB will collect the full - 26 - annual debt service. Part of the reason for this progress is the NWSDB's aggressive approach to collection of accounts in arrears. 56. There appears to be considerable potential to improve the performance of the sector by increasing the role of the private sector. The private sector could reduce cost by improving the utilization of labor and resources in areas such as meter repair and reading, repair of pumps and billing and collection. It can reduce capital costs through BOO and BOT arrangements. In the foreseeable future the NWSDB could change its role and become a regulatory agency. 57. The Public Expenditure Program. The PIP for 1993-97 does not include any new projects in water supply because funds were not available when the plan was formed. Some foreign funding for the NWSDB became available after the PIP was formed. In addition, the NWSDB intends to finance several new projects in its internal investment program from several sources of financing. For instance, the NWSDB intends to start at least one major new project, the 'Towns South' of Colombo project in 1994 by persuading the Cabinet to approve financing through a supplementary budget. In addition, the NWSDB will, spend about Rs. 400 million of its own funds, starting in 1994, on new small projects in areas which have no alternative source of supply and for rehabilitation. 58. The NWSDB's investment program places a welcome emphasis on maintenance and rehabilitation (Box 11.1). The program shows some investments in the Colombo area that are part of the proposed Fourth Urban Water Supply and Sanitation Project. The private sector should participate more fully in providing water supply through contracting out of specific tasks and through BOO and BOT arrangements for installation of infrastructure. One candidate for a BOO and BOT arrangement is the water treatment and distribution system planned for the Free Trade Zone in the Southern Province of Kogalla. Since the majority of the consumers will probably be private businesses, revenue should be considerably higher per cubic meter than those of a typical water supply scheme. 59. The Bank's assistance has historically concentrated in the Greater Colombo Area, where the system serves over one million people, and where demand is growing fastest. Recently the Bank has initiated a program to assist the rural sector in three districts with a combined population of nearly 2.5 million people. The Asian Development Bank is concentrating its efforts on the rehabilitation of 27 schemes. Other donors, such as Finnida and Dannida have been concentrating their effort in the central region in one relatively large town and in several small local governments. Investment on these small schemes should be made with the participation of the local community in all phases of project preparation and implementation. H.C. Power 60. Background. Two ministries supervise most power generation in Sri Lanka: (i) the Ministry of Power and Energy oversees the Ceylon Electricity Board (CEB) and the Lanka Electricity Company (LECo); and (ii) the Ministry of Lands, Irrigation and Mahaweli Development oversees the Mahaweli Authority. The CEB is a large public utility that is responsible for most of the generation, transmission and distribution of power in Sri Lanka. LECo is a small public utility that is taking over some small distribution systems from local authorities. The Mahaweli Authority runs multi-purpose projects in the Mahaweli river basin that generate electricity. Hydro generation predominates because the country has abundant hydro resources and no known reserves of fossil fuels. 61. Growing demand is straining existing generation capacity and creating shortages. In years of normal rainfall, almost the entire energy supply is generated by hydro power plants. In dry years the CEB is either forced to shed load or to run standby thermal plants. At present, the installed - 27 - Rehabilitation and Maintenance in Water Suply Because of a significant improvement in its financial health starting in late 1990, the NWSDB has been much more aggressive in rehabilitating the country's potable water infrastructure. Expenditures for rehabilitation jumped 500 percent between 1990 and 1991 (from Rs. 12.6 million to Rs. 75.6 million) and remained at about the same level in 1992. This amount includes funds for regular water system maintenance and is about seven percent of total revenues. Approximately Rs. 50 million was spent on the water system to arrest further deterioration in its aging water storage, transmission and distribution network, much of which dates to the first decade in the 20th century. Another estimated Rs. 25 million was spent on buildings, vehicles and non-water infrastructure and equipment and a modest amount for the waste-water system (which is mainly a municipal responsibility). Assuming that the financial condition of the NWSDB remains hopeful, its maintenance and rehabilitation strategy for the rest of the decade will be: (i) hold more or less constant the proportion of its budget spent on water system-related maintenance; and (ii) to significantly increase the capital amounts spent on rehabilitation of the water system.' Totaling ongoing and planned IDA and ADB projects, NWSDB-funded programs and other donor grants and loans, annual estimates for rehabilitation and maintenance for the water system over the next four years are: Capital Recurrent Percentage of Revenues (Rs. millions) (Rs. millions) 1993 374 967 7 1994 548 1308 8 1995 393 1689 9 1996 389 2111 10 The NWSDB is likely to increase its overall spending for rehabilitation and maintenance to about 10 percent 'of revenues over the rest of the 1990s. After most of the major system rehabilitation is completed in about 2002, it is expected that expenditures on maintenance would increase and that, together with those for rehabilitation; the combined total would be about 3 to 5 percent of revenues. Box H.1 generating capacity of the CEB's grid system is 1,247 MW of which 997 MW is hydro-power. The Mahaweli system generates two-thirds of the hydro-power and the Laxapana system generates the remaining third. The CEB is planning to enhance the role of thermal generation in order to improve reliability and to supply more energy. It is estimated that by 2000, about one-third of all generation will be thermal. 62. Sector Finances. However, the CEB would not be able to undertake the expansion program without continued large transfers from the budget or large credits from the state-owned banks. The reason is that the CEB's finances are constrained by inadequate tariffs, high operating costs and high financial charges. As a result, since 1991 the share of internal cash generation in financing investment has been grossly insufficient. The CEB has been obliged to resort to heavy borrowing - 28 - from the state-owned banks (Rs. 2.0 billion in 1992) and to deferment of debt service obligations to the treasury. In the 1993 budget, the CEB absorbs 1.2% of GDP in capital transfers and in on-lending for its capital investments. 63. Electricity tariffs are low in Sri Lanka. The CEB took a positive step by raising its average electricity tariff by 30.5%, effective July 1, 1993. However, this places the average tariff at about 6 US cents per KWH and at around two-thirds of the long-run marginal cost of producing electricity. Moreover, there is cross-subsidization of household consumers, some of whom pay only 15% of long-run marginal cost. Further tariff adjustments, starting with a 30% increase this year, are required to raise tariffs for all consumer categories to the economic cost of supplying electricity. Higher tariffs would also help to constrain demand and limit the investment necessary to meet demand. The CEB could then finance internally a greater part of its capital investment and reduce the need for transfers from the budget and borrowing from state-owned banks. 64. The CEB's financial condition is being negatively affected by several other factors. The CEB pays high financial charges because of the high interest rates on short-term borrowing from domestic banks. Through much of 1993, the financial condition of the CEB is also being weakened by the Government's unwillingness to meet its contractual obligations to reimburse the CEB for the increase in the Rupee cost of servicing foreign debt caused by depreciation of the exchange rate. The Government contracted to pay these costs in compensation for the significant interest differential that it charges the CEB for on-lending over the interest rate charged by the foreign lenders. 65. The rural electrification program also imposes an important financial burden on the CEB. For reasons of social policy, the Government plans to supply electricity to all 25,000 villages by 2000. Rural electrification schemes involve dis-proportionately high operating and maintenance costs because of the low consumer density, low consumption per consumer, price subsidization of consumers and long transmission and distribution lines. These are generally uneconomic schemes which financially burden the CEB. A capital transfer from the budget only partly compensates the CEB for the investment costs. Because of the financial burden, the CEB requested a direct subsidy from the budget for its rural electrification operations. The Monetary Affairs subcommittee appointed by the cabinet is now considering that request. Since rural electrification is justified mainly by social rather than economic considerations, the entire cost should be an explicit subsidy from the government budget. Given its low economic return, investment in rural electrification should not expand: keeping investment at the 1993 rate of about Rs. 650 million would save over Rs. 1 billion from the investment program during 1994-95. 66. The Public Expenditure Program. Investment programming in the power sector suffers from certain weaknesses. In particular, there is uncertainty over what expansion plans actually are, because of differences between the PIP (published by the Ministry of Policy Planning) and the corporate plans of the CEB and LECo. Annex Table 9 presents the investment program of the CEB as of May 1993 and shows which projects are included in the PIP. Annex Table 10 presents LECo's investment plan, which is not included in the PIP. 67. As it currently stands, the PIP in the power sector appears over ambitious. The projected growth of demand for electricity is one of the key factors for determining investment in generating capacity. The CEB investment program of the spring of 1993 (Annex Table 9) is based on an annual growth of demand of 10% per year through 2000; this was recently lowered to 8.4% which is high compared with the 6.6% annual growth of demand experienced in 1990-92 and with the 7.8% projected annual growth of demand estimated by the Bank in connection with the appraisal of the Second Power Distribution and Transmission Project (IDA Credit 2297-CE) in 1991. Therefore, the CEB's existing investment plans for generating capacity are higher than necessary to meet the CEB's - 29 - demand growth projection of 8.4%. The excess planned capacity included in Annex Table 9 consists of about 60 MW of thermal generation plant (diesel and gas turbines), with an estimated cost of Rs. 4 billion, including the cost of related transmission lines. The CEB is presently revising its investment program; the revised version is expected to be ready by the end of 1993. 68. Expanding the Role of the Private Sector. The Government is seeking to lower the fiscal cost of power investments and to increase the efficiency of investment by encouraging the participation of the private sector in several future power generation projects on a BOO or BOT basis. Several projects have been moved off CEB's investment program and are now intended to be built by the private sector (Annex Table 9). The Government advertized for proposals from local and foreign investors for coal fired thermal plant, gas turbines, diesels, the Broadlands hydropower station and mini-hydro schemes. The Government wishes to retain the large hydro projects (Kukule and Kotemale) in the public sector. These plans are in principle consistent with Bank policy toward the power sector, although the rapid implementation of the BOO and BOT schemes will eliminate the immediate need for additional hydro schemes. The Government needs to establish a regulatory framework covering private sector investment in power that would deal with tariffs. Under a Japanese grant, the Government is engaging consultants to assist in setting up a regulatory framework. ILD Transportation 69. Background. The transportation sector consists of roads, rail, ports and shipping, airports and air travel, and bus lines. Road transport is the backbone of Sri Lanka's economic activities, accounting for more than 80% of total freight movements and carrying about 85% of total passengers in 1991. The Railway has been run as a government department since its inception in 1864. Its network consists of nearly 1,500 kilometers of track and it employs about 22,000 people. However about one third of the routes are not operating because of the conflict in the North. The Port Authority is responsible for serving the shipping; the main port at Colombo has a deserved reputation for efficiency. There is a modern airport at Katunayake that operates well under capacity. The government participates in air travel through its ownership of Air Lanka. Air Lanka can be characterized as a niche carrier, with short and medium-haul routes to the Indian sub-continent and the Middle East and with some long-haul routes to Europe, Australia and East Asia. The airline's passenger volume has benefitted from the expansion of the region's traffic and has grown by about 6% a year over the past decade. However, growth has been uneven and affected by political disturbances in Sri Lanka. The Government also participates in bus travel through partial ownership of the bus companies and regulation of fares. 70. Sector Finances. Most of the institutions in the transportation sub-sector suffer from financial stress caused by inadequate tariffs, high operating costs and expensive investment programs. For instance, the Railway has incurred operating losses which have been financed from the General Treasury for forty years. In the 1993 budget, the Railway absorbs transfers amounting to 0.8% of GDP. The main reason for the losses are inadequate passenger tariffs, high operating costs (which consist mainly of wages and salaries), increased competition from buses, and uneconomic investment decisions. Air Lanka's finances are weak even though the airline's load factors have been high by industry standards in recent years. While covering variable costs, yields have been insufficient to cover fixed costs on several routes, especially to Europe. Overall, operating margins have been poor and cash flow has been barely sufficient to service debt and lease payments. 71. The Public Expenditure Program. The Government has been correcting gradually its past neglect of the transportation sector during the 1980s (Table II.1). While the transport sub-sector receives 20 to 30% of public investment in many developing countries, in contrast in Sri Lanka it - 30 - received only 6% of the total PIP during 1981-85. As a result of the low priority given to rehabilitation and maintenance in the transport sector, transport infrastructure deteriorated and vehicle and rail operating costs were much higher than those of its neighbors. Beginning in the PIP for 1988- 92, the transport sector began to receive more adequate funding, partly because of recognition of the need to rectify the past neglect and partly because of the assistance extended by the OECF, ADB, ODA and the Bank. The share of the transport sector in the PIP for 1992-96 is nearly 18% of the total PIP, or Rs. 47 billion. This is an increase of 77% over the PIP of 1990-94. The largest increase (425%) was registered by the ports and shipping sub-sector, followed by the railway (70%), and highways (25%). 72. With regards to roads, the public expenditure program is in general well conceived. It properly emphasizes maintenance and rehabilitation of the existing road system (Box 11.2). In addition, most new investments are in line with sectoral priorities. The proposed expressway from Colombo to the main airport at Katunayake is a special case in that it may displace a large number of people and its phasing may be too ambitious. The economic and social costs and economic feasibility of this project should be re-examined by updating the study carried out by Japan International Cooperation Administration. 73. There are several issues regarding public expenditure in rail. Over 50% of the PIP in rail, about Rs. 9 billion, is allocated to rehabilitation, mostly of tracks, locomotives, workshops and bridges. Considering the deteriorated state of the existing railways facilities, this amount appears reasonable. However, there are several questionable investments in the current PIP for Rail. The PIP calls for extension of the railway line from Matara to Kataragama (Rs. 0.8 billion) and from Batticcaloa to Pottuvil (Rs. 0.7 billion) as well. The new lines will probably not prove economically and functionally profitable. The plan to lay a broad gage in the Kelani Valley line is also questionable. The railway extensions and the broad-gaging are not supported by economic and technical feasibility analyses, and therefore, they should be suspended until such studies are carried out. 74. On the other hand, on the basis of a transport system analysis of the Colombo urban area, the proposed railway commuter electrification project merits careful consideration. A study of the Galle corridor was undertaken to examine the competitive position of intercity rail versus highway service and identified those commodities and passenger services which would be attracted to rail, if both road and rail services were priced at their economic costs. This analysis indicated that the improvements, with a capital cost estimated at about Rs. 2.2 billion, would have a benefit/cost ratio of 1.97 using a 10% opportunity cost of capital. 75. In the area of air travel, there are two questionable investments. First, The PIP for 1992-96 includes a project to develop a second international airport in Hingurakgoda, which is in the North-east near the historic town of Polonnaruwa. The project does not appear in the PIP for 1993-97 and it is not clear whether it will be built. The project would require a Rs. 950 million extension of an existing airstrip. Given current air-traffic levels, the project appears unnecessary, as the existing international airport at Katunayake is operating well below capacity. The extension would be under the control of the Secretary of Defense and Civil Aviation. The Government argues for the second airport mainly on the grounds that civilian aircraft need an alternative to the emergency airport at Madras. A closer field would allow incoming flights to carry ,less fuel against the contingency of an emergency and more cargo. The second airport should be canceled or delayed until a cost-benefit study is completed. - 31 - Road Rehabilitation and Maintenance Most roads and bridges, built many years ago, were not designed to carry the present heavy traffic volumes and the vehicles, with heavy chassis and axle loads, that have emerged following the initial liberalization of the economy in the late 1970s. With the growth of the economy, the demand for transport services increased, and during the 1980s, transport output grew at an average annual rate of 8% which placed a severe burden on already run-down road infrastructure. However, funding constraints deterred the new infrastructure investments that are needed to keep pace with rising transport demand, and adversely affected the maintenance and rehabilitation of road infrastructure. Road-related expenditures prior to 1979 amounted to less than 1% of total public expenditures. This situation led to accelerated road deterioration, pavement failures and substandard and weak bridges on the nation's road network. In response to this critical situation, the Government sought and received multilateral and bilateral assistance to embark on a major road rehabilitation effort to address the immediate needs. Expenditures for road maintenance and rehabilitation increased from Rs. 797 million in 1985 to over 3 billion in the first years of the 1990s. However, a recent study indicated that about Rs. 7 billion would be needed annually to arrest the decline of the main road network and to raise it to a maintainable state. Responsibility for rehabilitation and maintenance of 900/ of the country's 97,400 kms of road network lies with a combination of Provincial Councils (16%), local governments (53%) and specialized agencies (21%). However, these agencies are only allocated about 3% of the overall rehabilitation and maintenance budget. In addition, these departments are generally poorly organized, trained and equipped for road repair and rehabilitation tasks and require substantial amounts of technical assistance and institutional strengthening. Further aggravating this situation is that the Central Government disburses road funds to them as late as six months into the fiscal year and, thereby, prevents the formulation of a,predictable and smoothly operating road repair program. For this reason, many provincial authorities have often only been able to spend 80% of their already insufficient allocation from the Center. Box II.2 76. In air travel, the major public investment issue is Air Lanka's planned purchase of Airbuses. Air Lanka's financial standing will be damaged, with attendant consequences on the government budget and the country's balance of payments, by the recent contract with Airbus Industry for the purchase of five A340 aircraft. These aircraft are part of the fleet renewal to meet the strong market growth it projects. Its operating fleet now comprises the A320 recently put into service in replacement of a leased Boeing 737-20, four owned Lockheed Tristar LIOlls and three leased Ll0 11s. The fleet is overstretched operationally, with insufficient turn-around times between flights and too many stops on European routes while punctuality and reliability are poor. The A320s appear reasonably suited to the airline's medium-haul route system and economically competitive with other options. In contrast, the A340s are designed for long routes and are inappropriate for an airline with Air Lanka characteristics. The purchase of the A340s, valued at around $US 600 million, will substantially erode Air Lanka's net worth, invariably require government support in the form of guarantees and debt service, and impact considerably on the country's balance of payments. - 32 - 77. The Government is in the process of renegotiating the contract with Airbus Industry and examining all position options. These include reconfiguration of the existing fleet, and leasing of planes suitable for Air Lanka's route structure. In parallel, the Government is considering privatizing the airline. International experience indicates that undertaking large new investment before privatization is practically always unwise, which reinforces the above arguments against the A340s. 78. The major public investment issue in bus transport is the purchase of bus kits. The Government arranged for the import and assembly of kits for some 4,650 buses at a cost of Rs. 7.7 billion on the grounds that the privatized bus sector is financially unable to buy buses to meet a perceived shortfall. The assembly program is to last two years, but appears to be conceived as the first phase of a longer-term program. Assembly was justified in preference to the importation of finished buses on employment grounds, particularly bearing in mind the future of the 1,500 or so workers at the workshops of the former Ceylon Transportation Board (CTB). As these operations are not economically viable, they should be discontinued once present contracts have expired. Moreover, the costs of these operations should be borne directly by the budget, rather than by the state-owned banks. This would place the investment in the context of a transparent and hard budget constraint, and would keep potentially risky loans off the balance sheet of the state-own banks. Finally, the quality of investment could be best assured by fully privatizing the bus companies. - 33 - III. AGRICULTURE III.A. Overview 79. Background. Sri Lanka is richly endowed with agricultural resources. The country covers a total area of 6.5 million ha of which 2.2 million ha are under permanent cultivation and about 1.1 million ha are under shifting cultivation. Climatic factors divide Sri Lanka into three major zones: the dry zone, the intermediate zone and the wet zone, with an average annual rainfall of 50 to 75 inches, 75 to 100 inches and over 100 inches, respectively. In the dry zone, which covers two-thirds of the country's land area, rainfall is inadequate to sustain the cultivation of a seasonal crop without supplementary irrigation, and, hence, irrigation is the key factor determining agricultural intensity. In the wet zone, rainfall distribution is relatively uniform. Nearly two-thirds of the wet zone is under permanent crops, mainly tea, rubber and coconut. As a consequence of its diverse climatic and other associated features, Sri Lanka is capable of producing a wide range of crops, both tropical and temperate. They include tea, rubber, coconut, various grains, yams, legumes, vegetables, spices, condiments, fruits and oilseeds. Notwithstanding this potential diversity, (rice) paddy occupies 700,000 ha, or almost one-third, of the land under permanent cultivation, and about two-thirds of this is irrigated. The major plantation crops (tea, rubber and coconut) occupy a further 800,000 ha, or over one-third, of land under cultivation. 80. The agricultural exploitation of these natural resources has significant economic and social importance. Agriculture generates more than a quarter of GDP and employs nearly half of the labor force. Agriculture is the main income source for about 70% of the country's population who live and work in rural areas. Despite the recent very rapid expansion in textile exports, agriculture still accounted for some 25% of total exports of goods in 1992. 81. Main Policy Issues. The fundamental issue in agriculture is the sector's faltering growth performance (0.5% p.a. on average over the past 5 years). Negative growth rates in paddy and plantation crops lie behind this weak performance, despite enormous amounts of public resources directed at these sectors, especially in the past. During the period 1988-1992, value-added in paddy (25% of sectoral value-added) and plantation crops (24% of sectoral value-added) had declined by 2% p.a. and 2.9% p.a. on average, respectively. While these results are partly due to adverse weather conditions in 1991-1992, inappropriate sectoral policies are the key factor. 82. Achieving food self-sufficiency, mainly through import substitution, has been a major goal of all development plans of Sri Lanka since the 1940s. There has been a particular emphasis on self-sufficiency in rice. Currently, about 90% of the total rice consumed in Sri Lanka is produced domestically. In contrast, in the early 1950s, about two-thirds of total consumption was imported. Public policy promotes rice and food crop cultivation through import protection, subsidy of irrigation, cropping and land use restrictions, and a concentrated research and extension effort. In the past, a fertilizer subsidy also promoted rice cultivation. Rather than encouraging diversification, policy has been preoccupied with sustaining production of the main crops. 83. The drive for self-sufficiency in rice was also a key motivation for an intensive government effort to develop irrigation infrastructure, mainly in the dry zone. The most significant part of this investment was for the Accelerated Mahaweli Development Program which started in the 1970s. This massive program, when completed, will provide irrigation to about 120,000 ha. of new - 34 - lands and about 85,000 ha of existing lands in the dry zone, provide nearly half the country's electricity generation, and settle some 100,000 families. 84. Irrigation and land development, in particular the Mahaweli program, has required massive expenditures and continues to absorb the lion's share of public expenditures in agriculture (Table 111.1). During 1980-1985, public expenditure in agriculture accounted for 7-8% of GDP; public expenditure on Mahaweli alone absorbed over 7% of GDP in 1982. As major Mahaweli works were completed, public expenditures in agriculture have declined sharply. On average, expenditures in agriculture accounted for some 2% of GDP during 1990-92 and Mahaweli was budgeted to amount to less than 1% of GDP in 1993. Notwithstanding this decline, public resources directed to the sector are still significant, and irrigation and land development were budgeted to account for 64% of these resources in 1993. Without an improvement in performance of irrigated agriculture, the returns on this enormous investment will be marginal, an outcome that Sri Lanka can ill afford. Table 111.1: SECTORAL COMPOSITION OF PUBLIC EXPENDITURE IN AGRICULTURE, 1980-93 Sub-sector 1980 1985 1990 1991 1992a' 1993" (Percentage Distribution) Food Crops (%) 12 9 17 33 16 16 Irrig. & Land Devt. bl(%) 80 82 55 62 58 64 Minor Export Crops (%) - - I - 1 1 Plantations Crops"(%) 2 6 21 2 22 13 Livestock (%) 1 1 3 1 1 1 Forestry (%) 1 1 2 1 1 1 Fisheries 4 1 1 1 1 4 Total 100 100 100 100 100 100 For Memorandum: Public Expenditures in Agr./GDP 9.0 5.6 2.0 2.0 1.5 1.8 a/ Estimates. b/ Includes Mahaweli. dl Does not include credit from the State Banks. Source. World Bank Staff Estimates 85. While further analysis is needed, to enhance growth in irrigated agriculture, the Government should review its food self-sufficiency policies, and undertake fundamental reforms in the management and operations of Mahaweli. With respect to the policies, price and trade policies and land-use and cropping restrictions are believed to result in significant efficiency losses. Food crop prices and trade barriers should be liberalized to promote competition and efficient resource allocation, while land-use and cropping restrictions should be lifted to allow diversification away from rice to - 35 - higher value-added crops. With regard to Mahaweli, public resources should be directed primarily into enhanced maintenance and rehabilitation to improve operational efficiency and avoid expensive rehabilitation expenditures in the future, cost recovery should be strengthened by transferring responsibility for maintenance to farmers' organizations, and commercial activities of the Mahaweli Authority should be privatized while public support services should be handed back to national line agencies. 86. The other main government intervention in Agriculture was through the large and very inefficient nationalized plantation corporations: the Janatha Estate Development Board (JEDB) and the Sri Lanka State Plantations Corporation (SLSPC). These plantations operated about 250,000 ha of land and employed about 250,000 persons. Between 1970 and 1992, two-thirds of all tea and one third of all rubber were produced by the two state corporations. The performance of the state-owned plantations has been poor, especially when compared to the privately-owned estates. In tea, for example, output in the state sector declined by 1.2% per year between 1981 and 1991, while the private sector registered an annual increase of 12.6%. Weak management, high export taxes and government interference in marketing and labor markets largely explain these results. As a result, the two large state-owned plantations have absorbed large current transfers from the budget and excessive credit from the state-owned banks. 87. There have been major changes in government policy regarding the plantation sector. First, the Government has decided to increase the role of the private sector in the industry. In June 1992 the JEDB and SLSPC were broken up into 22 new publicly-owned plantation companies and their management was contracted out to private agents on a profit sharing basis. However, the assets of the plantations are still owned the Government and financing comes from the state-owned banks. The Government is currently negotiating long-term leases-type arrangements with the management companies that would transfer commercial risks fully to the private sector, and motivate inflows of private funds for rehabilitation and replanting in the plantations. Quick progress needs to be made on completing these discussions. Second, export taxes on tea and rubber have been largely eliminated to the benefit of both the plantations and small-holders. 88. The remainder of the chapter discusses in more detail these issues regarding Mahaweli and the plantations sector. It also reviews expenditure issues regarding Sri Lanka's research and extension system. H.B. Restructuring Mahaweli 89. Public Expenditures Issues. With the completion of most major construction and settlement activities, the Mahaweli program has reached a new stage, in which the main emphasis should be to restructure rather than to embark on new investments. This applies with particular force to the older areas of the system, which account for a substantial share of planned total expenditures on Mahaweli over 1993-97. Evidence from studies undertaken by the International Irrigation Management Institute and other research agencies in Sri Lanka indicate that further investment in developing new irrigation systems to increase the irrigable area cannot be justified on economic grounds, except in isolated cases. Instead, the efficiency of operation of Mahaweli should be improved through enhanced maintenance and rehabilitation (Box 111.1). In fact, studies show that inadequate funding of operations and maintenance is causing a premature deterioration in the infrastructure. Unless the situation is remedied, costly rehabilitation works will be required in the future. - 36 - Operations and Maintenance of Mahaweli The funds currently allocated for operations and maintenance (O&M) in the irrigation sector are estimated to cover only 40 to 60% of the actual requirement. Because of inadequate funding, the infrastructure is deteriorating prematurely. In the Mahaweli irrigation system, the Central Government currently carries out most of the maintenance. This has worked reasonably well for the headworks and main and branch canals. But downstream facilities have not been adequately maintained. Farmers often do not maintain contributory and field canals. The bulk of O&M expenditures in Mahaweli are overhead costs of building and vehicle maintenance, leaving little for true maintenance. Around 15 to 20% of the O&M funds are spent for maintenance of tertiary canals, which should be handled by the farmers who benefit. It has been shown that Sri Lankan farmer organizations established at the grass root level can effectively mobilize greater farmer participation in irrigation schemes. One of the causes of inadequate maintenance is the failure to fund maintenance by recovering the cost from farmers. In 1984, farmers in all major schemes in Sri Lanka, including those in Mahaweli, were legally obligated to pay water charges covering the O&M costs of their scheme. During 1985, collections reached 50% of the amount due, but the payments ceased almost completely in 1989, mainly because of: (i) the civil disturbances in the country, (ii) the weak legal framework for implementation of the fee collection program (which was subsequently corrected); and (iii) the low level of farm income which affected farmers' capacity to repay. The fee collection exercise was finally abandoned in 1989. Box I.1 90. In this regard, the largest potential public investment project in agriculture is in the Mahaweli system, where the PIP for 1993-1997 records plans to build the Kalu Ganga project, but does not provide any funds. The project is designed to develop water resources within Mahaweli and adjoining basins and is estimated to cost Rs. 12.0 billion over a period of five years. It aims to stabilize the cultivation of about 29,990 ha. of already irrigable lands in the Mahaweli and adjoining areas and to develop 1,300 ha. of new lands in the dry zone. The project would displace about 3,000 persons and is likely to have negative environmental impacts. The project does not seem to conform to the agreed strategy of restructuring Mahaweli by improving existing systems rather than undertaking new heavy capital investment. As mentioned above, returns on additional investments to expand the irrigation system are estimated to be low. In view of these unresolved issues, the Government's decision to postpone funding the project is appropriate. Plans for the Mahaweli System L and Udawalawe, Right Bank are other examples of massive expansions of Mahaweli that are likely to have low return. 91. Policy and Institutional Issues. In the Mahaweli area at present, paddy is the only major crop and the area under cash crops is limited, even though paddy cultivation is associated with relatively low net returns. The small size of holdings (1 ha) in Mahaweli and the incentives provided by subsidies, trade policy and research and extension contribute to the excessive concentration on paddy, as do land-use and cropping restrictions. The difficulty of obtaining large tracts of land also hampers the development of agro-enterprises, which might lead the sector in diversification to higher - 37 - value crops and livestock products. Artificial constraints to diversification should be eliminated while trade and price policy should be reviewed to eliminate efficiency losses. 92. There is evidence that irrigation water in Mahaweli areas is not used optimally, partly because of weaknesses in the physical conveyance system and partly because of other factors, such as a strong farmer preference for cultivating paddy. Water conservation is important because hydro- power, industry, and households compete with agriculture for scarce water and because seasonal droughts occasionally induce acute scarcity. Since paddy makes the heaviest demand on irrigation water, diversification is an effective means of reducing consumption of irrigation water. 93. The Mahaweli Authority has taken responsibility for the construction and management of all social and regulatory activities and other governmental services in the Mahaweli. All such services should be transferred to the appropriate national agencies. Such transfers have been under way for some time with activities such as schools, main roads, health services, and police. 94. A final issue concerning Mahaweli is its extensive investment in commercial operations which should be handled by the private sector. For instance, the Mahaweli Authority is engaged in commercial poultry production, milk processing and rice milling. The Mahaweli Authority includes the Mahaweli Engineering and Construction Agency, the Mahaweli Livestock Enterprise, the Draught Animal and Dairy Development Program, and the Mahaweli Venture Capital Company. The Authority has tied up a large amount of capital in these commercial activities and investments. These activities should be turned over to the private sector, which could operate them more efficiently. This should permit some restructuring of the Authority's work force of 9,600. 95. Although the Authority should withdraw from certain activities, it would retain several important roles. Drawing on its highly talented staff built up over many years, the Mahaweli Authority has an essential long-term role to play in the management of the Mahaweli River basin. This includes operation and maintenance of the headworks, regulation and provision of water to the various irrigation systems, townships and industry as well as for power generation. It also has a vital role to play in protecting land and water resources in the Mahaweli basin from environmental degradation. The Mahaweli Authority's responsibilities for environmental protection would extend to natural resources coordination and management of the land and water to ensure the protection in the long-term of water quality, provide water to meet environmental needs of flora and fauna and of sustainable agriculture. In order to avoid duplication and to conserve scarce resources, it should also coordinate all basin related affairs. The Authority would deal with issues of catchment management, minimum stream flows below reservoirs, salinity and drainage. IIB.C. The Plantations Sector 96. Background. The tree crops sub-sector in Sri Lanka has been under-performing for over twenty years, causing not only a loss of national income but also absorbing scarce resources from the national budget. The Government has tried unsuccessfully to restore economic health to the sub- sector. On three major indicators of performance, output growth, productivity and cost of production, the state sector in Sri Lanka has lagged behind international competitors for both tea and rubber production (Box 111.2). The small holders, however, have performed better than the international market average in tea but not in rubber. From 1985 onwards, the two state plantation corporations reported huge losses. By 1990, cumulative losses reached Rs. 3 billion and the debt ballooned to Rs. 5 billion, suggesting that the plantations required restructuring. In contrast, the private sector managed to operate quite well in spite of many restrictions, increasing both its production level and productivity. The cause of this failure are to be mainly found in the domestic operating environment: - 38 - Performance of the State-Owned Plantations Ten years ago, tea, rubber and coconut provided nearly a quarter of the national revenue and over half the total export of Sri Lanka. Since then, the revenue contribution has declined to one twentieth and exports to one quarter of the total. Sri Lanka also lost world market share which, in the case of tea, fell from 22% in 1981 to 18% in 1989, while newcomers like Kenya and China gained share. During this period, world production of tea increased by 33% while output from Sri Lanka's state plantations, which account for two-third of the national production, decreased. Production from the private smallholder sector, however, grew by almost 300%, which enabled the country to achieve modest growth in export volumes. Similarly, world rubber production increased by 25% between 1983 and 1988, whereas Sri Lanka's production declined by 26%. Sri Lanka is not a major supplier of coconut products to the international market because of its huge internal demand, although lately it has achieved higher levels of export in some products like desiccated coconut. In any case, its international market share is less than 3% and the importance of the crop is derived almost entirely as a domestic food product. Productivity and productivity growth in tea for Sri Lanka have been low compared to the major competitors. Sri Lanka has lower yields than India, Kenya, Indonesia and Malawi and the yield growth is about half that of Kenya and Indonesia, because of poor performance of the state plantations. In contrast, the private small holder sector in Sri Lanka has not only outperformed the state sector by a margin of 100% in yield per hectare, but has also outperformed major competitors. In rubber, over the past ten years private sector yields have risen steadily, whereas in the state sector yields and output have declined. The cost of production of tea in Sri Lanka is about 25% higher than South India and Kenya. Production costs of private small holders are slightly lower than in the state sector (13%) due to lower wages in the latter. In rubber, the small holder has done far better in controlling the cost of production, keeping it about 70% lower than the state sector. Box 111.2 (i) The nationalization of the tea, rubber and coconut industries in 1975 placed a large part of the tree crops sub-sector in the hands of a relatively inexperienced state bureaucracy, which failed to maintain the market orientation of the sub-sector. (ii) The State imposed direct taxes on tree crops, such as export duties, ad valorem taxes and cesses, reducing profit margins. (iii) The State intervened in the labor market by increasing wages and benefits and drove the cost of production upwards in the state sector every year, despite no offsetting increase in productivity. Labor laws such as the mandatory six day work rule for plantation workers and the restriction on moving labor from one plantation to another have established a sizable fixed labor cost component. Increasing labor costs, which today account for two-thirds of total costs, squeezed producers' margins to a point - 39 - where they became negative in many years. The difference had to be met from public borrowing, which created a heavy debt overhang by 1990. In contrast, the private small holder sector, which was not directly affected by state sector wages, was able to maintain positive margins in all but one year between 1980-1990. (iv) Productivity did not improve because of inflexible labor use, inefficient agronomic practices, and inappropriate inputs of growing stock and fertilizer. Productivity in the state tea sector remained stagnant, while in rubber, the state sector's productivity declined continuously every year between 1980 and 1990. (v) The marketing of tree crops was severely constrained by a series of outdated regulations that controlled everything from expansion of production to the method of crop disposal. And, (vi) The rubber and tea research institutions did not sufficiently improve the product or process technology to increase yield and quality, especially for rubber. 97. Rehabilitating the Plantations Sector. The first effort to rehabilitate the tree crops sector was made in the 1984 when export taxes were reduced and investment increased with the help of the World Bank and Asian Development Bank. For rehabilitation of the state sector, the Fourth Tree Crops Project, co-financed by the two banks and Netherlands, Norway and the United Kingdom, provided $211.8 million between 1986-1991. Two similar projects, the Small Holders Tea Development Project and the Small Holder Rubber Rehabilitation Project, financed by the Asian Development bank and the World Bank respectively, were directed towards the private sector in 1989 for meeting medium term investment needs. These projects, although they did address some of the investment needs of the sub-sector, did not look into the other problem areas such as labor productivity, cost of production and the restrictions on marketing. 98. As early as 1984, many senior officials in the Government believed that adequate efficiency could not be achieved under public sector management and that a segment of the public managed estates should be returned to private sector management. The Government finally acted upon this idea in June 1992, by removing 455 tea, rubber and coconut estates from the two state corporations and placing them under 22 newly created companies under the Company's Act. These companies are fully state owned and they have complete right of exploiting the estate resources under a 99-years lease from the state. The management of these 22 companies has been transferred under individual management contracts to private management firms, under a profit sharing plan. The management contracts run initially for 5 years and are renewable thereafter depending upon satisfactory performance. The expectation was that under private managers, the plantations would be run more efficiently and private investment would be attracted to fill the liquidity gap. However, while it appears that during the past year the private sector has introduced better management practices, doubts still remain whether private investment will be forthcoming. 99. A major concern facing the private sector is that past and present ownership arrangements have not provided an incentive for investment in replanting. There was no incentive to invest when plantation owners expected nationalization. After nationalization there was no profit motive for investment, while wage and other recurrent costs crowded out investment. The current private sector management contracts will also probably result in suboptimal investment levels. Replanting tea and rubber trees requires large investments and at the same time involves a gestation period of five years or more, so private management companies operating on five-year contracts are - 40 - unlikely to replant adequately. In the plantations, hence, an important step to encourage private investment would be to introduce long-term lease arrangements. 100. Recent government efforts to negotiate arrangements of this type are a very positive step. The Government should continue to negotiate the establishment of management contracts (in place of the present 5-year contracts) that would be renewable for a further 30 year period. For the contracts to be viable, the private management companies should have virtual ownership of the land. Furthermore, the government plantation companies should subordinate their right over all immovable and movable assets to the private management companies, and should also transfer the management of all labor employed in the plantations. Such contracts would provide the basis for raising term loans and equity finance from the market. The availability of financing from the market is unlikely to be a problem so long as the management companies: (i) can display strong balance sheets with substantial equity in the enterprise; (ii) the profitability of the projects are good; and (iii) the company's strategy to cope with labor costs, productivity and diversification is sound. The successful negotiation of such extended contracts should enhance investment in replanting, in planting higher yielding vegetatively propagated tea and in upgrading factories to produce highly demanded teas. 101. Another concern facing the private sector is the perceived risk of government intervention in the labor market and the possible consequence for productivity and profitability. The 30% mandatory wage increase imposed in early 1993 and the continued insistence on the six day work rule have worsened these perceptions. The Government should allow labor and management to resolve their differences through collective bargaining without interference from outside forces; including itself. Ways need to be found to modify the guaranteed six-day workweek and the restrictions on transferring labor from deficit to surplus estates, and to design work programs with a strong focus on productivity enhancement. 102. The Government can also contribute to the viability of the plantations by removing remaining regulatory constraints on the marketing of tea and abolishing remaining discriminatory taxes. The notional quota of 10% direct sales of bulk tea outside the auctions should be lifted to provide freedom to the producers to seek the best prices wherever they can be obtained. With regard to discriminatory taxes, the Government has already taken an important step forward by abolishing all trade taxes, including the export duty and ad-valorem sales tax on coconut, in 1992. The Government should now revise the cess-subsidy scheme to eliminate all that do not provide common benefits to every tree crop producer. In particular, it should replace subsidies to smallholder with market finance for development. M.D. Research and Extension 103. Sri Lanka has the advantage of having a well established national agricultural research system, with the essential infrastructure already in place. Agricultural research is mainly the responsibility of the Ministry of Agricultural Development and Research and the Ministry of Plantation Industries. Of the 5,000 persons employed in the national research system in 1989, 472 were scientists 1,664 support staff and 2,804 were laborers. 104. In the 1980s, real expenditures declined and by 1991 accounted for less than 0.5% of agricultural GDP while expenditures on extension amounted to about 1%. Funding for research appears insufficient (Table 111.2). Moreover, expenditures on research are low relative to other countries (Table 111.3). Specifically, there is a major shortage of recurrent expenditures to maintain the staff, infrastructure, vehicles and equipment. In the national research system, salaries, wages and benefits alone account for about two-thirds of the research budget. - 41 - Table 111.2: CROP AGRICULTURE RESEARCH BUDGET BY MINISTRY IN SRI LANKA, 1989 % of Ministry (Rs. Million) Total Ag. Research Agriculture 100.7 51.3 Plantation Industries 83.2 42.4 Lands & Mahaweli 12.2 6.3 Total 196.1 100.0 Source Agriculture in Sri Lanka: Programs and Resources, G W Gysbers et al, Council for Agriculture Research Policy Table 111.3: COMPARISON OF EXPENDITURES ON RESEARCH AND EXTENSION IN SRI LANKA WITH OTHER COUNTRIES Percent of Value of Agricultural Product Research Extension Low-Income .50 .44 Mid-Income .81 .92 Semi Industrialized .73 .54 Industrialized 1.50 .63 Sri Lanka (1985-90) .30 .90 Source. Robert E Evanson The Economics of Extension in Investing in Rural Extension Strategies and Goals, ed Gwyn Jones, 1986. 105. Inadequate funding has damaged the quantity and quality of research staff, reducing efficiency of the research system. The shortage of experienced research staff is particularly severe in crop research stations located in remote areas. For lack of facilities, scientists are unwilling to serve in these stations. Low remuneration is also a major problem for attracting and retaining experienced research personnel. In the long-run, an increased level of funding needs to be combined with efficient management and adequate staff at the senior levels. 106. Another issue in research is that the Department of Agriculture has historically concentrated its research efforts on paddy as part of its attempt to achieve self-sufficiency. Sri Lanka has already gained substantial returns from paddy research in the past three decades and the level of technology adopted by farmers is relatively high. Thus, it is likely that the marginal returns for research in rice will be low. Currently, however, resource allocation has improved through - 42 - theintroduction of the program budgeting system under the Agricultural Research Project (Credit 1776- CE) supported by IDA. 107. In extension, the main issues are the number and quality of extension workers and the role of the Provincial Councils. So increasing the available funds alone would not provide the answers to the problems in extension. The Second Agricultural Extension Project addresses the personnel issue by relying on fewer but better qualified extension agents. These agents should be supported by a mass-media effort. To save fiscal resources, the private sector should be encouraged to provide extension service. Agricultural extension came to a near stand-still after 1987 because of the devolution of responsibilities to the Provincial Councils. The devolution has resulted in the transfer of about 2,480 village level extension workers to the administration of Provincial Councils. The delinking of extension by village level extension worker from the center has resulted in a sharp discontinuity in the extension system, introducing a major weakness in the extension-research linkage. - 43 - IV. THE SOCIAL SECTOR 108. This section reviews government expenditures which aim to improve human welfare, that is, expenditures in the social sector defined as the social transfer and poverty-alleviation programs, education and health. Sri Lanka has established an excellent record in social development, as evidenced by indicators that compare favorably with much higher income-level economies, especially in terms of life expectancy, infant mortality, school enrollment and adult literacy. It is impressive that Sri Lanka has targeted women as well as men: currently boys and girls are enrolled in secondary education at about the same rate. This effort appears to have paid off in the form of a decline in fertility and an accelerated demographic transition. These remarkable achievements reflect the high priority and long-standing commitment of public expenditures to these sectors. Sri Lanka's experience in the social sector is evidence that a developing country can afford expenditures for basic health care and education before it becomes rich--and can achieve a high degree of success.' 109. However, there are some areas of insufficient progress or slippage. For instance, several large surveys show that the nutritional status of young children is poor. The studies found a high incidence of stunting (low height per age), which reflects chronic malnutrition of children, and low birth weight, which reflects malnutrition of mothers while pregnant. Human welfare may well have deteriorated in the zones most heavily effected by the civil conflict, although firm data is not available. Weak income growth, especially in agriculture, raises concerns about the sustainability of the poverty alleviation effort. Moreover, the limited availability of resources and policy constraints threaten to undermine the achievements to date. 110. The tension between resource availability and the achievement of desirable goals is perhaps nowhere more evident than in the social transfer and poverty alleviation programs. These are henceforth called social transfer program because of their emphasis on transfers. The programs in Sri Lanka are impressive for a country at a relatively low level of per capita income. They demonstrate the existence of a tradition of concern with helping the poor. In the present time of fiscal stringency, however, even programs of this nature have to be carefully scrutinized, since they may have a high opportunity cost in terms of foregone public goods and services. The main social transfer programs together amounted to about 2.5% of GDP in 1992 and are described, in brief, in Box IV. 1. This section evaluates each of these programs, starting with the largest: Food Stamps. IV.A. The Food Stamp Program 111. Sri Lanka introduced a food stamp scheme in September 1979 to replace its three- decade old universal food ration scheme for subsidizing food. The intent was to target benefits to the poor, and, as a result, the number of recipients was reduced by about one half. The change formed an important part of the package of reforms for economic liberalization initiated during the late 1970s. By 1992, there had been two major efforts to improve the screening of benefits, to add newly-eligible households and to remove ineligible households. However, political constraints limited the effectiveness of these efforts. Also, even though the nominal benefits provided by FSP were adjusted in 1988, the real value of the transfer eroded substantially between 1979 and 1992. The result is a poorly-targeted program that does not serve as an adequate safety-net for the truly impoverished. 2' See Amartya Sen. "The Economics of Life and Death." Scientific American May 1993. - 44 - Social Transfer Programs The Poor Relief Food Stamp Program (FSP: 0.7% of GDP in 1992). The program was introduced in 1979 to replace an earlier Food Ration Scheme. It is intended to improve the nutritional status of poor households by increasing their disposable income with monthly coupons that can be exchanged for food in special cooperative stores. The Janasaviva Program (JSP: 0.7% of GDP). The program was introduced in 1989 to provide poor households with fixed monthly cash grants for two years, in the expectation that some recipients would exit poverty by setting up small businesses, training, or improving their fields. Once a household enters the JSP, it loses FSP benefits, except for households receiving Public Assistance. Monthly grants under the JSP are Rs. 1,458: a cash 'consumption' component of Rs. 1,000 and a 'savings' component of Rs. 458 which is only available to the households at the end of the two-year period and for investment projects. Moreover, JSP beneficiary households were originally supposed to get a grant of Rs. 25,000 at the end of the two-year period. This was not fiscally feasible, so instead, the Government is paying 'interest' of Rs. 250 per month for an indefinite period, to households that have completed the two-year period (unless a member of the household gets a government job, or a job in the government-sponsored 200 garment factory program). The JSP is being executed in successive, geographically-defined rounds. A total of eleven rounds are planned; four have already been launched. The Mid-Day Meal Program (MDMP: 0.4% of GDP). This program was introduced in 1989 together with the JSP. It covers all children attending primary and secondary school and is not targeted to the poor. Its objectives are: (i) to improve nutrition among children, and (ii) to enhance education by increasing student attendance and alertness. The program was initially intended to involve on-site feeding of children. However, in May, 1990 the program was changed to provide a subsidy of Rs. 3 per day per child to the families of students bringing a meal to school which is nutritious and not made from wheat flour. The subsidy is given in the form of monthly coupons that can be redeemed in special cooperative stores--like the FSP coupons. The teachers are responsible for assuring that students who do not attend school or who do not bring a meal with the required characteristics lose a corresponding amount of their monthly entitlement. Transfers to the Truly Needy (less than 0.1% of GDP). The programs were operated by the Department of Social Services and the Department of Child Care and Probation of the central government until 1989. Then they were mostly devolved to the Provincial Councils. The largest of these programs is Public Assistance payments to households,with incomes of less than Rs. 300 per month, where the heads of household are disabled, elderly and orphans. Assistance to Displaced Population (0.6% of GDP). This is assistance to families and individuals who were displaced from their homes by the civil conflict in the North and East. The government operates 463 camps for displaced persons, and provides dry food rations to the population of the camps and many living outside the camps. At present, there are 615,000 registered displaced persons inside and outside the camps. In addition, displaced families returning to their original place of residence are eligible for a range of benefits under the Unified Scheme of Assistance, introduced in 1983. Box IV.1 - 45 - 112. The role of FSP has been changing as large numbers of households moved from that program into JSP. Once JSP has begun in a given Assistant Government Agent (AGA) division, the only households remaining on FSP rolls should be the truly destitute households receiving Public Assistance. In this sense, FSP is increasingly serving as a narrowly-targeted transfer that reinforces the Public Assistance safety net. Nevertheless, targeting problems persist in areas where JSP has not yet arrived. 113. Past and Current Program Operation. At its inception, the food stamp program provided benefits to 1.55 million households. During the first year, the number of beneficiaries steadily increased and a freeze on new entries was put in effect from March 1980. The number of recipients subsequently tended to decrease, so that there were 1.48 million households at the start of 1986. By that time, several major problems had become apparent. There were sizeable leakages of program benefits to the non-poor, and yet there were inadequate opportunities for the newly-poor to enter the program. In addition, inflation had cut the real value of food stamps by more than half. In response to these factors, the government launched a restructuring program in 1986 with the aim of reducing the number of beneficiaries by almost one third, adding newly-poor households, and increasing the value of the benefits from the savings. 114. The new program had several important features. First, the program was moved to the Department of Social Services and renamed the Poor Relief Food Stamp Program. By housing the program in this department and by explicitly naming it "poor relief," the government hoped that a certain amount of self-screening would take place. Second, the FSP implemented a new strategy for screening that allowed the community to participate in the selection process. A committee of government officials and village residents reviewed applications and posted lists of selected households throughout the village. The public was then asked to assist in identifying any households that were erroneously included or excluded from these rolls. This re-targeting was on the verge of being completed when a Government intervention led to its virtual abandonment and the adoption of less stringent procedures. As a result, a large number of households were added to the rolls, but few were removed, and participation rose to 1.84 million households. 115. Another targeting exercise was undertaken in 1991. The 1991 re-screening was essentially based on the JSP screening process. One important element of the JSP approach is that it focuses on non-income criteria, such as productive assets owned, sources of income, and durable goods consumption. In addition, the JSP approach relies even more heavily than the 1986 FSP on community-based screening. The survey that was completed by the end of 1991 did not include the North and the East. It identified 901,877 families in the poorest category, of which 132,373 were not receiving food stamps. A total of 634,258 households were found to be ineligible (or 45% of total recipients prior to screening). If a similar pattern were to exist in the North and East, an island-wide survey might have reduced the number of eligible households to about 1.13 million. However, as a result of political pressures and because of a drought during this period, the newly eligible households were added to the FSP, but the ineligible households were not removed. 116. Eligibility for FSP is based on household income, assets, and consumption, and therefore is not adjusted for family size or composition. However, on the benefit side there is a partial adjustment insofar as the number of household members eligible to receive food stamps varies according to household income level as follows: - 46 - Table IV. 1: ELIGIBILITY CRITERIA FOR THE FOOD STAMPS PROGRAM Monthly Income Maximum Number of Household Eligible Rs. 300 or less All members. Rs. 301 - 40 4 members Rs. 401 - 600 3 members Rs. 601 - 700 2 members Source. Poor Relief Commission. In addition, the value of food stamps is adjusted by age in order to target benefits more effectively at households with young children. Following the doubling of benefits in 1988, the monthly stamp values are currently: (i) Rs. 50 for children below 8 years; (ii) Rs. 40 for children between 8 and 12 years; and (iii) Rs. 30 for individuals above 12 years. 117. FSP Budget. Table IV.2 presents recent spending on FSP, along with numbers of participating households and projections for spending and participation. The average monthly benefit per household between 1989 and 1992 was approximately Rs. 167. Total administrative costs of the program, including printing, distribution, salaries, and overhead, have been less than 1% of the total value of benefits. In forming the projections, it is assumed that both the timing of JSP rounds and the efficiency of its screening remain fairly constant, while the size of subsequent rounds declines moderately (excluding the North and East). It is also assumed that beneficiaries in the North and East are transferred from FSP to JSP starting in 1996. Finally, the projections are computed both with and without a cost of living adjustment. 118. Table IV.2 shows that the FSP budget declines steadily, even if one indexes benefits Table IV.2: BUDGETARY PROJECTIONS FOR POOR RELIEF FOOD STAMP PROGRAM Actuals Projections YEAR 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 A. Historical Budget (Rs. Mn.) 3634 3812 2999 3047 B. Projected Budget (Rs. Mn.) 2721 2490 2270 1962 1640 1318 997 675 C. Projected Budget with Inflation Adjustment (Rs. Mn.) 2721 2689 2648 2472 2232 1937 1581 1156 D. FSP with Inflation Adjustment as a Percentage of GDP 1.4 1.2 0.8 0.7 0.5 0.5 0.4 0.4 0.3 0.2 0.2 0.1 E. Number of Households ('000) 1899 1733 1502 1478 1358 1243 1133 979 819 658 497 337 Source - World Bank - 47 - by the cost of living. This occurs as a result of: (i) households being moved from FSP to JSP; (ii) ineligible households being removed from FSP rolls during the JSP screening process; and (iii) households in the North and East leaving FSP. The overall reduction is particularly pronounced if one expresses FSP as a fraction of aggregate GDP. 119. Targeting Efficiency. FSP is intended to be a safety-net program for the poor. However, in order to assess targeting efficiency, it is first necessary to define the target population. The official target population currently consists of those households below Rs. 700 of income. However, the 1990/91 Household Income and Expenditure Survey shows that income cut-off for the bottom 30% of the households is closer to Rs. 1,700. Similarly, note that at the program's inception in 1979, the eligibility ceiling of Rs. 300 for a family of five corresponded to a poverty line drawn at about 60% of nutritional adequacy (assuming that 75% of expenditures go for food). Adjusting for inflation, this corresponds to a poverty line in 1992 of approximately Rs. 1,700."o 120. While the official eligibility criteria fall well below the current poverty line, it is also true that the published statistics on program participation reflect a substantial under-reporting of income. Reports from the 1991 screening exercise, based on asset ownership, income sources, and consumption patterns show that 41% of the beneficiary households to be above the official threshold of Rs. 700, while only 15% of the households were above the Rs. 1,500 threshold. At the same time, however, these data also show the under Rs. 700 category to include 901,877 households. Using the Household Income and Expenditure Survey data, puts the cut-off for the bottom 850,000 households at approximately Rs. 1,700. Given that the total number of beneficiaries in 1991 was 1.3 million households (excluding North and East), it can be concluded that FSP leakage was about 35%. 121. Indeed, this is consistent with earlier calculations. For instance, data from the Central Bank's Consumer Finance and Expenditure Survey 1986/87 (which was conducted after the commencement of the FSP) have indicated that nearly one-third of the spending units in the bottom 20% of the income range had not received food stamps while nearly 40% of the spending units in the third and the fourth quintiles did receive food stamps." Thus, from both perspectives it is clear that targeting based on self-declared incomes and the past administrative arrangements for the selection of beneficiaries have not been able to assure proper targeting of the costly transfers. 122. Estimates of regional distribution of poverty are not available to evaluate regional equity in the supply of food stamps. Figures on the percentage of population receiving food stamps in the districts in 1989 do not indicate a regional bias in the food stamps distribution. All the districts in the North and East have participation rates above 60% (except Jaffna, for which data are not available, and Trincomalee, which had 49% participation). Matara and Hambantota in the South also had similar levels of coverage. Since 1989, however, there has been a steady decline in the number of beneficiaries in the North and the East--from 380,000 in 1989 to about 275,000 in 1992. Reasons for this decline could be a combination of a number of factors including out-migration, receipt of disaster relief and a break-down in the district administration. LO' Based on the Colombo Consumer's Food Price Index. Alternatively, if one uses 1992 rice prices to approximate the average cost of calories for the poor, the Rs. 1,700 poverty line would correspond to a poverty line drawn at 80% of nutritional adequacy. See, Anila Dias Bandaranaike, "Aspects of Food Intake and Nutritional Levels of the Poor" (mimeograph, June 1989). - 48 - 123. Efficiency: Incentives and Disincentives. The question of whether the food stamp scheme has a poverty trap associated with reduced hours worked needs to be examined in the light of the size of the transfers relative to beneficiary incomes and the effective benefit reduction rate. In 1981/82, food stamps income accounted for around 17% of the average total income of the beneficiary households in the lowest quintile. The value of the food stamps has been doubled since then, but the poverty line has risen nearly five-fold. Viewed from another perspective, the present average monthly food stamps benefit of approximately Rs. 170 is equal to less than three days wages. Combined with the fact that benefits have rarely been removed for households experiencing increases in their incomes, it seems highly unlikely that food stamps discourage wage earnings by the recipient households. 124. Program Effectiveness. FSP is intended to increase nutritional welfare of poor households by increasing disposable income of the households. The value of food stamps to these households was estimated at Rs. 36 per capita, twice the amount received in 1981/82. A study by Edirisinghe estimated the marginal propensity to consume out of additional income from food stamps to be 0.63 in the lowest quintile, so that the additional calorie consumption from food stamps for a household in the lowest quintile can be estimated to be 106 calories per capita per day in 1992." Relative to 1981/82, the marginal addition of calories from food stamps in 1992 shows a decline by about 40% in spite of the doubling of food stamp values. In the absence of household expenditure data, changes in household calorie consumption including substitutions due to inflation cannot be ascertained. The food price index increased by a factor of 3.25 during the last ten years. Some substitution in calorie consumption is likely to occur with price increases. Allowing for this, average food stamp transfers have to increase at least by a factor of 3 relative to 1981/82 to compensate for the erosion in their real value. In other words, the transfer has to be at least Rs. 250 per month to a family of five. The role of food stamps receives greater significance from this possible scenario of a decline in the real incomes of the poor households. There is an urgent need to eliminate leakages of transfers to upper income households and strengthening the programs meant for the real needy. 125. Children are given higher valued food stamps with the expectation that they will derive greater benefits from these transfers. The evidence is that all transfers are viewed as additional income to the household and that benefits to the children are guided by the intra-household allocation practices of the households which tend to favor adults when distributing relatively low calorie supplies. Additional costs of giving age-specific food stamps can be eliminated without harming the intended benefits. 126. Finally, note that a generalized income transfer as in the case of public assistance payments (as opposed to food stamps) would minimize the transaction costs to the recipients and increase the utility from the transfers. In particular, FSP has faced persistent problems associated with short-weighing at the village Cooperatives. Some of the advantages of FSP with regard to childhood nutrition may be preserved by making the transfers an entitlement of mothers. IV.B. Janasaviya Program 127. The JSP or "people's strength" program provides households with monthly payments that are intended to free households from the urgency of obtaining daily sustenance, thereby enabling them to engage in training, private investment projects, and community infrastructure development. L2 Neville Edirisinghe, The Food Stamps Scheme in Sri Lanka: Costs, Benefits and Ontions for Modification ( IFPRI Research Report No.58 March 1987). - 49 - The program also includes savings and investment components that are designed to assist households in implementing these plans. The intent is to make JSP a true poverty alleviation program, rather than a cash transfer program to maintain the poor at a subsistence level. 128. Despite its laudable goals, there are a number of important problems with JSP. There has been virtually no planning effort undertaken to identify enterprises or strategies that households might pursue. At the local level, resources have often not been mobilized to enhance the program. The JSP has, however, funded an improvement in the housing stock of the poor. Finally, JSP may have succeeded in bringing many households into the formal banking sector, both as loan recipients and as savers, while also enhancing the formation of informal community-level banks. The following subsections present the operational details of JSP, our budgetary projections for the project, and an appraisal of the program's impacts. 129. Past and Current Operation of JSP. During implementation of the first three rounds, JSP underwent important operational changes. To qualify for JSP, households officially must have less than Rs. 700 of income per month from all sources and must have at least one household member able to contribute 20 days per month toward a mix of public infrastructure development and private training and investment projects. During the 24 month program, beneficiaries receive a "consumption" transfer of Rs. 1000 per month (which must be spent in the state-run cooperative stores) and a "savings" component of Rs. 458 per month (deposited in a savings account at the Cooperative Rural Banks (CRB)). The savings component is only released to households at the end of the program--and then only to be used for investment projects. As of Round 3, these savings can also serve as collateral for bank loans received during the program. A final component of the program is that graduates receive a certificate nominally entitling them to Rs. 25,000 (the accumulation of 24 monthly payments of Rs. 1,042). Due to budgetary pressures, however, only the interest on this grant is currently being paid (i.e., Rs. 250 per month)." Neither the eligibility criteria nor the benefit levels are adjusted for family size. 130. In return, beneficiaries must provide 20 days of labor per month, some of which is spent on training and private projects, and some of which is spent on community infrastructure projects. These efforts were poorly supervised prior to 1991. Since 1992, a large portion of JSP's Rs. 50 million in administrative funds have been transferred to the divisions, enabling payments of Rs. 750 to 1,000 per month to "social change agents" who are assigned at the rate of approximately one per 100 households to assist community mobilization and project implementation. In addition, these administrative funds may help finance other divisional resources, such as extension agents and road- making equipment and personnel, that have been increasingly allocated toward JSP projects to enhance the environment for sustainable development. 131. JSP Budget. JSP is being implemented in eleven rounds composed of AGA divisions selected on the basis of need (as measured by the proportion of FSP beneficiaries in the population). Round I began in October 1989 and included 161,000 households (of which 42,000 in the North and East were removed in mid-1990). Round 2 began in December 1990, but was implemented in "phases," with some of the 104,000 Round 2 households entering the program as late as August 1991. .L3 The Government thereby achieves a considerable discounted savings, because at 12%, the nominal interest rate translates into a real return of negative seven percent. Nevertheless, discounting at a real anpual rate of 5%, the present value of this "annuity" over the next ten years is still Rs. 11,045 (assuming the principle is never paid) or Rs. 13,954 (if the principle is paid). - 50 - As of Round 3, the phase-in system was largely removed, so that most of the 101,000 households in this round started together in February 1992. Round 4 started in March, 1993 and is likely to cover another 100,000 households. 132. Table IV.3 presents budgetary projections for JSP from base-year data for 1991. Three critical elements of these projections are: (i) the timing of subsequent rounds; (ii) the numbers of households in subsequent rounds; and (iii) the timing for extension of benefits to households in the North and East. The tables assumes that the rounds are implemented on an annual basis, and that the number of households in each round declines slightly (due to the greater prosperity of the remaining AGA's). With regard to the North and East, once stability is restored in these areas, there may be a rapid increase in the number of JSP beneficiaries. The budgetary projections assume that rapid implementation begins in 1996. Table IV.3: BUDGETARY PROJECTIONS FOR JANASAVIYA YEAR 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 A. Consumption and Savings Components (Rs. Mn.) Excluding North & East 521 2082 3229 3439 3377 3143 3135 3171 3149 3149 3149 3149 North & East 184 735 0 0 0 0 0 1338 1968 2406 2449 2449 B. Interest Payments on Certificates (Rs. Mn.) 0 0 89 357 643 947 1222 1517 1886 2291 2763 3243 C. Administration (Rs. Mn.) 10 50 50 50 50 50 50 50 50 50 50 50 D. Total Budget (Rs. Mn.) 714 2867 3369 3846 4070 4140 4407 6077 7053 7895 8412 8892 E. JSP as Percentage of GDP 0.3 0.9 0.9 0.9 0.8 0.7 0.7 0.9 0.9 0.9 0.9 0.8 F. JSP + FSP as Percentage ofGDP 1.7 2.1 1.7 1.6 1.4 1.2 1.1 1.2 1.2 1.2 1.1 1.0 Source. World Bank 133. Off-budget JSP Costs. There are three main sources of off-budget costs associated with JSP. The first consists of resources diverted from other line ministries to assist JSP-related projects. For instance, community road-building projects initiated under the community infrastructure component of JSP are frequently augmented by skilled labor, materials, and equipment that are diverted from other road-building projects. No accounting for such inputs has been made; however, our estimates indicate that the total for all community projects is less than Rs. 200 million per year-- - 51 - some of which may be funded out of JSP's administrative budget." The amounts funded by other ministries may be viewed as off-budget expenditures for JSP--or one might view the labor supplied by JSP beneficiaries as an off-budget contribution to these other ministries. In any event, it is likely that divisional infrastructure spending is being allocated more equitably, but perhaps also less efficiently, than would be the case in the absence of JSP. 134. The second off-budget component of JSP involves the provision of subsidized credit by the banking sector. Late in the first round, Government pressure to extend loans resulted in what banking officials humorously term a loan "mela" (or festival). Overall, Rs. 624.9 million in loans were made to Round I beneficiaries, with Rs. 251 million in credit extended by the Bank of Ceylon alone. There was evidently a belief that these loans would be collateralized, either by the Rs. 458 savings component or by the Rs. 25,000 investment component; however, this was not the case in Round 1. As a result, loan default rates exceeded 33%, according to Bank of Ceylon sources, and associated banking sector losses may have been as high as Rs. 290 million.'s To help alleviate this problem, and to confront households with more appropriate incentives, the CRB was allowed to use the Rs. 458 savings component as collateral as of Round 2. However, other banks were not similarly collateralized, with a result that other sources of credit declined substantially. Total lending to Round 2 households as of October 31, 1992 was Rs. 322.5 million, of which Rs. 188.3 were extended by the CRB. As of Round 3, all banking loans are collateralized by the Rs. 458 savings component. 135. The final off-budget component of JSP involves the subsidized purchases of output from household projects by Cooperatives and other government institutions. The magnitude of such subsidies is probably quite small. However, there are numerous examples of investments that were successful while subsidized outlets were available, but which failed once protection was removed, suggesting that household losses associated with these temporary distortions may be quite important. 136. Targeting Efficiency. From a regional perspective, the selection of AGA's for participation in JSP has been conducted on an equitable basis, with benefits being provided to the poorest divisions first (with poverty measured by participation rates in FSP). From a distributional perspective, however, JSP was initially targeted only slightly more efficiently than FSP. The L"/ An upper bound on average monthly labor contributions to community projects is 4 days per household, with a market value of Rs. 260 (at the current wage rate). Thus, the total annual unskilled labor component of these projects is less than Rs. 625 million (assuming 200,000 participating households on an annual basis). Unskilled labor components of low-tech projects such as these typically exceed 75% of total costs. Thus, the imputed upper bound on skilled labor, material, and equipment contributions is approximately Rs. 200 million. This upper bound estimate greatly exceeds the Janasaviya Commission estimate that community projects totalled Rs. 275 million in value as of October 1992--which would suggest that spending on skilled labor, materials, and equipment was closer to Rs. 70 million over a three-year period. L"/ However, the widespread practice of delaying loan disbursement while requiring interest payments on the entire loan amount may have reduced these losses somewhat. For our estimate, we have assumed: (i) defaulted loans generate no revenue for the banks; (ii) the opportunity cost of banking funds is 16%; (iii) the interest rate charged on loans is 20%; and (iv) all loans have three-year maturities. (Bank of Ceylon officials indicated 16% was the rate at which low-cost, default-free loans could be made during this period.) Finally, we have assumed that all banks faced administrative costs of Rs. 243 per loan (a Bank of Ceylon estimate). - 52 - Women's Bureau conducted a random survey of 299 Round I households shortly after they had completed the 24 months of JSP. Nearly 30% of these JSP participants had household incomes above Rs. 1,500 per month prior to entering JSP.`6 137. Subsequent rounds, however, have made increasing use of community participation to distinguish between eligible and ineligible households. The official eligibility threshold of Rs. 700 has become less of an issue than a set of detailed, region-specific criteria issued by the Janasaviya Commission specifying thresholds for ownership of land and consumer durables and for sources of income. Village-level committees, largely composed of community leaders, prepare eligibility lists that are publicized throughout each village, with households ranked in descending order of poverty according to these criteria. Anonymous complaints are then gathered and evaluated by the selection committee. This may create a moderate amount of social tension, yet it appears to be a method with widespread acceptance in Sri Lanka--and it appears to yield substantial improvements in targeting. Moreover, the criteria used have the advantage of being fixed in real terms, avoiding the longstanding confusion over enforcement of the Rs. 700 per month guideline. 138. While JSP screening has improved upon the targeting of FSP, there are still leakages of about 10 to 20%. Also, 5 to 10% of eligible households may have been excluded, and perhaps 10% of beneficiaries are elderly or have chronic disabilities, limiting their prospects for sustainably exiting poverty. These magnitudes were confirmed during a Rapid Rural Assessment (RRA) of JSP, conducted in two villages during February 1993, whereby 5 pairs of researchers held informal conversations with both JSP and non-JSP households. 139. The implementation of JSP has resulted in a reduction of 473,000 households from FSP by mid-1992, over 149,000 of which were found ineligible for JSP. Because of this linkage between the two programs, it is important to assess the combined budgetary implications of JSP and FSP. The bottom row of Table IV.3 demonstrates that the combined total is a steadily declining fraction of GDP. 140. Impacts of JSP: Income. The Rural Rapid Assessment (RRA) uncovered modest income gains and highlighted the constraints on income generation. JSP operates primarily in rural areas, so agricultural projects constitute the primary avenue for advancement. Given widespread underemployment, increases in capital intensity generally appear unprofitable. Thus, the major ways to increase income are to introduce more valuable crops for labor-intensive cultivation and make better use of existing output. For instance, many households increased tea and fruit cultivation, and many others increased the marketed output of betel leaves, cashews, and pan reeds (for weaving into mats). The RRA also observed households expanding cropped acreage. However, this often involves squatting on government forest lands, raising issues of deforestation. 141. Although the consumption component of JSP has been able to free household labor for self-employment in productive ventures, success typically also requires asset ownership (e.g., land) and is therefore not directly reaching the poorest JSP households. Only a few households managed to successftilly implement nonagricultural projects, such as marketing fish, producing reed mats or coir mattresses, and preparing and marketing baked products. There is also a serious lack of guidance concerning what might be successful ventures. There simply has been no formulation, at any level of government, of strategies households might follow to increase their incomes. .6/ Women's Bureau of Sri Lanka, "Role of Women in Janasaviya" (December, 1991, mimeograph). - 53 - 142. Impacts of JSP: Housing. The record is far brighter on housing investment. In many cases, beneficiaries have added concrete floors, plastered walls, and tiled roofs as a result of JSP. RRA results suggest that at least half of the households have made substantial improvements--many spending over Rs. 5,000 on materials and skilled labor. A study by the Women's Bureau found about one third of the households have made substantial improvement, with average market value of Rs. 4,500. Of course, housing is not a productive investment insofar as it does not increase the productivity of labor. Nevertheless, JSP's broader objective is to sustainably increase the living standards of the poor, and the flow of housing benefits from these improvements will be an enduring legacy of JSP. 143. Impacts of JSP: Credit and Savings. JSP has had important impacts on credit markets. As of October 1989, the Rs. 458 per month component of JSP had led to new savings in formal sector banks totalling Rs. 2.3 billion (or roughly Rs. 7,000 per household). In many cases, these savings accounts are the households' first encounter with the formal banking sector. 144. This has not always been a smoothly functioning aspect of JSP. The Marga Institute report indicates that banks are often not prepared to provide services to the poor in such great numbers.'7 Notwithstanding some improvements, banking remains a full day activity for many households due to long lines and the difficulty of travelling to the bank offices. Another issue is that savings balances earn only 14%, which is lower than existing rates of inflation. Despite these impediments, however, the generation of savings has proven to be one area in which JSP has achieved a remarkable social transformation. 145. There has also been a substantial increase in credit extended to the poor. Evidently, a large portion of this lending resulted from government pressure on the commercial banks to extend credit. Many of the loans were to finance housing improvements, and nearly 33% of the Round 1 loans will not be paid back. However, since Round 3 the Rs. 458 savings component has served as collateral for loans. This should: (i) decrease the default rate; and (ii) improve the quality of the loans (at least insofar as households may be more judicious concerning their borrowing). 146. A final impact on credit markets involves the widespread formation of informal community banks. Savings in these institutions as of October 1992 totalled approximately Rs. 60 million and were growing rapidly. As demonstrated by the Grameen banks in Bangladesh, community credit institutions can play an important role in grassroots development. However, the formation of these banks in Sri Lanka is occurring in an institutional vacuum. Little or no guidance is being offered, either by the government or the formal sector banks, with the unfortunate consequence that easily-avoided errors are being made with excessive frequency. For instance, accounts are rarely kept, and in many cases community bank officers are simply absconding with the money. In- virtually none of the cases are these informal banks holding their cash reserves in formal sector savings accounts, so that their idle capital is eroding at the rate of inflation. 147. Impacts of JSP: Labor Supply Effects. JSP has undoubtedly caused at least a temporary decline in the supply of labor for wage employment as individuals have implemented their own projects. This was clearly confirmed during the RRA. Indeed, households who do not engage in such projects risk losing their JSP benefits. The longer-term effects of such declines depend largely on the successes of households' income-generating projects. The Women's Bureau study finds that the 7/ Marga Institute, "A Framework for a Comprehensive Social Welfare Policy" (December 1991, mimeograph). - 54 - importance of casual employment as a primary source of income is the same before and after JSP. This was confirmed by the RRA, insofar as many income-generating projects are viewed as requiring only part-time attention. The impact of JSP on rural wages has not been rigorously evaluated; however, we do not expect reductions in labor supply to have serious impacts within JSP AGA's, because of the availability of excess labor supply in neighboring areas. 148. The possibility that JSP has caused an increase in leisure was also examined. The RRA revealed cases in which husbands used JSP to increase their alcohol consumption and reduce hours worked. At the same time, most households reported no increases in leisure whatsoever. Indeed, some income generating activities appear to have increased hours worked, by reducing travel time to distant job sites and by increasing the marginal returns on labor. For instance, the ability to purchase building materials has lead to substantial expenditures of time on home improvement. Thus, while the supply of labor for casual employment may have temporarily declined, the overall impact is more difficult to assess. 149. Impacts of JSP: Effects on Women. The Women's Bureau study suggests that JSP may have had a differentially positive impact on women. Tabulated data in this study reveal that a substantial fraction of the credit extended to JSP households is being issued directly to women, and an overwhelming 76% of JSP women indicate that they experience a "greater recognition" within their families. In addition, women play prominent roles in many of the more successful informal sector banks that have been created. However, the same data also reveal only small increases in the percentage contributions of women to household income--and women's control of household assets may have even declined slightly (though the difference may not be statistically significant). 150. Impacts of JSP: Assessing the Net Benefits. Any program of JSP's magnitude can be expected to have profound impacts throughout the village-level, and even the national, economy. A benefit-cost analysis that incorporates all of these effects is clearly beyond the scope of our analysis. Nevertheless, it is useful to trace through for 1993 at least some of the main components one would like to include in such an analysis. 151. A useful starting point is to assume that the money spent on JSP would have been used to reduce taxes (net of any savings on FSP). Factoring in the off-budget expenses listed above, we have that the net expenditures on JSP total Rs. 3 billion in 1993. On the cost side, the Tax Commission Report indicates that the overall taxation system in Sri Lanka is mildly regressive. Because of income inequality, it is reasonable to speculate that the bottom 30% of the population receives only 10% of aggregate income, while bearing perhaps 15% of the tax burden. Assuming a 30% excess burden on each rupee of revenue collected, the burden of JSP for this segment of the population is Rs. 0.6 billion in 1993. Assuming further that 75% of the JSP funds are targeted at this sector, and that each rupee transferred results in Rs. 0.9 of benefit (due to restrictions on use and low returns on some projects), benefits to the bottom 30% were Rs. 1.9 billion in 1993, net of lost FSP benefits and increased taxes. Of course, more optimistic assessments of the sustainable increases in income would argue in favor of a higher multiple than 0.9; nevertheless, even this estimate shows that the poor benefitted substantially from JSP. 152. It also seems clear that the rest of the population has incurred substantial costs. Following the assumptions above, the richest 70% of the population may have sustained net losses of Rs. 2.9 billion (including the excess burden of taxation and the removal of FSP benefits to 189,000 households as of March 1993 due to JSP screening). Thus, one can say that JSP has entailed a loss of nearly three rupees among the upper 70% for every two rupees of benefit provided to the poorest 30%. Whether this results in a positive social net benefit depends, as is typically the case for transfer - 55 - programs, on one's aversion to inequality. Weighing benefits to all individuals equally, JSP creates a net loss of Rs. I billion; whereas, JSP creates a net benefit of over Rs. 2 billion if one uses normalized welfare weights that are inversely proportional to income. Of course, any net benefit calculation depends critically on the counter-factual one assumes. Net benefit would be greater if the relevant counter-factual is not a general reduction in taxes, but instead spending on projects with poor returns or with regressive patterns of benefits. Similarly, the net benefit of JSP would be smaller if the alternative were investment in projects with large returns or implementation of transfer programs that spread their benefits more evenly among the poor. 153. Unfortunately, there has not been an in-depth study of the success of the Janasaviya Program in alleviating poverty. However, the scattered available evidence suggests that the program has not been effective in sustainably lifting recipients out of households, and in particular, that the expensive transfer component has not been adequately supported by poverty-alleviation efforts. Given the new concept embodied in Janasaviya, it seems that it would have been preferable to launch a small pilot program and to learn from the outcome before beginning transfers on a large scale. IV.C. Mid-Day Meal 154. The Mid-Day Meal Program was begun in May 1989 with the objective of improving nutrition among children and increasing the returns to education (through increased attendance and through increased student alertness during classes). The program was initially intended to involve on- site feeding. However, by May 1990 the program was adapted to provide Rs. 3 per day of subsidy to students who have brought a meal to school. In order to qualify under the current guidelines, the meal must be nutritious and non-flour based (thereby encouraging import substitution). Students failing to attend school or failing to bring a non-flour meal are marked down on their MDM cards by their teachers. The unmarked portions of the MDM cards can then be exchanged for food at the Cooperatives. 155. Data supplied by the Ministry of Education show that students currently receive MDM compensation for nearly 80% of the maximum possible school days at the primary and secondary levels. Non-receipt is less at the primary level than at the secondary level. Absences from school account for approximately 3/4 of all non-receipts, with most of the remainder due to students not bringing nutritious, non-flour meals. 156. From the perspective of targeting efficiency, non-receipts probably occur disproportionately among the poor, according to Ministry sources, making the program mildly regressive. Regressivity is exacerbated at the secondary school level, because enrollment rates are lowest among the poor. With regard to the program's nutritional impact, it appears that Rs. 3 is substantially below the cost of a rice-based meal. In addition, many teachers are not enforcing guidelines on meal content (and may not even be recording students who do not bring meals). For these reasons, the program probably has only a small impact on the nutrition of children. One argument on behalf of the program is that it may help to reduce absenteeism. Indeed, there is some evidence that absenteeism has declined since 1980 from 20 to 15%. However, MDM's contribution to this decline is unclear. Overall, MDM has smaller net benefits than many other poverty programs. 157. In view of these concerns, the Government recently decided to take a step toward targeting the program. In September 1993 the Cabinet announced a decision to deliver the Mid-day Meal to Food Stamp recipient households with school age children in the form of an extra Rs. 50 monthly stamp. The improvement in targeting will be limited because the Cabinet also decided to allow households outside the Food Stamp program to apply for the additional stamp. This change is - 56 - expected to reduce the fiscal cost of the program from Rs. 1.8 to Rs. 1.3 p.a. in 1994. Furthermore, the management of the program will be shifted from the Ministry of Education to the more efficient Commissioner of Poor Relief in the Ministry of Reconstruction, Rehabilitation, and Social Welfare, which manages the Food Stamp Program. IV.D. Transfers to the Truly Needy 158. We examined a wide range of programs that can be grouped under the heading of transfers to the truly needy. Many of these programs are operated by the Department of Social Services (DSS), including: (i) public assistance payments of approximately Rs. 160 per month to households with less than Rs. 300 of income per month, with household heads who are disabled, elderly, or widows; (ii) custodial care for destitute elderly who are without families and living in nonprofit or public homes; (iii) rehabilitation of physically and mentally disabled; rehabilitation of socially disabled; and (iv) rehabilitation of detoxified addicts. In addition, before 1993 DSS ran the country's main disaster relief programs (including payments for drought relief and a large component of the refugee expenditures for the North and East). 159. The total budget in 1989, net of payments for disaster relief, was Rs. 189 million, of which over Rs. 156 million was for public assistance payments. Unfortunately, the DSS accounts cannot be reconciled after 1989, reflecting the disruption caused by devolution of programs (i) through (iv) to the provincial councils. The central government office continues to operate the DSS drug program. Also, the central office is nominally responsible for policy formation and evaluation, though in reality the central office has no data on how the money is being spent, so that meaningful policy analysis is impossible. 160. Despite the lack of data, it is clear that the real value of spending has declined substantially over the past decade. On average, public assistance payments have risen to Rs. 160 per month from Rs. 100 per month. However, this represents a real decline of nearly 50%. If it is assumed that these households, which on the average may have about three persons, receive about a two-third of their income from other government sources (food stamps) and personal transfers and that they spend 90% of their income on food, their total income inclusive of public assistance payments may not allow a calorie consumption level of more than 50% of the recommended level even if relatively cheap calories are purchased. Public assistance transfers have to increase to about Rs. 400 per month if at least a 60% of the recommended calorie consumption is to be expected. 161. Under funding of the Department of Child Care and Probation is even more dramatic. In particular, monthly support payments (or "maintenance grants") to nonprofit homes for orphans, abandoned children, and destitute children without relatives have been held at a constant Rs. 100 per month for well over a decade. Also, ad hoc capital grants to these homes have remained at a low level, as have funds for state orphanages and remain homes. At the same time, the number of orphans, street children, child prostitutes, and child drug users has risen dramatically. Our recommendations in the next section stress the urgent need for increased funding of these programs. IV.E. Recommendations for Changes in the Social Transfer Programs 162. Regardless of one's conclusions concerning the net benefits of Sri Lanka's poverty programs, it seems clear that substantial improvements are possible. At least a part of the resources targeted at JSP and Food Stamp households could be redistributed more equitably distributed among the truly needy in a way that substantially strengthens Sri Lanka's safety net. In particular, the payment of Rs. 250 per month, associated with the Rs. 25,000 certificates, appears to run counter to - 57 - the initial design of JSP as an intensive, fixed-term intervention to move households off the welfare rolls. The Rs. 250 per month payments are not even means-tested in principle, so that this is even more of a "dole" than FSP, which JSP was intended to replace. Table IV.4: COST PROJECTIONS FOR REFORM OF THE SOCIAL TRANSFERS (Millions of Rupees) YEAR 1994 1995 1996 1997 1998 1999 2000 A. Poor (FSP) 2263 3388 3659 3951 4268 4609 4978 B. Very Poor (Public Assistance) 1300 2106 2274 2456 2653 2865 3094 C. Total 3563 5494 5933 6408 6921 7474 8072 (As Percentage of GDP) 0.6 0.9 0.9 0.8 0.8 0.8 0.8 D. Savings 5948 4423 5673 6146 6409 6262 5982 (As a Percentage of GDP) 1.1 0.7 0.8 0.8 0.7 0.7 0.6 Source. World Bank 163. The following is a proposal for the reform of the transfer components of Sri Lanka's existing poverty programs. The proposal generates substantial savings while strengthening the country's safety-net for the poorest of the poor. The savings could be devoted to well designed and effectively targeted poverty alleviation and safety-net programs. 164. The proposal is to consolidate the transfer payment components of the social transfer and poverty alleviation programs into a single safety-net payment: Proposal: (i) Consolidate the transfer components of JSP, FSP, MDM, and Public Assistance into a single cash transfer program. (ii) Set two classes for income eligibility: the 'poor' are those that would be eligible for benefits under the current FSP and the 'very poor' are those that would be eligible under current Public Assistance program. (iii) Replicate the screening process used in JSP to increase targeting efficiency of FSP. The screening should be completed by the end of 1994. (iv) Once the screening is successfully completed 1994, restore some of the real value of transfers. Increase transfers to recipients in the 'poor' category to Rs. 200 per month - 58 - in 1995, Rs. 250 per month in 1996, and then index them to the inflation rate. Increase transfers to households in the 'very poor' category to Rs. 400 per month in 1994 and Rs. 600 per month. Payments at a slightly lower rate should be made to orphans in voluntary homes. (v) As in (iii), eliminate the under-funding of rehabilitation of the disabled and drug abusers and of homes for the disabled and elderly. (This expenditure is included with "Public Assistance" in Table IV.4). The net savings from adoption of this option average Rs. 5.2 billion per year in 1994 and 1995, some or all of which could be used to strengthen poverty alleviation programs such as nutritional interventions and community development initiatives. IV.F. Education 165. Sri Lanka has an impressive record of performance in Education, which it should sustain with the continued adequate support of expenditures from the government budget. Total Government expenditure on Education as a percentage of GDP, excluding transfers under the Mid-Day Meal program, averaged about 2.6% in 1980-88, reached a high of 3.2% in 1989, and declined subsequently to 2.3% in 1992. Government expenditure on Education as a percentage of GDP does not appear excessive, especially taking into account the high enrollment rates achieved in general education (primary and secondary). The constraints on fiscal spending over the next several years mean that the funds available for Education will have to be wisely spent. 166. For this reason, the Government should address the rapid expansion of the wage and salary bill, which is crowding-out other expenditures and harming the quality of education. Salaries absorbed 97% of recurrent expenditure on education of the Provinces in 1992, which now run most of the general education schools. Less than 1% of these expenditures were for quality-enhancing expenditures such as teaching materials (excluding books, which are centrally financed) and maintenance of facilities (Box IV.2). School Maintenance Expenditure on school repair and maintenance in 1992 averaged about Rs 2,000 per school. A rough estimate of an adequate expenditure per school is about Rs 20,000 per school (in 1992 prices). The almost negligible expenditure on school repair and maintenance in recent years has left many school buildings, facilities as well as furniture and equipment in deplorable condition, affecting the quality of education. The Government recognizes the inadequacy of expenditure on school maintenance, and has set up a School Board for each school to attempt to develop a source of local non-government funds for repair and maintenance. If the experiment is successful, it is plausible that perhaps one-half of required repair and maintenance expenditure could be met with School Board funds. In that case, the total government budget allowance required for adequate repairs and maintenance of the existing 10,000 schools would be about Rs 100 million (Rs 10,000 times 10,000 schools) in 1992 prices, which would be equivalent to about 1% of the total government budget for education (at all levels) in 1993. Box IV.2 - 59 - 167. Real expenditures on salaries in general education increased at an annual rate of 6% over 1985-92. Large salary increases in 1988 and 1993 and recruitment were a major source of the rising bill. The other major source was recruitment to fill vacancies for necessary staff but also to absorb white-collar unemployment. The number of school teachers has increased rapidly from about 141,000 in 1988 to 176,000 in 1992 and to over 180,000 in 1993. As the number of teachers increased, there was a parallel decline of the student-teacher ratio from about 28 in 1985 to 22 in 1992, which is low by international standards. Nevertheless, recruitment damaged the quality of education as the new recruits did not adequately replace the departing teachers. Many of the teachers that retired during the attempted reform of the public administration in 1990 were experienced and taught subjects such as math and English which are in demand while the new recruits are typically young, have at most a secondary school education, and are often not qualified to teach subjects in high demand. The recruits are hired as trainees who qualify for full salaries after four years of experience, so the full effect of hiring on salaries has yet to be seen. If the recent teacher recruitment policies were to continue, the 1992 salary bill for general education would double by 2000 and the overall sector budget would increase to an unsustainable annual growth rate of 6.4%. 168. As the number of teachers increases, demographic developments are reducing the need for teachers. The decline in fertility is lowering the number of students entering primary school, and this effect will progress up the primary grades (1-5) and then will gradually reduce the number of students who can enter secondary school. At the upper-secondary school level the effect of the decline in the entering cohort is balanced by students that wish to enter grades 12-13. These favorable demographic trends open up the possibility for the country to increase the secondary enrollment rate and upgrade the quality of general education without substantially increasing spending on general education relative to GDP. Furthermore, resources freed by lowering the student-teacher ratio should be used to increase the quality of education through better expenditures on educational materials and maintenance. 169. These are the reasons why gross recruitment of teachers should be frozen for the time being. Meanwhile, the Ministry of Education and Higher Education needs to establish priorities for current and emerging shortages and develop a staffing plan. Even then, hiring should be confined to the natural attrition rate of about 2% per year. If teacher recruitment could be confined to attrition, savings in salary would reach over Rs. 1.2 billion (17% of the 1992 bill). 170. Another expenditure issue is the wide discrepancy in resource allocation per student between small and large schools, and this is only partly explained by the tendency of larger schools to have a higher proportion of students at the secondary level. There are also wide discrepancies in the allocation of education resources among provinces. In general, poorer provinces have proportionally more unqualified teachers, higher teacher shortages, and tend to spend much less on instructional and teaching materials. 171. The remaining expenditure issues, aside from salaries and recruitment, is the Government's strategy toward the development of higher education. Only 2.5% of students of university age are enrolled in universities, which is low compared to other low and middle income countries. The Government is now expanding higher education; real spending grew at an annual rate of growth of 12% in the 1987-92 period, and would continue to grow at about 9% per year up to the end of the decade under present plans. The Government needs to set feasible targets for university enrollment. Moreover growth needs to be directed and linked to quality control by focussing on areas which need improvement: math, science, technology, and the command of practical English. - 60 - 172. The Government should reconsider its policies concerning financing of university education. One reason is that the fast-growing East Asian countries followed a strategy of concentrating public education spending on the primary and secondary levels. At the post-secondary level, they focussed public spending on science and technology education (including engineering), while allowing the demand for other types of higher education to be met by a self-financed private university system. To assure the feasibility of the expansion program for higher education, the Government should place greater emphasis on cost-recovery and the role of the private sector. 173. The Public Investment Program in Education. According to the PIP for 1993-97, public investment in education is expected to remain stable at its 1991-92 share of about 0.6% of GDP. Six items compose most of public investment in education: general education (national schools), 4.4%; general education (other schools), 26.6%; teachers' education, 2.0%; technical education, 9.8%; the University Grants Commission, 26.9%; and provincial councils, 18.1%. No functional breakdown of the provincial councils allocation is given, but most likely nearly all of it would be for general education. 174. The amounts planned for general education appear reasonable overall, since the secondary system is still growing and the quality of primary and lower secondary education need upgrading. Within this broad category, the construction of new buildings for national schools may be of lower priority, given that these schools are already well endowed compared to other schools and that maintenance is a higher priority. Another concern is that investment by the provincial councils is close to one-fifth of the total, yet no details on the composition of these investments were available to the mission. 175. The increase in planned investments in technical education does not appear justified, since existing technical colleges are under-utilized and have often inappropriate programs. Indeed, the PIP for 1993-97 shows a large increase for this type of investment compared to the PIP for 1992-97. This matter will be analyzed in more detail in the context of the preparation of the proposed IDA- assisted Vocational Training Sector Reform Project, a major objective of which would be the rationalization of the technical education sub-sector. The allocation for teachers' education is small, but it may be adequate since existing training centers are under-utilized. IV.G. Health 176. Total Government expenditure on health has averaged around 1.5% of GDP since 1980-87 and is budgeted to remain at that level in 1993. This is about average compared with most South Asian and East Asian countries. Public plus private expenditure in health-related goods and services as a percentage of GDP was an estimated 3.1% in 1986-87, which is also about average when compared with the same group of countries. 177. Govdrnment health facilities already provide good coverage throughout the country. As with education, the decline in fertility provides an opportunity for using future increases in per capita Government health expenditure to improve the quality of services. In the last decade of the century the population of children under five will decline; at the same time, however, the population aged 60 and over will rise from 1.4 million in 1991 to 2.3 million in 2001. This will increase the load of diseases with high treatment costs, such as cancer and cardiovascular diseases. 178. These changes in epidemiological patterns, together with the strong growth of private health services in recent years, call for a thorough policy review in the sector. In 1979, private expenditures for health related goods and services represented 38% of overall national health - 61 - expenditures; by 1986-87 this figure had risen to 51%, and the trend is likely to have continued since then. The growth in private services is highlighting the need to formulate an adequate regulatory and supervisory framework for these services as no regulatory body has been set up. It may also be desirable to redefine the division of labor between the public and private sectors in both the provision of services and the financing of access to services. For example, the public sector could concentrate more on preventive care, health education and communicable disease control, leaving most curative services to the private sector. However, this would raise difficult questions of financing access to curative services in general for the poor, and for the non-poor with relation to chronic and catastrophic illness. 179. Still, prevention remains a priority activity for the Government even though mortality rates continue to decline. The incidence of morbidity from preventable infectious and parasitic diseases are on the rise, including malaria, which has again reached epidemic proportions, and other diseases such as dysentery, typhoid, amoebiasis, and infectious hepatitis. Additional public expenditure in reducing the incidence of these diseases would have a high economic pay-off, through reduction of the burden these diseases place on the curative health services and through reduction of the days of productive work lost. Only a part of this additional expenditure ought to be in health services. Most of these diseases are transmitted through contaminated food and water, which is often the result of poor sanitation facilities; hence Government assistance in the provision of adequate latrines to households would be one of the most important interventions to reduce incidence of these diseases. 180. Higher level facilities such as teaching and specialized hospitals and provincial hospitals are overcrowded, while smaller facilities such as rural hospitals and dispensaries are under- utilized. This imbalance could be reduced by providing more adequate funding to the under-utilized facilities. Such increased funding could be achieved at the expense of slowing down the expansion of higher level facilities, which continue to be emphasized in the public investment program. That expansion has clearly been driving the priorities in the current budget: in the 1981-91 period, when current Government expenditure in health services increased in real terms at an average annual rate of 5.2%, almost all of the total increase went to finance the provision of curative services at tertiary and secondary facilities. In contrast, current expenditure for facilities at the primary level and preventive programs actually declined in real terms. 181. There has been a large decline in the ratio of annual health current expenditure to the value of the health capital stock, from an average of 0.9 in 1975-77 to 0.4 in 1990-92. The decline raises the concern that an imbalance may have developed between spending on expanding facilities and spending on operating and maintaining facilities. Indeed, expenditures on repairs and maintenance of health facilities appear inadequate (Box IV.3). 182. The Public Investment Program in Health. In the 1993-97 PIP, public investment on health is expected to increase from a base of Rs. 1.8 billion in 1992 to an annual average of about Rs. 2.5 billion (in constant 1993 prices) in 1993-97. Table IV.5 below shows the composition of the Public Investment Program in Health for 1993-97. The main question concerning composition is how it relates to the functional classification of services. Items 4, 8, 10, 11 and 12 generally relate to primary and preventive care; these items add up to about 26% of the total. Items 5, 6, 7 and 9 generally relate to higher level curative facilities; these four items comprise a further 26% of the total. Items 1, 2, and 3 sum to about 40% of the total and relate to all levels of the health system, but it is likely that the major portion of it relates to secondary and tertiary level facilities. If this is so, the 1993-97 Public Investment Program would further reinforce the current emphasis on higher level curative services. - 62 - Repairs and Maintenance of Health Facilities An internationally accepted norm for expenditure on repairs and maintenance of health facilities is that an adequate annual expenditure level would be approximately 1.5% of the value of the health facilities capital stock. In Sri Lanka, the value of the capital stock in public health facilities in 1992 was approximately Rs 10 billion in current (1992) prices. Thus, an adequate allowance for repairs and maintenance of these facilities in 1992 would'have been about Rs. 150 million. In recent years, however, actual current expenditure for repairs and maintenance has amounted to only about one-fourth of the above norm, or about one-third if the rehabilitation and improvements component of the public investment budget is also considered. This low level of expenditure on repairs and maintenance of health facilities, if allowed to continue, will inevitably result in deterioration of these facilities and a consequent decline in the quality of services rendered. Box IV.3 Table IV.5: COMPOSITION OF THE HEALTH PUBLIC INVESTMENT PROGRAM, 1993-97a Items Percentages 1. Rehabilitation and Improvement of Capital Assets 13.6 2. Furniture, Equipment and Vehicles 21.9 3. Staff Quarters 4.7 4. Water Supply and Sanitation 3.1 5. Development of Colombo General Hospital 3.1 6. Construction of Bio-Medical Engineering Services Plant 3.8 7. Development of LURE Hospital 2.1 8. IDA/ADB Health and Family Planning Projects 12.0 9. Grants to Provincial Councils 16.7 10. Indigenous Medicine 1.1 11. Implementation of Action Plan for Children 5.3 12. Building for STOOD/AIDS Program 4.2 13. Miscellaneous 8.4 Total 100.0 a/ Calculated from figures expressed in constant 1993 prices. Source Public Investment 1993-97, Department of National Planning, Ministry of Policy Planning and Implementation, May, 1993 - 63 - V. PUBLIC ADMINISTRATION AND THE CIVIL SERVICE V.A. Introduction 183. The public sector in Sri Lanka is large and elaborate. The public administration consists of the Central Government, with 31 ministries, eight Provincial Councils, 25 District Administrations, several urban and city councils, and 268 Divisional Secretariats. There are also about 100 statutory agencies which are formally state-owned enterprises, but which provide a public service (i.e., libraries, museums, etc.) and are mainly supported by transfers, and about 17 commercially- oriented state-owned enterprises. The size of the Government can be better understood through examination of its employment structure (Table V.1). The public sector (excluding the plantations) employs about 800,000 persons, of which 168,000 are teachers employed by the Provincial Councils. Employment by sector in the public administration reflects the re-deployment of a large number of teachers and some medical staff to the Provincial Councils. The large staff in agriculture reflects the Government's heavy commitment in irrigation and the staff in transport and communications is large because the railway and postal services are departments of the government. A further 400,000 persons work on the state plantations at wages set in the public sector, but often under the management of private companies. 184. Since 1990, the Government has been implementing, with mixed results, a program to increase the effectiveness of the civil service and to control the fiscal cost of wages, salaries and pensions. Sri Lanka's civil service, like most in South Asia, has a long tradition of independence, integrity, and high technical and analytical skills. However, factors such as excessive political interference in hiring and promotion, the impact on morale of ethnic and political conflict, and misguided remuneration policies have eroded its effectiveness. At the same time, the fiscal cost of the civil service has been increasing as a proportion of the budget; wages an pensions increased their share in current expenditures (net of interest) by 7 percentage points between 1980 and the budget for 1993. Increasing fiscal pressures together with the need to reverse the perceived decline in public confidence in government convinced the Government that a thorough reform of the civil service was necessary. 185. In order to increase the cost-effectiveness of the civil service, various interrelated issues need to be addressed. Among the most important are over-staffing, imbalanced incentives, and the existence of units in the administration with overlapping responsibilities and units which have lost their function. Over-staffing is partially the result of political pressures that have led the Government to use the civil service as employment of last resort, especially for young graduates. While over- staffing has raised the wage cost of the civil service, resource constraints have resulted in un- competitive remuneration for the higher echelons of the civil service. In addition, as explained below, the civil service pension scheme is a significant and growing fiscal burden. Finally, as a result of the redefinition of the role of the pubic sector, some administrative units lost, at least in part, their function. This is a formidable list of problems that need to be tackled concurrently. 186. The chapter first reviews past experience in civil service reform in Sri Lanka and attempts to draw some lessons. On the basis of this analysis, it then makes some recommendations. This chapter also reviews certain aspects of a relatively new area of concern in public administration that has emerged in Sri Lanka: decentralization. In 1987, the responsibility of delivering crucial public services was transferred to newly created Provincial Councils. However, in 1992, the Government set up parallel and potentially competing local administrations called Divisional Secretariats that report to the Central Government. In addition, inter-governmental financial - 64 - arrangements and the efficiency of the Provincial Councils in delivering public goods and services are also of concern. Little work has been done in the past in this area in Sri Lanka, and while there is some diagnosis of the main issues, the sections dealing with decentralization are mostly descriptive. Table V.1: EMPLOYMENT IN THE PUBLIC SECTOR Organizational Civil Service Semi-Autonomous Total % Unit (1) Bodies (2) Staff No. of No. of Units Staff Units Staff General Administration 6 28,093 22 34,750 62,946 8 Defense/Police 1 149,995 2 1,699 151,694 19 Sub-Total 7 178,008 24 36,499 214,640 27 Productive Sectors Agriculture 5 30,853 46 51,677 82,530 10 Industry/Commerce 3 2,096 44 37,983 40,079 5 Tourism 1 854 8 1,754 2,608 0 Sub-Total 9 33,803 98 91,414 125,217 16 Infrastructure Transport 2 18,241 17 58,965 77,106 10 Communications 1 17,506 17,506 2 Energy 1 95 4 20,867 20,962 3 Construction 1 2,073 15 20,567 22,640 3 Sub-Total 5 37,915 36 100,299 138,214 17 Social/Religious Education 1 5,149 24 8,456 13,605 2 Health 1 17,530 6 2,022 19,552 2 Others 5 7,791 15 7,495 15,286 2 Sub-Total 7 30,470 45 17,973 66,416 8 Total Central Government 28 280,276 203 246,185 526,461 66 Provincial Councils 8 241,903 241,903 30 of which: teachers 168,298 168,298 21 Other Local Governments (3) 33,000 33,000 4 Consolidated Total 555,179 246,185 801,364 100 (1) Ministries other than Defence: Ministry of Policy, Planning and Implementation, Quarterly Census of the Civil Service, September 1992. Defence: Civilian Staff (15,478), Quarterly Census of the Civil Service, September 1992. Military and Police Staff: Ministry of Policy, Planning and Implementation. (2) Department of Census and Statistics, Census of Employment in the Public Sector, 1990 Excludes state-plantation employees. (3) Ministry of Policy Planning and Implementation, rough estimate. Source: Public Expenditure Review Mission estimates V.B. The Administrative Reform Committee and the Reform of 1990 187. The recent history of civil service reform began with the appointment of the Administrative Reform Committee (ARC) by the President in 1986. In 1987 and 1988, the ARC published 10 reports recommending substantial reforms. The reports were founded on the following assessment: - 65 - "...the administrative system of Sri Lanka is afflicted by a twin malady --namely, a high level of over-staffing and an unrealistically low salary structure. The two are, in reality, linked. They have, in turn, led to poor morale and discipline; to an erosion in standards of integrity; to public servants turning away from training and self improvement; to low productivity and consequent high costs of government operations; and, finally, to an erosion of public confidence." (Report No. 2, p.1). 188. The most striking recommendation was to reduce the central government civil service by more than 72,000 persons over 18 months. This would have been a 20% reduction from a base for the civil service that excludes the education and health sectors, the military, and the police. Reductions in police and health services were to be considered separately. The total reduction was to be achieved by a freeze on new hiring, attrition by retirement and by offering separation packages to induce voluntary departures. 189. Furthermore, the ARC proposed specific increases in the wages or salaries of each grade of civil servants, with relatively large increases for skilled workers whose wages have been well below private sector levels. It also proposed the indexation of wages on non-executive grade employees to inflation. Moreover, it favored the establishment of staffing norms, for instance, pointing out that the Malaysian railroad operated the same length of track as Sri Lanka with half the staff. In addition, the ARC recommended numerous measures to discipline the hiring and personnel process. Among these were planning of staffing levels and the creation of a Cadre Review Commission to make final decisions on the staffing level of ministries, departments and offices. This Commission was meant to decide how to distribute the recommended cut in staffing among various parts of the Government. The ARC also sought to limit the growth of Government by supporting the adoption of "sunset clauses" that would terminate any new government program and dismantle its organization after a fixed period unless a review led to a decision to continue the program for another fixed period. The recommendations of the ARC also provided for the decentralization of services and programs to the provinces. 190. In 1990, the Government launched an attempt to reduce the civil service by offering generous severance packages--including enhanced retirement benefits under Circular 44/90. An estimated 43,000 staff retired in 1991, including staff who would have retired under normal conditions. An additional 25,000 retired in 1992. The retrenchment program was coupled with a freeze on new recruitment into the civil service, except in education, health and the security forces. 191. The reform appears to have had mixed results. The retrenchment program did not ease the fiscal burden but deprived the civil service of some of its most qualified senior staff members. The immediate cost of the scheme was very high, resulting in more than a doubling of the pension bill to Rs. 9 billion (2% of GDP) in 1991. Savings on wages and salaries were reduced as the Government replaced departing teachers with about 14,000 new trainees, who were confirmed as permanent staff in early 1993, and recruited between 2,000 and 3,000 contractual employees to replace departing essential staff. 192. The Government's ability to monitor the number of actual staff through quarterly surveys has been impeded by faulty reporting by spending agencies and local governments. Also, the survey does not record staff in the military and the police nor cover semi-autonomous bodies, state- owned enterprises and urban councils, which largely depend on government transfers. Thus, the freeze is less than fully effective. In fact, there is evidence of an increase in the approved cadre and there probably has been continued recruitment in semi-autonomous state bodies since the freeze was - 66 - imposed. On balance, a freeze on employment, which was meant to ensure that retrenchment targets were achieved, does not appear to have been successful so that the size of the civil service has increased since 1991. In particular, the large recruitment of new teachers (trainees) in 1992 and 1993, despite the already relatively high teacher/student ratio, has significantly eroded the gains made. 193. There was also an.effort to compress the size of the administration through the consolidation of cabinet ministries. This measure helped streamline the government but was less than fully effective because additional layers of management were subsequently created within the new ministries, thereby maintaining the high overhead costs that had prevailed before structural rationalization. As explained below, the problem was compounded when the decentralization program stalled, while recruitment in the Provincial Councils was only belatedly arrested. The total impact is the maintenance of a large central administration together with large staffs in the provinces which are only partly employed. 194. Finally, the reform brought some partial improvements in hiring practices. Recruitment has been centralized for staff officers and is handled by the new Public Service Commission at the national and provincial levels. Lower level staff are still hired directly by line ministries, agencies and Provincial Councils--but the approved cadre is monitored by the Ministry of Policy, Planning and Implementation. Requests for new positions are investigated by a Salary and Cadre Committee which reports directly to the Ministry of Policy, Planning and Implementation. Nevertheless, this Committee would need to be strengthened given the very large agenda for cadre review in the future. In general, the institutional set-up for managing the civil service is still weak and lacks the strong central focus to make it effective. V.C. The Civil Service Today 195. The Wage Bill and the Structure of the Civil Service. Over the past 10 years, despite the 1990 reform, the central government's civil service wage bill has increased more rapidly than current expenditures (Table V.2). The figures include allowances, which account for an average of Table V.2: TRENDS IN THE COSTS OF WAGES AND PENSIONS (In Rs. Billions and in Percentage Points) 1983-85 1986-88 1989-91 1992 1993 Wage Bill (Rs.) 5.7 8.7 16.1 20.8 23.0 Wages/Current Exp. (%) ' 29.2 23.8 20.1 32.7 34.6 Wages/Total Govt. Revenues 17.7 21.5 24.4 24.2 22.6 Wages/GDP 3.7 4.0 4.8 4.9 4.6 Pension Bill (Rs.) 2.2 3.5 6.0 8.0 9.1 Pensions/Current Exp. (Rs.) ' 11.2 9.5 9.1 10.0 13.7 Pension/GDP (%) 1.5 1.7 1.8 1.9 1.8 Wages & Pensions/GDP (%) 5.2 5.7 6.6 6.8 6.4 a/ Current expenditures less interest payments. Source. Ministry of Finance, Budget Estimates and Actuals. - 67 - 13% of the civilian wage bill and about 50% of military wages. They range from cost-of-living allowances, railway warrants to pay for travel during annual leave for civilians, to uniforms, meals and lodging for the military. In the 1993 budget, wages an salaries account for 4.6% of GDP; the pension bill has increased to 1.8% of GDP, as a result of adjustments to the pension system in 1985 and the generous benefit package offered in 1990 for early retirees. 196. One of the important recommendations of the ARC was to reform the compensation system. Salaries could not assure civil servants an adequate standard of living. Moonlighting was rife, leading to divided loyalties and poor performance. The differential between higher-level staff compensation in the private and public sectors is about 4:1, even after taking into account the tax-free status of civil servants. This discourages qualified candidates from entering and remaining in the civil service. The compensation package for lower-level staff, including the generous pension scheme, nonetheless is attractive relative to the private sector. The compression ratios between higher and lower grade staff is at a relatively low 6:1, which further discourages performance at managerial levels. 197. The ARC emphasized reform of the salary structure, but since then, the situation has worsened. The 1991-92 retrenchment of staff failed to provide the savings necessary to improve the salary structure by increasing salaries of higher grade staff toward private sector levels. A recent study by the Salary and Cadre Committee indicates that real wages and salaries for staff officers have declined by more than 60% since 1988, while remuneration for lower grade staff, including cost-of- living adjustments, remains competitive with the private sector. Since the bulk of the wage bill is paid to lower grade employees, the margin to improve salary levels for higher level officials without exacerbating the budget deficit is severely constrained by over-employment at lower grades. 198. In early 1993, the Government raised salaries across-the-board by 30%--the first general salary adjustment since 1987. Staff in the military and the police force also received increases, while pensioners received a lump sum adjustment of Rs. 300 per month. The 1993 salary adjustment adds Rs. 4.3 billion to the wage and pension bills. The salary adjustment exacerbates the budget deficit and further weakens the salary structure by narrowing compression ratios. It is also unleashing wage increases in semi-autonomous government bodies and state-owned enterprises. The effect on lower-grade staff is to make government employment even more attractive, compared to the private sector. If the government sector is the wage leader, competing staff at lower grades in the private sector will seek similar adjustments in pay, thereby undermining international competitiveness. 199. Hence, a key lesson from the 1991-1992 reform process is that retrenchment should focus on specific categories of employees. The bulk of civil service employment is in low-skill grades C & D, consisting of minor officials, clerks, secretaries, and teacher trainees --although Grade C includes some professional and technical staff (Table V.3). Efficiency in these grades is low because of featherbedding, the slow introduction of office technology, and, in the case of teachers, low education and training levels upon entry. With a large share of employees occupying lower level positions and the bulk of the wage bill, there is a strong case for focussing on rationalizing staffing within this cohort. 200. A second important lesson is that, rather than attempt to retrench civil servants throughout the Administration, efforts should be concentrated on those parts of the Administration that have lost their function and are being eliminated. Similarly, the consolidation of units that have over- lapping responsibilities may also open opportunities for retrenchment. For example, many government agencies which, hitherto, had been responsible for directing and controlling economic activity are still in place (such as the Food Commission), while commercial entities such as the railway and the postal - 68 - services (with 25,000 employees) remain in the administration. Furthermore, the Government should consider consolidation of small ministries, such as Energy, Tourism, Environment, Plantation Industries. Table V.3: CENTRAL GOVERNMENT APPROVED CADRE BY GRADE, 19938/ Sector Staff Grades D C B A Ministers Total Guidance Cluster 16,688 98,961 4,918 378 12 120,957 of Which: Defense 8,478 73,709 761 0 6 82,954 Production 16,979 12,067 2,381 56 30 31,513 of Which: Rural Sector 16,054 10,185 2,024 31 16 28,310 Economic Infrastructure 25,663 28,055 844 20 12 54,594 of Which: Transport 18,896 11,561 311 4 2 30,774 Postal/Telecom 6,204 15,877 311 5 2 22,406 Social Sectors 13,871 20,559 3,382 35 18 37,865 General Total 73,201 159,642 11,525 489 72 244,929 Percentage Share in Total 29.9 65.2 4.7 .2 .0 100.0 a Employment in the military is not included. Source Sri Lanka, 1993 Budget Estimates 201. Pensions. Sri Lanka's pension scheme for civil servants is generous compared to international norms. It is a liability-driven system which obliges the Government to pay retirement benefits out of the budget, without contributions from employees. A separate widow and orphan scheme, to which employees contribute between 3 to 6% of base salary, provides coverage for surviving spouses and children under legal age. Pension payments are not indexed but the Government has provided ad hoc adjustments of Rs. 100 to 300 per month to pensioners with monthly pensions below Rs. 3,000 per month, for an average 3.0% per year increase in the pension bill. 202. Two changes introduced in 1985 and 1990 to the pension scheme have increased its cost even further. Until 1985, the pension payable to a retiring employee (age 55 for men and 50 for women) was based on the number of years in service. Staff had to work at least 10 years in the civil service before being entitled for a pension, and the maximum pension paid represented 80% of the average salary over the last three years of employment times a factor linked to length of employment. The 1985 modification weakened the link between pension payments and years in service. Staff with a minimum of 10 years of service could retire with a pension representing about 80% of final earnings, averaged over their last three years of service, with the maximum pension being pegged at 90%. Furthermore, pensioners could opt for a reduced pension for a period of 10 years, in exchange for a lump sum payment of 24 months of pension benefits. Most pensioners exercise this option - 69 - because it increases the value of benefits by an estimated 14%. The impact of the 1985 modification increased the cost of pensions to the Government by about 17% of payroll. 203. In September 1990, Circular 44/90 was introduced as a temporary measure to induce staff to leave the civil service. However, regular retirees can opt for its benefits, and since it has been in place for two years, it is de-facto the regular retirement scheme. It significantly increased incentives to retire while expanding the number of eligible pensioners. Civil servants are eligible to retire after 20 years of service with a pension equivalent to 80% of their salary in their final year of service; those with more than 20 years of service receive up to 90% of their final salary. The lump sum payment of 24 months of pension does not entail a reduction in monthly pension payments. Furthermore, notice that pension payments are calculated on the basis of the last year's salary, instead of the average over the three preceding years. The 1993 adjustment in pensions has further increased benefits. 204. The voluntary retirement plan under Circular 44/90 resulted in a surge of retirements, about 35,000 retirements in 1991 and a further 25,000 in 1992. Despite this surge, there were 33,500 employees still eligible to retire under the scheme at end 1992. The estimated additional cost relative to the regular retirement plan is about Rs. 5.0 billion over the two years 1991-92. As part of the 1993 salary adjustment, pensioners will receive a lump sum adjustment of Rs. 300 per month which will add another Rs. 1.0 to the pension bill. 205. Hence, a third lesson that could be drawn from the 1991-1992 reform, concerns the need to reduce the cost of future retrenchments. The system offers extraordinary incentives for voluntary retirement. Civil servants acquire these as rights, thereby making civil service employment salary poor and pension rich. The incentive system provides little reward for current work but rather rewards passive employment in the civil service. Should the voluntary departure scheme remain in place, the system will continue to distort incentives and raise the total wage and pension bill beyond what would be necessary to attract and retain qualified staff. In addition, a reform should ensure that the pension cost of securing voluntary retirement does not exceed the savings in salaries. V.E. Reforming the Civil Service 206. The objectives of a reform program would be to reduce the fiscal cost and improve the efficiency of Government. A reform which attempts to cut costs should start by slowing and then stopping the expansion of the wage and pension bill of the civil service. The single most important step, in the short-term, is to impose strict recruitment limits in the civil service, including Health, Education, the Provincial Councils, and the statutory agencies (such as the Mahaweli Authority). There should be no replacement of lower-grade staff, replacement of only one out of every two retiring professional and technical staff, and full replacement of senior managers. The freeze should remain in effect until the completion of administrative and functional review to determine the necessary cadre. At the same time, the Government should identify a short-list of units which have lost their function or have over-lapping responsibilities, and then eliminate or merge redundant units. Excess staff would then be obliged to accept available transfers or severance packages. As a result of these measures, Government employment would contract by an estimated 48,000 over 1993-97--mostly through retirement. Rationalizing job contents and descriptions would ensure that the remaining staff are fully employed. Another urgent step is to withdraw Circular 44/90 as a means of balancing incentives and limiting the pension bill. 207. As the size of the civil service decreases, incentives should be improved to enhance the efficiency of Government. A reform of salaries should integrate allowances in the civilian - 70 - administration into personal emoluments and strengthen performance incentives at the higher staff grades. The savings generated from attrition of the size of the civil service would finance the salary reform while also contributing to deficit reduction. As salaries increase, civil servants should begin to pay income taxes to conserve the unity of the tax system. 208. Improvements in the cost and efficiency of the Central Government will be sustained only if the agencies charged with managing and monitoring the civil service are strengthened. This is particularly the case for the Civil Service Commission and the Salary and Cadre Committee--which play a crucial role in auditing requests for new positions and in making salary adjustments. A comprehensive census of the civil service would support adjustment of the civil service. The data would then form the basis of a computerized personnel management data base. To ensure that changes in personnel files are systematically reported by spending agencies, the data base would be maintained by the Ministry of Finance and be used to manage the wage bill. 209. A final means to improve costs and efficiency over the medium- and long-term is to establish a capitalized pension fund with employee contributions. As a first step, the Government should start making contributions to a fund that would cover future pension liabilities under a new scheme. Several proposals to reform the pension scheme have been developed in the past, including a plan that resembles the private sector's Provident Fund and a new capitalized pension plan. Assuming stability in civil service employment, the introduction of a capitalized pension fund over time would significantly reduce the expected cost of the pension bill. A census of pensioners and computerization of the pension system would greatly aid the monitoring and achievement of the reform. Furthermore, there is no reason why the pension system should discriminate against women, who, under the present regime, must retire at age 50 while men must retire at 55. The Government should also consider increasing the retirement age to age 65. 210. Table V.4 compares the projected outcome of the proposed reform program with the projected outcome without reform. With a reform program the fiscal burden of wages and pensions declines steadily relative to GDP while staffing quality improves. Through the accelerated departure of lower-level staff, staffing ratios between lower and higher-level grades markedly improve. The non-replacement of teaching staff who retire increases the student-teacher ratio. There is a marked improvement in the salary structure, thereby strengthening incentives for managers and enabling the Government to attract higher caliber staff. The margin created by administrative downsizing allows the Government to raise salaries so that average real incomes achieved following the 1993 salary adjustment are maintained till the end of the decade. With structural rationalization being undertaken concurrently, obsolescent and non-performing programs are abandoned, reducing the need for public servants and allowing the administration to be down-sized, without affecting the quality of public programs and services. At the end of the decade, government employment is reduced by a total 13%. The scenario assumes that the size of the country's security forces remain at the 1992 level till the year 2000. 211. Without reform, no further attempts are made to retrench excess staff, and public employment continues to expand. Without savings, salary reform is not possible: the government's margin to improve incentives decreases so that the compression ratio is not allowed to rise. Past trends suggest that nominal salaries cannot be sustainably frozen, so the no-adjustment scenario assumes salaries and allowances expand by about 13% a year (including wage drift). The combined effect of new recruitment and salary increases somewhat raise the wage and pension bills relative to the reform scenario. However, as a share of GDP, wages and pensions would fall from 6.4% in 1993 to 5.1% in 1997 and 4.8% in 2000. The main reason for this decline in the share of wages and pensions in GDP is that the size of the civil service grows more slowly than real GDP. - 71 - Table V.4: PROJECTIONS OF CENTRAL GOVERNMENT WAGES AND PENSIONS 1992 1993 1994 1995 1996 1997 2000 With Reform Total Staff (In thousands) 555.2 552.3 546.0 534.7 522.3 510.1 485.6 Compression ratio (in percent) 7.0 6.5 6.8 7.2 7.7 8.0 9.1 Wage + Pension Bill (Rs. billions) 27.9 30.9 33.5 34.8 36.6 37.7 43.7 Wage + Pension Bill/GDP (in percent) 6.4 6.2 6.1 5.6 5.3 4.8 4.0 Without Reform Total Staff (in thousands) 555.2 563.4 573.0 582.1 590.6 598.7 621.0 Compression ratio (in percent) 7.0 6.5 6.5 6.5 6.5 6.5 6.5 Wage + Pension Bill (Rs. billion) 27.9 31.1 34.3 35.0 38.8 39.5 52.7 Wage + Pension Bill/GDP (in percent) 6.4 6.4 6.2 5.6 5.6 5.1 4.8 Source: Public Expenditure Review Mission 212. The additional cost of the no-reform scenario versus reform averages 0.2% of GDP throughout 1993-97 and 0.8% of GDP in 2000. The main cost of not undertaking reform is not fiscal. Without reform, Sri Lanka would risk suffering a persistent downward spiral in the effectiveness of government services as the government traded-off improved incentives in favor of the continued expansion of the civil service. The absence of progress in adjusting the size and composition of the civil service and the incentive regime, in turn, undermines administrative and management reforms. Moreover, the policy of infrequent but large salary adjustments can temporarily destabilize public finances and create inflationary conditions which undermine real civil service incomes. V.D. The Provincial Councils 213. The Legal Basis for Provincial Government. The 13th Amendment to the Constitution that was passed in November 1987, in connection with the India-Sri Lanka peace accord, devolved certain administrative powers to nine newly created Provincial Councils.' Under the 13th Amendment, the powers of government were divided into three categories: (i) powers devolved exclusively to the Provincial Council (the Provincial Council List); (ii) powers retained by the center (the Reserved List); (iii) shared powers exercised in consensus between the provincial government and the center (the Concurrent List). 1LO Subsequently reduced to eight when the North and East provinces were provisionally merged while awaiting the outcome of a still to be held referendum on their permanent union. - 72 - 214. Among the powers that appear in the Provincial Council List, the more important are education, health, and law and order. The Councils also received the power to supervise local government institutions (subject to parliamentary provisions), implement and supervise provincial housing and construction, develop and maintain minor roads and bridges (i.e., "C" class and below) and small irrigation projects, regulate road passenger services, agrarian services (except on interprovincial schemes, state lands and plantation), and corporations (except financial and trading enterprises), implement provincial plans, and to collect specific categories of taxes. 215. Three appendices to the Provincial Council List explain Council powers in law and order, land and land settlement and education in detail. Appendix 1 provides for a provincial police force while clearly maintaining the dominance of the national level and the president. In Appendix II, the center and president are given control over state land and interprovincial river, irrigation and land development projects. For example, before Provincial Councils can use any land, they must secure permission from the center. Appendix III provides the Councils with the power to manage preschools and all state schools not designated as national or special schools by the center. The designated schools cover all the most prestigious educational establishments and additional institutions can be designated at the will of the center. 216. Under the Reserved List, the center retains exclusive and broadly defined control over an array of subjects including defence and national security, foreign affairs, posts and tele- communications, judiciary and court structure, foreign trade, ports, harbors and aviation, national transport and other items of national importance or interest. The Reserved List also reserves the right to revenue gained from taxes on land, agriculture and agricultural income for the center. Administration and elections (including those to local authorities) fall in the Reserved List, giving the center the power to demarcate local administrative and electoral districts and divisions. 217. The Concurrent List includes the formulation of provincial plans and many of the details of provincial planning and implementation. Furthermore, the Concurrent List covers the acquisition of property, interprovincial irrigation schemes, soil conservation, electrification schemes, private health facilities, printed media and information, employment, tourism, trade and production of any product deemed by law to require the center's attention. 218. Structure of the Provincial Councils. The titular head of the Provincial Council is a Governor appointed by the national President as his representative. The Provincial Council itself consists of representatives chosen in a provincial election. According to rules laid down in the 13th Amendment, the Governor chooses a Chief Minister that can form a government from among the members of the majority party. The government consists of the Chief Minister and ministers for finance, planning, education, health, etc. 219. The Governor can exercise significant power over a provincial council government. For instance, he is responsible, as custodian, for the Provincial Fund, through which the Center makes most transfers to the Provincial Council. He also can exercise powers of appointment, dismissal and disciplinary control over all provincial cadres, either directly or through delegation to the Provincial Public Service Commission appointed by him (on the advice of the Chief Minister). The Governor also can report to the President on any failure of the Provincial Council administration and grant or withhold assent to Provincial Council statutes. 220. A Provincial Council may pass legislation (called statutes) relating to any matter set out in the Provincial Council List but the Governor can send the statute back to the Provincial Council for further consideration. In the event that the governor continues to withhold approval when a statute - 73 - returns to him a second time, he can send it to the President who would refer it to the Supreme Court. The court would then determine whether the statute is consistent with the Constitution and hence whether it becomes law. 221. The Provincial Councils are served by a cadre of civil servants, of every grade, transferred from the central services and line ministries to the provincial cadre. The provincial civil service is headed by the Chief Secretary appointed by the President. However, many provincial civil servants fall under the management of Divisional Secretaries. 222. The Provincial Councils and the Divisional Secretariats. The Divisional Secretariats cover about 300 Pradesiya Divisions, which were formerly called Districts." Each Division has a Divisional Secretary, formerly called an Assistant Government Agent, who oversees the village divisions headed by Village Officers (Gram Niladharis). The Divisional Secretaries were formerly appointed by the Provincial Councils. However, since 1992 the Divisional Secretaries have been appointed directly by the President. 223. Under the current organizational arrangement, the Divisional Secretary executes many programs of the Provincial Council at the local level. The most potentially powerful authority exerted by the Divisional Secretary is control over making payments on behalf of the Provincial Councils. Yet, because the Divisional Secretary also represents the center at the local level, he is often in a position to exert authority over the Provincial Council. This arrangement makes the Divisional Secretariats and the Provincial Councils parallel and potentially overlapping and competing local administrations. 224. Provincial Government Revenues and Expenditures. Provinces have a legal right to a share of national revenues and a large measure of autonomy in the use of resources. The Provinces collect turnover taxes, license fees, and various other fees within their jurisdiction, although their revenues only account for a fraction of their needs (except for Western Province which collects a substantial portion of its current expenditure). Therefore, the Central Government continues to finance most Provincial Council expenditures. Provincial budget requests are reviewed annually by a Finance Commission which reports to the President. However, the size and distribution of the transfer ultimately depends on the outcome of the debate on the budget in Parliament. 225. The Central Government transfers funds to the Provincial Councils through several means. It funds the recurrent expenditures of the Provincial Councils through block grants that pass through the Ministry of Provincial Councils and Home Affairs to Provincial Funds and on to the Provincial Councils. The central government line ministries transfer some funds for capital investment directly to the Provincial Councils. The capital transfers to the Provincial Councils consist of: (i) criteria-based grants; (ii) medium-term investment program (MTIP) grants; and (iii) rehabilitation of properties and industries and reconstruction of infrastructure. The criteria-based capital grants are transfers that are specifically related to that particular province's per-capita income and social and economic conditions, such that the relatively disadvantaged provinces received relatively high allocations in relation to their populations. The MTIP grants, are channelled to Provincial Councils via the relevant line ministries. These grants cover investments in areas defined by the Provincial Council List. The third and final category of capital transfers are related to the cost of reconstruction E' / The Government also established new local authorities for rural areas, designated Pradesiya Sabhas. These approximately 300 locally elected bodies are largely restricted in their scope and function to those of the former Village Councils. - 74 - judged necessary in a particular province following the JVP insurrection in 1988/89 and the ongoing conflict in the north and east. 226. Besides the amount administered by the Provincial Councils, the central government line ministries make capital expenditures for purposes within the constitutional domain of the Provincial councils through transfers to the Divisional Secretariats or by directly undertaking larger projects. These funds never enter into the revenues of the Provincial Councils. 227. The adequacy of monitoring of expenditures of the Provincial Councils is a major concern, especially since the Councils have responsibilities in the key health and education sub-sectors. The Councils are accountable for their expenditures to the National Parliament and to the Finance Commission, where the Treasury and the Ministry of Policy Planning are represented. Moreover, the Councils are accountable to the Governor and the central Secretary of Provincial Councils for funds spent from the Provincial Funds. The central Auditor-General's Office has oversight over financial management but audits are infrequent, focus on financial management rather than value-for-money, and carried out long after the end of the budget year. While, the Councils do make regular financial reports to the Treasury, the key breakdown is in physical monitoring of expenditures to assure that they are used for their intended purposes. The expenditure monitoring unit in the Ministry of Provincial Councils and Home Affairs is understaffed and thus unable to adequately spot-check activities in the provinces. 228. There are additional weaknesses in the financial reporting of the Councils. At the beginning of each month the Treasury releases an imprest amount to the Provincial Councils for their immediate cash needs. By the fifteenth of the month, each Provincial Council is required to send a statement regarding the expenditure of the previous month. This statement only gives expenditure under very broad heads, namely establishment costs, block grant recurrent expenditure, and criteria based grant capital expenditure.12 Often this information is not comprehensive and is simply based on the previous months figure. This is especially the case with the education and health sectors, where there are a large number of paying out units. Several Provincial Councils have failed to furnish quarterly (and sometimes even annual) retums. Where information has been furnished, especially in the case of criteria based capital expenditure, it is often considered unreliable. For example, in the case of construction items, there has frequently been no physical verification of the investment. 229. Furthermore, there are effective limits on the Finance Commission's oversight of the Provincial councils. Once a lump sum grant has been allocated in the Budget and voted by parliament, the Finance Commission has no authority to modify this in any way. In the event of a lump sum grant being inadequate to meet the needs of a Provincial Council, supplementary funds can be voted for it by parliament, and the Finance Commission does not play any role in this process. The current basis of apportionment among the provinces does not seem to reflect the provisions of the 13th Amendment, which requires monies to be distributed according to population, per capita income and need. In addition, the current practice of reducing block grants according to the estimated yields of local revenues, further limits the financial autonomy of Provincial Councils and acts to remove the incentives for local resource generation. 2/ More detailed statements with a breakdown of recurrent expenditure into: (i) personal emoluments; (ii) travelling expenses; (iii) supplies and requisites; (iv) repair and maintenance of capital assets; (v) transportation, communication and other services; and (vi) grants, contributions and subsidies appear in quarterly retums. - 75 - 230. Table V.5 provides a consolidated financial statement of the Provincial Councils and shows the dependence of the councils on transfers by the Central Government to finance their expenditures. The Councils collect a relatively small revenue from taxes, despite the number of taxes devolved to them. The Western Province, which includes Colombo, has the largest expenditure program and mobilizes the largest provincial revenues. The table reveals that the Councils do not absorb the available resources, reflecting a combination of late disbursement and organizational difficulties in implementing programs. Table V.5: FINANCIAL OPERATIONS OF THE PROVINCIAL COUNCILS (Rs. Billions) 1990 1991 1992 1993 Actuals Actuals Actuals Estimate Revenue Collection .. 2.1 3.3 3.0 Of which: Western Province .. 1.3 2.2 1.9 Block Grants Recurrent 13.4 9.6 9.0 10.5 Capital 3.8 4.3 3.2 3.0 Total available resources 17.2 15.9 15.5 16.5 Current Expenditures 12.5 10.7 10.7 13.5 Of which: Western Province 2.5 2.3 2.4 2.6 Capital Expenditure 2.8 2.5 1.8 N.A. Total Expenditure 15.4 13.2 12.6 N.A. % Use of Resources 89.2% 82.9% 81.0% N.A. Source: Finance Commission 231. The limited information that is available suggests that devolution has been unnecessarily costly. Provincial Council staffs are about equal to the central government civilian staff. However, the Councils spend 1/8 to 1/10th of the central government expenditure and the disbursements are mostly on wages and salaries. With the exception of staff in the educational sector, it would appear that Provincial Council staffs were recruited in anticipation of the transfer of programs, which has not yet fully materialized. 232. The increase in spending on labor costs appears to be crowding out critical expenditures on supplies and maintenance. For example, the budget of the Ministry of Education of the Central Provincial Council shows a rapid increase in labor costs that appears to be constraining expenditures on supplies and maintenance (Table V.6). In the Province's Department of Health, labor costs are growing rapidly because of general wage increases and overtime payments to doctors and nurses while expenditures on medicines and other supplies remain very low and stable. Expenditures on medicines were nil in 1993 as the Central Government transferred medicine in kind to the Central Provincial Council and the Council withheld payment. These are examples of a broader potential problem inherent in funding formerly centralized activities through transfers of block grants: there is no way of assuring key supplies and maintenance are funded so that services are maintained. - 76 - Table V.6: RECURRENT EXPENDITURES OF THE CENTRAL PROVINCIAL COUNCIL: MINISTRY OF EDUCATION (Millions of Rupees) 1992 prov. 1993 est. 1994 budget Recurrent exp. 861 965 1,179 Labor costs 808 948 1,064 Supplies 8 4 4 Maintenance 21 8 72 Other 24 5 39 Source: Central Provincial Council Table V.7: RECURRENT EXPENDITURES OF THE CENTRAL PROVINCIAL COUNCIL: MINISTRY OF HEALTH (millions of rupees) 1992 prov. 1993 est. 1994 budget Recurrent exp. 304 298 482 Labor costs 190 237 319 Medicines 43 0 47 Other supplies 40 40 14 Maintenance 4 2 4 Other 27 19 98 Source Central Provincial Council. Reform of the Provincial Councils. 233. Thus, the relationship between the Government and the Provincial Councils have basic weaknesses that need to be addressed: (i) the provinces have achieved little fiscal autonomy because they depend on large block grants from the Central Government; (ii) the Councils and District Secretariats are potentially competing local administrations; and (iii) the Central Government has a limited ability to monitor policy implementation through spending programs. Reform of the current arrangement dealing with the Provincial Councils and the Divisional Secretariats is urgent to assure that local administration is effective in delivering public goods and services such as education and health, and to contain costs. Moreover, reform of the Councils is also a necessary counterpart to reform of the central government civil service. 234. The current constitutional arrangement points to a goal of establishing fiscally independent and managerially able Provincial Councils. Whatever system Sri Lanka may ultimately choose, it should establish a single system of regional government to enhance administrative efficiency and minimize fiscal cost: there is no need to retain both the Provincial Councils and the Divisional Secretariats. If the Government wishes to research alternative forms of government structure and - 77 - service delivery mechanisms, an IDA-financed or administered study to enhance service delivery might bring to light additional options for local government financing and operation. In the meantime, the Central Government should improve both financial and physical monitoring of expenditures of funds that it transfers to the Provincial Councils. - 78 - STATISTICAL ANNEX OF TABLES - 79 - LIST OF ANNEX TABLES AND BOXES TABLE NOS. TITLES Table 1: Summary of Budgetary Operations, 1981-1993 Table 2: Economic Classification of Expenditure and Net Lending (In Million of Rupees) Table 3: Economic Classification of Expenditure and Net Lending (Percentage Point Shares in GDP) Table 4: Functional Classification of Total Expenditure and Lending (In Million of Rupees) Table 5: Functional Classification of Total Expenditure and Lending (Percentage Point Shares in GDP) Table 6: Functional Classification of Current Expenditure Table 7: Functional Classification of Capital Expenditure Table 8: Sri Lanka: Identified Public Investment by Sector Table 9: The CEB's Investment Program, 1993-1996 Table 10: LECO's Investment Program, 1993-1996 Table 11: Overview of Evaluation of Troubled Investments Table 12: Overview of Safety-Net Programs TABLE 1: SUMMARY OF BUDGETARY OPERATIONS, 1981 TO 1993 (In millions of rupees) Prov Budget 1980 1981 1982 1983 1984 1985 1966 1967 1988 1988 1990 1991 1992 1993 Total Revenue and Grants 15641 17496 19588 26790 37354 39558 40991 46822 48337 60386 74661 84049 94398 110927 Total Revenue 13022 14775 16210 23317 34061 36249 37238 42145 41749 53979 67964 76179 86118 101977 Tax Revenue 12158 13890 14737 19912 29939 30442 31272 35119 35946 47513 61206 68157 76692 94377 Non-tax Revenue 884 1079 1473 3405 4122 5807 5966 7026 5803 6466 6758 8022 9425 7600 Grants 2619 2721 3376 3473 3293 3307 3753 4677 6588 6407 6697 7870 8280 8950 Expenditure and Lending minus Repayments 28388 28014 33512 39637 47837 65234 59194 63894 76532 82164 99813 119527 117627 141916 Current 12319 14649 18339 21999 24630 32645 33967 39560 46132 56884 71770 83756 88726 97437 Capital 12123 11252 15431 15866 19915 21530 23236 22816 22878 20750 19161 25968 29690 37270 Lending minus Repayments 3946 2113 -258 1772 3292 1059 1991 1518 7522 4530 8882 9803 -789 7209 Current Account Surplus/Deficit (-) 703 126 -2131 1318 9431 3604 3272 2585 -4383 -2905 -3806 -7577 -2608 4540 Budget Deficit (before grants) -15366 -13239 -17302 -16320 -13776 -18985 -21956 -21750 -34783 -28185 -31849 -43348 -31509 -39939 Budget Deficit (after grants) -12747 -10518 -13926 -12847 -10483 -15678 -18204 -17072 -28195 -21778 -25152 -35478 -23229 -30989 Financing 12746 10518 13926 12846 10483 15678 18204 17072 28193 21776 25152 35477 23229 30989 Foreign Borrowing 3516 4880 4744 6312 6492 7109 9061 5716 7128 5926 11644 19329 7984 19489 Domestic Borrowing 9230 5638 9182 6534 3991 8569 9143 11356 21065 12372 16986 16148 15246 11500 Non-Bank 2201 1791 6506 6081 6086 3858 6096 7815 10912 15659 16728 16114 -2321 1300 Bank 7029 3847 3676 473 -2095 4711 3047 3541 10153 -3287 258 34 17567 -1500 0o Arrears 0 0 0 0 0 0 0 0 0 3478 -3478 0 0 0 As a % of GDP Total Revenue and Grants 23.5 206 197 220 243 244 22.8 23 8 21.8 24.0 23.2 226 22.3 22.4 Total Revenue 109 17.4 16.3 19.2 22.2 22.3 20.7 21.4 18.8 21.4 21.1 20.4 20.3 20.6 TaxRevenue 183 161 149 16.4 19.5 187 17.4 17.9 16.2 18.9 19.0 18.3 90 191 Non-tax Revenue 1.3 1 3 1.5 2.8 2.7 36 3.3 3.6 2.6 2.6 2.1 2.2 22 1 5 Grants 3.9 32 3.4 2.9 2.1 20 2.1 2.4 3.0 25 2.1 2.1 2.0 1.8 Expenditure and Lending minus Repayments 42.7 330 338 32.6 31.1 340 33.0 32.5 34.5 32.6 31.0 32.1 27.8 28.7 Current 18.5 172 18.6 18.1 16.0 20.1 18.9 20.1 20.8 22.6 22 3 22 5 21.0 19.7 Capital 18.2 13.2 15.5 13.0 13.0 133 12.9 11.6 10.3 8.2 60 7.0 7.0 7.5 Lending minus Repayments 5.9 25 -0.3 1.5 2.1 0.7 1.1 0.8 3.4 1.8 28 20 -0.2 1.5 Current Account Surplus/Deficit(-) 1.1 0.1 -2.1 1.1 6.1 2.2 1.8 1.3 -2.0 -1.2 -1 2 -20 -0.6 09 BudgetDeficit(before grants) -23.1 -15.6 -17.4 -13.4 -9.0 -11.7 -12.2 -11.1 -15.7 -11.2 -99 -11.6 -7.4 -81 Budget Deficit (after grants) -19.2 -12.4 -14.0 -10.6 -6.8 -9.7 -10.1 -8.7 -12.7 -8.6 -7 8 -95 -5 5 -6.3 A Financing 192 124 14.0 10.6 68 9.7 10.1 8.7 127 8.6 78 95 5.5 63 Foreign Borrowing 5.3 57 4.8 52 4.2 4.4 5.0 2.9 3.2 2.4 36 52 19 39 0 Domestic Financing 139 66 9.3 5.4 2.6 5.3 5.1 58 9.5 4.9 53 43 36 2.3 Non-Bank 3.3 2.1 55 5.0 4.0 2.4 3.4 4.0 4.9 6.2 52 43 -0.5 03 Bank 10.6 4.5 37 0.4 -1.4 29 1.7 1.8 4.6 -13 0 1 0.0 4.2 -03 Source- Central Bank and World Bank TABLE 2: ECONOMIC CLASSIFICATION OF EXPENDITURE AND NET LENDING (in millions of rupees) Ptro9 Budget 1980 1981 1982 198S 1984 1985 1986 198 1988 1989 1990 1991 1992 1993 Current Expenditure 12318 14647 18339 21999 24631 32645 33966 39559 46132 56884 71771 83755 88726 97436 Operational Expenditure 5018 5221 6498 7691 9195 16288 15154 18466 20654 25508 30167 35832 39135 42894 Salaries and Wages 3301 3579 4561 4811 5553 6878 8027 8006 10016 14446 15748 17985 20772 22995 Other Goods and Services 1717 1642 1937 2880 3642 9410 7127 10460 10638 11062 14419 17847 18363 19899 Interest 2241 3738 5104 6606 6738 7428 8762 10157 12590 14352 20668 22073 25283 30921 Foreign 413 713 915 1270 1623 1970 2209 2564 2896 3337 3678 4113 4686 4602 Domestic 1828 3025 4189 5336 5115 5458 6553 7593 9694 11015 16990 17960 20597 26319 Current Transfers 5059 5688 6737 7702 8698 8929 10050 10936 12888 17024 20936 25850 24308 25636 Transfers to Public Corporations 399 371 982 1452 1221 504 1786 1169 1569 2283 2639 1759 1722 1468 olw Transfers to Railways 146 194 308 453 629 72 408 348 399 425 276 616 341 423 Transfers to Postal 68 0 79 40 52 58 6 110 226 245 156 292 590 316 00 Transfers to St. Bodies & Inst. 226 187 251 316 542 933 1032 910 1377 1696 1811 1405 1479 1912 - Transfers to Local Governments 160 182 317 388 444 501 600 609 918 902 1237 1050 1103 1123 Transfers Abroad 17 26 28 106 58 29 60 117 89 97 192 166 148 196 Transfers to Private Institutions 199 67 125 170 106 0 3 0 4 4 125 1227 300 17 Transfers to Households 4058 4855 5034 5270 6327 6962 6569 8131 8931 12042 14932 20243 16738 20920 Food and Kerosene Stamps 1664 1685 1646 1710 1801 1728 1799 1900 2069 3634 3812 2999 3047 2900 Mid-DayMeal 28 39 48 40 38 31 25 61 147 653 1644 1402 1899 1700 FreeTextbooks 61 57 67 73 62 154 102 108 314 135 134 146 175 275 NSB Subsidy 257 388 401 500 498 495 199 174 77 110 256 0 0 130 JanaSaviya 0 0 0 0 0 0 0 0 0 705 2912 3407 3068 5150 Pensions 718 903 1416 1813 2180 2728 2984 3242 4128 4735 4490 8832 8041 9148 Sugar Production Subsidy 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Fertilizer Subsidy 657 1070 893 642 1006 788 614 511 600 796 0 0 0 0 Other 673 713 563 492 742 1038 846 2135 1596 1274 1684 3457 504 1617 Under Expenditure Provision 0 0 0 0 0 0 0 0 0 0 0 0 0 -2015 Source: Central Bank o TABLE 2: ECONOMIC CLASSIFICATION OF EXPENDITURE AND NET LENDING (CONTINUED) (in millions of rupees) Prov Budget 18o 1961 19M 198 1$84 1985 1985 1987 low: 1989 I99 19al 1992. 1993 Capital Expenditure 12123 11255 15432 15866 19915 21530 23236 22817 22879 20750 19161 25968 29690 37270 Capital Expenditure by Ministries& Dept 4358 3174 4224 4830 4371 5520 5620 8194 8710 10460 10311 13101 14257 16792 Capital Transfers 7765 8081 11208 11036 15544 16010 17616 14623 14169 10290 8850 12867 15433 20478 Transfers to Public Corporations 3211 3409 3541 4140 6618 5257 7505 5970 6019 4108 2602 4669 6062 11462 o/w Railway 503 460 377 365 392 969 1046 1062 1755 1431 765 1746 1146 3640 Postal 18 16 11 11 11 20 18 18 16 12 21 41 61 430 Telecom 400 277 348 470 1075 444 696 1338 745 1166 567 0 0 0 Janatha Estate Development Board 179 30 173 75 68 111 138 54 4 3 0 0 0 35 State Plantations Corporation 118 80 242 148 436 98 72 2 0 0 0 0 0 35 SriLanka SugarCorporation 8 47 75 123 410 508 131 185 131 10 35 0 0 0 Air Lanka 0 1 300 900 935 800 2404 892 809 0 0 0 0 0 Ceylon Shipping Corporation 0 0 0 156 200 275 274 54 320 0 0 0 0 0 Sri Lanka Transport Board 6 1 0 0 0 0 0 0 1 238 196 0 0 0 Sri Lanka Port Authority 0 0 17 3 213 307 597 230 0 0 0 0 0 0 Ceylon Electricity Board 117 673 247 111 1016 153 109 46 704 7 0 20 233 1528 Nat'l. Water Supply & Drainage Bd. 0 343 316 967 854 966 994 1233 922 754 658 1579 2769 2977 Nat'l. Housing Dev. Authority 976 1020 703 571 480 360 465 374 380 352 242 323 289 267 Transfers to St. Bodies & Inst. 3795 4518 7660 6830 8541 10034 9537 8005 6962 5354 5296 6948 6959 8265 o/w Mahaweli Dev. Authority 3381 3816 7142 6096 7057 9082 6652 5578 3430 2475 1810 2410 2293 276 Road Development Authority 0 0 0 1 346 679 876 773 650 1204 1870 2373 2898 2999 Airport and Aviation Authority 0 0 0 100 336 955 1423 910 1617 917 110 150 21 61 Transfers to Other Levels of Govt. 131 150 168 317 198 481 473 571 1123 799 843 950 708 708 TransfersAbroadtoHHs&Inst. 628 4 -161 -251 187 238 101 77 65 29 109 300 1704 43 Less: Prov. for Underexpenditure 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Lending Minus Repayment 3946 2113 -258 1772 3292 1059 1991 1518 7522 4530 8882 9803 -789 7209 AdvanceAccounts 3240 1716 -879 1120 2917 314 -170 -1180 2732 -993 1699 1819 -150 500 On-Lending 807 569 718 801 901 997 2669 3159 5093 5901 7587 6622 4337 8274 o/w Ceylon Electricity Board 449 195 165 433 507 371 1371 1919 3628 3775 5410 4010 2109 4540 Port Authority 0 0 0 0 0 0 0 162 0 630 98 925 848 2124 0 Repayment on On-Lending -101 -172 -97 -149 -526 -252 -508 -461 -303 -378 -404 -1132 -2511 -1665 UQ Restructuring Cost 0 0 0 0 0 0 0 0 0 0 0 3335 750 600 > Privatisation Proceeds 0 0 0 0 0 0 0 0 0 0 0 -841 -3115 -500 0 Total Expend.&Net Lending 28387 28015 33511 39636 47838 55234 59193 63893 76532 82164 99814 119526 117627 141916 Source: Central Bank TABLE 3: ECONOMIC CLASSIFICATION OF EXPENDITURE AND NET LENDING (percentage point shares in GDP) Prov Budget 1980 1981 1982 198 1984 1985 1986 198 1988 1989 1990 1991 1992 1993 Current Expenditure 18.5 17.2 18.5 18.1 16.0 20.1 18.9 20.1 20.8 22.6 22.3 22.5 21.0 19.7 Operational Expenditure 7.5 6.1 6.5 6.3 6.0 10.0 8.4 9.4. 9.3 10.1 9.4 9.6 9.2 8.7 Salaries and Wages 5.0 4.2 4.6 4.0 3.6 4.2 4.5 4.1 4.5 5.7 4.9 4.8 4.9 4.6 Other Goods and Services 2.6 1.9 2.0 2.4 2.4 5.8 4.0 5.3 4.8 4.4 4.5 4.8 4.3 4.0 Interest 3.4 4.4 5.1 5.4 4.4 4.6 4.9 5.2 5.7 5.7 6.4 5.9 6.0 6.2 Foreign 0.6 0.8 0.9 1.0 1.1 1.2 1.2 1.3 1.3 1.3 1.1 1.1 1.1 0.9 Domestic 2.7 3.6 4.2 4.4 3.3 3.4 3.7 3.9 4.4 4.4 5.3 4.8 4.9 5.3 Current Transfers 7.6 6.7 6.8 6.3 5.7 5.5 5.6 5.6 5.8 6.8 6.5 6.9 5.7 5.2 Transfers to Public Corporations 0.6 0.4 1.0 1.2 0.8 0.3 1.0 0.6 0.7 0.9 0.8 0.5 0.4 0.3 o/w Transfers to Railways 0.2 0.2 0.3 0.4 0.4 0.0 0.2 0.2 0.2 0.2 0.1 0.2 0.1 0.1 Transfers to Postal 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.0 0.1 0.1 0.1 00 Transfers to St. Bodies & Inst. 0.3 0.2 0.3 0.3 0.4 0.6 0.6 0.5 0.6 0.7 0.6 0.4 0.3 0.4 Transfers to Local Governments 0.2 0.2 0.3 0.3 0.3 0.3 0.3 0.3 0.4 0.4 0.4 0.3 0.3 0.2 Transfers Abroad 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.1 0.0 0.0 0.1 0.0 0.0 0.0 Transfers to Private Institutions 0.3 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.1 0.0 Transfers to Households 6.1 5.7 5.1 4.3 4.1 4.3 3.7 4.1 4.0 4.8 4.6 5.4 4.0 4.2 Food and Kerosene Stamps 2.5 2.0 1.7 1.4 1.2 1.1 1.0 1.0 0.9 1.4 1.2 0.8 0.7 0.6 Mid-Day Meal 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.3 0.5 0.4 0.4 0.4 Free Textbooks 0.1 0.1 0.1 0.1 0.0 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.3 NSB Subsidy 0.4 0.5 0.4 0.4 0.3 0.3 0.1 0.1 0.0 0.0 0.1 0.0 0.0 0.0 Jana Saviya 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.9 0.9 0.7 1.0 Pensions 1.1 1.1 1.4 1.5 1.4 1.7 1.7 1.6 1.9 1.9 1.4 2.4 1.9 1.8 Sugar Production Subsidy 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Fertilizer Subsidy 1.0 1.3 0.9 0.5 0.7 0.5 0.3 0.3 0.3 0.3 0.0 0.0 0.0 0.0 Other 1.0 0.8 0.6 0.4 0.5 0.6 0.5 1.1 0.7 0.5 0.5 0.9 0.1 0.3 Under Expenditure Provision 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 -0.4 Source: Central Bank o TABLE 3: ECONOMIC CLASSIFICATION OF EXPENDITURE AND NET LENDING (CONTINUED) (percentage point shares in GDP) Provt Budget q1980 191 982 198a 84 l98? 98 1ast 98 1ow lo0 1991 1992 199 Capital Expenditure .18.2 13.2 15.6 13.0 13.0 13.3 12.9 11.6 10.3 8.2 6.0 7.0 7.0 7.5 Capital Expenditure by Ministries & Dept 6.6 3.7 4.3 4.0 2.8 3.4 3.1 4.2 3.9 4.2 3.2 3.5 3.4 3.4 Capital Transfers 11.7 9.5 11.3 9.1 10.1 9.9 9.8 7.4 6.4 4.1 2.8 3.5 3.6 4.1 Transfers to Public Corporations 4.8 4.0 3.6 3.4 4.3 3.2 4.2 3.0 2.7 1.6 0.8 1.3 1.4 2.3 olw Railway 0.8 0.5 0.4 0.3 0.3 0.6 0.6 0.5 0.8 0.6 0.2 0.5 0.3 0.7 Postal 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 Telecom 0.6 0.3 0.4 0.4 0.7 0.3 0.4 0.7 0.3 0.5 0.2 0.0 0.0 0.0 Janatha Estate Development Board 0.3 0.0 0.2 0.1 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 State Plantations Corporation 0.2 0.1 0.2 0.1 0.3 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sri Lanka Sugar Corporation 0.0 0.1 0.1 0.1 0.3 0.3 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 Air Lanka 0.0 0.0 0.3 0.7 0.6 0.5 1.3 0.5 0.4 0.0 0.0 0.0 0.0 0.0 Ceylon Shipping Corporation 0.0 0.0 0.0 0.1 0.1 0.2 0.2 0.0 0.1 0.0 0.0 0.0 0.0 0.0 Sri Lanka Transport Board 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.0 Sri Lanka Port Authority 0.0 0.0 0.0 0.0 0.1 0.2 0.3 0.1 0.0 0.0 0.0 0.0 0.0 0.0 Ceylon Electricity Board 0.2 0.8 0.2 0.1 0.7 0.1 0.1 0.0 0.3 0.0 0.0 0.0 0.1 0.3 Nat'l. Water Supply & Drainage Bd. 0.0 0.4 0.3 0.8 0.6 0.6 0.6 0.6 0.4 0.3 0.2 0.4 0.7 0.6 Nat'I. Housing Dev. Authority 1.5 1.2 0.7 0.5 0.3 0.2 0.3 0.2 0.2 0.1 0.1 0.1 0.1 0.1 Transfers to St. Bodies & Inst. 5.7 5.3 7.7 5.6 5.6 6.2 5.3 4.1 3.1 2.1 1.6 1.9 1.6 1.7 o/w Mahaweli Dev. Authority 5.1 4.5 7.2 5.0 4.6 5.6 3.7 2.8 1.5 1.0 0.6 0.6 0.5 0.1 Road Development Authority 0.0 0.0 0.0 0.0 0.2 0.4 0.5 0.4 0.3 0.5 0.6 0.6 0.7 0.6 Airport and Aviation Authority 0.0 0.0 0.0 0.1 0.2 0.6 0.8 0.5 0.7 0.4 0.0 0.0 0.0 0.0 Transfers to Other Levels of Govt. 0.2 0.2 0.2 0.3 0.1 0.3 0.3 0.3 0.5 0.3 0.3 0.3 0.2 0.1 Transfers Abroad to HHs & Inst. 0.9 0.0 -0.2 -0.2 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.1 0.4 0.0 Less: Prov. for Underexpenditure 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Lending Minus Repayment 5.9 2.5 -0.3 1.5 2.1 0.7 1.1 0.8 3.4 1.8 2.8 2.6 -0.2 1.5 Advance Accounts 4.9 2.0 -0.9 0.9 1.9 0.2 -0.1 -0.6 1.2 -0.4 0.5 0.5 0.0 0.1 On-Lending 1.2 0.7 0.7 0.7 0.6 0.6 1.5 1.6 2.3 2.3 2.4 1.8 1.0 1.7 olw Ceylon Electricity Board 0.7 0.2 0.2 0.4 0.3 0.2 0.8 1.0 1.6 1.5 1.7 1.1 0.5 0.9 Port Authority 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.3 0.0 0.2 0.2 0.4 Repayment on On-Lending -0.2 -0.2 -0.1 -0.1 -0.3 -0.2 -0.3 -0.2 -0.1 -0.2 -0.1 -0.3 -0.6 -0.3 Un Restructuring Cost 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.9 0.2 0.1 > Privatisation Proceeds 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 -0.2 -0.7 -0.1 o Total Expend. & Net Lending 42.7 33.0 33.8 32.6 31.1 34.0 33.0 32.5 34.5 32.6 31.0 32.1 27.8 28.7 Source: Central Bank TABLE 4: FUNCTIONAL CLASSIFICATION OF TOTAL EXPENDITURE AND LENDING (in millions of rupees) Prov Budget 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 General Services 2753 2987 4104 4227 5158 9807 9897 12789 16196 16593 21560 22233 26286 27902 Civil Administration 1902 2070 3140 2631 3038 3954 3748 3813 8278 8783 8221 7292 8992 10746 Defense, Public Order & Saf 851 917 964 1596 2120 5853 6149 8976 7918 7810 13339 14941 17294 17156 (olw Public Order & Safety) 393 438 478 617 845 1239 1798 2975 3186 3737 6603 4624 5158 4844 Social Services 7432 7565 8394 9793 10320 12278 13671 15864 18791 23643 27468 34558 35375 42539 Education 1935 2117 2599 2944 3209 4581 5115 5186 6385 8141 9571 9129 9664 13782 Health 1358 1020 1163 2087 1796 2091 2257 3380 3931 4639 4964 5229 6591 7589 Welfare 2929 3136 3602 3881 4538 4900 5269 6151 7620 9760 11998 19281 17807 19134 Housing 1056 1144 869 696 583 472 588 518 594 804 513 535 990 854 Community Services 154 148 161 185 194 234 442 629 261 299 422 384 322 1180 Economic Services 10976 11448 16218 16865 21563 22339 26290 24235 25782 24233 24296 29893 23745 38106 Agriculture & Irrigation 5704 6333 10314 9396 10911 8937 8968 9212 7938 6210 6197 6975 5940 8055 (o/w Mahaweli) 3381 3816 7333 6395 7057 7233 5952 5101 3430 2475 1810 2410 2293 2676 Fisheries 280 173 188 185 190 139 261 334 180 330 179 245 146 1092 Manufacturing & Mining 816 520 1001 633 1191 1364 1236 1303 527 879 301 402 332 415 Energy & Water Supply 568 1210 1005 2084 2426 1499 2597 3334 5544 5135 6845 6926 5805 9468 Transport & Communicatio 2243 1847 2741 3529 5571 5397 9316 6833 7877 7804 6912 12118 7895 11740 Trade & Commerce 187 225 303 329 211 347 724 580 433 1000 364 371 499 525 Other 1178 1140 666 709 1063 4656 3188 2639 3283 2875 3498 2856 3129 6811 Other 4089 4472 5775 7782 8404 10749 10015 12650 13337 19068 25193 29662 35600 36449 (o/w Interest Payments) 2241 3738 5104 6606 6738 7428 8762 10157 12590 14352 20668 22073 25283 30921 ( Unallocable) - - - - - - - - - - - - - - UnderExpenditure 0 0 0 0 0 0 0 0 0 0 0 0 0 -2105 Total Expenditure 25250 26472 34491 38667 45445 55173 59873 65538 74106 83537 98517 116346 121006 142891 Source: Central Bank Z TABLE 5: FUNCTIONAL CLASSIFICATION OF TOTAL EXPENDITURE AND LENDING (percentage point shares in GDP) 190 19$1 1 982 19$3 1984 1985 108s 1987 198M 1989 190 1001 1902 10ge General Services 4.1 3.5 4.1 3.5 3.4 6.0 5.5 6.5 7.3 6.6 6.7 6.0 6.2 5.6 Civil Administration 2.9 2.4 3.2 2.2 2.0 2.4 2.1 1.9 3.7 3.5 2.6 2.0 2.1 2.2 Defense, Public Order & Saf 1.3 1.1 1.0 1.3 1.4 3.6 3.4 4.6 3.6 3.1 4.1 4.0 4.1 3.5 (olw Public Order & Safety) 0.6 0.5 0.5 0.5 0.5 0.8 1.0 1.5 1.4 1.5 2.1 1.2 1.2 1.0 Social Services 11.2 8.9 8.5 8.1 6.7 7.6 7.6 8.1 8.5 9.4 8.5 9.3 8.4 8.6 Education 2.9 2.5 2.6 2.4 2.1 2.8 2.8 2.6 2.9 3.2 3.0 2.5 2.3 2.8 Health 2.0 1.2 1.2 1.7 1.2 1.3 1.3 1.7 1.8 1.8 1.5 1.4 1.6 1.5 Welfare 4.4 3.7 3.6 3.2 3.0 3.0 2.9 3.1 3.4 3.9 3.7 5.2 4.2 3.9 Housing 1.6 1.3 0.9 0.6 0.4 0.3 0.3 0.3 0.3 0.3 0.2 0.1 0.2 0.2 Community Services 0.2 0.2 0.2 0.2 0.1 0.1 0.2 0.3 0.1 0.1 0.1 0.1 0.1 0.2 Economic Services 16.5 13.5 16.3 13.9 14.0 13.8 14.6 12.3 11.6 9.6 7.6 8.0 5.6 7.7 Agriculture & Irrigation 8.6 7.5 10.4 7.7 7.1 5.5 5.0 4.7 3.6 2.5 1.9 1.9 1.4 1.6 (olw Mahaweli) 5.1 4.5 7.4 5.3 4.6 4.5 3.3 2.6 1.5 1.0 0.6 0.6 0.5 0.5 Fisheries 0.4 0.2 0.2 0.2 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.0 0.2 Manufacturing & Mining 1.2 0.6 1.0 0.5 0.8 0.8 0.7 0.7 0.2 0.3 0.1 0.1 0.1 0.1 Energy & Water Supply 0.9 1.4 1.0 1.7 1.6 0.9 1.4 1.7 2.5 2.0 2.1 1.9 1.4 1.9 Transport & Communicatio 3.4 2.2 2.8 2.9 3.6 3.3 5.2 3.5 3.5 3.1 2.1 3.3 1.9 2.4 Trade & Commerce 0.3 0.3 0.3 0.3 0.1 0.2 0.4 0.3 0.2 0.4 0.1 0.1 0.1 0.1 Other .1.8 1.3 0.7 0.6 0.7 2.9 1.8 1.3 1.5 1.1 1.1 0.8 0.7 1.4 Other 6.1 5.3 5.8 6.4 5.5 6.6 5.6 6.4 6.0 7.6 7.8 8.0 8.4 7.4 (o/w Interest Payments) 3.4 4.4 5.1 5.4 4.4 4.6 4.9 5.2 5.7 5.7 6.4 5.9 6.0 6.2 ( Unallocable) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Under Expenditure 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 -0.4 Total Expenditure 38.0 31.1 34.8 31.8 29.6 34.0 33.4 33.3 33.4 33.2 30.6 31.2 28.6 28.9 Source: Central Bank o TABLE 6: FUNCTIONAL CLASSIFICATION OF CURRENT EXPENDITURE (in millions of rupees) Prov Si*cet lose Iasi 1982 1983 t1es less 19w 1987 1988 1989 Imen 1991 tesw 1993 General Public Services 2126 2648 3064 3510 4257 8768 8680 11113 13424 14090 17943 19811 23213 24359 Civil Administration 1329 1613 2163 1971 2288 3095 2884 2943 6145 6781 6622 5612 6618 7942 Defence 458 479 486 979 1275 4614 4351 6001 4732 4073 6736 10317 12136 12312 Public Order and Safety 339 376 415 560 694 1059 1445 2169 2547 3236 4585 3882 4459 4105 Social Services 5207 5766 6789 7739 8897 10437 11075 12927 15593 19976 24449 31594 29003 34809 Education 1477 1695 2127 2437 2672 3636 3834 4279 5371 6612 8529 7951 7657 10725 Health 756 879 989 1310 1580 1773 1852 2401 2427 3381 3685 4110 4571 5318 Welfare 2904 3117 3576 3871 4518 4889 5254 6098 7607 9750 11973 19266 16620 17786 Housing 7 7 9 11 12 13 13 13 15 17 0 0 0 0 CommunityServices 63 68 88 110 115 126 122 136 173 216 262 267 154 980 00 Economic Services 1656 2095 2719 3034 3457 2994 4363 3148 4190 4580 5137 3940 1661 4879 Agric.&Irrigation 965 1392 1398 1144 1511 1322 1340 1090 1412 976 982 912 579 1332 Fisheries 13 19 41 23 24 31 38 40 40 47 46 98 51 806 Manufacturing and Mining 212 134 189 197 158 255 443 141 197 408 181 231 188 288 Energy and Water Supply 2 0 276 572 49 7 50 73 83 416 345 659 133 289 Transport & Communication 306 303 608 780 1342 592 1668 1200 1613 1138 2600 1308 129 1355 Trade & Commerce 46 120 67 102 87 191 279 271 313 950 159 153 136 159 Other 112 127 140 216 286 596 545 333 532 645 824 579 446 650 Other 3330 4318 5768 7716 8019 10447 9849 12373 12927 18240 24241 28412 34849 35406 O/W Interest 2241 3738 5104 6606 6738 7428 8762 10157 12590 14352 20668 22073 25283 30921 Under Expenditure Provision 0 0 0 0 0 0 0 0 0 0 0 0 0 -2015 Total 12319 14827 18340 21999 24630 32646 33967 39561 46134 56886 71770 83757 88726 97438 Source: Central Bank o TABLE 7: FUNCTIONAL CLASSIFICATION OF CAPITAL EXPENDITURE (in millions of rupees) .............___....._.. __._.._..1980 1981 1982 1983 1984 1985 1986 198? 1I988 1989 1990 1991 1992 1993 General Public Services 627 519 1040 717 901 1039 1217 1676 2772 2503 3617 2422 3073 3543 Civil Administration 573 457 977 660 750 859 864 870 2133 2002 1599 1680 2374 2804 Public Order and Safety 54 62 63 57 151 180 353 806 639 501 2018 742 699 739 Social Services 2225 1799 1605 2054 1423 1841 2596 2937 3198 3667 3019 2964 6372 7730 Education 458 422 472 507 537 945 1281 907 1014 1529 1042 1178 2007 3057 Health 602 141 174 777 216 318 405 979 1504 1258 1279 1119 2020 2271 Welfare 25 19 26 10 20 11 15 53 13 10 25 15 1187 1348 Housing 1049 1137 860 685 571 459 575 505 579 787 513 535 990 854 CommunityServices 91 80 73 75 79 108 320 493 88 83 160 117 168 200 00 Economic Services 9320 9353 13499 13831 18106 19345 21927 21087 21592 19653 19159 25953 22084 33227 00 Agric. & Irrigation 4739 4941 8916 8252 9400 7615 7628 8122 6526 5234 5215 6063 5361 6723 o/wMahaweli Project 3381 3816 7333 6395 7057 7233 5952 5101 3430 2475 1810 2410 2293 2676 Fisheries 267 154 147 162 166 108 223 294 140 283 133 147 95 286 Manufacturing and mining 604 386 812 436 1033 1109 793 1162 330 471 120 171 144 127 Energy and water supply 566 1210 729 1512 2377 1492 2547 3261 5461 4719 6500 6267 5672 9179 Transport & Communucation 1937 1544 2133 2749 4229 4805 7648 5633 6264 6666 4312 10810 7766 10385 Trade & Commerce 141 105 236 227 124 156 445 309 120 50 205 218 363 366 Other 1066 1013 526 493 777 4060 2643 2306 2751 2230 2674 2277 2683 6161 Other 759 154 7 66 385 302 166 277 410 828 952 1250 751 1043 Under Expenditure Provision 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total Cap. Expand Lending 12931 11825 16151 16668 20815 22527 25906 25977 27972 26651 26747 32589 32280 45543 Source: Central Bank 0 TABLE 8: IDENTIFIED PUBLIC INVESTMENT BY SECTOR (in millions of Sri Lankan Rupees) Actual Planned* 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1 Agriculture 7910 7739 8161 6870 7291 6578 6737 5556 7217 7537 9646 8991 8267 (a) Mahaweli 5338 4934 4525 3430 4625 3028 2695 2340 2676 2604 4170 4238 4315 (b)OtherIrrigation 819 1022 1526 1012 710 905 753 908 1216 1209 1511 1195 732 (c) Forestry 429 582 812 484 312 490 739 728 369 451 600 654 669 (d) Land 0 0 0 0 0 0 0 0 920 874 499 529 557 (e) Fleld,Export Crops 661 413 725 970 687 686 683 685 927 1099 680 465 299 (f) Livestock 45 66 82 100 93 100 134 23 41 81 76 66 55 (g) Fisheries 110 214 193 182 243 120 161 137 377 555 669 575 446 (h)Plantations 511 508 299 693 621 1250 1572 735 691 664 1441 1270 1196 2 Industries,Tourism &Trade 0 0 603 600 1276 1728 2621 3524 1049 1047 1105 1216 1338 3 Human Settlements 1582 1443 2062 1847 1870 3600 4299 6287 8229 8923 8904 8173 6873 (a) Housing 504 500 410 428 373 262 263 355 555 657 830 588 582 (b) Urban Infrastructure 0 0 203 276 112 494 178 665 1282 1594 1927 1716 870 (c) Other Construction 280 254 206 267 280 136 190 130 227 263 280 292 280 (d) Environmental Management 7 3 3 4 1 4 26 49 146 152 274 335 295 00 (e) Water Supply& Sanitation 791 686 1233 864 754 656 1614 2882 3046 3432 2386 1694 1376 (Q Provincial/Regional Development 0 0 7 8 186 1412 1408 1601 2015 2019 2253 2453 2551 (g) IRDPP 0 0 0 0 164 636 620 605 958 806 954 1095 919 4 Economic Infrastructure 1805 3024 4295 7998 6470 9274 11633 8043 16897 22716 24365 17456 11911 (a) Transport 512 1061 1064 1469 1436 3405 5699 4381 6630 8666 11404 10467 8515 (b) Power & Energy 540 1040 1986 4348 3758 5051 3987 2451 5955 5550 2839 3132 2178 (c)Posts&Telecommunications 479 630 949 1671 1181 613 846 292 2114 4500 3000 1917 892 (d) Other Economic Overheads 274 293 297 510 95 205 1101 919 2198 4000 7122 1940 326 5 Social Infrastructure 1272 2096 2974 2347 2305 2857 4434 4707 5298 6500 7934 7510 7667 (a) Education 905 1180 1143 1096 991 1043 2357 2660 2932 3430 3667 3897 4065 (b)Health 248 398 1313 1110 1195 1677 1827 1767 2027 2660 3769 3087 3045 (c)Others 119 518 518 141 119 137 250 280 339 410 498 526 557 6 Administrative Overheads 7793 11269 9510 8950 7493 8852 9223 11149 8193 9000 9900 10890 11979 (a) Administrative Overheads 4139 6686 7349 7455 3101 6020 5739 5904 5337 6716 7387 8126 8939 (b) Head Miscellaneous 3654 4583 2161 1495 4392 2832 3484 5245 2856 2284 2513 2764 3040 co TOTAL IDENTIFIED 20362 25571 27605 28612 26705 32889 38947 39266 46883 55723 61854 54236 48035 *From Public Investment Plan, 1993-97 (May 93) Source: Ministry of Policy Planning and Implementation -90- ANNEX Page 11 of 14 TABLE 9: THE CEB'S INVESTMENT PROGRAM, 1993-1997 (Rs Millions) PIP a/ Project 1993 1994 1995 1996 1997 PIP Kukule 120 592 1691 3144 Coal 1 x 150 MW (2002) 621 1618 PIP New A.D.B. Rural Electrification Sche.-A 995 Canyon Hydro Project 5 Transmission IV & VI (1996) 741 200 Diesel 3 x 20 MW (1996) 418 2899 1651 D.C.B.R.E. 147 PIP Third Rural Electrification Project 649 1775 577 Coal 2 x 150 MW (2004) Coal 1 x 15 (2007) Other Transmission 297 766 1267 1123 Upper Kotmale 479 2099 4967 PIP Rantambe Hydro Power Project 308 61 PIP Samanalawewa Hydro Project 255 Service Connection 351 781 866 954 1052 Augmentation 100 274 303 335 369 PIP Other Dist. (MV DEP & Third R.E.) 0 0 PIP Power System Ex. Sapu 2 x 20 MW 800 985 250 160 and Kelanitissa Rehabilitation 10 PIP Broadlands Hydro 135 1518 2935 1310 PIP New Diesel Station 20 MW 120 716 388 PIP Second Power Dist. & Trans. Project (SPDIP) 1225 900 700 502 PIP Tr. Sys. Aug. & Devt. Pr. (TSADP) 750 350 180 Bulk Supplies 183 D.E.P.P. (Ninth Power Project) 669 PIP Medium Voltage DEP 512 PIP Trans. Lines & Grid Sub. Pr. Samana (TGSP) 15 PIP Second Towns Power Dis. Pr. Project (A)(STPSP) 91 PIP Tr. Lines & G/S Devt. Project (TGSDP) 530 410 280 56 PIP Power Dist. & Trans. Project (PDTP) 995 490 318 Subtotal 10002 11776 10285 8995 12373 Projects Removed b/ Diesel 3 x 20 (1997) 0 484 3060 1732 G.T. 1 x 22 MW (1995) 224 576 Coal 1 x 150(2000) 433 1092 3538 6378 Coal 1 x 150 (2001) 441 1145 3716 G.T. 1 x 22 MW (1999) 302 G.T. 3 x 22 MW (1998) 869 2009 Total 10226 13269 14878 16279 24778 a/Projects marked PIP appear in Public Investment: 1993-97 published by the Ministry of Policy Planning b/Projects removed are proposed for construction by the Private Sector Source: Ceylon Electricity Board -91 - ANNEX Page 12 of 14 TABLE 10: LECO'S INVESTMENT PROGRAM, 1993-1996 (Rs Million) Capital Works Plan 1993 1994 1995 1996 Distribution Net Work Expenditure 415 137 818 633 Other Capital Expenditure 35 25 76 65 Total 415 162 894 698 Source: Lanka Electricity Company - 92 - ANNEX Page 13 of 14 ТаЫе 11: OVERVIEW OF EVALUATTON OF MARGINAL INVESTMENTS Есаlюпис Таддкаl Fии1 Аддиписппгvе Effiarnry �arnry SustunaЫbry Suawnabdlry EqwtaЫe АиЬш Э10в 1 No i No 1 No 1 N1t i 7 и No ti 1 и NR ш Роог1У 1it NR rv NR rv NoflMfpvmL У � Ка1и Gиga i Уа 1 Tb 1 No 1 Уе 1 Миед и. Tb и No и Уе т Рооду т No rv No rv Тпиврагтг У Т'IO TdCд1il EIECtllOty P1mta 1 Уа 1 NA 1 No 1 No 1 NA i1 Уа д Not yet и No ш Pa0r1y ii1 Уа rv Ун rv Тгwрвппг v Уе Ru1wy СоагЬа в No " 1 Уа 1 No 1 T/R 1 SuMldun poor и Уе и No и NR рттgи i1i. РооАу 1i1 W n No rv Tnиpnmt У NO Raalway Futmarooa 1 No 1 Уе 1 No 1 NR 1 Уе и Уе д No п NR т Рооду iд No rv No rv Тгатритг У NO Seoood Аироп 1 No i NA 1 No 1 NR 1 1 i1 Уа и No 1i. No ш NA � ш No N No 1V Not авшрагтг v NA Вговд-gвglдg tbe Ке1ап1 Vd1ry Loc 1 Уа 1 Уе 1 No 1 NR 1 SиЬвlдве роог i1 Yn 1i. No и Mt pwmgm ii� Роогlу и1 No rv No 1У Тлгорите У NO вш к1о 1 No 1 No 1 ио 1 NR 1 п No и No и Уа 1и NA ш No rv Модавгеlу rv Not Тгиврааlt v МодсамJу Kaamayake Expnsaavвy i Уа 1 NA 1 МодавWу 1 NA 1 Towrd п�1 и. Ул п Модаиду ti Ут iu NA 1it Tb� rv СодмЬига rv Тnпвриедг v No Non: NR таав oot пievant; NA тЧпf lnfortnatюa пог rvulaЫe -93 - ANNEX Page 14 of 14 Table 12: OVERVIEW OF SAFETY-NET PROGRAMS Economic Techn*-c2! Fiscal Administrative Efficiency Efficiency Sustainability Sustainability Equitable Food Stamps i. No 1. No i. No 1. Yes 1. To poor ii. Yes ii. NR ii. Yes iii. NR iii. Yes iv. ? iv. Yes V. NR v. Some problerns Janasaviya i. No i. NR i. No i. Yes i. To poor ii. Yes ii. NR ii. Yes iii. NÅ iii. Yes iv. ? iv. Some problerns V. NR Mid-day Mcal i. No i. ? i. No i. Yes i. ? ii. Yes ii. NR ii. Yes iii. NÅ iii. Yes iv. ? iv. Traxuparent V. NR vi. NR Transfers to the i. No i. NR i. ? i. No i. To poor Truly Needy ii. Yes fi. NR ii. ? (Dept. of Social iii. NÅ iii. ? Services) iv. 9 iv. Transparent V. NR Assistancc to i. Yes i. NR i. No i. Some problenLi i. To displaced Displaced ii. Yes ii No ii. Some problems population Populations fli. NR iii. Yes iv. No iv. Transparent V. NR Notc- NR mcam not relevant; NÅ means information not available. - 94 - IBRD 20879 alf. SRI LANKA --- a tioal roods ' Jafn - Rahoys - s 0 Selected towns and villages -~» N x * Nolional coprial DisWafkat b,oundae Province bondaries e,.9 J A F F N A i mU L A I T I V U Manrm~ Tojo mar 1 - u / NO,1?HE R/ N.. ~VA V N IY A Jm IOMfIEs MANN A R NlY- l 2p 2 40 50 . r MILUS ln d i a Vvuny TRINCOMALE Oc e anr. Tnn~ae Medawachchyo A N -U R A D A P U A Anurodhopura IVN R H C(EN R A L PUTTA LA,,Hb POLONNARUWAar Al OR-v Dambulo \l /e RATT CAAAOA MAT A L E K U R U N GALI A MA iKurnegoaaMt T R A JrKandy BADUILLA AMPAR AI Negombo -Jl - KNtunoya GAM A KEGALLA Bodulla7- COLO.BO ..&rwaro COL ~='. p.."'.." E1LYA M ONE AGALA Poftuvil .. as... -.. - auwy APUTARA R N AP R A ( 5 CHINA PAKISTAN,i -1ý - H AMB ANT OT A INDIA - S O- T NE - G A L L E ATAR IMATARA mGlleban= Motora 1 \SRI LANKA APRIL 1988