Nepal - Selected Issues in Infrastructure Development (Vol. 2 of 2) : The Annexes

DRAFT CONFIDENTIAL Report No. 11800-NEP CONFIDENTIAL 11800-NEP VOL. 2 NEPAL SELECTED ISSUES IN INFRASTRUCTURE DEVELOPMENT (In Two Volumes) Volume II: Annexes February 2. 1994 South Asia Country Department 1 Energy & Infrastructure Operations Division FOR OFFICIAL USE ONLY Document of The World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Page No ANNEXES TABLE OF CONTENTS I. AN OVERVIEW OF THE NEPALESE ECONOMY . . . . . . . . . . . . . . 1 A. The Spatial Distribution of Economic Activity . . . . 1 Regional Contributions to Gross Domestic Product 2 Agriculture . . . . . . . . . . . . . . . . . . 3 Manufacturing . . . . . . . . . . . . . . . . . 4 Investment and Inter-Regional Capital Flows . . 5 Employment . . . . . . . . . . . . . . . . . . . 6 B. Regional Distribution of Population and Urbanization 7 Trends in the Distribution of Population . . . . 7 Migration Patterns . . . . . . . . . . . . . . . 8 Urbanisation . . . . . . . . . . . . . . . . . . 9 The Growth of Urban Centers . . . . . . . . . . 11 C. Household Demand . . . . . . . . . . . . . . . . . . 13 Household Consumption and Incomes . . . . . . 13 Regional Shares of Household Demand . . . . . . 14 Rural and Urban Sources of Household Income . . 16 D. Projections on Population Distribution and Urbanization 17 The Growth of Cities . . . . . . . . . . . . . . 18 II. GEOGRAPHIC INFORMATION AND ANALYSIS . . . . . . . . . . . . . . 19 A. Demographic Patterns . . . . . . . . . . . . . . . . 19 B. Land Use . . . . . . . . . . . . . . . . . . . . 25 C. Agriculture . . . . . . . . . . . . . . . . 25 Agricultural Output Per Capita . . . . 27 Agricultural Cropping Patterns . . . . . . . . . 27 Agricultural Productivity . . . . . . . . . . 30 D. Forest Degradation . . . . . . . . . . . . . . 34 E. Impact of Infrastructure Development . . . . . . 34 III. EXTERNAL AND INTERNAL TRADE . . . . . . . . . . . . . 38 A. Trends in Nepal's External Trade . . . . . . . . . . 38 Composition of Exports . . . . . . . . . . . 39 Composition of Imports . . . . . . . . . . . . 40 Regional Shares of Production for Exports . . . 40 B. Trade with Third Countries . . . . . . . .. 41 C. Trade with India . . . . . . . . . . . . . 42 D. Internal Trade within Nepal . . . . . . . . . . . . . 43 IV. QUALITY OF INFRASTRUCTURE SERVICES FOR INDUSTRY . . . . . . . . 46 A. Survey objectives and Sample Profile . . . . . . . . . 46 Selection of a Representative Sample . . . . . . 46 Some Characteristics of the Industrial Sample 47 Profile of Industrial Expenditures . . . . .. . 48 Paae No B. Water Supply . . . . . . . . . . . . . . . . . . . . 50 Use of Boreholes . . . . . . . . . . . . . . . . 50 Water Supply Costs . . . . . . . . . . . . . . . 50 C. Transport . . . . . . . . . . . . . . . . . . . . 51 Costs Incurred in Own Account Trucking . . . . . 52 Impact of Transport Problems . . . . . . . . . . 53 D. Telecommunications . . . . . . . . . . . . . . . . 53 E. Drainage, Sewerage and Solid Waste Disposal . . . . . 54 F. Ranking of Infrastructure Problems . . . . . . . . . 55 V. TRANSPORT . . . . . . . . . . . . . . . . . . . . 57 A. Road Transport . . . . . . . . . . . . . . . . . . .7 Road Network Distribution . . . . . . . . . . . 59 The Trail Network . . . . . . . . . . . . . . . 60 Condition of Roadt . . . . . . . . . . . . . . . 61 Road Traffic . . . . . . . . . . . . . . . . . . 61 Road Transport Services . . . . . . . . . . . . 64 Road Sector Revenues and Expenditures . . . . . 65 Maintenance Expenditure Needs . . . . . . . . . 67 Future Demand for Roads and Trail Improvements . 69 Estimated Financial Requirements for the Road Sector 72 B. Air Transport . . . . . . . . . . . . . . . . . . . 72 Existing Aviation Infrastructure . . . . . . . . 73 Air Transport Services . . . . . . . . . . . . . 76 Air Traffic Growth and Projections . . . . . . . 77 Air Transport Costs and Tariffs . . . . . . . . 79 DCA Revenues and Expenditure . . . . . . . . . 80 Investment Needs . . . . . . . . . . . . . . . 81 C. Other Transport Modes . . . . . . . . . . . . . . . . 82 VI. WATER SUPPLY AND SANITATION . . . . . . . . . . . . . . . . . . 83 A. Existing Facilities and Coverage . . . . . . . . . . 84 Coverage . o . . . . . . . . . . . . . . . . . 84 B. Level of Service . . . . . . . . . . . . . . . . . . 86 Technologies Used . . . . . . . . . . . . . . . 88 Water Quality and Environmental Impact . . . . . 89 C. Sector Policies and Strategies . . . . . . . . . . 89 Sector Budget Allocation . . . . . . . . . . . . 91 Future Plans and Targets . . . . . . . . . . . . 92 D. Financial Requirements . . . . . . . . . . . . . . . 93 VII. TELECOMMUNICATIONS . . . . . . . . . . . . . . . . . . . . 95 A. Existing Facilities . . . . . . . . . . . . . . . . . 95 B. Access to Service . . . . . . . . . . . . . . . . . . 97 C. Service Usage . . . . . . . . . . . . . . . . . . . 99 D. Demand for Telephone Service . . . . . . . . . . 100 E. NTC Finances . . . . . . . . . . . . . . . . . . . 102 F. Telecommunications Development Expenditures . . . 104 G. Future Plans . . . . . . . . . . . . . . . . . 106 Development Plan . . . . . . . . . . . . 106 Pace No VIII. INFRASTRUCTURE DEMAND AND RESOURCE MOBILIZATION . . . . . 109 A. Infrastructure Demand in the Economy . . . . . . . . 109 Aggregate Demand for. Transport & Communications 109 Aggregate Demand for Water Supply & Sewerage . 111 Regional Distribution of Infrastructure Demand 111 Rural -ve- Urban Infrastructure Demand . . . . . 113 B. Investment Levels . . . . . . . . . . . . . . . . . . 114 C. Trends in Infrastructure spending . . . . . . . . . . 117 Central Government Expenditure . . . . . . . . . 117 Local Government . . . . . . . . . . . . . . . . 118 D. Resource Mobilization . . . . . . . . . . . . . . . . 118 Central Government Tax Effort . . . . . . . . . 118 Local Government Resources . . . . . . . . . . . 120 E. Options for Domestic Resource Mobilization . . . . . 121 Pricing of Infrastructure Services . . . . . 121 Improving Local Revenue Systems . . . . . . . . 124 Poverty Concerns . . . . . . . . . . . . . . 125 ANNEX TABLES 1.1 Gross Domestic Product by Industrial Sectors 1980-1990 . . 1 1.2 Gross Domestic Product by Regions 1984/85 . . . . . . . . 2 1.3 Indicators of Agricultural Production 1984 . . . . . . . . 3 1.4 Production and Consumption of Edible Cereal Grains by Region 4 1.5 Regional Shares of Manufacturing Activity 1981/82 & 1986/87 5 1.6 Savings and Credit by Region . . . . . . . . . . . . . 6 1.7 Regional Shares of Economically Active Population by Sectors in Nepal, 1981 . . . . . . . . . . . . . . . 7 1.8 Regional Shares of Total Population 1971-1991 . 8 . . . 1.9 Origin and Destination of Inter-Regional Migrants in 1981 9 1.10 Urbanization 1952/1987 . . . . . . . . . . . . . . . . . 10 1.11 Regional Shares and Growth Factors of Urban Population 1971-1991 . . . . . . . . . . . . . . . . . . . . 11 1.12 Growth of Urban Centers 1971-1987 . . . . . . . . . . . 12 1.13 Composition of Household Consumption by Category and Type 13 1.14 Composition and Elasticity of Household Consumption by Level of Per Capita income . . . . . . . . . . . . 14 1.15 Regional Shares of Household Demand 1981 . . . . . . . . . 15 1.16 Composition of Household Income by Sources and Region . . 17 1.17 Regional Shares and Projected Total Population . . . . . . 17 1.18 Actual and Projected Population of Urban Centers . . . . .18 3.1 Nepalese Foreign Trade . . . . . . . . o. . . . . . . . 38 3.2 Exports Classified by Major Commodity Group . . . . . . . 39 3.3 Imports Classified by Major Commodity Group . . . . . . . 40 3.4 Regional Shares of Selected Exports 1987/88 . . . . . . . 41 3.5 Overseas Imports and Exports through Main Gateways, 1989 . 42 3.6 Imports and Exports to India through Main Gateways, 1989 . 42 3.7 Sources of Procurement by Wholesale Unit . . . . . . . . 43 3.8 Regional Variations in Retail Price Indices, 1988 . . . . 44 3.9 Profile of Rural Trading Activity, 1990 . . . . . . . . . 45 4.1 Distribution of the Sample of 150 Industrial Establishments . . . . . . . . . . . . 47 Pace No 4.2 Infrastructure Costs in Industrial Cost Structure . . . . 49 4.3 Source and Consumption of Water for all Establishments . 51 4.4 Use of Different Shipping Methods for Transportation of Inputs and Outputs, 1990 . . . . . 52 4.5 Own Account Fleet Operations, 1990 . . . . . . . . . . . . 52 4.6 Loss of Output Due to Transportation Bottlenecks . . . . . 53 4.7 Telephone Installations and Payment to NTC, 1990 . . . . . 54 4.8 Proportion of Successful Calls on First Attempt in 1990 . 54 4.9 Public Sanitary Sewer Connection and Total Cost for Sewage Disposal, 1990 . . . . . . . . 55 4.10 Solid Waste Collection and Disposal System . . . . . . . . 55 4.11 Comments on Public Provision of Infrastructure . . . . . . 56 5.1 Road Densities in Nepal and Other Countries . . . . . . . 59 5.2 Nepal Road Network by 1991-92 . . . . . . . . . . . . . . 60 5.3 Suspension Bridge Construction 1970-85 and Target 1985-90 61 5.4 Indicative Traffic 1986 and Projections 1990-2010 for Key Roads . . . . . . . . . . . . . . 63 5.5 Estimated Vehicle Fleet, 1986 .. . . . . . . . . . . . . 64 5.6 Road Budget Allocations and Expenditures in Current Prices 66 5.7 Average Road Maintenance Expenditures in Nepal . . . . . . 67 5.8 Estimated 1991 Road Maintenance Cost per km per Year . . . 68 5.9 Estimated Financial Requirements for Road Subsector . . . 72 5.10 Regional Distribution of Airports by Category . . . . . . 74 5.11 Aircraft Utilization and Fleet Age Structure . . . . . . . 77 5.12 Domestic Airport Traffic Flows . . . . . . . . . . . . . 78 5.13 RNAC Domestic Airfares and Estimated Aircraft Operating Costs 1986/87 . . . . . 80 5.14 Summary of Proposed Future Works for Aviation Development 1992-2010 . . . . . . . . . . . . . . . . 82 6.1 Estimated Water Supply 1991-92 . . . . . . . . . . . . . . 8 6.2 Drinking Water Availability and Population Served . . . . 86 6.3 Estimated Sanitation Coverage in Nepal 1990-91 . . . . . . 87 6.4 on-Going and Completed Water Supply Projects under DWSS . 90 6.5 WSS Sector Investments - 6th & 7th Five Year Plans . . . . 91 6.6 Projected Coverage and Additional Population Served . . . 94 7.1 Telecommunication Facilities in Nepal . . . . . . . . . . 95 7.2 Projected Number of Telephones by the End of IVth Project 98 7.3 Actual Telephone Coverage in Nepal 1991 . . . . . . . . . 99 7.4 Profile of Telephone Subscribers in Nepal - 1984 . . . . . 99 7.5 Telephone Demand and Supply 1980-1989 . . . . . . . . . . 101 7.6 NTC Revenues and Expenses 1984-92 . . . . . . . . . . . . 103 7.7 Investment in Telecommunications . . . . . . . . . . . . . 104 7.8 Telephone Demand and Supply 1980-96 . . . . . . . . . . . 105 7.9 NTC's Future Financial Requirements . . . . . . . . . . . 108 8.1 Projected Growth in Transport & Communication Demand . . . 111 8.2 Average Annual Household Income & Expenditures . . . . . . 112 8.3 Projected Distribution of Discretionary & Total Cash Incomes 113 8.4 Sectoral ICORs for Nepal and India . . . . . . . . . . . . 115 8.5 Investment Needs Implied by Growth in Demand . . . . . . . 116 8.6 Estimated Investment Levels and Five Year Plans . . . . . 117 8.7 Sources of Investment Spending . . . . . . . . . . . . . . 118 Pace No ANNEX FIGURES 2.1 Distribution of Population 1971, 1991, 2010 . ., . . . . . 20 2.2 Spatial Distribution of GDP 1971, 1991, 2010 . . . . . . . 22 2.3 Historical and Projected Migration Patterns . . . . . . . 24 2.4 Spatial Land Use Patterns, 1980 . . . . . . . . . . . . 26 2.5 Spatial Distribution of Agricultural Output - 1986 . . . 28 2.6 Per Capita Agricultural Production - 1986 . 29 2.7 Paddy, Wheat, Maize Cultivation - 1986 . . . . . . . . . 31 2.8 Crop output and Roads, Irrigation and Fertilizer Use . . 32 2.9 Forest Degradation . . . . . . . . . . . . . . . 35 2.10 Transport and Value of Agricultural Output . 36 2.11 Transport and Industries . . . . . . . . . . . . 37 2.12 Transport Infrastructure and Degradation 37 5.1 Roads and Main Trails . . . . . . 58 5.2 Regional Distribution of Airports in Nepal . . . . . . . 75 8.1 PCGDP and PC Transport & Communication Cross-Section of LDCs 110 Page 1 of 18 AN OVERVIEW OF THE NEPALESE ECONOMY 1. In order to derive pointers to guide a strategy for infrastructure development, this Annex provides an understanding of emerging trends in the Nepalese economy. of particular interest is the future spatial distribution of economic activities and population since these have a direct bearing on the provision of infrastructure facilities in various parts of the country. This analysis examines three elements of the macro-economy. First, it reviews the geographic dimensions of production, how each region contributes to output, and how capital and labor is distributed among the regions of the country. Second, it examines regional trends in the growth of rural and urban population, the growth of urban centers and labor migration. Third, it looks at demand, as represented by the composition of household consumption and income. The final section contains projections on urban and rural population distributions, and the growth of cities. A. The Spatial Distribution of Economic Activity 2. Agriculture has always been and still remains the backbone of the Nepali economy, accounting for around 60% of gross domestic product during the period 1980 through 1990 (Table 1.1). While there have been few signs in recent years that agriculture is losing its importance there have been significant shifts among non-agricultural activities. Manufacturing has made significant advances, its share increasing from 4.3% of the total in 1980 to 5.4% in 1990. The growth in manufacturing is clearly associated with increases in exports of textiles, clothing and leather goods. The social services sector has also increased its share of the total, from 6.8% to 8.7%. These gains among non-farm sectors have been offset by relative declines in transport and communications, down from 7.0% to 5.6%, and in the financial sector, down from 8.4% to 8.3%. Table 1.1 GROSS DOMESTIC PRODUCT BY INDUSTRIAL SECTORS 1980-1990 (in 1990 NRs Million) Gross Domestic Product Sector 1980 1985 1990 1980/90 NRs Percent NRs Percent NRs Percent Annual % I Share Share Share Agriculture 32329 61.8% 38200 57.6% 45848 58.6% 3.56% Mining 100 0.2% 224 0.3% 107 0.1% 0.64% Manufacturing 2238 4.3% 3190 4.8% 4200 5.4% 6.50% ELectricity 143 0.3% 313 0.5% 657 0.8% 16.43% Construction 3754 7.2% 5720 8.6% 5620 7.3% 4.23% Comerce 2126 4.1% 2933 4.4% 4059 5.2% 6.68% Transport & Coamunications 3685 7.0% 4413 6.7% 4413 5.6% 1.82% Financial and Real Estate 4383 8.4% 5460 8.2% 6524 8.3% 4.06% Social Services 3575 6.8% 5893 8.5% 6787 8.7% 6.62% Total 52335 100.0% 66345 100.0% 78275 100.0% 4.11% Source: Ministry of Finance, Economic Survey, 1990-91 Tables 1.1 and 1.2 U/ Page 2 of 18 Regional Contributions to Gross Domestic Product 3. Unlike some developing countries where a major part of national production is concentrated in a single primate city region, production in Nepal is relatively dispersed. Nevertheless, in 1984/85 some 60% of total national output came in roughly equal shares from just three of the country's fifteen geographic regions V, the three being the Hills in the Central region (19%), and the Terai in the Central (21%) and Eastern (20%) regions (see Table 1.2). Other zones of importance include the Western Terai (8.7%), the Western Hills (7.4%), the Eastern Hills (6.4%) and the Mid-West Terai (5.2%). Collectively, the Terai accounted for 58% of total output, substantially more than the Hills with 37%, even though both geographic regions had similar size populations at the time. 4. Among the fifteen geographical regions of the country, there was an unusually wide variation in estimated per capita income levels, with the poorest regions in the Mountains of the Mid and Far-West recording indices of little more than a third of the national average, compared to around 130 in the Terai of the Western, Central, and Eastern regions. (The index of 186 for the Mountains in the West may be associated with the concentration of tourist activities among a small population.) In general, per capita income levels in the Terai exceed those in the Hills and Mountains, which is one factor accounting for the sustained migration towards the Terai. The main exception to this pattern is the Hill zone of the Central region, which includes the Kathmandu Valley, with an index of 141 compared to 129 in the Terai. Table 1.2 GROSS DOMESTIC PRODUCT BY REGIONS 1984/85 (1990 NRS) DFr- Mid- West CentraL Eastern Total EcotogicaL ReinWst IWest Mountain GDP NRs MilLion 464 398 133 1128 995 3052 GDP % Share 0.7% 0.6% 0.2% 1.7% 1.5% 4.6% GDP per Capita NRs 1503 1542 6358 2603 2839 GDP Capita Index 36 36 186 64 73 Hill GDP NRs Million 929 1990 4910 12871 4246 24879 GDP % Share 1.4% 3.0% 7.4% 19.4% 6.4% 37.5% GDP per Capita NRs 1435 1767 2121 5625 3218 GDP Capita Index 36 44 53 141 80 Terai GDP NRs Million 1990 3450 5772 13800 13402 38414 GDP % Share 3.0% 5.2% 8.7% 20.8% 20.2% 57.9% GDP per Capita NRs 3475 4265 5079 5128 5335 GDP Capita Index 87 108 129 129 134 Total GDP MRs Million 3317 5838 10814 27201 18643 66345 GDP % Share 5.0% 8.8% 16.3% 41% 28.1% 100% GDP per Capita NRs 2210 2688 3130 5031 4489 3977 GDP Capita Index ' 56 67 79 129 113 100 SOURCE: REGIONAL GDP SHARES FROM - BASIC FEATURES OF THE NATIONAL DEVELOPMENT MASTER PLAN (1990-2000). National Planning Comission, 1989. For the purposes of spatial disaggregation the country is generally divided, In this analysis, into 5 development regions (Far Western, Mid Western, Western, Central and Eastern) which are each then divided into 3 ecological zones (Mountain, Hill, Terai) of which the Mountain is the northernmost and the Terai is the southernmost. -03 - Annex I Page 3 of 18 Agriculture 5. In agriculture, the Central region is the main area of production, accounting for about a third of output and employment in the sector in 1984 (Table 1.3). The Eastern region follows in second place, accounting for a fourth of production and employment, and in third place comes the Western region with a fifth part of each. The Mid-West and Far-West regions provide the remainder, about a fifth part between the two of them. The regions follow the same rankings when measured in terms of land and labor productivity, with the Central region above the national average, the Eastern close to the average, and the western regions well below the national average. Over the ten year period 1976 to 1986, there has been a small shift in regional contributions to aggregate production of the four main agricultural commodities rice, maize, wheat and millet. The Eastern region's share fell 2.1 percentage points, from 26.9% in 1976/77 to 24.8% in 1985/86, while the other regions have gained. Table 1.3 INDICATORS OF AGRICULTURAL PRODUCTION 1984 Region Production Percent share Percent share Index of Labor Index of Land metric tons of production of Emptoyment Productivity Productivity Eastern 1099500 26.1% 24.7% 106 99 CentraL 1520140 36.1% 32.7% 110 115 Western 837390 19.9% 20.8% 95 95 Mid-Western 457090 10.9% 13.0% 83 83 Far-Western 296950 7.1% 8.8% 80 86 TotaL 4211070 100.0% 100.0% 100 100 SOURCE: BASIC FEATURES OF THE REGIONAL DEVELOPMENT MASTER PLAN (1990-2000). 1989 6. Estimates of production and consumption of edible cereal grains indicate in general terms that the Eastern region and usually the Central and Western regions produce surpluses, while the two most westerly regions are almost invariably in deficit, meaning production is insufficient to meet local demand (Table 1.4). Whether the Western regions satisfy demand from domestic suppliers or from India is not known, but it certainly indicates a potential market for Nepalese producers. Local suppliers, however, will only be able to compete with Indian suppliers in those markets if they are integrated into Nepalese grain marketing networks, and the costs of transportation to the western-most regions are not prohibitively expensive. -04 - Annex I Page 4 of 18 Table 1.4 PRODUCTION AND CONSUMPTION OF EDIBLE CEREAL GRAINS BY REGION (000's Metric Tons) Year Far- Mid- Western Central Eastern Nepal Western Western Production 1981/82 182.7 309.5 480.8 908.7 626.9 2508.6 1982/83 173.2 282.8 467.4 742.5 530.7 2196.5 1983/84 201.3 315.1 508.1 1017.0 700.9 2742.5 1984/85 183.7 293.3 516.5 937.3 664.1 2594.9 1985/86 200.9 328.8 581.2 959.9 681.5 2752.2 1986/87 205.1 324.0 542.3 885.0 632.5 2588.8 1987/88 208.1 377.1 631.9 1015.8 773.0 3005.9 1988/89 266.0 457.1 733.5 1110.6 850.5 3417.6 Consumption 1981/82 179.4 333.5 478.6 778.3 477.9 2247.6 (Requirement) 1982/83 184.1 342.6 490.9 799.6 490.3 2307.5 1983/84 216.7 368.8 527.2 828.1 558.5 2499.3 1984/85 210.2 348.7 510.2 850.3 580.4 2499.8 1985/86 217.4 359.7 524.9 877.2 599.7 2578.9 1986/87 237.9 370.2 544.4 886.0 613.9 2652.5 1987/88 246.8 380.3 556.6 911.5 631.1 2726.2 1988/89 261.4 407.4 632.1 953.5 666.8 2921.2 Balance 1981/82 3.3 -24.0 2.2 130.5 149.0 261.0 1982/83 -10.9 -59.8 -23.5 -57.1 40.4 -111.0 1983/84 -15.4 -53.7 -19.0 188.9 142.4 243.2 1984/85 -26.6 -55.4 6.3 87.1 83.7 95.0 1985/86 -16.5 -31.0 56.3 82.7 81.8 173.4 1986/87 -32.9 -46.3 -2.1 -1.0 18.6 -63.6 1987/88 -38.7 - 3.1 75.3 104.3 142.0 279.7 1988/89 4.6 49.7 101.4 157.1 183.6 496.4 1981-1989 -133.0 -223.7 196.8 692.4 841.5 1374.0 SOURCE: AGRICULTURAL STATISTICS OF NEPAL, 1990, TABLE 7.1 Manufacturing 7. As is the case in agriculture, the Central region is the major player in manufacturing, and has substantially increased its dominance in recent years. In 1981/82 the region accounted for 48% of total employment in the sector, 58% of fixed assets and 51% of value added (See Table 1.5). Five years later, the region had increased its shares of employment by 11%, fixed assets by 4%, and value added by a remarkable 17%. The greatest part of this increase occurred in the Bagmati zone, which includes the city of Kathmandu. The zone's share of all value added in manufacturing rose 10.3%, and its share of employment 17.5%, even though its share of fixed assets moved up only 1.6%, implying most of the increase was in labor intensive activities. Some part of this increase, perhaps a large part, is associated with the contemporaneous increase in exports of manufactures (see Annex III) much of which is produced in the zone. 8. The Eastern region fills second place in manufacturing, though growth lagged far behind the Central region in the period 1981/82 through 1986/87, with the result that its share of total employment in the sector fell 12%, share of fixed assets by 6%, and of value added by 4%. The Western region was third with about one eighth of total manufacturing activity. Overall, the data reveal a shift of manufacturing jobs from the Eastern region to the Central and Western regions, while investment (as reflected by fixed assets) favored the Central. In terms of value added, the Central region registered major gains, while all other regions lost shares. In interpreting the data on manufacturing it is -05 - Annex I Page 5 of 18 important not to lose sight of the fact that manufacturing is still a very small part of the Nepalese economy, contributing only about 6% of GDP in 1990, and less than 1% of employment. Table 1.5 REGIONAL SHARES OF MANUFACTURING ACTIVITY 1981/82 AND 1986/87 1981/82 1986/87 1981/82-1986/87 I_ I Changes in Shares Region Zone Enployees Fixed Value EnpLoyees Fixed Value EupLoyees Fixed VaLue Assets Added Assets Added Assets Added % % % % % % % % % Eastern 34.1% 25.1% 21.1% 22.2% 19.3% 17.2% -11.9% -5.8% -3.9% Central 47.8% 58.2% 51.3% 58.4% 62.6% 68.4% 10.6% 4.3% 17.1% Western 10.2% 11.3% 17.1% 12.9% 11.2% 8.7% 2.7% -.1% -8.4% Mid-Western 4.4% 2.2% 3.9% 3.6% 2.1% 3.2% -.9% -.2% -.7% Far-Western 3.5% 3.1% 6.6% 3.0% 4.9% 2.4% -.5% 1.8% -4.1% Total 100.0 100.0% 100.0% 100.0% 100.0% 100.0%1 SOURCE: CENSUS OF MANUFACTURING ESTABLISHMENTS 1986-1987, TABLE 1 DEVELOPMENT ATLAS OF NEPAL, 1988 Investment and Inter-Regional Capital Flows 9. Data from the banking system on the volume of credit issued to various regions between 1980 and 1989 reinforces the trend observed from the manufacturing census that investment has been favoring the Central region at the expense of the East. At the beginning of the decade, the Central region accounted for some 60% or more of total credit, and this share had edged up to 72% by 1989 (see Table 1.6). Meanwhile, the Eastern region's share of credit slipped from around a quarter to about one sixth of the total. However, one zone alone, Bagmati, accounts for more than half of total credit issued (60%). Given the relatively modest increase in fixed assets in manufacturing observed in Table 1.5, it must be inferred that these credits are being used mainly for purposes unrelated to manufacturing. 10. Data on the volume of deposits and credits are recorded by financial institutions located for the most part in the larger cities. Thus flows based on this data reflect the activities of banking institutions rather than the depositors or investors, whose activities may be located elsewhere. Nevertheless, banks transfer resources from one branch to another, and this provides a first approximation of the flow of capital between regions. Far and away the largest volume of deposits (around 66% or more) are made in the Bagmati zone, most of them presumably in Kathmandu, where major banks and business enterprises are headquartered. Other main sources of deposits are Koshi zone (around 7%) with the second largest city Biratnagar, and Narayani zone (around 8%) with the third largest city Birganj. Most of credit is issued in these same three zones, which together account for 75% of the total credits. In terms of net capital flows, total deposits in each year exceed the credits issued (about 83 percent of deposits, 1980-89), implying the balance is invested in other ways (perhaps lending to government and transfers overseas). Deposits exceed credits -06 - Annex I Page 6 of 18 in three of the five regions; only in the East and the Far-West is the reverse true. Table 1.6 SAVINGS AND CREDIT BY REGION Region 1980 1985 1989 Net FLow Average (Percent) (Percent) (Percent) (NRs mition) Ratio of Credit to Deposits Credit Deposits Credit Deposits Credit 1980 1985 1989 Deposits Eastern 12.8 22.4 11.0 19.0 8.6 14.9 214 339 541 1.49 of which: 7.8 14.3 7.1 13.5 5.1 10.2 179 299 510 1.59 Koshi Central 70.9 64.2 73.2 66.7 75.8 72.3 -1162 -1616 -3608 0.76 of which: 58.7 49.4 62.2 48.1 67.0 59.4 -1311 -1926 -3835 0.68 Bagmatt Western 9.3 6.6 9.5 8.1 9.4 7.9 -241 -245 -610 0.62 Mid-Western 4.3 3.3 4.0 2.7 4.1 2.3 -121 -140 -420 0.58 Far-Western 2.7 3.5 2.3 3.8 2.1 2.6 -32 55 -23 1.14 Total () 100.0 100.0 100.0 100.0 100.0 100.0 .----__ VoLune (NRs 3262 3181 8092 6464 18340 14220 -1342 -1627 -4120 0.83 Million) SOURCE: CBS, Statistical Yearbook of NepaL, Kathmandu, 1991. Employment 11. An analysis of the sectoral composition of employment emphasizes perhaps more than any other statistics that Nepal has still barely started the transition from an agricultural economy. At the time of the 1981 census, the great majority (92.9%) of the economically active population was still occupied in agriculture. Among non-farm sectors, services accounted for the next largest proportion, 4.7%, commerce employed 1.6%, and manufacturing still a tiny 0.5%. Remaining sectors made up the balance of 0.3%. In regional terms, the East accounted for about a quarter of the total, the Central region a little more, and the West a little less, while the two most westerly regions made up the remaining quarter (Table 1.7). The lack of data for other years makes it impossible to measure shifts and changes over time. Since agriculture accounts for most of the working population, it is distributed among regions more or less in the same proportions as the total workforce. Non-agricultural employment, however, is concentrated in the Central and Eastern regions which account for over 70 percent of employment in manufacturing, commerce and services. In particular, the Mid- Western and Par-Western regions together account for less than 10 percent of employment in the non-agricultural sectors even though their share of population is over 20 percent. -07 - Annex I Page 7 of 18 Table 1.7 REGIONAL SHARES OF ECONOMICALLY ACTIVE POPULATION BY SECTORS IN NEPAL, 1981 REGION/ZONE Share of Total Agriculture Manufacturing Conmerce Services Other popuLa- eupLoyment tion Eastern 24.8% 24.0% 23.2% 36.1% 29.5% 37.4% 26.4% Central 32.7% 31.1% 30.2% 45.5% 43.5% 41.8% 55.8% Western 20.8% 21.4% 22.1% 10.2% 15.6% 11.6% 11.3% Mid-Western 13.0% 14.1% 14.7% 4.9% 8.0% 5.9% 4.2% Far-Western 8.7% 9.3% 9.8% 3.2% 3.4% 3.4% 2.3% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Nunber (0001s) 15023 6723.6 6244.3 33.0 109.4 313.6 23.3 Percent 100.0% 100.0% 92.9% .5% 1.6% 4.7% .3% SOURCE: PoeuLation Census - 1981, Vol. II, 1984, Table 27 B. Recional Distribution of Population and Urbanization 12. Nepal's population has grown from just under 10 million at the time of the 1961 census, to 18.5 million (1991 census). Population growth rates which averaged 1.7% p.a. in the 1960s accelerated to about 2.7% in the early 1980s, and is now believed to be at around 2.08% based on the 1991 census. More remarkable than this increase in the overall population however, are changes that have taken place in the distribution of population among various parts of the country. Trends in the Distribution of Population 13. Between 1971 and 1991, there was a pronounced shift of population down from the Mountain and Hill areas to the Terai, and thir occurred consistently across all five development regions. Most of this shift t iginated in the Hills, whose share of the total population fell 6.9 points from 52.5% to 45.6% during the period (Table 1.8) . Another shift involving a further 2.1% of the population moved out of the Mountains. Together these shifts raised the Terai population by 9 points from 37.6% to 46.6% of the nation's total. Shifts between the regions, on the other hand, have been much smaller. Over the same 20 year period, it is estimated that the Central and Eastern regions lost 0.1% share of the total, and the Western region lost 0.8%. These outward shifts translated to a gain of 0.8% for the Far-West region, with a small residual of 0.1% accruing to the Mid-West. The most pronounced change occurred in the Far-Western Terai where the population in 1991 was about 3.5 times the population in 1971 (see Gr. Fct. of 3.48), similarly the Mid-Western and Western Terai also experienced rapid population increases with growth factors of 2.4 and 2.2 respectively. Overall the Terai population just about doubled during this 20 year period. 14. Given that about 90% of the population is still rural, the story for regional changes in rural population is almost the same as that for the entire population. Between 1971 and 1991 similar shifts occurred out of the Mountains (1.8%) and Hills (7.5%) down to the Terai, the largest part coming from the Western Hills (3.0%) and to a lesser extent the Central and Eastern Hills (1.7% each). The main destination for these rural shifts were the Eastern Terai (3.3%) -08 - Annex I Page 8 of 18 and the Western Terai (2.3%), with relatively little accruing to the Central Terai (0.6%). As was the case with total population, the Central and Western regions lost rural population shares, mainly, to the East and the Far-West. Table 1.8 REGIONAL SHARES OF TOTAL POPULATION 1971-1991 Ecological Year and Far- Mid- Western Central Eastern Total Zones Gr.Fct. ? Western Western Mountain 1971 2.1% 1.8% 0.3% 3.1% 2.6% 9.9% 1991 1.8% 1.4% 0.1% 2.6% 1.9% 7.8% Gr.Fct. 71-91 1.37 1.24 0.53 1.34 1.17 1.26 HiLL 1971 4.5% 7.7% 15.7% 15.1% 9.6% 52.5% 1991 3.6% 6.6% 13.1% 14.5% 7.7% 45.6% Gr.Fct. 71-91 1.28 1.37 1.33 1.53 1.28 1.39 Terai 1971 1.7% 3.4% 5.1% 15.3% 12.0% 37.6% 1991 3.7% 5.0% 7.1% 16.4% 14.4% 46.6% Gr.Fct. 71-91 3.48 2.35 2.22 1.71 1.92 1.98 Total 1971 8.3% 12.9% 21.1% 33.5% 24.2% 100.0% 1991 9.1% 13.0% 20.3% 33.4% 24.1% 100.0% Gr.Fct. 71-91 1.75 1.61 1.54 1.59 1.59 1.60 Total (000s) 1971 958 1488 2446 3866 2798 11556 1981 1320 1955 3129 4909 3710 15023 1991 1681 2406 3752 6174 4448 18462 SOURCE: CBS, Population Monograph 1987 and Preliminary Results, 1991 census. a' Gr. Fct: Growth Factor is the ratio of 1991 Population to 1971 Population. Migration Patterns 15. Analysis of inter-regional migration flows indicates (Table 1.9) that in order of importance the four most preferred destinations, among the 15 geographical regions, have been the Eastern, Central, Western and Far-Western Terai which together were the destination for about 70 percent of migrants!/ Also favored destinations were the Hills in the Central and Western regions which accounted for another 12 percent of destinations. Zones with the largest exodus have been the Eastern Hills and Mountains (together accounting for 48 percent of out-migrants), followed by the Western and Central Hills which accounted for another 26 percent of out-migrants; these four zones therefore accounted for almost three quarters of all out-migrants. The five largest zone- to-zone flows have all been intra-regional: in the east from the Hills and Mountains to the Terai; likewise in the Center; and in the West from the Hills to the Terai. In general the Terai attracted the majority of migrants (74 percent). Migration estimation : District-wise migration for a given period was estimated by finding the difference between actual population at the end of the period and the population that might have resulted from natural increase since the beginning of the period, i.e. by assuning a growth rate equaL to the national average. -09 - Annex I Page 9 of 18 Table 1.9 ORIGIN AND DESTINATION OF INTER-REGIONAL MIGRANTS (000'S) IN 1981 Destinations Origin Mountain Hill Terai Out- % FWMW W C E FW MW W C E FW NW W C E Migrants Mountain Farwest (FU) - 0.6 - - - 1.8 - - - - 9.9 - - - * 13.2 1.3 Midwest (MW) 0.7 - - - - 0.8 1.2 0.5 - - 1.5 0.7 - - - 5.9 0.6 Western (U) - - - - - - 0.1 5.5 1.4 - 0.7 1.2 22.7 6.0 - 38.2 3.7 Central (C) - - - - 0.1 - - - 10.2 - - - - 8.7 1.1 21.7 2.1 Eastern (E) 6.2 4.7 - 5.6 - 8.0 18.3 33.9 32.6 18.2 5.7 6.4 9.6 16.9 69.9 236.0 22.7 Hill Farwest (FU) 1.3 - - - - - 2.4 - - - 41.3 1.0 - - - 47.4 4.6 Midwest (MW) 0.7 0.9 - * 0.6 0.9 - 1.5 0.6 - 12.5 31.1 4.5 0.7 - 54.5 5.2 Western (U) - - 0.6 - - - 1.6 - 5.5 - 4.1 7.0 64.3 63.2 2.6 150.1 14.4 Central (C) - - - 3.8 1.0 - 0.8 4.4 - 3.8 2.5 2.0 4.0 83.3 12.8 119.3 11.5 Eastern (E) 1.4 - - - 21.3 0.9 2.6 5.4 9.8 - 2.3 1.4 1.7 15.3 202.6 265.4 25.5 Terai Farwest (FW) - - - - - - - - - - * - - - - 1.2 0.1 Midwest (MW) - - - - 1.0 - 1.0 - 9.5 - - 0.7 - 13.2 1.3 Western (W) - - - - - - - 1.2 0.8 - - 0.6 - 0.8 - 4.7 0.4 Central (C) - - - - - - 0.7 6.9 5.1 0.7 0.7 1.9 2.7 - 9.7 29.1 2.8 Eastern (E) - 0.1 - - 0.6 - 2.4 2.4 4.2 7.4 1.1 0.7 1.0 18.1 - 39.0 3.7 In-Migrants 10.9 7.3 1.1 10.4 23.9 14.6 31.5 62.6 71.9 31.4 92.2 54.4 111.4 214.5 300.8 1038.9 (%) 1.1 0.7 0.1 1.0 2.3 1.4 3.0 6.0 6.9 3.0 8.9 5.2 10.7 20.7 29.0 100 SOURCE: CBS, Population Census-1981, Vol. II, IV, Kathmandu, 1984 Gurung, H., Regional Patterns of Migration in Nepat, 1989 t ALL origin-destination fLows of Less than 500 have to be neglected. Urbanization 16. As is to be expected, given the small proportion of the workforce occupied in activities outside of agriculture, the level of urbanization in Nepal is still very low, among the lowest in the world. At the time of the 1952/4 census, only 2.9% of the population lived in urban areas (Table 1.10). By 1971, the proportion had risen to 4.0%, and by 1991, the urban population was up to 9.1%. As in neighboring India, the rate at which the country is urbanizing is slow, though it accelerated between 1971 and 1981 with the rate of growth of urban population rising to 7.6% per year in the decade to 1981, subsequently however, it appears to have dropped back to 5.8% p.a. between 1981 and 1991. Urban growth in Nepal is ultimately restricted by the slow growth of demand for goods and services. Exports and tourism, despite their high rates of growth, are not yet large enough to affect the urban sector as a whole. The domestic market is constrained by low incomes, poor domestic transport, poor infrastructure, a high import content to domestic output, high import transit costs and labour force of poor skills. 17. Increasing urbanization is reflected by the rising importance of in- migration as a contributing factor in the growth of urban population. Based on the assumption that the natural rate of population increase in urban areas is similar to the growth rate of the total population, natural increase during the decade 1961-71 is estimated to have accounted for nearly two-thirds of the increase in urban population, and in-migration the remainder of a little more than one third. (Since no new urban areas were established during that decade, the increase from that source was zero.) In the ten years after 1971, however, - 10- Annex I Page 10 of 18 in-migration accounted for nearly 40% of urban growth, and in the period 1981- 1991 this proportion rose even higher to 45%. With almost a third of the urban additions since 1971 being due to newly classified urban areas, the contribution of natural increase has been a relatively minor 30% share of the total, down from a high of over 60% in the decade 1961-71. Table 1.10 URBANIZATION 1952-1987 Variable 1952 1961 1971 1981 1991 Population (000s) Total 8257 9413 11556 15023 18462 Rural 8018 9074 11094 14066 16780 Urban 238 339 462 957 1682 Population Shares Total 100.0% 100.0% 100.0% 100.0% 100% Rural 97.1% 96.4% 96.0% 93.6% 90.9% Urban 2.9% 3.6% 4.0% 6.4% 9.1% Annual Growth Rates 2.1% Total 1.5% 2.1% 2.7% 1.8% Rural 1.4% 2.0% 2.4% 5.8% Urban 4.0% 3.1% 7.6% Urban Increase (000s) Natural increase* 33 77 139 219 New urban centers N/A 0 161 180 (Number of centers) (10) (16) (16) (23) (33) In-migration N/A 46 195 326 Total 101 123 495 725 Shares of Urban Increase Natural increase 32.7% 62.6% 28.0% 30.2% New urban centers N/A .0% 32.5% 24.8% In-migration N/A 37.4% 39.5 45.0% Total 100.0% 100.0% 100.0% 100.0% * Assuning the same growth rate as for total population Source: CBS Population Monograph (1987) and Preliminary Results, 1991 Census. 18. Urban population is concentrated in the Central Hills, and in the Central and Eastern Terai, which together account for more than 70% of the total urban population (Table 1.11). None of the Mountain zones has any urban areas. Over the 1981-91 decade, urban population in Nepal has increased by 75 percent (Gr. Fct. of 1.75), with the Terai experiencing the strongest increase. Urban areas in the Central Terai more than doubled their population (Gr. Fct. 2.12) in these ten years, while in Eastern and Western Terai the increases were about 80 and 70 percent respectively. The main features of shifts in the urban population shares is the marked decline of the Central Hill area, where the share of total urban population fell 17 points from 58% to 41% in the period 1971-91, and the corresponding gains in the Terai, mainly in the Center (7.6%), but also to the East (1.3%), and surprisingly in the Far-West (6.4%) due to the rapid growth of two towns, Mahendranagar and Dhangadhi. -11 - Annex I Page 11 of 18 Table 1.11 REGIONAL SHARES AND GROWTH FACTORS OF URBAN POPULATION 1971-1991 Far- Mid- Western Central Eastern TotaL Western Western HiLL 1971 0.0% 0.0% 5.9% 57.5% 1.6% 65.0% 1981 0.0% 1.5% 6.3% 41.6% 2.5% 51.8% 1991 0.7% 1.4% 6.4% 40.9% 1.8% 51.3% Gr. Fct. 81-910 0.0 1.64 1.79 1.73 1.27 1.74 Terai 1971 0.0% 5.1% 6.5% 5.9% 17.5% 35.0% 1981 7.4% 7.2% 5.6% 11.1% 18.4% 48.2% 1991 6.4% 4.6% 5.4% 13.5% 18.8% 48.8% Gr. Fct. 81-91 1.51 1.12 1.68 2.12 1.79 1.77 TOTAL 1971 0.0% 5.1% 12.4% 63.4% 19.1% 100% 1981 7.4% 5.7% 11.9% 52.7% 20.8% 100% 1991 7.1% 4.6% 11.9% 54.4% 20.6% 100% Gr. Fct. 81-91 1.68 1.42 1.75 1.81 1.74% 1.75 TOTAL (000s) 1971 0 24 57 293 89 463 1981 71 69 114 504 199 957 1991 120 101 200 915 346 1682 Gr.Fct.: Ratio of 1991 urban population to that in 1981. The Growth of Urban Centers 19. Unlike many developing countries, the urban settlement system in Nepal is not dominated by a single large metropolis, but is characterized by a hierarchy of cities and towns that adhere to a smoothly declining curve of city size. In 1987, according to a sample survey of urban centers made that year, Nepal had two cities with populations greater than 100,000, seven cities in the range 50,000 to 100,000, and twenty-three urban centers in the range 10,000 to 50,000 people. Kathmandu-Lalitpur with a population of just over 400,000 was two and a half times the size of the second city, Biratnagar, and four times the size of the third city, Birganj. A measure of primacy, based on the four largest cities, declined from 0.66 in 1971 to 0.57 in 1987, which is well below the figure of 0.8 to 1.2 typical of a primate city structure. Of the 10 largest cities, five are in the Central region, two in the East and West, and one in the Far-West region. 20. The mutual interdependence of agricultural production and urban growth, especially at early stages of urbanization, is well illustrated in the case of Nepal, where areas with prosperous agriculture have also experienced rapid growth of urban population. Of the 16 urban centers which grew as fast or faster than the median rate of 5.8% per year during the period 1971 through 1987, 11 were located in zones where agriculture was strong (Table 1.12), i.e., those with above average increases in output, as measured by shifts in the value of production of four major commodities in the period 1975 through 1986. The four exceptions were Rajbiraj, (which may have been subject to a boundary change), and three Terai towns - Tansen, Butwal and Siddarthanagar, all in the Western region. On the other hand, thirteen of the sixteen slower growing urban centers were located in zones where agricultural output also increased more slowly. The three exceptions in this group were Kalaiya and Jaleswor in the Central Terai, and Dhankuta in the Eastern Hills. It should also be noted that of the 16 towns growing faster than the median, 14 were located in the Terai; whereas, of the - 12 - Annex I Page 12 of 18 towns growing at rates below the median, more than half are towns located in the Hills. Also interesting is the relationship of gateway (for imports/exporte) towns with urban growth; of 15 towns categorized as gateways, 10 grew faster than the median rate. Table 1.12 GROWTH OF URBAN CENTERS 1971-1987 Town Geog. Gate- Dev. Zone Ag Change Mfg 71-87 Region Way Region 76-86 Change Growth 81-86 Rate Mahendranagar T G FW * Mahakatt + (-) 13.9% Birganj T G C Narayani + (+) 13.5% Baratpur T C * Narayani + (+) 11.4% Rajbiraj T G E Sagarmatha - (-) 11.3% Janakpur T G C Janakpur + (-) 9.71 Tansen H W Lumbini - (-) 8.1% Pokhara H G W Gandaki + C-) 8.01 Tribhuvannagar T MW * Rapti + - 7.91 Dharan T E Koshi + (+) 7.8% Biratnagar T G E Koshi + (+) 7.7% Netauda T C Narayani + () 7.6% ButawaL T U Lumbini - (*) 7.2% Inaruwa T G E ** Koshi + (+) 7.1% Siddarthanager T G U Lumbini - (-) 6.7% MaLangawa T G C ** Janakpur + (-) 5.9% Dhangadhi T G FW * Seti * (.) 5.8% Damak T E ** Mechi - - 5.0% Bhadhrapur T G E Mechi - - 4.2% Kathmandu-LaLitpur H G C Bagmati - + 4.2% KaLaiya T C ** Narayani + (+) 4.2% Lahan T E * Sagarmathe (-) 3.8% Nepalganj T G MW Bheri - C-) 3.8% I1am H E Mchi - - 3.2% Banepa H C ** Bagmati + 3.2% Dipayal H FU ** Seti - C-) 3.1% Dhankuta H E * Koshi + (+) 3.1% TauLihawa T G W 00 Lumbini - C-) 2.8% Bidur H C ** Bagmati - + 2.3% Bhaktapur H C Bagmati - + 2.2% JaLeswor T G C 00 Janakpur + (-) 2.2% Dhutikhet H C B Bagmati -+ 1.5% Birendranagar T NW * Bheri - C-) N/A TOTAL I 1 1 6.1% SOURCE: 1. Ministry of Housing and PhysicaL PLanning, Recommended PoLicies and Strategies for Urban Development, 1991 2. CBS, PopuLation-1981, VoL. III, Kathmandu, 1984 3. USAID, NepaL Urban DeveLopment Assessment, 1984 NOTES: 0 Not Urban Areas in 1971 ** Not Urban Areas in 1971 and 1981 E Eastern, C: CentraL, W: Western MW Mid-Western, FW: Far-Western + = Positive shift in 3 indicators - negative shift in 3 indicators (+) = Positive shift in 2 indicators (-) = negative shift in 2 indicators 21. Given the incipient state of manufacturing in Nepal today, it is not surprising to find that the performance of the sector has little impact on the growth of urban population, even in zones with major shares of manufacturing activity. The most noticeable examples are the five cities in Bagmati zone. Even though these cities collectively increased Bagmati's share of total - 13 - Annex I Page 13 of 18 manufacturing employment by 17 percentage points in the five year period 1981/82- 87/88, all five experienced population growth rates below the median. 22. It appears, then, that at cNepal's present stage of development, agriculture more than manufacturing is still the primary local economic base that determines the growth of population in urban centers, and implicitly therefore the growth of so called "secondary" activities, i.e., those such as commerce and services which are driven by income initially earned in agriculture. C. Household Demand 23. In order to assess the demand for infrastructure services, it is necessary to understand the determinants of demand in the overall economy. This section analyses demand in the Nepalese economy by examining the two facets of domestic demand - the household demand for goods and services as evidenced by household consumption, and household incomes. Household Consumption and Incomes 24. As might be expected in a nation where incomes are still low, the bulk of household spending goes to the basics of food, shelter, and clothing (see Table 1.13). Taking all households together, over 60% of spending goes to food and drink, 17% for housing and furnishings, and 12% for clothing and footwear. Three inter-related factors affect patterns of consumption. One is the shift of population out of the Mountains down to the Terai. Another is the movement of population from rural areas to cities and towns. The third is the rise in income levels. Future trends in consumption can be inferred from spending patterns by different types of households, and the relative changes in the proportions of various types of households in the total population. Table 1.13 COMPOSITION OF HOUSEHOLD CONSUMPTION BY CATEGORY AND TYPE (PERCENT) Expenditure Category ---------- Rural---------- ----------Urban---------- ALL Terai HiLLs Mountain AlL Terai HitLs Nation Food Drink & Tobacco 62.2 62.5 61.9 62.2 51.0 55.1 48.8 61.2 Clothing & Footwear 11.9 11.5 12.2 11.8 10.1 10.1 10.1 11.7 Housing and Furnishings 16.2 14.8 16.5 19.2 27.6 23.9 29.6 17.3 Medical & Personal Care 3.8 4.8 3.3 1.7 3.9 4.9 3.4 3.7 Education & Recreation 2.7 3.1 2.9 1.6 4.1 3.2 4.6 2.9 Transport & Comnunications 1.1 1.5 .9 .8 2.2 1.8 2.4 1.2 Other 2.1 1.8 2.3 2.7 1.1 1.0 1.1 2.0 ALL Spending 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Total MonthLy Spending 1092 1137 1068 1041 1575 1211 1887 1126 (1985 NRs) I I SOURCE: Multipurpose HousehoLd Budget Survey, Section G, 1988. 25. For rural households, the 1985 data indicates that there was no great difference in spending levels between households in the Mountains, Hills and Terai, the average monthly amounts rising only slightly from NRs 1041 in the Mountains to NRs 1137 in the Terai. Neither was there much change in the proportions spent on food and drink and clothing, which remain fairly stable at around 62% for the former and 12% for the latter. There is however a drop in the -14 Annex I Page 14 of 18 amount spent on housing and furnishings, down from 19% in the Mountains to 15% in the Terai, perhaps because the need for sturdy construction in the low lands to protect against the elements is less critical. Instead, spending increases on medical care, education, and transportation, which was almost double in the Terai no doubt in part because it is more accessible there. 26. The second factor affecting household consumption is the process of urbanization. Compared with rural households, urban households spent relatively less on food and drink, down from 62% to 51% of total spending, but substantially more on housing, which rises from 16% to 28%, due both to higher costs of shelter in urban areas, and rising expectations (see table 1.13). Urban households also spent more on education, up from 2.7% to 4.1%, and transportation, which doubles from 1.1% to 2.2%. While urban households spent relatively less on food than rural households, they also consumed different kinds of food. Urban households spent relatively less on cereals (50% versus 64%), particularly maize (down from 11.5% to 1.0%) and millet (down from 3.5% to 0.3%). Instead they spent relatively more on other kinds of food, especially fruit and vegetables (up from 9.2 to 12.4%), meat and fish (up from 4.2 to 6.0%), sugar, tea and coffee (up from 1.5 to 3.7%) and meals away from home (up from 3.5 to 8.3%). 27. The third factor affecting household consumption is rising incomes. While food as a group has a low income elasticity of demand for both rural and urban households, 0.38 and 0.32 respectively, meat fish and eggs are superior goods with elasticities close to 1.5 (see Table 1.14). As a group non-food items are also superior goods, with income elasticities of 1.12 for rural households and 1.20 for urban households. Within the group of non-food goods and services, however, there is wide variation between rural and urban households. While Table 1.14 COMPOSITION AND ELASTICITY OF HOUSEHOLD CONSUMPTION BY LEVEL OF PER CAPITA INCOME (PERCENT) Expenditure Category ------Rural----- ------Urban------ All Second Ninth Income ALL Second Ninth Income Decite Decile Elasticity DeciLe DeciLe Elasticity Food Drink & Tobacco 62.2 69.1 56.6 .38 51.0 62.0 47.0 .32 Grain & Cereal Products 40.8 48.2 35.0 .26 26.5 40.0 19.1 -.06 Vegetables Fruits & Spice 7.3 8.5 6.4 .30 7.5 8.5 7.2 .44 Meat Fish Eggs & Milk 2.5 4.2 7.0 1.46 2.9 4.3 7.1 1.51 All Other Food & Drink 11.6 8.2 8.2 .60 14.1 9.2 13.6 1.33 Other Goods & Services 37.8 30.9 43.4 1.12 49.0 38.0 53.0 1.20 Clothing & Footwear 11.9 8.2 15.3 1.71 10.1 8.4 10.0 .91 Housing & Furnishings 16.2 15.5 16.8 .72 27.6 21.7 29.3 1.13 Education & Recreation 2.7 1.3 3.3 2.53 4.1 1.8 5.8 3.71 Other 7.0 5.9 8.0 1.07 7.2 6.1 7.9 1.06 All Spending 100.0 100.0 100.0 100.0 100.0 100.0 TOTAL MONTHLY INCOME 1176.0 693.0 1733.0 1785.0 1062.0 2538.0 (1985 NRs) TOTAL MONTHLY SPENDING 1092.0 766.0 1467.0 1575.0 1103.0 2105.0 (1985 NRs) SOURCE: Multipurpose Household Budget Survey, Section G, 1988 education and recreation is a superior good for both, clothing is a superior good for rural households (1.71) but not for urban ones (0.91), while the reverse is true for housing and furnishings, 0.72 compared to 1.13. -15 - Annex I Page 15 of 18 Regional Shares of Household Demand 28. Table 1.15 provides an estimate of the regional distribution of household demand in 1991. The estimate is based on the number of households in Table 1.16 REGIONAL SHARES OF HOUSEHOLD DEMAND 1981 VariabLe De eLopment Regions Total Farwest Midwest West Central Eastern Mountain Rural HousehoLds (000s) 62 48 5 95 68 276 HH Spending Index* 93 93 93 93 93 93 Purchasing Power Index 6 5 0 9 6 26 Percent Total 1.8% 1.5% 0.1% 2.7% 1.8% 7.8% NHits Rural Households (000s) 122 218 445 361 257 1402 HN Spending Index* 95 95 95 95 95 95 Purchasing Power Index 12 21 42 3 24 133 Percent Total 3.6% 6.3% 12.5% 0.9% 7.2% 39.7K HilLs Urban Households (000s) 2 5 23 128 6 165 MH Spending Index* 168 168 168 168 168 168 Purchasing Power Index 0 1 4 22 1 28 Percent Total 0.1% 0.3% 1.2% 6.6% 0.3% 8.4% Teral Rural Households (000s) 84 133 203 500 436 1355 HN Spending Index* 102 102 102 102 102 102 Purchasing Power Index 9 14 21 51 45 138 Percent Total 2.7% 4.2% 6.3% 15.2% 13.4% 41.2% Terai Urban Households (000s) 17 14 17 40 59 147 NH Spending Index* 108 108 108 108 108 108 Purchasing Power Index 2 2 2 4 6 16 Percent Total 0.6% 0.6% 0.6% 1.2% 1.8% 4.B% Total Rural Households (00s) 268 399 650 870 761 3033, NH Spending Index* 96 97 97 98 98 98 Purchasing Power Index 26 39 63 85 75 297 Percent Total 7.8% 11.6% 18.8% 25.4% 22.4% 88.6% Total Urban Households (000s) 20 19 40 168 65 312 NH Spending Index* 108 108 144 150 116 136 Purchasing Power Index 2 2 6 25 8 42 Percent Total 0.6% 0.6% 1.8% 7.5% 2.4% 12.5% TotaL Region Households (00s) 287 419 693 1124 826 3345 NH Spending Index* 97 97 99 103 99 100 Purchasing Power Index 28 41 69 116 82 335 Percent Total 8.4% 12.2% 20.1% 34.6% 24.5% 100.0 *Household Spending is expressed by an index estimated using total household spending as the base (100) SOURCE: Household - Population Census - 1981 Vol. I, III, Kathmandu, 1984, Table 1 Household Spending: Multipurpose Household Budget Survey, 1988. -16- Annex I Page 16 of 18 each geographical zone weighted by household spending in each zone as reported in the 1984/85 Multipurpose Household Budget Survey. Since the survey distinguishes between only five categories - rural and urban, Mountain, Hill and Terai - it is assumed that spending levels are constant across all development regions in the same geographical regions. In practice, data on regional domestic product (see Table 1.2) suggest that per capita income levels and therefore consumption levels are higher in the Central and Eastern regions. This assumption, therefore, will tend to overstate the shares of the poorer regions to the west. 29. The analysis shows that rural households accounted for about 88% of final demand in 1981, and urban households account for 12%. Given the low level of urbanization of the country, the rural sector will clearly dominate final demand for some time to come. Households in the Hills and Terai generate approximately equal shares of demand, 48% and 46% respectively, the balance of 6% coming from households in the Mountains. But as the center of gravity of population gradually shifts towards the Terai, these shares can be expected to change significantly. In terms of development regions, the Central region dominates the picture with at least a third of final demand (35%), followed by the Eastern region with a quarter (25%). The Western, Mid-West, and Far-West regions account for 20%, 12% and 8% respectively, though as mentioned before this is probably an overstatement. By comparison, it has been estimated that in 1984/85 these three western regions accounted for only 16%, 8% and 5% of gross domestic product. Rural and Urban Sources of Household Income 30. The importance of rural-urban linkages in Nepal is underlined by an analysis of the composition of household incomes. This reveals that no matter where they are located, households derive income from both rural and urban sources. This is particularly evident among urban households, which gain almost a fifth of their total income from farming, a proportion which rises to as much as a quarter among urban households in the Terai (Table 1.16). 31. Among rural households, farming provides 64% of total income, but 36% comes from other sources. Although the Household Survey does not explicitly identify the location of these other sources, it may be inferred that a good part must come from urban activities. Wages and salaries, for example, make up nearly 16% of total rural household income. Based on experience in other countries, this typically comes from working on other people's farms, from domestic service, and from teaching, but also from employment in nonfarm businesses and the public sector, much of which is located in small towns nearby. Income from nonfarm enterprises contributes a further 5.9% of rural incomes. Part of this no doubt comes from activities located in rural areas, such as food processing and handicrafts, but typically it also includes small businesses located in urban areas, such as shops, cafes, commodity trading, and transport services. The last category "other" includes income from property, and again evidence from other countries suggests part of this represents rents from premises owned by farmers in nearby towns. - 17 - Annex I Page 17 of 18 Table 1.16 COMPOSITION OF HOUSEHOLD INCOME BY SOURCES AND REGION (%) Source of Income Rural Rural' RuraL Rural Urban Urban Urban All Terai Hills Mountains All Terai HilLs Wages and salaries 15.7 14.2 17.2 15.9 31.8 30.0 33.0 Enterprise Income: Farm 63.8 67.9 59.1 63.9 19.3 25.1 15.4 Non-Farm 5.9 6.3 7.2 4.0 21.4 23.3 20.1 Transfers 6.6 3.7 8.7 7.8 4.8 4.4 5.1 Other 8.0 7.8 7.7 8.3 22.7 17.2 26.4 ALL income 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Total Monthly Income 1176 1287 .1125 1116 1758 1408 2108 (1985 NRs) I I I I SOURCE: Multipurpose Household Budget Survey, 1988, Table 1, P.82 D. Projections on Population Distribution and Urbanization 32. In projecting the distribution of population among the five development regions and three ecological belts, the overall population was first projected based on a growth rate of 2.08 percent p.a., which was the rate recorded for 1981 to 1991. Various methods for spatial disaggregation of these projections were tried, and the one adopted was based on first projecting for each of the 15 zones the rural and urban populations based on their individual historical growth rates in each zone over the period 1981-1991. The individual populations were then adjusted proportionately to conform with regional and national total populations. Table 1.17 presents the results of this projection. Table 1.17 REGIONAL SHARES AND PROJECTED TOTAL POPULATION Region Far-Western Mid-Western Western Central Eastern Zonal Share % TotaL Mountain 1991 1.8 1.4 0.1 2.6 1.9 7.8 1995 1.7 1.3 0.1 2.5 1.8 7.4 2000 1.6 1.2 0.1 2.3 1.7 6.9 2010 1.3 1.0 0.0 2.1 1.4 5.8 Hill 1991 3.6 6.6 13.1 14.5 7.7 45.6 1995 3.4 6.3 12.5 14.4 7.4 44.0 2000 3.1 6.0 11.9 14.2 6.9 42.1 2010 2.4 5.3 10.6 13.8 6.0 38.1 Terai 1991 3.7 5.0 7.1 16.4 14.4 46.6 1995 4.2 5.4 7.5 16.6 14.9 48.6 2000 4.9 5.8 8.0 16.9 15.4 51.1 2010 6.6 6.7 8.8 17.6 16.4 56.1 Total 1991 9.1 13.0 20.3 33.4 24.1 100.0 1995 9.3 13.0 20.1 33.5 24.0 100.0 2000 9.6 13.1 19.9 38.5 24.0 100.0 2010 10.3 13.0 19.4 33.5 23.8 100.0 TotaL (000) 1991 1681 2406 3752 6174 4448 18462 1995 1867 2616 4038 6707 4821 20049 2000 2132 2901 4425 7439 5329 22226 2010 2814 3558 5304 9147 6491 27314 Since this method is based on an extrapolation of past trends, the projected figures reflect this. Assuming past trends continue, the Terai will absorb an increasing share of the nation's population, rising from 47% in 1991 to 51% in -18- Annex I Page 18 of 18 2000, and 56% by the year 2010. Likewise, the Hills will continue losing its share of population, down from 46% in 1991 to 38% by the year 2010, as will the Mountains, down from 7.8% in 1991 to 5.8% in 2010. In reality, these trends may slow down further, as countervailing factors gain momentum, and gradually lessen the comparatively better opportunities for employment and farming in the Terai. Projected shifts among regions are likely to be minor, the main trend being further movements out of the Western region to the Far-West. The Growth of Cities 33. The figures for projected urban and rural population in each region and zone based on the method described above were used to estimate the projected population of individual cities in the year 2000. The procedure used is as follows. The future population of each city is first calculated using its historic growth rate for the period 1971-1991. The increase in population in each city is then adjusted such that the sum of increases of all the cities, within each zone matches the projected increase in total urban population for that zone. The results suggest that in nearly every city the rate of population growth over the period 1991 to 2000 will be lower than that experienced between 1971 and 1987. Taken together, the population of the 32 urban centers is projected to grow 4.8% per year, a little less than the 5.2% per year recorded for the period 1971-91. The growth rate of the urban population of the nation as a whole may be higher, due to the reclassification of rural places crossing the threshold as urban centers. Table 1.18 presents population projections for the 12 largest urban centers. Table 1.18 ACTUAL AND PROJECTED POPULATION OF URBAN CENTERS P pulation (000s) Annual Growth Rates (%) Town Region 1971(A) 1981(A) 1991(A) 2000(P) 1971-91(A) 1991-2000(P) Kathmandu-Latitpur C 209.5 375.0 531.5 752.7 4.8 3.9 Biratnagar E 45.1 93.5 130.1 198.7 5.4 4.8 Pokhara U 20.6 46.6 95.3 172.7 8.0 6.8 Birganj C 13.0 43.6 68.8 124.8 8.7 6.8 Dharan E 20.5 42.2 68.2 110.4 6.2 5.5 Mahendranagar FW * 6.9 43.8 62.4 141.7 11.6 9.5 Shaktapur C 40.1 48.5 61.1 71.3 2.1 1.7 Janakpur C 14.3 34.8 55.0 88.6 7.0 5.4 Bharatpur C * 6.8 27.6 54.7 116.7 11.0 8.8 Netauda C 16.2 34.8 54.1 82.5 6.2 4.8 Nepalganj MU 23.5 34.0 48.7 70.2 3.7 4.2 Dhangadhi FU * 11.8 27.3 45.1 73.2 7.0 5.5 Other 20 Urban Centers 181.2 241.8 407.3 564.4 4.1 3.3 TotaL for 32 centers 609.5 1093.5 r682.3 1 2,567.9 5.2 4.8 Notes: *!Not an urban area in 1971, E - Eastern, C - Central, U - Uestern, MW - Mid-West, FU - Far-West 34. Based on past performance, Kathmandu-Lalitpur will reach a population of 752,000 in the year 2000, up from 531,000 in 1991. According to this method of projection, Mahendranagar will emerge as the fourth largest city with 142,000 inhabitants, reflecting its rapid growth between 1971 and 1981. In order of size, other major cities with populations over 100,000 will include: Biratnagar with close to 200,000, Pokhara (173,000), Birganj (125,000), Bharatpur (117,000), and Dharan (110,000). Of the 12 largest cities in the year 2000, four would be in the western parts of the country, two in the Eastern region, and the remaining six would be located in the Central region. -19- ANNEX II Page 1 of 19 GEOGRAPHIC INFORMATION AND ANALYSIS 1. Nepal is a country undergoing significant change in the spatial dimensions of economic activity and population distributions. In order to comprehend these changes and their implications for infrastructure development it is especially useful to examine the economic and demographic geography of the country. This annex presents information on the spatial distribution of various economic and demographic parameters with a view to identifying patterns of development or activity which could guide infrastructure development strategy. The analysis presented here was carried out using a Geographic Information System based on the TerraSoft program. A. Demographic Patterns 2. There are 75 districts in Nepal of which 20 districts form the Terai and lie as a band along the southern border of the country. The 16 Mountain districts are located in the northern section of Nepal and 39 Hill districts lie between Mountain and Terai districts. The Terai comprises about 23% of total land area, the Hill region 42% and the Mountain region 35% (see also IBRD Map No. 23949). Figure 2.1 shows the distribution of population in 1971 and 1991, and a projection for the year 2010 based on historical trends both in the natural increases in population as well as observed migration patterns in the past. The maps show the spatial distribution of population densities per square kilometer by district. 3. In 1971, the population was fairly evenly distributed over the Hill and Terai districts, with the only significant concentrations occurring in the Kathmandu Valley. Only the valley districts of Kathmandu and Bhaktapur had population densities greater than 500 persons/sq. km. A few additional districts, such as Lalitpur in the Valley, and Mahottari in the Terai had densities greater than 300 persons/sq. km. Most parts of the country, including the Terai had densities below 200/sq. km. The Far-Western Terai in particular had densities considerably less than 50 persons/sq. km. 4. By 1991, the population distribution had undergone major shifts, in general favoring the Terai. Most Terai districts in the Eastern and Central development regions had crossed the threshold of 300 persons/sq. km. In the Western Terai population densities had increased to over 100 persons/sq. km. This has been a particularly significant change in the Far-Western Terai, with districts such as Kanchanpur and Kailali almost quadrupling their populations from densities in the range of 40 persons/sq. km. to densities in the range of 130 to 150 persons/sq. km. in 1991. Similarly, in the Eastern Terai, districts such as Morang and Jhapa more than doubled their populations in these 20 years, with densities increasing from around 160 persons/sq. km. to around 380 persons/sq. km. Growth in the Hills was much less pronounced and took place in only a few locations. The Kathmandu-Lalitpur-Bhaktapur complex which forms the Kathmandu Valley increased its population by about 70 to 80 percent. The only other Hill districts with significant increases were the districts in the Central region near Kathmandu, and a few districts in the Western and Eastern Hills. - 20- ANNEX i Page 2 of 19 Figure 2.1 Distribution of Population 1971r 1991. 2010 IBRD 24679 NEPAL DISTRIBUTION OF POPULATION 1971, 1991, 2010 Population Density, 1971 J, POPULATION DENSITY - PERSONS/KM2 <10/KM2 Population Density, 1991 10-50/KM2 50-1 00/KM2 100-200/KM2 200-300/KM2 - 300-500/KM2 - >500/K- Population Density, * >500/KM2 -(Projected), 2010 Th-s mop hon beon p ayd by T he Word Bonk s o euweiy hor Boheconemn.e el.oders and,ab I. Bo ensemoJ us e The Warid Bonk Grou,p Th. dénon.nanon, used ond me boundone.s shon on k . d no, on h. pai Th. Word Bonk Gro-pwony p,dgmnt on e legl sensus al ony earna ry ony ondorsmont or oceponC 19 .uCh bon,e MARCH 199 i -21 - ANNEX II Page 3 of 19 5. Projecting these past trends into the future, would indicate very substantial shifts in population into the Terai. The map showing population distribution for the year 2010 should perhaps be viewed as an exaggeration of the likely future distribution, since some of the momentum behind past migration is likely to reduce and therefore moderate the extreme changes shown in the map. However, an equally convincing case could be made that the forces behind this migration have only begun to take effect and therefore the projections understate the impact of migration on future population distribution in Nepal. Based on past trends it would appear that the Terai will increasingly become a continuous population belt with the lower density Western Terai growing rapidly to equal the densities in the Eastern and Central Terai. By 2010, these projections indicate that most of the districts in the Eastern and Central Terai will cross 500 persons/sq. km., and districts in the Far-Western Terai will cross 300 persons/sq. km., while other Terai districts will generally have population densities in excess of 200 persons/sq. km. The other noteworthy feature of the projected 2010 population distribution is the formation of a contiguous belt of high population densities in the Terai and the Central Hill districts indicating further integration of the Kathmandu Valley with the Terai. 6. To understand the shifts in population it is useful to examine the spatial distribution of output in Nepal. Figure 2.2 shows the distribution of GDP as an index. The index was constructed to show output of the economy in relation to land area, since this allows representation of the spatial dimension of output. The total GDP of the economy divided by the total land area was taken as the base (Index = 1.00). Each district's GDP was computed based on per capita GDP for the zone in which the district is located Y, times the population of the district, and this GDP was divided by the land area of the district. The index for each district is obtained as the ratio of the GDP/Land Area for that district to the national average. Thus, assuming relative levels of GDP per capita remain the same (as in 1984/85) among the 15 regions, Figure 2.2 shows the distribution of GDP output for 1981, 1991 and 2010. 7. For 1991, the indices range widely from about 0.01 in Dolpa to over 17 times the national average in Kathmandu. By 2010, these projections indicate even greater concentration of GDP, with the highest index reaching 27 (in Kathmandu). The trend would therefore appear to be that development, in the future, rather than expanding spatially, is more likely to localize in the Central Hill and in the Terai. It should be noted however that these observations are based on the strong differences in per capita GDP among the 15 zones, that existed in 1984/85. Migration from areas of lower output to areas of higher relative productivity could reduce these differences in per capita GDP levels. Whether or not such changes in relative regional per capita GDP levels occur, will partially depend on the extent to which new opportunities in the Terai and Central Hill regions continue to exist and be exploited by Per capita GDP for the 15 regions was taken from *Basic Features of the National Development Master Plan (1990-2000)", National Planning Commission, HNGN, Kathmandu, October 1989. - 22 - ANNEX II Page 4 of 19 Figure 2.2 spatial Distribution of GDP 1971. 1991. 200 IBRD24680 N E PAL SPATIAL DISTRIBUTION OF GDP 1971, 1991, 2010 Spatial GDP Index, 1971 GDP INDEX 1< 0 5 --Spatial GIDP Index, 1991 0.5______ 0 k--- - ~ 11 0-20 20-30- >5.0) Spatial GDP Index, 2010 Note Spatial GDP Index is constructed as ratio of output per unst area for a district to output per unit areo for Nepal, e Average p5ptial GDP Index for Nepol 1 0 ih.'flwpho~benprwor byThemw&Or id a a cnL~ kwyfo # =_ &i -d--d., . # rne30-50'Smence & edr'nwslrte nra s h ol MARCH 199 3 - 23 - ANNEX II Page 5 of 19 the new migrants into these areas. With the Western Terai, in particular being relatively undeveloped, and increasing diversification of economic activity in the Central Hills (industrialization in the Kathmandu Valley) such opportunities would continue to exist and the differences in per capita GDP could persist for some time into the future. 8. Spatial concentration, as opposed to dispersion, of economic activity in the future has strong implications for the provision of infrastructure. A spatially expanding economy poses fewer risks in the provision of infrastructure in various parts of the country. Future concentration of activity in some regions poses greater risks since this could cause infrastructure provided in response to present needs to be under-utilized or abandoned in the future as the loci of economic activity shifts to other parts of the country. In Nepal, this appears to be the case, with economic activity shifting out of most of the Hill and Mountain areas into the Central Hills and the narrow southern band of Terai districts. 9. Migration patterns appear to strongly correspond to the spatial distribution of GDP. Figure 2.3 shows the distribution of origins and destinations of migrants. The number of migrants in each district was estimated by taking the difference between the actual population and the population which would have resulted from the district population growing at the natural rate of population increase for the country as a whole. 10. Between 1971 and 1981, there were 44 districts which appear to have experienced out-migration according to the above estimation procedure. In the following decade, the number of districts experiencing out-migration increased to 49. This expansion in the areas from which people are migrating contrasts with a decline in total migrants from 1.3 million (8.7% of 1981 population) in the 1971-81 decade to 0.83 million (4.5% of 1991 population) in the 1981-91 decade. The expansion in areas with out-migration, despite the decrease in overall numbers of migrants, further strengthens the conclusion that the Nepalese economy is concentrating in a few areas of the country, primarily the Terai and the Central Hills south of the Kathmandu Valley. 11. Also noteworthy, is the dominance of the Far-Western Terai in the share of total in-migration, and the relatively lower increases in in-migration in the already densely populated Eastern and Central Terai districts. The three districts of Kanchanpur, Kailali and Bardiya in the Far and Mid-Western Terai attracted about 25 percent of total migrants during the period 1981-91. By comparison, the three districts in the Eastern and Central regions which attracted most in-migrants were Sunsari, Chitwan and Parsa and they accounted for only about 12 percent of total migrants. Terai districts in the Western region and some of the Mid-Western region attracted fewer migrants, and these are the districts where the East-West Highway remains to be completed, and infrastructure provision is the least developed compared to rest of the Terai. In the future, with completion of the East-West Highway these Terai districts may be expected to receive larger numbers of migrants. The only other area, apart from the Far- Western Terai, which has been a major destination of migrants is the Kathmandu Valley which received a little over 20 percent of total migrants. - 24- Page 6 of 19 Faure 2.3 Eistorical and Prolected migration Patterns IBRD 24681 NE PAL HISTORICAL AND PROJECTED MIGRATION PATTERNS 11 *;~i~ . -i Migration, 1971-81 Migration, 1981-91 1 Mgration (Projected), - - 1991-2010 Out Migration- in M9gration Note Mop show distncts vccounting For mnore ?han- I perCeYnl Of toto[ igranis BkGpThl. d~hWn- ori ~-- d nd Th. WM BkuNdo -t, sh 4 o -kf fud -y-ndyr-y MACH 1993 - 25- ANNEX II Page 7 of 19 B. Land Use 12. Based on aerial photographs taken during 1978/79, the Land Resource Mapping Project (LRMP) V prepared a data base on land utilization for the entire country. Figure 2.4 presents data on land use extracted from the LRMP and incorporated into the GIS. The maps in the figure show land use by four categories: agriculture, forest (excluding lands classified as shrub), grazing, and other uses which includes shrubland (less than 10 percent forest cover), urban areas and other lands which do not belong in the above categories. The maps show by district, the percentage of total land in each of the four land use categories. 13. As would be expected, the highest proportion of land devoted to agriculture is in the Eastern and Central Terai, where more than 50 percent of land is in agricultural use. The Hill districts generally have between 25 percent to 50 percent of land in agricultural uses, while in the Mountain districts this percentage falls to below 10 percent. 14. Forest land appears to be complimentary to agricultural land use. The Western Hill and Terai districts, and some Central Hill districts appear to have the largest proportion of land still under forests. Comparing the maps for agriculture and forest land use clearly shows that expansion in agricultural land use has been usually at the expense of forest lands. Future increases in population in the Far-West and the concomitant increase in agriculture would therefore pose a threat to the forests in the Far-West, and it is likely that significant changes have already occurred as the data presented here is more than 10 years old and there has been very significant in-migration into the Far- Western Terai since then. 15. Land specifically devoted to grazing appears to be relatively small in proportion to total land use, with most districts having about 10-15 percent of land in this category. The Mountain region, especially Western Mountains, and the Hill districts devote a larger proportion of land to grazing, the highest being in the Mid-Western and Western Mountains where between 25 and 50 percent of land is used for grazing. In the Terai, land devoted to grazing in each district is less than 10 percent without exception. C. AGRICULTURE 16. Agricultural production being the mainstay of the Nepalese economy, providing over 90 percent of total employment and about 60 percent of GDP, will continue, in the foreseeable future to be a major focus of infrastructure strategy. The spatial dimensions of agricultural production and its implication for infrastructure provision must therefore be understood. Figure 2.5 shows the spatial distribution of the value of output for the major crops, meat and milk production. The data is from 1986. Major crops included are paddy, maize, wheat, millet, barley, potato, oil seeds, sugar cane, tobacco, jute and pulses. 21 "Land Resource Mapping Project*, HMGN, Kathmandu, NepaL, 1986. - 26 - ANNEX II Page 8 of 19 Figure 2.4 Sipatial Land Use Patterne. 1980 IBRD 24682 NEPAL SPATIAL LAND USE PATTERNS, c. 1980 _- Agricultural Land Use = - - -- --- ------- Forest Land Use PERCENT OF TOTAL LAND -- < 10% 10-25%-- 25-50% j7 >50% Grazing Land Use -7-- -- Other land Use homa h ooe p'.pored by Th Warid Bonk's ssIx luey or 00h 0connc ohf h rea der . lor the mornaol oso of Th Word.,- -4nk Group . denoon .sed antho boundones sho.n en Zh. n.op do notnpy on the pon of T. WordBonk Group an~y udgmC 19 on bo no o nyrryo..nonmmrcpns MARCH 1993 - 27 - ANNEX II Page 9 of 19 The value of output was computed using (approximate) average national prices, and the distribution shown in the figure is intended to be a rough representation of the spatial distribution of agricultural output. The Kathmandu Valley and the Central Terai are clearly the areas of highest agricultural output. The Eastern Terai and parts of the Western Terai also have high contributions, whereas the Mid and Far-Western Terai are only slightly above the levels found in the Hills. Agricultural Output Per Capita 17. Figure 2.6 shows per capita agricultural output for various districts, and disaggregates this output by the major crops and milk and meat production per capita. On a per capita basis the concentration of output in the Terai is not evident since areas with low production also have small populations thereby raising per capita production levels. Crop output per capita in the Mountains is unexpectedly higher than in the Hills (though the data for the districts of Mustang and Manang, which have the highest values in the Mountains, is somewhat suspect), and this is accounted for by the lower populations in the Mountains, and by the high production of potatoes in the Mountain districts. Milk production per capita is generally highest in the Hills, and the lowest in the Terai. Meat production per capita is higher in the Mountain districts and appears to be fairly well correlated with the distribution of population. 18. The indices used to represent per capita agricultural output by district in Figure 2.6 were calculated using the national average per capita as the base (Index = 1). The index for total agricultural output per capita, ranged from 0.20 to 2.94 times the national average. For crop production, these indices ranged from as low as one fourth the national average to almost four times the national average. Milk production per capita similarly ranged from one quarter to 2.5 times the national average. Meat production per capita, being more uniformly distributed, had a narrower range of 0.44 to 1.86 times the national average. There were a few apparent anomalies in these per capita output data; for instance per capita milk production in Kathmandu was only 0.4 times the national average and one would expect that the higher incomes here would generate above average per capita demand. One possible explanation is the greater availability of dried milk as part of food aid received from overseas. Agricultural Cropping Patterns 19. Figure 2.7 shows the distribution of paddy, wheat and maize cultivation, by showing the percentage of total land used for each crop. The Figure does not show information for districts which have less than 10 percent of total land area devoted to agriculture. The differences in cropping patterns are clearly identifiable with paddy predominating in the Terai, particularly the Eastern and Central Terai, with very little paddy cultivation in the Western Hills and Mountains. Wheat is primarily a Terai crop, and almost everywhere, it is secondary to paddy, and very little land is devoted to wheat in the Hills and Mountains. Maize is grown primarily in the Hills, and the Western Terai, and appears to be the inferior choice, giving way to paddy and wheat in areas with better infrastructure. - 28- A ii Page 10 of 19 Figura 2.5 Svatial Distribution of Agricultural Output - 1986 IBRD 24683 NEPAL SPATIAL DISTRIBUTION OF AGRICULTURAL OUTPUT - 1986 _- - - - --- - -- - ----Agricultural Value of Output index AGRICULTURAL VALUE INDEX - < 2 2-4 4-6 6-8 8 o-dh b- 0. n. W. d0de &rd T 199w MARCH 1993 - 29 -X ii Page 11 of 19 Figure 2.6 Per Capita Acricultural Production - 1986 IBRD 24684 e, NEPAL PER CAPITA AGRICULTURAL PRODUCTION - 1986 'Agriculturaira tuction PER CAPITA __Per Capita Crop Production PER CAP-T--------- PRODUCTION INDEX -_-_ Z<05 05-10 1 0.1.5 15.2.0Per Capita Milk Production 2.0-2.5 > 2 5 - Note Average for Nepal 1 0 Per Capito Meat Production Th,n mW o h .bn hm bo pr d by Wmäo & ~u y Iw Bankd Group The. "denommnn used and n. bonndon.. .lho,-m n map do n wonply o pwr of %he world sank Group ny g do h log&.! s olo o n. 0'm0ry onno n.0 0r oceepionce 1993 MARCH 1993 - 30- ANNEX II Page 12 of 19 20. The largest proportion of land devoted to paddy, was about 40 to 60 percent in Eastern and Central Terai. In the Terai, while wheat generally commanded about 10 percent of land, around Kathmandu there was greater reliance on wheat, with Bhaktapur showing over 40 percent of land in use for wheat production. In the Hills, larger proportions of land were devoted to maize rather than to wheat. Aaricultural Productivity 21. As would be expected, the data shows fairly strong relationships between the productivity of agricultural lands and the availability of infrastructure such as roads and irrigation facilities, and inputs such as fertilizer. Figure 2.8 presents four maps showing distribution of the value of crop output per hectare of agricultural land, density of the road network, irrigated lands as a percentage of agricultural land, and fertilizer use per unit of agricultural land. A regression of the crop value index against road density, irrigation and fertilizer use for the 75 districts yielded a fair correlation with an R Square of 0.65. Crop value was computed as discussed earlier (see para 16), and was divided by the total land devoted to agriculture; this value of output per hectare was divided by the national average to obtain the index. 22. Availability of irrigation, defined here as percentage of agricultural land irrigated, appears to be the most significant factor of the three considered, followed by fertilizer use, and finally road density. The latter is almost uniformly low and perhaps nowhere, other than in the Kathmandu Valley, has the extent of the road network made a significant impact on agricultural production. There does, however, appear to be a fair correlation between fertilizer use and the density of the road network (R Square of 0.56), which would mean that roads, in addition to impacting the value of agricultural output directly, also influences it indirectly through its impact on fertilizer use. 23. The value of crop output per hectare of agricultural land area varied from three times the national average (Bhaktapur) to one quarter for some of Far- Western Hill districts. In the Terai, all districts were above the national average ranging between 1.03 (Saptari) to 2.81 (Bara) times the national average. Most of the districts with very low values were located either in the Far and Mid-Western Hills or in the Eastern Hills. Interestingly, Rasuwa district in the Mountains has a level of agricultural productivity comparable to districts in the Terai and in the Kathmandu Valley, and the area appears to have benefitted from the good road link established with Nuwakot in the Hills and with the Kathmandu Valley. 24. Fertilizer use is extremely concentrated, with the use in the Kathmandu Valley being 7.5 times the national average. In the Terai, the other region with high fertilizer use, fertilizer use ranges from 4.5 times the national average (Parsa) to as low as half the average (Bardiya). In general, the Central region has the highest fertilizer use. The Western and Eastern extremities of the Terai both appear to use about half the national average. 31 ANNEX II Page 13 of 19 Figure 2.7 Paddy. Wheat. laize Cultivation - 1986 IBRD 24685 NEPAL PADDY, WHEAT, MAIZE CULTIVATION - 1986 Area Under Paddy PERCENTAGE OF TOTAL LAND AREA -i - Area Under Wheat 5-10% 10-15% 15-20% 20-25%~- >25% ) Area Under Maize N, ..mp ho. b,.n p byTh Wor-d Bmnk b & o m lu d»4y lo. he con.ensence of reader and, o or tm rnl seo .l iho Word Snk Group The donaan . sd -nd ~o åoun ). , hoon, , m.oo do not nmp-y on eso Pon o37. hewardd Son Grou.p any ,odgment o w legol -ynus e0 any 0mery o ndonseen 0r o.pnce e nuch onoone9 MARCH 199 i - 32. ~X 1 Page 14 of 19 Figurs 2.8 C_rol) output and Roads. Irrication and rertilizser Use - 1986 1BRD 2468< NEPAL CROP OUTPUT AND ROADS .__r_-- Crop Output Index =kJ < 0.5% 0.5-10% 10-1.5% 1.5-2.0% 2.0-2.5% > 2.5% No- C-pop utinde ,j consuced as aorhw o oo d o e of croImof p otPr of rci ngo~ nMm ewch d,iffc fo © e p oI se Arage frNepol- l O Road Density/Km2 ROAD Km/KM2 0 .05 0 05-0 10 0 10-0 25 0.25-1 00 >1 00 Book C-op The denm~ o~n ue -nd - bon#esson n ,r m w d. no py por Th & rdBn G~pn d fld -iv - b-J, son he . c iony ~rns y or any sk endrsmentr T nce W R& 9 MARCH 1993 -33- ANNEX II Page 15 of 19 Figure 2.8 Crop Outiput and Roads. rrigation and Pertilizer Use - 1986 (cont.) IBRD 24687 NEPAI. IRRIGATION AND FERTILIZER USE, 1986 Irrigated Land PERCENTAGE OF AGRICULTURAL LAND IRRIGATED <10% 10-20% 20-30% 30-40% 40-50% > 50% Note Only distrcts with more than 10% agncultural land ncluded Fertilizer Use Index --- <05 0.5.1.0 10-20 20-30 30-50 50 Note Fertizer use mndex is constructed as a raho of value of fertslizer use per unt of agricultural land in each district to that for Nepal overall, ge Average for Nepal - 10 N, -p ., b- wo by, tw B-&, o-t, .o éc---,e of ~.. -od, , ,o 0»wo ~. of T4o WM,d 8o- Gp Vw d,, de. d #» bounden., owni- on rå oog,o o ,,, , w. pof. W, d ww Bank Groop oany waeen n d lgao n .'y ien .Iy or ky c~dn.rneno e o on MARCH 1993 - 34 - ANNEX II Page 16 of 19 In the Hills and Mountains outside the Central region, only two districts appear to use anywhere near the national average (Kaski: 0.57, Parbats 0.59). Generally, the Hill districts appear to have between 10 to 20 percent of the national average for fertilizer use. As noted earlier there appears to be a fair correlation (R Square of 0.56) between fertilizer use and accessibility and this may explain the low usage in the Hill districts which are far removed from Kathmandu. The two Hill districts, Kaski and Parbat, which have higher fertilizer use are located close to Pokhara which is well connected to Kathmandu by road. D. Forest Degradation 25. Figure 2.9 shows the extent of forest degradation in the early 1980s based on Land Resource Mapping Project data as the percent of forests degraded in each district. The most severe degradation appears to have been in the Central and Western Terai with districts such as Chitwan and Rupandehi having as much as 27 percent of their forests classified as degraded in the early 1980s. Adjacent districts such as Bara and Nawalparasi showed 16 to 18 percent of forests as degraded. The Western and Eastern extremities of the Terai appear to have suffered less degradation. In the Western region forest degradation seems to have been more severe in the Hills than in the Terai. It is possible that in the two areas, Central Terai and Western Hills, forests have been degraded for different reasons. In the Central Terai, the presence of a better road network has facilitated greater access to forests, whereas in the Western Hills, degradation may be caused by local population with poor access to other types of fuel, and thus having to rely more heavily on fuel wood as the main source of energy. 26. Regressions run on the extent of forest degradation with the availability of roads, population density and per capita agricultural output (a measure of rural poverty) showed weak correlations; however, forest degradation was best explained by rural incomes which had a negative coefficient, indicating that greater rural poverty increased the reliance of the population on forest resources. Road density also had a negative correlation with forest degradation, indicating that where there was improved access to alternative fuels, the reliance on forests for fuel wood decreases. E. Impact of Infrastructure Development 27. Relationships between infrastructure endowments and the agricultural, industrial and environmental situation are difficult to establish in Nepal owing to a general lack of appropriate data. The foregoing analysis provides some insight into the influence of infrastructure on aspects such as agricultural productivity, forest degradation, etc. With a view to examining more systematically the relationship between infrastructure and agriculture, industry and forest degradation, use was made of the detailed central services maps produced for the Hill and Mountain districts under the Swiss Trail Bridge Project. These maps, covering 55 Hill and Mountain districts show the location 35 -ANNEX II Page 17 of 19 Figure 2.9 Forest Decradation IBRD 24688 NE PAL FOREST DEGRADATION Degraded Forest Land - -J PERCENTAGE OF FOREST LAND DEGRADED 7 < 5% F 5-10% 10-15% 15-20% 20-25% > 25% Th,, m~p ho3 been poored by The Word Book , ,mff endeunrel for Ihe conln..nce of rooers ond, for h i,enno us of Th, Word Bank Gop The o -,nnmnon ..sd ond he boundones .hown on k, m cp do nonpy on ~ h1 pon of The word Bonk Group 8 ny C- dgmpnt on moe d n o any gerneory or ony .ndon.Mrsmt or oc 1nc 9 of MARCH 199 3 - 36 - ANNEX II Page 18 of 19 of settlements in relation to the road and trail network, and also provide information on the types of services available at each settlement. This information was analyzed to examine the relationship between transport on agricultural output, extent of industrial (mainly cottage and agro-processing) activities, and forest degradation. 28. Availability of transport in a district was quantified as the percentage of settlements in a district within 1 km. of a road or main trail. This parameter was regressed against index of agricultural output to examine the relationship between agriculture and the availability of transport. Similarly, industrial activity was quantified as the percentage of settlements in a district with some form of cottage or small scale industry, and this was regressed against the availability of transport and. Forest degradation in a district was quantified as the ratio of degraded forest land area to the total forest land. Figure 2.10, 2.11 and 2.12 present the results of these regressions. 29. From Figure 2.10 it is evident that there is a fairly strong correlation between the availability of infrastructure and agricultural output in the Hills and Mountains. Eliminating the strong outlier the R Square of these regressions improved considerably with transport showing an R Squared of 0.24. Figure 2.10 Transport and Value of Aqricultural Output TRANSPORT WIWM me ML 11020ft WWL 2.4 X=RD-RATIO W .Y=RATIO-DISTINAT VALUE AGR 0/P Regression Outpu: 2 Constant 0.728798 I Std Err of Y Est 0.315025 R Squared 0.336312 a No. of Obsarvations 33 a Degrees of Fredow 31 1.230 0 1.a0 a X Cocfficient(s) 1.201777 s a G o a Std Err of Cod. 0.303216 066- LS* LI 0 03 L34 0.0 MW 9ML low mD / T SUEMi 30. Similarly, Figure 2.11 shows strong correlation between transport and industrial activity. The impact of infrastructure on development of rural industry appears to be even stronger than on agricultural output. - 37 - ANNEX II Page 19 of 19 Figure 2.11 Transport and Industries TRANSPORT X=RD-RATIO Y=INDR_RATIO CasIn Regression Output: a Constant 0.054440 av a Std Err of Y Est 0.107661 R Squared 0.623694 No. of Observations 33 Degrees of Freedom 31 -44 - a6* a a X Coefficient(s) 0.742790 . oa a ap o Std Err of Cotf. 0.103626 a a 1 a 0 o6 a as a1 0 01 GA GA O Mi - SEuM tm / T RTRlMDIE 31. The impact of transport on forest degradation is somewhat surprising. Figure 2.12 presents the scatter plots and regression results and as can be seen, transport availability appear to be negatively, though weakly, correlated with forest degradation. It should be noted that this result is for Hill and Mountain districts where there has not been significant in-migration pressure on forest resources. An explanation for the negative relationship could be that with improved transport access, the attractiveness of fuel wood as a source of energy diminishes thereby reducing the reliance of existing populations on forest resources. Figure 2.12 Transport Infrastructure and Degradation TRANSPORT X=TRFT-RATIO Y=DEOR(D) LAND / FOREST AREA I.4 c Regression Output: Constant 0.25327 Std Err of Y Est 0.057051 R Squared 0.234929 No. of Observations 51 a Degrees of Freedom 49 a a a aa a a Std Err of Cof. 0.064625 a a o a& o aQ A as O I t=- Sus t I / WT EIMIS - 38 - Annex Il Page 1 of 8 EXTERNAL AND INTERNAL TRADE 1. Infrastructure in support of production will count for little unless it includes adequate facilities to integrate markets and ensure that products can reach their intended markets. Therefore a major component of infrastructure development strategies and policies must be improvements in services and facilities that strengthen trade and marketing links. This section examines the current situation with regard to three aspects of Nepalese trade: internal trade among various regions of the country; Nepal's border trade with India; and Nepal's third country trade with the rest of the world. A. Trends in Nepal's External Trade 2. Table 3.1 provides data on Nepal's trade with India and with third countries. Nepal's external trade has grown from NRs 2.7 billion in 1974/75 to NRs 20.5 billion in 1988/89, which represents an annual growth rate of 16 percent in nominal terms. Exports have grown at an annual rate of 12 percent from NRa 0.9 billion in 1974/75 to NRe 4.2 billion in 1988/89, while imports have grown somewhat faster at 17 percent p.a., from NRs 1.8 billion to NRs 16.3 billion over the same period. The trade deficit which was NRs 0.9 billion at the beginning of the period and NRs 12.1 billion by 1988/89, grew at around 20 percent p.a. over the period. Inflation during the 1980s has been around 8 percent p.a., which indicates that in real terms exports grew at around 4 percent, imports grew at around 9 percent, and the trade deficit has been growing at around 12 percent p.a. over the 1980s. This large and growing trade deficit has been financed primarily through external assistance in the form of grants and loans. Table 3.1 NEPALESE FOREIGN TRADE Trade Volumes (NRs Billion) Nominal Growth Rates % p.a.) 1974-75 1984-85 1988-89 1975-85 1985-89 1975-89 Total Trade 2.7 10.4 20.5 14 17 16 - India 2.2 5.5 5.3 9 -1 6 - Rest of World 0.5 4.9 15.2 23 29 28 Exports 0.9 2.7 4.2 11 11 12 - India 0.7 1.6 1.0 8 -12 2 - Rest of World 0.2 1.1 3.2 17 27 25 Iports 1.8 7.7 16.3 15 19 17 - India 1.5 3.9 4.2 10 2 8 - Rest of World 0.3 3.8 12.0 26 29 27 Deficits 0.9 5.0 12.1 17 22 20 - India 0.7 2.3 3.2 12 8 11 - Rest of World 0.2 2.7 8.9 26 30 27 Source: Statistical Year Book, HMGN, 1990 - 39 - Annex III Page 2 of 8 3. The most interesting development in Nepal's external trade, aside from the rapidly growing trade deficit, has been the shift away from Indian trade, to trade with the rest of the world. Overall trade with India, as well as imports and exports separately, has either declined or remained static (imports) in real terms since 1975. The second half of the 1980s shows a decline even in nominal terms, and this data pre-dates the trade and transit impasse which occurred in 1989/90 following which trade would have diminished even further. Recent developments in India which has led to significant liberalization of India's trade regime could further reduce trade levels with Nepal since some of Nepal's trade with India was due to businesses in India using Nepal's more liberal trade regime to circumvent Indian controls. Composition of Exports 4. The past decade has seen a significant shift in the composition of Nepal's exports. Whereas in the late 1970s the main exports were raw agricultural materials, foodstuffs and live animals, today they are manufactured goods (especially carpets and to a lesser extent leather goods) and miscellaneous manufacturing, primarily clothing (See Table 3.2). In the three year period prior to the imposition of the Indian trade embargo, exports of food and live animals actually declined in value while exports of crude materials increased barely at all. On the other hand, the current value of both categories of manufactured exports increased sharply, at the rate of 25% per year for the former (carpets), and 40% per year for the latter (clothing). In 1990, manufactured goods accounted for over 80 percent of total export value, as compared with 30 percent in 1980. While this extreme dominance of manufactured exports in 1990's trade figures may be partly due to the Indian trade embargo, (in 1988 manufacturing accounted for 63%) there is no question that the manufacturing sector has emerged as the primary export base for the economy. This shift in the composition of exports is attributable to two main factors. Table 3.2 EXPORTS CLASSIFIED BY MAJOR COMMODITY GROUP (NRS MILLION) Coamodity Group 1979/80 (a) Percent Total 1989/90 (b) Percent Total Annual X Change 1980-1990 Food/Live Animals 306.5 26.6 647.7 12.5 7.8 Tobacco/Beverages 2.6 0.2 21.2 0.4 23.3 Crude Materials 469.6 40.8 226.0 4.3 -7.1 Mineral Fuels 0.5 0.0 0.0 0.0 0.0 Animet/Vegegable OiLs 20.4 1.8 15.0 0.3 -3.0 Chemical/Drugs 1.3 0.1 20.6 0.4 31.8 Manufactured Goods 291.8 25.4 2724.8 52.1 25.0 Machinery/Transport 3.2 0.3 0.1 0.0 -29.3 Misc. Manufacturing 54.0 4.7 1580.1 30.2 40.2 other Comodities 0.6 0.0 0.0 0.0 0.0 TOTAL 1150.5 100 5235.5 100.0 16.4 SOURCE: (a) 1989 Statistical Year Book of Nepal (b) 1991 Statistical Pocket Book of Nepal, Table 14 - 40 - Annex III Page 3 of 8 Agricultural exports have declined due in part to rising domestic consumption, which has meant not only less surplus available for export but also a widening trade deficit in farm goods, and secondly to the rapid growth in the carpet manufacturing and garment industries. Composition of Imports 5. In comparison with exports, the composition of imports has remained fairly stable over the 1980s, as shown in Table 3.3. Though imports have grown rapidly, at a 17 percent p.a. nominal growth rate, this growth appears to have taken place across the board, with no single or group of commodities causing the increase. The fastest growing import categories appear to have been animal/vegetable oils, followed by crude materials; the latter is probably related to expansion in carpet and garment manufacturing both of which are fairly import intensive. The slowest growing categories are food and live animals, and fuels and lubricants which grew at an average annual nominal rate of 13 percent. Table 3.3 IMPORTS CLASSIFIED BY MAJOR COMMODITY GROUP (NRS. MILLIONS) Comnodity Group 1979-80 Percent Total 1989-90 Percent Total Annual % Change Food & Live Animals 413 12 1,549 8 13 Tobacco & Beverage 26 1 264 1 23 Animal/Vegetable Oils 26 1 477 3 29 Crude Materials 101 3 1,475 8 27 Fuels & Lubricants 410 12 1,522 8 13 Chemicals & Drugs 397 11 2,862 16 20 Manufactured Materials 1,090 31 5,163 28 16 Machinery & Transport 720 21 3,811 21 17 Misc. Manufactures 288 8 1,283 7 15 Not Classified 10 * - - I Total Inports 3,480 100 18,406 100 17 Recional Shares of Production for Exports 6. Though no definitive data is available, it is clear that most exports come from the Central region, and to a lesser extent the Eastern region. Table 3.4 estimates the regional shares of six commodity exports that made up the bulk of the total in 1987/88. For five of the six commodities in question, the Central region accounts for the largest share of production, ranging from 25% of livestock to 98.5% or virtually the entire production of apparel. Only in the case of leather products, 60% of which is concentrated in the Eastern region, does the Central region account for less than the largest share, though even here it still produces 40% of total value added in the sector. This method of estimation suggests that in 1987/88, the Central region accounted for 60% of the value of exports of six key commodities, the Eastern region 24%, the Western region 7%, the Mid-West 6% and the Far-West 3%. -41 - Annex III Page 4 of 8 Table 3.4 REGIONAL SHARES OF SELECTED EXPORTS 1987/88 (NRS MILLIONS) Commodity Deve lopnent Re ions Total Far-Western Mid-Western Western Central Eastern Live Animals 60 123 132 157 152 625 Ol Seed 25 4 19 60 21 168 Fruit 14 21 26 63 25 149 Textile Manufacturing 1 8 63 791 521 1384 Wearing Apparel 0 0 11 884 2 897 Leather Products 0 0 0 67 98 165 Sub-Total Value 100 196 251 2022 819 3388 Share of Total (%) 2.9% 5.8% 7.4% 59.7% 24.2% 100.0% Total Value of Exports 4115 SOURCE: Census of Manufacturing Establishments, Nepat 1986/87. Agricultural Statistics of Nepal, 1990, Table 5.2-5.6 7. The longer term shift in the composition of exports away from agricultural products to manufactures of textiles and clothing means that the Central and Eastern regions have become even more important in the production of exports, while the role of the three western regions, which has never been prominent, has declined even further. This trend is reflected in an exaggerated form by figures for "overseas" exports, leaving those to India out of account. According to these figures, the Central region's contribution to overseas exports rose from an already high 66% in 1980/81 to a virtual monopoly of 97% by 1987/88, displacing the Eastern region, whose share declined from 30% to only 2%. Given both the shift in the composition of exports from agricultural produce to manufactures, and the rising importance of overseas markets, it is clear that the Central region is playing the leading role in production for export markets today, and can be expected to continue doing so for some time in the future. B. Trade with Third Countries 8. While there are 17 customs points for Nepal's external trade flows, overseas trade (i.e., trade with countries other than India) is highly concentrated at 3 customs gateways. Table 3.5 shows the shares of these three gateways in total third country trade by value in 1979. Exports to third countries predominantly use air transport (88% of value), and move through the Tribhuvan Airport in Kathmandu. Imports generally arrive by sea at Calcutta Port in India, moving overland and entering Nepal either through Birgunj (59%) in the Central Terai or through Biratnagar (14%) in the Eastern Terai; another 18 percent of imports enter the country by air through the Tribhuvan Airport. -42 - Annex III Page 5 of 8 Table 3.5 OVERSEAS IMPORTS AND EXPORTS THROUGH MAIN GATEWAYS (1989) Exports to ROW Inmports from ROW NRs Million Percent WRs Million Percent Total Value 3219 100 9906 100 Main Gateways: - Tribhuvan Airport 2850 88 1757 18 - BirgunJ 172 5 5802 59 - Biratnager 47 1 1369 14 Total 3 Gateways 3069 95 8928 91 Other 150 5 978 9 Source: Foreign Trade Statistics, Dept. of Customs, HMG/N, 1990 9. Nepal's exports of carpets and garments, which have come to dominate the overseas export trade (90% of overseas exports), are mainly concentrated on Europe in the case of carpets and the United States in the case of garments. Germany accounts for about 50 percent of carpet exports, and Switzerland, U.K., and Belgium accounting for another 45 percent, which means that European destinations together account for over 95 percent of carpet exports. Garment exports are even more concentrated with the United States being the destination for 95 percent of garment exports. C. Trade with India 10. As in the case of overseas trade, the trade with India also tends to be concentrated at a limited number of customs gateways. As shown in Table 3.6, six gateway towns accounted for 86 percent of exports and 90 percent of imports Table 3.6 IMPORTS AND EXPORTS TO INDIA THROUGH MAIN GATEWAYS (1989) Exports to India Imports f om India NRs Million Percent MRs Million Percent Total Value 909 100 5759 100 Main Gateways: B Biratnagar 266 29 599 10 * Bhadrapur 150 17 176 3 - Birgunj 114 13 3276 57 - Nepatgunj 83 9 158 3 - Dhangadhi 122 13 98 2 - Shairahawa 48 5 881 15 Total 6 Gateways 783 86 5188 90 Other 126 14 571 10 Source: Foreign Trade Statistics, Department of Customs, HMG/N, 1990 - 43 -Annex III Page 6 of 8 by value in 1989. In exports, the Eastern region accounted for almost 50 percent of the total through the two gateways Biratnagar and Bhadrapur. Birgunj, the border crossing serving the Kathmandu Valley, accounted for almost 60 percent of imports reflecting the greater purchasing power of the Central region and in particular of the Kathmandu Valley. 11. As a result of the long open border of over 1000 km with India, the Terai markets in Nepal are said to be well integrated with markets in the bordering Indian states of West Bengal, Bihar and Uttar Pradesh. An unknown magnitude of unrecorded trade takes place from both sides of the border, and the official statistics include a 7 percent upward revision to account for this; however, it is generally agreed that unrecorded cross border trade is in excess of this estimate. For agricultural commodities such as paddy, wheat and pulses, prices in Nepal's Terai markets were found to be almost as strongly correlated as prices among markets in each country!/, indicating linkages between these cross border markets. Only rice prices were found to be weakly correlated, and this may be related to the ban on rice exports by Nepal which came into effect in 1987. D. Internal Trade within Nepal 12. Data on the sources of procurement by wholesale units in various parts of Nepal reveal the differences in economic and market activity between Terai and Hill/Mountain economies in Nepal. Table 3.7 presents data for 16 towns on the percentage of purchases by wholesale units from various sources (local, national and Indian) of supply. Most towns located in the Hills and Mountains procure Table 3.7 Sources of Procurement by Wholesale Units Development Region Local (%) National (%) India (%) Mountain Towns Junta Mid-Western 50 50 Khandbari Eastern 100 Hill Towns Birendranagar Mid-Western 100 Gorkha Western 100 Pokhara Western 47 50 Tansen Western 96 5 Banepe Central 89 11 Kathmandu Central 45 32 23 Dhankuta Eastern 100 Terai Towns Nepatgunj Mid-Western 46 22 33 Butwal Western 53 29 18 Bharatpur Central 68 3 29 KaLaiya Central 90 10 Birgunj Centrat 42 28 30 Rajbiraj Eastern 84 16 Biratnagar Eastern 73 9 19 Source: minter and Intra-Regional Trade Flows Study", CEDA, Tribhuvan University, Kathmani, 1988 Y' NAgricultural Pricing Policyu, ADB/HMG Study, Kathmendu, Septeuber 1991. - 44 - Annex III Page 7 of 8 very little from their respective local areas, the exceptions are major urban centers such as Kathmandu, Pokhara and Banepa (which is close to Kathmandu). By contrast, most Terai towns rely on local markets for a significant proportion of their procurements. This is indicative of the relative lack of integration between Hill and Mountain market towns and their economic hinterlands. Also interesting is the near absence of trading links between Hill/Mountain towns and Indian sources, whereas almost all Terai towns procured a significant portion from India. 13. Regional variations in prices also indicate the poor accessibility and lack of market integration in the Hill and Mountain zones. Table 3.8 presents price indices for farm and non-farm products in various parts of the country. The indices are based on simple averages of the prices of 5 farm commodities (rice, wheat, pulses, sugar, potatoes) and 5 non-farm commodities (soap, kerosene, sandals, pencils, paper). The average price for Nepal for each commodity is taken as 100. The commodities were selected based on the price information available to provide the beat geographic coverage and disaggregation. Since this is not a sophisticated price index, the results should only be treated as indicative of price variations. Table 3.8 REGIONAL VARIATIONS IN RETAIL PRICE INDICES 1988 Far- Mid- Western Central Eastern Total Western Western MOUNTAIN Farm Produce N/A N/A 152 N/A 148 150 Rice N/A N/A 203 N/A 152 178 Wheat N/A N/A 150 N/A 189 170 Non-Farm Products N/A N/A 155 N/A 116 136 ALL Products N/A N/A 153 N/A 131 142 HILL Farm Produce 95 102 96 83 119 99 Rice 91 106 101 87 86 94 Wheat 71 88 91 71 104 85 Non-Farm Products 83 108 97 64 92 89 ALL Products 96 105 96 75 104 95 TERAI Farm Produce 100 76 82 81 85 85 Rice 91 77 81 83 72 81 Wheat 109 68 83 83 103 89 Non-Farm Products 107 69 N/A 87 85 87 All Products 102 76 82 82 85 85 AVERAGE (Hill and Teral only) Farm Produce 97 89 89 82 102 92 Rice 91 92 91 85 79 88 Wheat 90 78 87 77 104 87 Non-Farm Products 95 89 87 76 89 89 ALL Products 99 91 89 78 95 90 Source: Weekly Prices Information Bulletin: Wholesale and Retail Prices of Essential Food and Non- Food Commodities by Districts Markets in Nepalr Vol. 1. 1990, PP.153-201. 14. Price variations for farm produce ranged from an index of 76 in the Mid-Western Terai to 152 in the Western Mountains. For non-farm products the index ranged from 64 in the central Hills to 155 in the Western Mountains. The average index for all products was 85 in the Terai, 95 in the Hills (12% higher), and 142 in the Mountains (67% higher than in the Terai). Rice prices in the Terai were lowest in the Eastern (72) and Mid-Western (77) regions, and highest - 45 - Annex III Page 8 of 8 in the Far-West (91) and Central (83) regions. Among regions, prices in the Central region appear to be the lowest, increasing progressively towards the western parts of the country, with prices in the Far-West being 25 percent higher. 15. A survey by the Central Bureau of Statisticsy of rural trading activity provides some information on the extent of commercial activity in various regions of the country. Table 3.9 summarizes some of the results. As is to be expected, the survey found that the number of trading units in relation to population, is highest in the Central and Eastern regions (about 6 units per 1000 of population), while the Far-Western region has very little commercial trading (0.4 units per 1000 of rural population). Average sales per unit was also the lowest in the Far-West (NRe 16,800 p.a.). The data shows relatively lesser development of commercial activity in the western parts of the country which corresponds to other indicators showing lower levels of economic activity in the west. Transport cost differences are also interesting. Whereas in the Eastern and Central regions transport costs were on the order of 1 percent of sales, in the western regions they were more than 3 percent, with the mid-west, which is the most poorly served with infrastructure, showing transport costs as high as 5 percent of sales. TabLe 3.9 PROFILE OF RURAL TRADING ACTIVITY. 1990 Far- Mid- West Central East Nepal Vest West Number of Trading Units (000's) 0.6 9.4 13.7 35.2 23.5 88.1 Trading Units per 1000 of Rural Population 0.4 4.1 3.9 6.7 5.7 5.3 Average Sales per Unit (NRs 000's) p.s 16.8 60.4 36.3 59.3 50.9 50.5 Average Gross Margin Per Unit (NRs) 3,039 8,346 5,705 7,798 7,084 6,997 Average Transport Cost Per Unit (NRs) 463 2,925 1,020 605 682 926 Transport as % of Sates 3% 5% 3% 1% 1% 2% Average Margin Net of Transport (NRs) 2,576 5,421 4,685 7,193 6,402 6,071 Net Margin as % of Sales 15% 9% 13% 12% 13% 12% Source: Multipurpose Production Survey (Rural), Central Bureau of Statistics, NMG/N. 1991. Multipurpose Production SurveV (Rural), Central Bureau of Statistics, MMG/N, 1991. - 46 - Annex IV Page 1 of 11 QUALITY OF INFRASTRUCTURE SERVICES FOR INDUSTRY 1. This annex reports the results of a survey of 150 industrial establishments which was conducted to assess the quality of infrastructure services available to industries located in various parts of the country. Whereas the primary object of the survey was to identify deficiencies in infrastructure affecting industrial activities, the survey also provides insight on infrastructure problems in general. Industrial establishments, being generally better staffed than other users of infrastructure services, constitute a superior source for information on the quality of infrastructure available in the country. In interpreting the results of this analysis, however, the reader is cautioned on the dangers of extending some of the results as these may in different ways be biased towards an industrial viewpoint on infrastructure deficiencies and needs. A. SURVEY OBJECTIVES AND SAMPLE PROFILE 2. The objective of the survey was to identify problems faced by industrial establishments with regard to the availability, reliability and cost of electric power supplies, water supply, transport and telecommunications services, and drainage and solid waste disposal services; to assess the extent of these problems in various parts of the country and their impact on industrial activity; and to determine the kinds of responses made by industry, such as substituting private provision for deficient public infrastructure services, and the coats of such responses. Selection of a Representative Sample 3. In order to ensure that the sample selected was both representative of Nepalese industry, as well as provided a reasonable geographical distribution of industry locations, certain criteria were established for selecting the sample of 150 industrial establishments. The criteria were as followes (a) the sample should include at least ten establishments from each of the five development regions (Eastern, Central, Western, Mid-Western and Far-Western); (b) at least 20 of the 75 districts and 33 municipalities should be represented; (c) at least ten locations should be in the Mountain ecological zone and at least 20 locations should be in the Hills; (d) the proportions of industries from the various sectors, i.e. manufacturing, agro-industry, mining and tourism, in the sample should roughly correspond to the share of these sectors in the overall universe of Nepalese industry; and (e) the sample should include a fair representation of industries by scale, i.e. small, medium and large scale industry. - 47 - Annex IV Page 2 of 11 In the sample eventually surveyed all of the above criterion were generally met with the exception of the one relating to the number of industries located in the Mountain zone; due to a lack of industries in the Mountain zone the sample included only four industrial establishments in this category. Table 4.1 summarizes the distribution of the sample by various categories. Table 4.1 DISTRIBUTION OF THE SAMPLE OF 150 INDUSTRIAL ESTABLISHMENTS No.) Industry Type Food, Beverages and Tobacco 61 41% Textiles, Weaving Apparel and Leather 22 15% Wood and Wood Products 5 3% Paper and Paper Products, Printing 3 2K Chemicals, Rubber and Plastic Products 20 13% Non-MetaLtic Mineral Products 7 5% Basic Metal Industries 10 7% Fabricated Metal Products, M/cy and Equipment 12 8% Tourism (Hotels) 10 7% Industrial Sector Manufacturing 89 59% Agro-Processing 48 32% Mining 3 2% Tourism 10 7% industry Scale Large 40 27% Medium 47 31% Small 63 42% Ecological Zones Mountains 4 3% Hill 69 46K Terai 77 51% Development Regions Far-Western 10 7% Mid-Western 10 7% Western 21 14% Central 75 50% Eastern 34 23% Some Characteristics of the Industrial Sample 4. Certain features of the sample of industries selected are noteworthys (a) of the 150 industries in the sample, 117 (78%) were established after 1971 and of these 63 (42%) were established after 1981; (b) of the 69 industries in the sample located in the Hills, 58 (84%) are located in the Central development region; (c) of the 41 industries in the sample located in the Western, Mid-Western and Far-Western development regions, 30 (75%) are in the Terai; and (d) of the 34 industries in the sample located in the Eastern development region, 32 (94%) were located in the Terai. - 48 - Annex IV Page 3 of 11 These observations indicate that the industries in the sample are relatively new, that industries in the Hills were primarily concentrated in the Central development region (Kathmandu and vicinity), whereas in the other development regions most industries were located in the Terai. These characteristics of the sample appeared to be fairly representative of Nepalese industry in general, with a large concentration of industry being located in Kathmandu Valley, and industrial activity elsewhere in the country being primarily located in the Terai. 5. In response to questions about their choice of location, the most common reasons cited were proximity to markets, infrastructure and transport facilities (about 50% of respondents). Government industrial policies (e.g. industrial estates, tax incentives) and availability of cheap labor were the least often cited reasons (about 11% of respondents). These findings indicate that infrastructure development strategies should be closely linked to assessments of the future spatial distribution of demand and markets. 6. With regard to markets for industrial output, the survey indicates that more than 80% of industrial output is sold within Nepal. Industries in the Eastern, Central and Western development regions tend to be less reliant on local (district level) markets (about 30% of output) and on India (less than 8% of output); whereas, in the Mid-Western and Far-Western regions local markets and India tend to be the destination of a larger proportion of output. The lower degree of integration of these western regions with the rest of the Nepalese economy, probably explains their higher reliance on local markets and India. 7. The industries in the sample, had an average of about NRs 40 million (about US$1 million) worth of assets employed, with large scale industries averaging NRs 63 million, medium scale NRs 30 million, and small scale NRa 12 million. Agro-processing industries tended to have the smallest average asset values (NRs 17 million), while tourism (hotels) on average had an asset base of NRs 53 million. Manufacturing industries employed on average about NRe 47 million in assets. Mining industries employed by far the largest asset base (NRE 143 million on average) but this may not be representative in view of the very small sample (three establishments). Profile of Industrial Expenditures 8. In order to understand the relative significance of infrastructure to industrial operations, the survey included collection of infrastructure related expenditures. Table 4.2 presents a breakdown of industrial production costs among various cost categories such as raw materials, labor, infrastructure costs, etc. The data is presented as percentages of total operating expenditures, and is classified in various ways; by development region, by ecological zone and by industrial sector. - 49 - Annex IV Page 4 of 11 Table 4.2 INFRASTRUCTURE COSTS IN INDUSTRIAL COST STRUCTURE Locatfon/Type of Raw Wages & Elect- Trans- Water Tele- Sewerage Total X infra- Industry Material Salaries ricity port Supply Cana & Waste Other Expenditure structure (C) () (%) (%) (%) (%) (%) (%) NRs 000s) (% Development Region Eastern (34) g/ 74.7 14.0 4.1 1.3 0.1 0.5 0.1 9.4 17,211 6.1 Central (75) 67.2 14.4 2.5 2.9 0.2 0.8 0.1 14.5 42,655 6.5 Western (21) 68.9 14.6 4.7 2.5 0.5 0.7 0.2 12.8 9,550 8.6 Nid-Western (10) 62.2 27.7 2.6 6.6 0.5 0.4 0.0 5.5 5,025 10.1 Far-Western (10) 81.5 7.5 3.0 3.7 0.2 0.5 0.2 6.5 17.916 7.6 Ecological Zone Teral (77) 71.5 11.1 2.5 2.2 0.2 0.4 0.1 14.8 32,150 5.4 Hills (69) 65.6 18.3 3.2 3.5 0.4 1.2 0.1 10.9 24,899 8.4 Mountains (4) 22.4 44.1 15.0 11.0 0.0 1.5 0.2 20.8 5.387 27.7 Industrial Sector Manufacturing (89) 69.2 14.2 2.5 2.4 0.2 0.5 0.1 13.4 27,544 5.7 Agro-Processing (48) 77.4 8.7 1.8 3.5 0.2 0.3 0.1 10.0 27,265 5.9 Tourism (10) 21.1 35.5 6.8 0.7 0.7 7.0 0.3 34.7 22,062 15.5 Mining (3) 62.9 24.1 8.1 3.8 0.2 0.4 0.1 8.6 77r924 12.6 Overall 68.8 14.2 2.8 2.7 0.2 0.7 0.1 13.2 28,095 6.5 a/ Numbers in parenthesis are the nunber of industrial establishments 9. Among development regions, infrastructure related costs appear to be the highest in the Mid-Western region (10.1% of expenditures) and the lowest in the Eastern (6.1%) and Central (6.5%). The Mid-Western region has in general the least developed infrastructure and therefore it is not surprising that industries located there need to spend relatively more on infrastructure services. In particular, transport related costs in the Mid-Western region appear to be high at 2.5 times the average. 10. Relative differences in infrastructure costs are even sharper among ecological zones. Whereas in the Terai, on average infrastructure accounts for only 5.4% of total expenditures, in the hills the figure is almost two times as much at 8.4%, and rises sharply to 27.7% of expenditures in the Mountains. The figure for the Mountains is based on a very small sample and, therefore, should be treated with caution. However, the high cost of transport in the Mountains is not surprising and, therefore, in general it is safe to say that infrastructure costs for industries located in the Mountains are likely to be considerably higher than elsewhere. 11. Among the various industrial sectors covered by the survey, tourism (hotels) appears to be the most infrastructure intensive (15.5% of expenditures). Mining is a close second. Both manufacturing and agro-procesing are considerably less dependent on infrastructure services. Between tourism and mining industries however, there appears to be a significant difference in the type of infrastructure demanded. For the hotel industry, telecommunication and electricity appear to be the dominant costs, whereas for mining these are electricity and transport. The importance of water supplies and transport to the tourism industry may have been underestimated; the former due to low water tariffs and the latter due to hotel expenditures not reflecting the overall transport costs of the tourism industry. These observations would lead us to - 50 - Annex IV Page 5 of 11 conclude that tourism is likely to be even more infrastructure intensive than the results of the survey would indicate. 12. On average, infrastructure costs appear to contribute 6-7% to the coat of industrial production. These figures, however, are likely to be a lower bound owing to the fact that some infrastructure related costs are built into the costs incurred on other accounts, e.g. the cost of transporting raw materials may sometimes be included in the cost of raw materials. On individual types of infrastructure, the most noticeable feature is the extremely low expenditures on water supply (0.2%) and sewage/waste disposal (0.1%). We have already commented on the low water tariffs; the low expenditures on sewerage/waste disposal probably reflects the almost total lack of facilities in most parts of the country. ELECTRICITY USE AND GENERATION 13. About one third (34%) of the industries surveyed have installed their own generating capacity to supplement power supplied by the Nepal Electricity Authority (NEA). The reliance on NEA is somewhat higher in the Central Hills. Among development regions the Eastern and the Far-Western appear to be least well served by NEA with about half of the industries in these areas having their own generators. Industries located in the Terai also appear to be somewhat less well served, with over 40 percent relying to some degree on their own generating capacity. B. WATER SUPPLY 14. Only 42 percent of the 150 sample establishments use water supplied by a public agency. The remaining for the most part either have their own boreholes (53 percent) or use natural sources. Only a very small proportion (2%) receive water from private tankers. Use of public water supply is most common in the Central Hills, which probably reflects such use by establishments located in the Kathmandu Valley. In the Terai, by far the most important source of industrial water supplies are boreholes (77% of establishments) with only 20% of establishments reporting use of public water supplies. Use of Boreholes 15. Table 4.3 shows the source and consumption of water by region, ecological zone and industry type. On average the establishments surveyed used 360 cubic meters (cum) of water per day, with tourism industries using the highest on average (923 cum/day). About 60 percent of total water consumed was from boreholes. On average, each establishment had two boreholes with a current market value of about NRs 0.7 million. Water Suipply Costs 16. Water supplied by public agencies was estimated to cost NRa 1.8 per cubic meter, while the operating costs of boreholes is estimated to be in the region of NRa 0.8 per cu.m. If the capital costs of boreholes (NRs 700,000 on average) is also included, amortised over 10 years at a 10% interest rate, the - 51 - Annex IV Page 6 of 11 costs of water from boreholes increases to over NRa 2.1 per cubic meter. However, given the poor reliability of public water supply, the capital costs of boreholes may be perceived as a cost of ensuring reliable production capacity. Clearly, the economic incentive for using boreholes exist even at the present low levels of water tariffs, and higher tariffs will only strengthen this incentive. Boreholes create numerous externalities for the community, such as depleting the aquifer and lowering the water table, and regulation of boreholes use on the one hand combined with improved (reliable) public water supply is needed to prevent further deterioration of the water situation, particularly in the urban areas of Nepal. Table 4.3 SOURCE AND CONSUMPTION OF WATER FOR ALL ESTABLISHMENTS Location/ Type Average Amount % Supplied % SuppLied % Supplied No. of Current Market of Industry of Water Used from Own from PubLic from Tanker Boreholes Value (cun/day) BorehoLes System Trucks in Use (NRs.1000) Development Region Eastern 212 84 16 0 4 105 Central 387 41 57 2 1 636 Western 742 53 47 0 1 4,783 Mid-Western 89 70 30 0 1 25 Far-Western 123 90 10 0 2 45 Ecological Region Teral 426 82 18 0 3 709 HilLs 304 29 69 2 0.3 668 Mountains 63 100 0 0 0 0 Industry tvM Manufacturing 381 52 47 1 2 1254 Agro-processing 215 69 31 0 1 104 Tourism 923 36 55 9 1 137 Mining 64 100 0 0 1 200 Total 360 58 41 1 2 701 C. TRANSPORT 17. Over 90 percent of transport requirements of the industries in the sample is met by road transport, and about a third of this is provided by trucks/vans owned by the firms themselves. The proportion of own-account trucking appears to be highest in the Terai and Mountains (see Table 4.4), which probably reflects the greater availability of common carrier fleets in the Kathmandu Valley (Hills). On average transport expenditures were just under NRs 0.7 million per year, however there was a wide range of variation: expenditures were highest in the Central (NRs 1,117,000 p.a.) and Par-Western (NRs 599,000 p.a.) regions, whereas in the Eastern, Western and Mid-Western regions expenditures were much lower (NRs 200,00 - 300,000 p.a.). These differences probably reflect the larger proportion of trade with India for industries located in the Central and Far-Western regions. - 52 - Annex IV Page 7 of 11 Table 4.4 USE OF DIFFERENT SHIPPING METHODS FOR TRANSPORTATION OF INPUTS AND OUTPUTS, 1990 Cost of Different Shipping Methods (in NRs.'000) Location/Type of No.of Own Hired Trucks/ Others Total Industry Cases Trucks/Vans Vans Only Develoment Reaion Eastern 34 12 83 5 251 Central 73 37 55 7 1,117 Western 21 18 79 3 199 Xld-Western 10 4 95 0 269 Far-Western 10 27 64 9 599 Ecolooical Reaon Teral 76 45 45 10 745 HiLls 68 17 81 2 646 Mountains 4 54 43 3 594 Industry Type Manufacturing 88 38 52 10 705 Agro-processing 48 26 73 9 653 Tourism 9 70 5 24 198 Mining 3 17 82 1 2r583 Total 148 33 60 7 695 Costs Incurred in Own Account Truckina 18. The high reliance on own account trucking results in higher than normal transport costs. As shown in Table 4.5, the industries using own account fleets, had on average four vehicles which operated about 20,000 km per year which is about a fourth of normal truck utilization rates. These Table 4.5 OWN ACCOUNT FLEET OPERATIONS * 1990 Location/Type of Percent No. of Own Distance Current Operating Cost per Industry Cases Trucks/Vans Travelled Market Costs Veh. Km with Own per Month Value of CNRs.1000) CNRs) Trucks/Vans (1000 KH) Fleet (NRs. '000) Devetopment Renion Eastern 18 3 6 1,260 443 106 Central 40 4 9 2,674 1,079 167 Western 24 3 5 574 236 65 Mid-Western 30 1 1 300 206 252 Far-Western 50 3 4 585 323 104 Ecological Region Teral 29 4 5 1,506 1,116 270 Hills 35 3 9 1,991 481 88 Mountains 50 11 0.3 5r079 649 4r823 Industry Tve Manufacturing 34 3 7 926 852 142 Agro-processing 25 4 6 3,370 448 163 Tourism 44 4 5 1,933 281 116 Mining 66 13 20 7,543 2r332 176 Total 32 4 7 1r896 772 153 53 - Annex IV Page 8 of 11 vehicles carry between 3 and 7 tons, and assuming an average S ton load, and a 70 percent load factor, the average transport costs are on the order of NRa 45 per ton/km (about US$1.00) which is a very,high cost of transport. The small sample (48 industries) makes it difficult to draw conclusions about individual regions or industry groupings. Impact of Transport Problems 19. There were 25 cases which reported a loss of output due to transport difficulties. These industries reported an average loss of 14 percent of output, valued at NRa 1.1 million in 1990. Highest losses were recorded in the Far- Western and Mid-Western development regions, which are also the ones with the poorest transport infrastructure. The trade and transit problem with India and the uncertain political environment in the country no doubt partly contributed to the transport problems affecting industries during 1990. Table 4.6 presents the survey results on industrial output losses. Table 4.6 LOSS OF OUTPUT DUE TO TRANSPORTATION BOTTLENECKS, 1990 Location/Type of Industry No. with Production % of Total Value of Lost Output Loss Value of (NRs.$000) Output Lost Develooment Reaon Eastern 0 0 0 Centrat 14 12 966 Western 6 13 437 Mid-Western 3 22 253 Far-Western 2 20 5,900 Ecologicat Reqion Terai 12 14 1,667 Hills 13 15 615 Mountains 0 0 0 Industry Type Manufacturing 18 13 974 Agro-processing 5 18 1,997 Tourism 2 30 220 Mining 0 0 0 Total 25 14 1,164 D. TELECOMMUNICATIONS 20. Table 4.7 summarizes data on telephone availability and use for 139 of the industries surveyed, the remaining 11 industries did not have telephone service. 21. Table 4.8 below presents data on the quality of telephone service experienced by the 139 establishments covered by the survey. The data appears to indicate that the Eastern and Far-Western regions receive better service than other parts of the country. Data for the Central region and the Hills is dominated by firms in the Kathmandu Valley, and the lower quality of service shown for these groupings probably reflects the inability of the Nepal Telecommunications Corporation to adequately meet the higher level of service demanded in the Valley. - 54 - Annex IV Page 9 of 11 Table 4.7 TELEPHONE INSTALLATIONS AND PAYMENT TO NTC, 1990 Location No. of Cases with No. of Lines Payment to NTC Total Cost for Waiting Tim NTC Telephone with (MRs. '000 per Installation for Last Service Different Nonth) (MRs/Line) Connection (No. Numbers of Weeks) Development Recion Eastern 29 2 8.8 7.5 39 Central 72 4 27.3 8.5 62 Western 18 2 7.6 7.7 32 Kid-Western 10 2 4.3 8.3 94 far-Western 10 2 8.4 5.3 5 Ecological Region Teral 73 3 9.7 8.6 38 Hils 65 3 26.9 7.0 68 Mountains 1 5 26.0 9.0 1 Total 139 3 17.8 7.9 51 Table 4.8 PROPORTION OF SUCCESSFUL CALLS ON FIRST ATTEMPT IN 1990 Location Within this City To Another City in To India To Countries Other Nepal then India Development Region Eastern 96 93 78 57 Central 89 78 62 48 Western 90 72 25 20 Mid-Western 99 64 50 35 Far-Western 100 88 86 84 EcMoicat Region 9 46 Tri96 84 62 46 Hills 88 76 60 50 Mountains 95 75 50 50 overalt 95 81 62 50 E. DRAINAGE. SEWERAGE AND SOLID WASTE DISPOSAL 22. Only 36 establishments, out of 141 responding, were served with public sewer connections, and only 7 had any sort of treatment facility. In general, the data indicate a very low utilization of any type of sanitary facilities by industries in Nepal. Expenditures associated with sanitation/waste disposal are correspondingly low. Table 4.9 presents the data for sewage disposal and treatment facilities. only 7 of the establishments surveyed had any form of treatment facilities. Such facilities also tend to be expensive, averaging a little over NRs 1.1 million per establishment with such facilities. - 55 - Annex IV Page 10 of 11 Table 4.9 PUBLIC SANITARY SEWER CONNECTION AND TOTAL COST FOR SEWAGE DISPOSAL, 1990 Location/Type of Saiple Total Cost No. Served Investment No. with Total Industry Distribution for Sewage with Public for Treatment Investment Disposal Sewer Connection Plant (NR6.1000) (NRs.'000) Connection (NRs.1000) Development Region Eastern 34 4.6 2 55 1 20 Central 67 8.0 32 15 4 1297 Western 20 4.9 1 0 1 1500 Mid-Western 10 2.4 1 0 1 0 Far-Western 10 20.7 0 .0 0 0 Ecolooical Region Terai 75 6.8 6 19 3 586 Hills 62 5.0 30 16 4 1650 Mountains 4 0.0 0 0 0 0 Industry Type Manufacturing 82 6.0 26 16 3 1000 Agro-processing 46 6.5 5 0.4 2 130 Tourism 10 3.2 5 37 2 2225 Mining 3 1.0 0 0 0 0 Total 141 5.8 36 17 7 1118 23. Details on the number of industries served with solid waste collection and disposal facilities are presented in Table 4.10. Only 13 establishments were served by the public system, 12 of them in the Central region and all of them in the Hills. Eighteen establishments contracted out for waste disposal, and a further two establishments owned their own trucks to dispose of solid waste. Table 4.10 SOLID WASTE COLLECTION AND DISPOSAL SYSTEM Location/Type of No. Served Charge/Year No. Payment/Year No. with Own Truck Industry with Public (NRs.*000) Contracting (NRs. 000) Trucks for Operating Solid Waste Out for Waste Cost per Collection Waste Disposal Year and Disposal (NRS.'000) Disposal Development Region Eastern 0 0 10 4 0 0 Central 12 9 2 0 1 77 Western 1 6 1 0.5 1 35 Mid-Western 0 0 4 3.3 0 0 Far-Western 0 0 1 50 0 0 Ecolonical Reqion Teral 0 0 16 6 1 35 Hills 13 8 2 0 1 77 Mountains 0 0 0 0 0 0 Industry Type Manufacturing 8 9 11 8 0 0 Agro-processing 1 22 6 3 1 35 Tourism 4 2 1 0 0 0 Mining 0 0 0 0 1 77 Total 13 8 18 6 2 56 F. RANKING OF INFRASTRUCTURE PROBLEMS 24. In order to assess and rank the problems industry faced with regard to various types of infrastructure, the survey requested the establishments to cite the types of infrastructure services which caused them the most significant difficulties. Table 4.11 presents a summary of the results. Clearly, unreliable electric power was the most frequently cited infrastructure problem, with fully one third of the establishments considering this a significant constraint. - 56 - Annex IV Page 11 of 11 Drainage problems and transport tended to feature in about 15 percent of responses, whereas only 8 percent cited water supply problems. Table 4.11 COMMENTS ON PUBLIC PROVISION OF INFRASTRUCTURE (% OF RESPONSES) Location/Type of No Comments Electricity Lack of Water from Difficult to Industry Voltage Drainage/Poor Boreholes not get Local Fluctuation Condition of Hygienic & Piped Transportation Drainage Supply System for Industry Insufficient Development Region Eastern 65 15 6 - 9 Central 41 25 12 4 15 Western 24 52 14 10 29 Kid-Western - 80 50 30 - Far-Western - 70 40 20 10 Ecological Zone Terai 32 44 23 6 14 Hills 48 23 7 7 12 Mountains s * * 50 Industry Type Manufacturing 36 31 12 7 19 Agro-processing 40 42 21 8 6 Tourism 60 10 20 0 0 Mining 33 33 0 0 33 Total 38 33 15 7 14 25. Regionally, however there were some considerable variations from the above overall statistics. Industries in the three western regions (Western, Mid- Western and Far-Western) appear least satisfied with infrastructure services. About 70 percent complained of electricity supply problems. In the Mid-Western and Far-Western regions more than 40 percent of the responses complained about drainage, and over 20 percent cited water supply problems. In the Western region, about 30 percent complained of poor transport facilities. Among ecological zones, industries located in the Terai were less satisfied with their infrastructure services than those located in the Hills. The latter is probably due to the generally better facilities in the Kathmandu Valley. The ranking of problems, excluding reliable supply of electricity, does not generally appear to change between development regions or ecological zones which in descending order of importance in the severity of problems appears to be: electricity supply reliability; drainage; transport; water supply. The exceptions being in the Hills where water supply is ranked second, and in the Central and Western regions where transport is ranked second. 26. Among industry groups, electricity supply appears to be the most critical problem with manufacturing, agro-processing and mining industries citing this most often. Manufacturing and mining industries cited transport second most often, whereas agro-processing industries indicated poor drainage as their second most frequent concern. None of the industry groups cited water supply as an important concern probably reflecting their ability as large users to invest in boreholes. 57 Annex V Page 1 of 26 TRANSPORT 1. Transport infrastructure in Nepal consists of: roads and trails; airports and airstrips; and railways and ropeways. Of these, road and aviation facilities carry the bulk of passenger and goods traffic. Railway and ropeway facilities are limited and their usage fluctuates. There is also an extensive network of trails which are often the only means of transport for many villages and communities located in the hills. A. Road Transport 2. The total length of the road network in Nepal increased more than tenfold over the last thirty-five years, from 624 km in 1956 to 7007 km in 1989 and to 8300 km in 1992, the latest year for which data are available. In fact, it is estimated that the road network in Nepal increased at an average rate of 250 km per year in the 1980s. In 1991, 3016 km (40.8 percent) of roads are paved, 1655 km (22.4 percent) are gravel and the remaining 2730 km (36.8 percent) are earth roads (DOR 1992]. Of the 7400 km of roads in 1991, Highways accounted for a total length of 2111 km (28.7 percent), Feeder Roads (24.9 percent), Urban Roads, 1098 km (15.0 percent) and District Roads, 2299 km (31.4 percent). The main network (about 1700 km) comprises the East-West Highway (Mahendra Rajmarg, about 1024 km), running mainly through the Terai, and the North-South highways. Figure 5.1 shows the network of main roads in Nepal. 3. The East-West highway, when completed, will traverse the entire length of the country and link major population centers in the Terai. Major North-South roads include the Tribhuvan Rajmarg which connects Kathmandu to Birganj to the south, at the Indian border (192 km); the Arniko Rajmarg (114 km) connecting Kathmandu to the Chinese border in the north; the Sidhartha Rajmarg (180 km) connecting Bhairahawa at the Indian border to Pokhara to the north; and the link between Biratnagar on the Indian border to Dharan and Dhankuta (100 km) to the north. The main network also includes Prithvi Rajmarg (200 km), the East-West link connecting Pokhara with Naubise on the Tribhuvan highway; and the north- south Gorkha-Narayangadh highway (61 km) which connects the Prithvi highway to the main East-West Highway (Mahendra Rajmarg) [Sharma and Shrestha 1992]. 4. In the Terai a number of feeder roads branch off towards the south linking the East-West Highway to rail and road heads in India. In the Mid and Far-West region several feeder roads joining district headquarters to the network system are in different stages of construction. His Majesty's Government of Nepal (HMGN) considers a mid-hill East-West Highway to be an important part of the basic road network system from a longer perspective. In fact, the Kathmandu- Pokhara section of it is already complete and is planned to be extended to Baglung. However, the rest of the sections, Terathum-Charikot in Eastern and Central regions and Pokhara-Doti in Western region through Far-Western region have not yet been considered. FICURES。1 ROADS AnD以IN TRAILS 一一 憤仔1斗戶乞一一－州雙―&－豐么－一－彎絮一f.- - 59 Annex V Page 3 of 25 5. The East-West Highway (942 km) is complete except for the section between Kohalpur and Mahakali (86 km) in the Mid-Western and Far-West regions of the country. Two important North-South roads under construction are Dadeldhura- Darchula section of Dhangadi-Dadeldhura-Darchula road and Ilam-Taplejung section of Charali-Ilam-Taplejung road. Road Network Distribution 6. Road densities (coverage) in Nepal are among the lowest in the world. Table 5.1 compares road densities in other countries (as of 1984) with those found in Nepal in the Terai and in the Hill and Mountain regions separately. Even countries with mountainous terrain, such as Afghanistan, have constructed considerably more roads on a per capita basis than Nepal. Considering the Terai separately, this relatively flat region has less than a third of the road densities found in neighboring India, and a little over 10 percent of the densities per square kilometer found in Bangladesh. Tabte 5.1 ROAD DENSITIES IN NEPAL AND OTHER COUNTRIES Km of Road per 1000 Poputation Km of Road per 100 sq.km Afghanistan 1.4 2.9 BangLadesh 1.8 112.0 Burma 0.7 3.3 China 0.9 9.0 India - TotaL 1.9 41.3 - Uttar Pradesh (1961) 0.8 20.9 Pakistan 1.1 11.9 Sri Lanka 1.6 37.7 United States 27.5 66.4 NepaL - TotaL 0.4 5.0 - Mountains & HILLs 0.5 4.0 - Terai 0.4 13.5 Source: Trends in the DeveLopment of Roads and Road Transport in Asia and the Pacific, ESCAP, Bangkok 1984; and Dept. of Roads, HMG/N, Kathmandu, 1992. J 7. In general, parts of the country situated to the north and towards the west have an extremely difficult terrain and are sparsely populated. Thus, out of a total of 75 districts in Nepal, 23 districts with a population of about 2.5 million, most of which are in Western and Far-Western regions, do not have any roads. Cost of road construction in the rugged terrain found in many parts of the country is high and thus access to these areas is limited mainly through trails used by human and animal traffic. Air travel is the only motorized mode of transportation available in some of these areas. Disaggregated data about the road network in 1991 are not available. Table 5.2 was prepared by adjusting the past data with information on appropriate lengths of roads under construction in 1987 [ADB 1988b] in respective categories in each region. It was assumed that all newly constructed paved roads are two-lane wide. - 60 - Annex V Page 4 of 26 Table 5.2 NEPAL ROAD NETWORK BY 1991-92 (ks) Paved Gravel Earth Region H, H2 F1 F, URB H F URB H F URB Eastern 27 110 239 228 99 117 226 71 95 102 104 Central 449 90 299 181 251 311 462 173 69 410 103 Western 181 138 69 209 105 4 159 29 171 188 110 Mid-Western 0 4 26 260 14 93 101 36 180 36 8 Far-Western 180 0 0 147 10 58 70 14 146 65 17 Total 873 342 633 1025 479 583 1018 283 661 801 342 There is some difference between Earth road Lengths in Nepal Road Statistics (NRS), 1989, and the ADB estimate. It seems that NRS tends to include same non-motorabLe tracks also in the category of earth roads. The difference is about 600 km. H, = hilly terrain - single Lane F, = flat terrain - single lane H2 = hilly terrain - two lane F2 = flat terrain - two Lane H = hilly terrain F = flat terrain URB = urban roads Source: Transport Sector Profile Study, Asian Development Bank, ManiLa, 1988 The Trail Network 8. Trails and tracks using human porterage were the only form of transport as well as the only means of communication in the hills until relatively recently. As early as 1814, there was a government transit system based on trails that used to work rather efficiently for the transit of official mail and other goods transport. The trail network is still very extensive and well used, and for those areas not connected to the road network human porterage remains the only form of transport. With the development of a road system, however, the nature of the trail network has changed, no longer providing long distance transport but feeding up into the hills from roads. With 23 out of the total of 75 districts having no road access and some others with a single earth road, many villages are accessible only on foot. The main trail and the relationship between trails and the road network is shown in Figure 5.1. The overall length of these trails is estimated to be around 15,000-20,000 km. Of the total 7,000 kma of the main trails, about 5,000 km are in the three western regions. Trails vary enormously in their standards, some being up to 2 m wide while others are barely discernible. Gradient can be as steep as 90 percent or even higher, and trails may become slippery water courses. Most trails now have drinking water facilities and trees planted at regular intervals for shady rest areas. 9. Over time trails have been, and are continuing to be, improved by construction of pedestrian suspension bridge crossings of rivers and gorges, which improve accessibility and reduce distances. Some 2500 trail bridges exist along various routes. Of these, around 400 are modern suspension bridges constructed by local districts and the suspension bridge division of the Ministry of Local Development. The condition of these bridges is highly variable and the older bridges are generally in poor condition. Accidents are not uncommon and in the worst one at Joljhibi some 100 people died when a bridge collapsed under load. There have been ambitious plans to increase the number of suspension bridges along the trails in each of the recent Five Year Development Plans. -61- Annex V Page 5 of 26 Achievement has been close to target but with wide variations in the performance of various agencies. Table 5.3 gives a brief account of suspension bridge construction history and target for the seventh plan (1985-90). 10. In all a total of NRs 121 million were spent in 1987-88 by various government agencies responsible for trails. The estimate for expenditures in 1988-89 was of the order of NRs 143 million. Several countries, such as Switzerland and the Netherlands, have been assisting the HHGN in its efforts for the improvement of trails and construction of suspension bridges. There is also a food-for-work program for trail improvement organized by the World Food Program. Table 5.3 SUSPENSION BRIDGE CONSTRUCTION 1970-85 AND TARGET 1985-90 Period Target (No.) Achievement (No.) Performance (%) 1970-75 27 10 37 1975-80 67 67 91 1980-85 315 300 96 1985-90 373 not known -- Source: Transport Sector Profile Study, ADB, Manite, 1988 Condition of Roads 11. In 1992, about 75 percent of the total network was estimated to be in poor to fair condition, a high percentage when compared to other low-income countries where road networks have been recently developed. The network's poor condition compounded by Nepal's difficult geography and weather, has resulted in high network operation costs which are four to five times higher than if adequate maintenence were implemented (Gray Cover Report on "Nepal - Expenditures in the Road Sector"). The Transport Sector Profile Study (ADB 1988b] estimated road conditions in Nepal. Out of a total of 1700 km of paved roads surveyed, about 650 km were in immediate need of pavement remedial works ranging from extensive patch work to rehabilitation and reconstruction. Some 476 kms were estimated to be in need of extensive shoulder repair works and numerous other cases of repair/replacement of drainage structures, correction of settled road sections and replacement/reconstruction of completely/partially washed out road sections were also indicated. A number of roads, although in otherwise good physical condition, had pavement of varying width and other such hazards. The conditions of gravel and earth roads were, in some cases, even worse than the paved roads. Regraveling is seldom done on graveled roads resulting in the development of ruts. Improper drainage is another problem and some roads, that are otherwise considered to be all weather roads, are almost impassable in wet season. Road Traffic 12. Traffic levels on roads in Nepal are relatively low, and most studies have assumed an annual growth rate of 5-7 percent for traffic movements in Nepal. A traffic data base study was conducted by ADB in 1986. The highest volume of two way traffic, 700-800 vehicles/day, was observed only on Hetauda-Birgunj, Kathmandu-Mugling and Biratnagar sections. A majority of key road sections had flows in the range of 200-600 vehicles/day, with some main roads having flows in - 62 - Annex V Page 6 of 26 the range of 50-100 vehicles/day. Truck and bus proportions of total flows were high on the trunk routes, 75-85 percent on the East-West Highway and Birgunj- Kathmandu section, and 50-75 percent on most other key roads. Traffic projections for various road sections are given in Table 5.4 for 1990, 2000 and 2010. 13. These projections are indicative and do not take into account particular road developments - for example, the completion of the East-West Highway which might alter the projected figures. In fact, traffic surveys carried out by the Department of Roads (DOR) in 1990 and 1991 indicate a much higher average daily traffic (ADT) than projected for 1990. The ADT on the Hetauda-Birgunj, Kathmandu-Mugling and Biratnagar sections is in the range of 1600 plus vehicles as compared to the projected ADT of 1048 vehicles per day. Traffic volumes on other important roads also appear to be somewhat higher than predicted. - 63 - Annex V Page 7 of 26 Table 5.4 INDICATIVE TRAFFIC 1986 AND PROJECTIONS 1990-2010 FOR KEY ROADS (AADT) Road/Section ActuaL 1986 1990 2000 2010 1990/91 East-Uest Highway (EWH) Kakarbhitta - Itabari 250 328 615 1004 Itahari - KameLa River (1234)d 350 459 862 1408 Keala River - PathLaiya 350 459 862 1408 Hetauda - Narayangarh (1685) 600 786 1474 2408 Narayangarh * ButwaL (1200) 400 524 984 1608 Sutual - Rapti River (297) 400 524 984 1608 Indian Border-Kathmandu-Chinese Border Birgunj - Pathlaiya 800 1048 1966 3211 Pathlatya - Hetuade (1685) 750 983 1844 3012 Hetauda - Naubise 200 262 492 804 Narayangarh - NugLing (881) 650 852 1599 2613 MugLing - Naubise (1600) 700 917 1721 2817 Naubise - Kathmani 800 1048 1966 3211 Kathmandu - DhutikheL 450 590 1107 1808 DhutikheL - Lamosangu (310) 400 524 984 1608 Lamosangu - Kodart 150 197 370 605 Muqling-Pokhara-ButwaL-India Border Mugling - Gorkha Road Jctn 250 328 615 1004 Gorkha Road Jctn - Pokhara (181) 150 197 370 605 Pokhara - SyangJa 200 262 492 804 SyangJa - Butwal 250 328 615 1004 Butwal - Bhairahawa 500 655 1229 2008 Other Roads(a) EWH Charali - FikaL 50 66 123 200 Fikal - ILam 20 26 48 78 Itaharf - Biratnagar (b) 700 917 1721 2812 Itahari - Dharan (b) 800 1048 1966 3211 Dharan - Dhankuta 150 197 370 605 Rajbiraj - Indian Border 200 262 492 804 Charikot - Jiri 70 92 172 282 Kathmandu * Trisuli (c) 70-200 92-262 173-492 282-804 Gorkha Road 50 66 123 200 Shairahawa - Lumbini 350 459 862 1404 Chandraute - Krishnagar 100 131 245 401 EWN - Ghorshi 100 131 245 401 a) From east to west b) Probably infLuenced by nearness to Itaherf c) Heaviest volumes near to Kathmmndu d) VaLues in parentheses are ADT observed in 1990/91 Source: ADS 1988a; DOR 1992 Projections based on pre-1990 traffic growth trends indicated that by 2010 traffic volumes on the East-West Highway will reach around 2500 vehicles/day on the busiest Betauda-Narayangarh section, assuming that a more direct Kathmandu- Hetauda road is not constructed. On other sections of the East-West Highway, the projected volumes by 2010 will be around 1400-1600 vehicles/day except in the Far West where the projected volumes are expected to be lower. 14. Traffic volumes on other currently busy roads such as Birgunj-Hetauda, Mugling-Kathmandu and Biratnagar-Dharan (all north-south roads heading towards Indian border) are projected to be around 3,000 vehicles/day, but volumes elsewhere on the system are typically not expected to exceed 1,000 vehicles/day, and for many roads, not to exceed 500 vehicles/day. Although some of the observed volumes in 1990-91 exceeded the predicted values, the projected 2010 traffic volumes can be expected to be well within the capacities of two lane roads in the relevant terrain, and the present roads in the network should have sufficient capacity to accommodate projected future traffic flows. - 64- Annex V Page 8 of 26 Road Transport Services 15. There has been a consistent increase in the operating vehicular fleet in Nepal. Table 5.5 shows details of various types of vehicles registered in Nepal during the period 1977-78 to 1986-87. The latest available data about the operating vehicle fleet date back to 1986, collected during the course of Road Transport Industry Study (RTI Study). About one fourth of the truck and bus fleet is operated on an own-account basis indicating the difficulty in obtaining reliable common carrier services. Table 5.5 ESTIMATED VEHICLE FLEET. NEPAL (1986) ConerciaL Private Use Other Of Total Trucks 2985 610 706 4301 Buses 1442 140 63 1645 Minibuses 724 185 43 952 Other Heavy Vehicles 33 144 295 472 Light Vehicles 2282 6675 3339 12296 Tractor 11 709 1655 360 2724 Motor Cycles 5 10195 3000 13200 Three Wheelers 704 65 54 823 Includes Goverranent, Public Sector and Diplomatic vehicles Dump trucks, tippers, cranes, etc. Includes powertitlers Source: Transport Sector Profile, ADS, Manila, 1988 16. Facilities for goods transportation by trucks and passenger services by buses and minibuses are provided both by the private as well as public agencies. The main public agency involved in the provision of transport services is the National Transport Corporation (NTC). It is a multimodal agency, but its share of road transport is, however, very limited. Based on statistics collected in 1986, NTC had a fleet of 59 trucks, 5 primemovers, 30 trailers and 21 container trucks. Fifty operating trucks were being used for internal transport; all the rest were used for transit trade through Calcutta. According to Nepal- India trade and transit treaty, only a government owned agency is entitled to operate container services. However, Indian trucks account for 70-80 percent of trade and transit because of lower rates. NTC also has a monopolistic control in the transport of goods imported by public agencies. NTC's Trolley Bus service operates in Kathmandu between Bhaktapur and Tripureswor (13 km) and provides daily service to about 10,000 passengers. By 1986, NTC had suffered a cumulative loss of NRs 78 million (includes losses from railway and ropeway services). 17. Another public agency, Sajha Yatayat under the Sajha Central Organization provides passenger transport facilities primarily in the Kathmandu Valley, besides providing some long distance services between Kathmandu and other towns. In 1986, out of its fleet of 100 buses only 30 were operational. About 13 buses operated on long distance routes. 18. The private sector has an extensive presence in road transport services. There are, however, only about half a dozen companies with more than - 65 - Annex V Page 9 of 25 25 trucks or buses. Nevertheless, collectively the private sector accounts for about 84 percent of the country's truck fleet and 96 percent of the bus and minibus fleet [ADB 1988a]. Private sector thus dominates the internal passenger and goods transport business. Private bus operation provides long distance services to almost all important towns in the country connected by the road network. In 1986 a total of 216 night bus services were operating between Kathmandu and other towns. Whereas the charges for the internal goods transport are determined by market forces, the bus fares are fixed by the Department of Transport (DOT). Bus services are also regulated. Road Sector Revenues and Expenditures 19. The 1986 Road Transport Industry (RTI) Study sponsored by ADB identified and calculated revenues related to road transport from the following sources: * Custom duties, sales tax and Indian Excise Duty Refund on vehicles, parts and tires; * Local government taxes and surcharges and sport council tax on imports of vehicles, parts and tires; * Vehicle registration and renewal fees, route permits, tolls, taxes on Indian vehicles in Nepal and driving license fees. The estimated total revenues for 1983-84 and 1990-91 were of the order of NRs 370.5 million and NRs 616.6 million, respectively. The projection for revenues for 1991-92 is about NRa 820 million. This indicates an annual increase of about 12 percent. Revenues by vehicle type in 1985-86 were as follows: light vehicles 54.6%; buses 18.4%; and trucks 27%. 20. Transport has historically been a major component of HMGN's development budget. In total development expenditures, social services have normally accounted for around one quarter of the total, and economic services and administration for around three quarters. Within economic services, transport accounted for around 27 percent of development expenditures in the Fifth Plan period (1975-76 to 1979-80) and was by far the largest sector in public expenditure terms. In the Sixth Plan period (1980-81 to 1984-85), transport accounted for a declining percentage of expenditure but overall averaged around 18 percent. The estimated allocation for the Seventh Plan period (1985-86 to 1989-90) for transport sector was 15.8 percent of the total expenditure on economic services. 21. Within the transport sector, development expenditures on roads normally account for about three quarters of the sector's total, bridges for about seven percent and aviation for about 20 percent [ADB 1988c]. The Seventh Plan allocations for roads and bridges are NRes 3381.5 and NRs 348.5 million, respectively. Implementation difficulties and financial constraints have caused the actual allocation and expenditure in the transport sector to be significantly different from the original plan. Table 5.6 gives details of these revised allocations and expenditures for roads. As is evident from the Table not only are development allocations for roads declining, but the expenditures as a proportion of the allocations declined steadily until 1990 with 46%. Details of -66- Annex V Page 10 of 25 the actual allocation and expenditure in the transport sector to be significantly different from the original plan. Table 5.6 gives details of these revised allocations and expenditures for roads. As is evident from the Table not only are development allocations for roads declining, but the expenditures as a proportion of the allocations declined steadily until 1990 with 46%. Details of the Eighth Five Year Plan (1992-97) background papers indicate that a sum of NR. 19.5 billion is estimated to be required to complete all presently on-going projects. This would, at present rates of development spending, imply a 15 to 20 year horizon for completion of ongoing projects. Table 5.6 ROAD BUDGET ALLOCATIONS AND EXPENDITURES IN CURRENT PRICES Allocation (NRs Million) E penditure (NRs Million) 1988 1989 1990 1991** 1988 1989 1990 1991 Development Budget* New Construction: 947.2 1205.4 1224.8 960.4 661.6 792.5 650.5 582.1 Highways 528.6 630.4 601.3 570.1 384.5 446.0 341.1 337.6 Feeder Roads 359.6 492.4 473.7 250.3 251.6 276.8 174.6 117.8 District Roads 29.0 52.5 119.9 140.0 26.6 69.7 134.9 126.7 Rehabilitation 158.7 344.9 346.2 521.2 20.7 108.0 110.3 671.6 Periodic Maintenance 121.2 169.4 149.5 182.9 110.6 139.3 104.5 145.1 Bridges 108.7 118.8 119.3 107.2 53.0 70.2 69.1 81.5 MiscelLaneous 59.0 79.1 32.2 36.4 17.5 35.0 15.5 33.4 Total Development 1427.7 1917.6 2071.9 1807.9 866.4 1145.1 949.8 1513.2 Expend. % of AlLocn. 61% 60% 46% 83% Regular Budget Routine Maintenance 40.0 50.0 60.0 50.0 39.7 47.5 56.1 48.9 District Road Office 5.3 5.9 7.0 6.6 4.8 5.9 7.0 5.5 Regional Road Office 9.4 10.6 13.7 18.7 8.9 10.4 13.7 15.2 Heavy Equipment Divn 2.4 2.5 2.6 3.9 2.1 2.3 2.5 2.2 15.5 25.7 25.9 14.6 17.4 21.7 17.7 14.6 Total Regular 75.6 94.7 109.2 93.8 72.9 88.1 97.0 86.4 Expend. % of allocn 96% 93% 89% 92% Total Road Sector 1500.3 2012.3 2181.1 1807.9 939.3 1233.2 1046.8 1600.1 * Includes estimates for district Level budget allocations. * Estimates from MOF. Source: "Resource Allocation, Budgeting and Resource Transfer System in the Road Sector of Nepal", N.D. Sharma and S.P. Shreshta, World Bank, Washington, D.C., 1992 22. During the period 1988/90, external assistance accounted for about 53 to 62 percent of development expenditures in the road sector, with loans accounting for about 28 to 32 percent and grants, including in-kind grants, accounting for 21 to 32 percent. In 1990, the road sector received about 90 percent of external assistance to the transport sector, which was up from about 75 percent in 1989. 23. The expenditure level for road maintenance in Nepal has been fluctuating. The expenditure level per kilometer in 1986-87 (latest for which record is available) was NRs 7600 (US$370, @ 1 US$ - NR. 20.5) for routine maintenance of paved and gravel roads and NRs 21,000 (US$ 1024) for periodic and emergency maintenance of these roads. In 1991/92, these expenditures levels declined to NRs 8305 (US$ 195, @ 1 US$ - NRs 42.6) and NRs 33,600 (US$ 787) respectively. The latter component, periodic and emergency maintenance, fluctuates considerably depending upon monsoon rains, landslides and other natural disasters but requires a minimum level of allocations which has not been -67- Annex V Page 11 of 25 guaranteed. The tentative allocations for 1992/93 confirmed the declining trend with NRs 10,000 (US$ 200, @ I US$ - 50.0) and NRs 21,000 (US$ 420), respectively. Table 5.7 gives details of maintenance budget allocations and the amounts spent on average per kilometer basis during the period 1981-82 to 1991-92. TabLe 5.7 AVERAGE ROAD MAINTENANCE EXPENDITURES IN NEPAL Year Road Length Maintained Routine Maintenance Periodic & Emergency (km) Cost/km (NRs) Maintenance Cost/km (NRs) 1981-82 2870 6969 24460 1984-85 3460 7225 30953 1985-86 3642 6864 19577 1991-92 4250 8305 33600 Notes: Road Length is for paved and gravel roads only, and no deduction is made for Lengths being maintained under deveLopment projects. Costs excLude adninistration costs. 1991-1992 figures are estimates. Source: Transport Sector ProfiLe, ADB, ManiLa 1988 and Sharma & Shreshta, op.cit Maintenance Expenditure Needs 24. The 1988 ADB Study estimated unit costs for routine, recurrent and periodic maintenance for different terrain and travelvay width combinations, in terms of 1987 prices. These costs were revised to reflect current prices (by using a 10 percent domestic and 5 percent international rate of inflation, 63 percent devaluation of Nepalese currency against US Dollar and a 90/10 split for domestic input/imported materials and services), and are shown in Table 5.8. A comparison of Tables 5.7 and 5.8 indicates that maintenance expenditures in the past have been considerably below levels needed to prevent a deterioration in the road network. 25. These inadequate levels of maintenance funding in the past has caused a sizeable portion of the road network to deteriorate too badly to be restored by regular maintenance activities. In 1987 a total equivalent of 302 km of roads were estimated to be included in a backlog maintenance program, over and above the 430 kms that were being improved or rehabilitated under various ongoing programs. There is no information about the implementation status, if any, of these programs. It is, however, believed that besides the above mentioned 430 km of roads, another 93 kms of paved roads have been rehabilitated or improved in the intervening period (1988-91) as planned [ADB 1988b]. The 302 km length recommended for inclusion in the backlog maintenance program along with the 430 km under other programs correspond to about 50 percent of the paved road network which was more than 10 years old by 1987. This provides a reasonably good yardstick for projecting the length of roads currently in need of backlog repair. - 68 - Annex V Page 12 of 26 Table 5.8 ESTIMATED 1991 ROAD MAINTENANCE COSTO PER KM PER YEAR (IN MRS 000's) Road Type Hilly Terrain Hilly Terrain Flat Terrain Flat Terrain Single Lane Two Lane Single Lane Two Lane (HI) (H2) (FI) (F2) PAVED: Routine 22 22 16 16 Recurrent 31 34 17 20 Periodic 44 41 41 41 Total 97 97 74 77 GRAVEL: Routine 22 22 16 9 Recurrent 118 118 33 9 Periodic 157 157 78 Total 297 297 127 18 EARTH: Routine 13 13 9 9 Recurrent 17 17 9 9 Periodic - - * Total 30 30 18 18 JV Estimated on the basis of 1987 estimates Source: ADB 1988b 26. Based on projected figures, there were about 3000 km of paved roads in Nepal by 1991 of which about 2167 km vere constructed before 1981 [ADB 1988b]. Using the 50 percent criterion and subtracting 523 kms (430 kme + 93 kms) out of it, the estimated road length in need of backlog repair is about 560 kms. The cost of the backlog maintenance program, in current prices, is estimated to be about NRs 910,000 per km. The cost was derived from the estimate prepared by the ADB Study consultants and updated to reflect devaluation of the Nepalese Rupee and domestic and international inflation, as explained before. The only difference here is that the cost split between domestic (labor and local supplies) and imported (materials and services) components being 75/25 instead of 60/40 as in case of construction, reflecting the more labor intensive nature of maintenance work. Thus, the total cost for the backlog maintenance program covering 560 kme, in terms of current (1991) prices, is about NRs 510 million (US$14.6 million). Assuming a 5-year backlog program, the annual cost would be NRa 102 million (US$2.92 million) per year in constant 1991 Rupees. 27. The resealing program under the periodic maintenance category is to cover 400-450 km of paved roads per year. The estimated cost for this work is NRs 185,000 per km in terms of current (1991) prices. This is based on the same criteria as for backlog maintenance except that the domestic/imported ratio was revised to 80/20. This indicates an yearly requirement of NRs 74 to NRs 83 million (US$2.1 to US$2.4 million) in terms of current prices. 28. Considerable amount of equipment support will be required to carry out the above mentioned construction, improvement, and maintenance programs. The ADB study assessed that most of the required equipment was either available with DOR or was due before 1990 under the Third IBRD Highway Project. However, some available equipment was in need of repair and rehabilitation, requiring about US$3 million in 1987 prices. An estimate for this work at present is US$4 to $5 million. - 69 - Annex V Page 13 of 26 Future Demand for Roads and Trail Improvements 29. As agriculture is the most important sector of Nepal's economy, any future road development will perhaps be closely associated with agricultural activities. The other aspect that needs attention is the access to remote areas, especially in the Western and Far-Western regions. HMGN had, and perhaps still has, plans to construct another East-West Highway in the mid-hills. However, given the state of economic development in that area and the high construction cost in rugged terrains expected to be encountered in the mid-hills, such a project may not be economically justified at present. 30. Farm to Market Roads: Future expansion in agricultural activities in Nepal are likely to be concentrated in the Terai. Its fertile plains and adequate supply of water for irrigation give it an edge over other areas and make it the largest producing area of the country. Surplus production from the area needs to be efficiently transported to deficit areas within the country. Similarly, fertilizer and other agricultural inputs need to be transported, mainly from India, to the production areas. A well planned network of low cost all weather feeder roads conveniently connected to the East-West Highway system, preferably at an interval of 10-20 km, is desirable. Of the 700 km or so of Terai, the eastern quarter has reasonable access to the East-West Highway (but not all at 10-20 km interval). A program of 50-60 such roads, typically 20 km long through selected corridors, means 1000-1200 km of new roads in all. To build an estimated 1000-1200 km network of all-weather low-cost farm-to-market roads primarily in the Terai region of Nepal over a period of 20 years, would have cost about NRa 430 to NRa 515 million in 1987 [ADB 1988c). Taking into consideration the effects of about 63 percent devaluation of Nepalese currency and domestic and international inflation, the equivalent cost of these roads in terms of 1991 prices is estimated to be around NRs 718 million to NRs 860 million (US$ 20.5 million to US$ 24.5 million). In view of the importance of such a program, particularly for agricultural development in the Western Terai, construction of such a network of feeder roads could be accelerated and completed within a 10 year period. This would require annual expenditures in constant 1991 prices on the order of NRs 70 to 80 million (US$2-2.5 million) in constant 1991 prices. 31. North-South Roads Linkini Terai with Hills: The key sections of North-South roads included in the Seventh Five Year Plan, and unlikely to be completed by the early 1990s, are as follows: - 70 - Annex V Page 14 of 26 NORTH - SOUTH RWMD PROJECTS Roads by Region TotaL Type Surface ERR Years Remaining Estimated Length Under Work Cost of Construction in Kas Remaining (1) (2) (3) Work in NRs Eastern 1. Chatara- Barah Chhetra- 97 F E+G 11 97 776 Bhojpur 131 N-S G 2 113 688 2. Kateri- 86 N-S G 9.3 +6 72 437 Okhaidunga 3. Phidim- 227 N-S G 17.1* 192 1,752 Taplejung 4. Dhankuta- Basantpur-Arun Central 1. Sindhuti- 58 N-S G 58 580 Manthati Western 1. Dunre- 115 F G 17.4* +10 115 532 Beiisahar-Chame 13 F E 13 92 2. BagLung-Beni I Mid-Western 1. ShaLubang- 44 F G +10 44 266 Pyuthan 2. SaLyan- 107 N-S E 7.4' 107 856 Musikot (Rapti 60 N-S E 5.0 +3 60 700 Roads) Surkhet- Dungeswor Far-Western 1. Dadeldhura- Baltadi- 215 N-S E 2.9 +10 215 1,088 Darchula_____ _ ___ TotaL 1archu1 __53 1,086 7.767 Source: EFYP Notes : (1) FzFeeder, N-S=North-South Highway; (2) E=Earth, GaGravel; (3) Based on Feeder Road Study (LBI, Renardet, East, CENAT) for 5 meter aLL-weather road, except where * denotes data from MoF. These roads involve a considerable length in difficult terrain, and are therefore expensive to construct. The estimated sum of NRs 7.7 billion needed to complete on-going projects is extremely high and reflects the high cost of these ambitious North-South corridor development programs. The economic justification of many of these roads is dubious at Nepal's present stage of development. 32. Kathmandu - Hetauda Direct Link: Another construction project under conpideration for the road network is a direct route south from Kathmandu to Hetauda. This project, although non-strategic, would offer substantial user benefits, as the existing 96-km long Tribhuvan Rajpath built in 1956 is too tortuous. For this reason most of the vehicles use the much longer (217 km) route via Mugling and Narayangadh. There are possible shorter routes. One such route, via Pharping, Kulekshi, Bhimpedi and Bhainse, is 67 km in length saving 150 km. Development of this route was estimated to coat, in 1991, at NRs 7000 million (about US$150 million)1'. Such a project however would require a five- fold increase in the present traffic levels, assuming that 80 percent of the 1' "Preliminary Study on Kathmandu-Hetauda Road Tunnel Project", Ministry of Construction, Tokyo, Japan, September 1991. - 71- Annex V Page 15 of 26 traffic from the Naubise - Hetauda road and 60 percent of the traffic from the Mugling - Narayangarh road would opt to be diverted to the shorter route. Therefore, during the next 5-10 years, the most likely justified investment would be improvements on the Thankot - Betauda link. 33. Trail Improvements and Remote Area Access: There are areas in Nepal, particularly in the Western and Far-Western regions, where the cost of construction of roads is prohibitive. The 1988 ADB Study estimated that at least for one such corridor (Jumla in Western region) the cost of road construction is almost seven times as high as that of providing regular air transport services to the area. Air travel will, therefore, continue to be the only form of motorized transport in these areas. The only means of surface transport for a long time to come will be trails, an extensive network of which already exists in such areas. It may be desirable to undertake improvements of trails, where possible, to make them usable for small tractor, motorcycle or carriage animals such as mules. A recent UNDP study estimated that about 20 percent of trail lengths in Mid-Western and Far-Western regions need or have potential for upgrading to "mule trails" of 2 m. width [UNDP 1990]. Installation of suspension bridges to replace existing bridges that are in poor condition along the trails is the major component of trail improvement works. Over the period 1990-2000, a continuing program of around 50 new or replacement bridges per year could be a very rough estimate of requirements [ADB 1988b; 1988c]. 34. Out of the existing 7000 kme of the main trail network, about 5000 kms are in difficult to access Western, Mid-Western and Far-Western regions. Furthermore, there are some 2500 trail bridges in Nepal of which about 400 are modern suspension bridges. It is estimated that 20Z of the trails and almost all of the remaining trail bridges need improvement. The cost of improvement of 1400 km of trails was estimated at NRs 75,000 per km in terms of 1987-88 prices. The cost for the construction of suspension bridges was estimated at NRs 15,000 per meter of span length (UNDP 1990]. The corresponding costs in 1991 prices are roughly NRs 100,000 per km for trail improvements (no foreign component) and NRs 24,300 per meter for suspension bridge construction respectively. These rates, as all others, are derived on the basis of an estimated 10 percent domestic and 5 percent international rate of inflation, respectively, and also reflect the devaluation (about 63 percent) of Nepalese currency against the US Dollar since 1987-88. Thus, the total cost for improvement of 1400 km of trails is NRs 140 million (US$4 million) and that for construction of 2100 suspension bridges with an average 50 m span is about NRs 2550 million (US$73 million). Suspension bridge construction has averaged 60-70 bridges per year in the recent past. Assuming that this rate could be increased to 100 bridges per year, this would imply a 20 year program, with an annual expenditure (in constant 1991 NRa) of about NRs 150 million. This program of trail improvement is a fairly expensive one in relation to the cost of feeder road improvement in the Terai (NRs 70-80 million p.a.). HMGN should carefully weigh the relative merits of these, in deciding priorities for allocation of scare resources. - 72 - Annex V Page 16 of 26 Estimated Financial Requirements for the Road Sector 35. Table 5.9 summarizes the above cost assessments for road and trail sector in Nepal. It should be noted that the Table reduces new construction of N-S roads by two-thirds to include only high priority economically justified projects (para. 31) connecting the Hills to the Terai, in view of the high cost of such roads. Further, a more effective maintenance strategy is needed and would mean increasing the sectoral share for maintenance from about 20 percent to about 60 percent of available resources, in order to cover adequately the requirements of essential roads in good and fair condition (50 percent of network) and minimum maintenance of the remaining roads to prevent closures. Table 5.9 FINANCIAL REQUIREMENTS IN THE ROAD SECTOR IN NEPAL Item FY 1993 FY 1994 FY 1995 FY 1996 FY 1997 Total EFYP USS USS ArteriaL Road Construction 8.6 9.6 10.8 12.0 13.6 54.6 174.9 Comector & LocaL Access Roads 4.1 4.1 4.1 4.1 4.1 20.5 33.2 Construction 7.4 7.4 7.4 22.2 22.2 On-Going RahabiLitation 5.5 8.2 12.0 19.5 20.0 65.2 17.7 Additional RahabiLitation 61.1 Upgrading (1) 8.3 10.2 12.5 15.3 18.5 64.8 37.1 Periodic Maintenance 2.0 2.4 2.9 3.5 4.2 15.0 Routine Maintenance 1.0 1.1 1.2 1.3 1.5 6.1 Emergency Repair 0.5 0.5 0.5 0.5 0.5 2.5 4.7 Trail Improvement 3.2 3.2 3.8 4.2 4.6 19.3 19.9 Suspension Bridges 1.0 1.5 2.4 2.8 3.2 10.9 11.7 Land Acquisition 2.0 2.0 2.0 2.0 2.0 10.0 10.8 InstitutionalDeveLopment TotaL 35.0 40.9 59.6 65.2 72.2 291.1 393.3 S o u r c : Nepal - Expenditures in the Road Sector, World Bank, 1992 N o t e a (1) Upgrading includes gravel to paved surface, but excludes rehabilitation. (2) Periodic and routine maintenance allocations reflect a shift from Lou-medium to medium-high absorptive capacity scenerio in order to cover maintenance needs of the network in 1997. B. Air Transport 36. Aviation, second only to roads, is an important mode of transport in Nepal. There is an extensive network covering almost the entire country, and penetrating many areas inaccessible by road. Airports in different parts of Nepal vary widely in their characteristics and services offered. Until 1992, Royal Nepal Airlines Corporation (RNAC), a wholly government-owned undertaking, exclusively provided scheduled air services to most of the domestic airports. More recently, the Government has licensed four private airline operators. Fares for domestic flights are low and need to be subsidized from surpluses generated by RNAC in its international operations. Demand for air transport is generally linked to the poor access provided by the road network, and with expansion of the road network demand on some routes has fallen considerably. -73 - Annex V Page 17 of 26 Existing Aviation Infrastructure 37. A 1989 study, "Development of Civil Aviation in Nepal" by the Japan International Cooperation Agency (JICA), reported a total of 43 airports under the control of the Department of Civil Aviation (DCA). Of these, Kathmandu's Tribhuvan Airport is the only international airport. Among the 42 domestic airports (excluding Kathmandu), standards and facilities vary considerably because of extreme variations in topographic and other physical conditions. The Department of Civil Aviation classifies the airports in four categories as shovn in Table 5.10. Figure 5.2 shows the location of airports in Nepal and their categories. There are 26 airfields located in hilly and mountainous areas of the country where terrain and approach conditions permit only short take off and landing (STOL) operations. Only the Tribhuvan International Airport (TIA), Kathmandu, qualifies for category A. Two recent air tragedies linked to human errors have put pressure on the Government to upgrade the navigational aids at Tribhuvan Airport. 38. The condition of the category B airports was reported to be generally good in the Transport Sector Profile Study for Nepal prepared by Asian Development Bank F. Runways and taxiways were found to be in sound condition, except for Nepalgunj. The terminal building and other structures were also in solid, though dilapidated, condition. The poor condition of Nepalgunj runway was a cause of concern for the Civil Aviation Department and studies for runway improvement and strengthening are planned. Concern about certain safety aspects of runway orientation at Pokhara has prompted the DCA to acquire a new site for the airport. 39. Category C airports are similar in size and layout to Category B airports; those on the Terai have excellent approaches but a few in the Hill regions are restricted. These airports are also located close to towns. In 1988, the condition of category C airport runways was generally good considering the fact that these runways are not paved. Exceptions to this were the airports at Dhangadi, Janakpur and Tumlingtar where inadequate strength, flooding or softness of the runway during monsoon is observed. The 15-25 year old buildings are in dilapidated condition primarily because of their age and lack of maintenance. 3' "Nepal Transport Sector Profile Study", Asian Development Bank, Manila, 1988 -74 - Annex V Page 18 of 26 Table 5.10 REGIONAL DISTRIBUTION OF AIRPORTS BY CATEGORY2 AIRPORT CATEGORY Region A B C D Total Eastern 1 2 8 11 Central 1 - 3 4 8 Western 2 - 5 7 Mid-Western - 1 1 6 8 Far-Western - - 1 8 9 Total 1 4 7 31 43 SCateary A: All weather airports with paved runways long enough to handle jet aircraft Cateaory B: ALL weather airports with paved runway, except in case of Pokhara which has a gravel runway Category C: Fair weather airports with unpaved runways suitable for operation of HS-748 type aircraft Catenory D: ALL other air weather airports with unpaved runways, 26 of these are STOL airfields and four are at attitudes exceeding 3000 m. Source: DCA 1990 40. Several of the Category D airports, especially the STOL airports, are characterized by steep sided valleys and short unsurf aced runways permitting only short take off and landing from one direction. These airports are situated either beside rivers, on level benches of land on valley sides or in some cases on the top of ridges. The condition of Category D airports varies considerably from one airport to the other. Some of them are well grassed and are easy to maintain while the others are in poor condition. Usually the bad condition is attributed to some specific problems, such as water runoff erosion, surface irregularities, visual obstructions and so on. Buildings at these airports are basic or in some cases nonexistent. The 1989 JICA study evaluated the airports in terms of the civil structure and navigational aid/telecommunication equipment condition, and identified various deficiencies at each. 41. Only Kathmandu and Nepalgunj airports have an approach lighting system. Runway edge lights and runway threshold lights are provided at Biratnagar and Bhairahawa airports but are not operational. There are Precision Approach Path Indicators (PAPI) at nine airports, the rest have no lighting facilities. Half-hourly weather observations are made in Kathmandu airport, and another four airports have facilities for hourly observations. Many airports in the mountains do not have electric power supply, and solar battery power is used. Some airports in Terai region do not have a reliable electric power supply and emergency generators are provided but in many cases are not maintained properly. Aircraft fueling facilities are provided at only five airports and flights to other airports must carry enough fuel for a round trip. Figure 5.2 REGIONAL DISTRIBUTION OF AIRPORTS IN NEPAL •SINIKO G3 DARCHUU. *PATAN OAJHANC BAJURA • SANKEBACAR j oCHAURAJHARI . 00 DRANCADHI DMORPATAN T ........ ROLP OPOKHAR -DANC LA1 AA NEPALCUNJ -t ..·.• ARAPUR KATIHANDU JIRI-. RAME RDATAR TAP .EJU DHAIRAHAtJAY •.A..I. •. ECHIAULt .....*IAMIDANDA . .4 . .8110JPUR o CATECORY A AIRPORT HICH HIHALAYAS . • o CATECORY 8 AIRPORT AN • CATECORY C AIRPORTILL RhECIONf • CATECORY D AIRPORT I-P-N t DCA Planned .a.,E,, ss. A CHAIDRACAD1IX $- 0< Source: JICA 1989 - 76 - Annex V Page 20 of 26 42. Air traffic control, navigation aids and telecommunications facilities are generally inadequate for safe and efficient operations. Most of the available equipment is obsolete. Categories B and C and most of Category D airports are manned and have radio communications equipment of varying age and condition resulting in poor point-to-point communications. Air traffic control arrangements in the form of aerodrome control are available at only eight airports. Others have only Aerodrome Flight Information Services. No domestic airport has approach control service or area control services. Non-Directional Beacons (NDB) are in place at 22 airports. VHF Omni-Directional Range/Distance Measuring Equipment (VOR/DME) is available at Tribhuvan International Airport (TIA) Kathmandu and Nepalgunj Airport. Instrument Landing Systems (ILS) are not in place at any of the airports. Control consoles with VHF and HF are provided as the minimum facilities at many local airports. Communication between Kathmandu Communication Center and distant places like those in Western regions is difficult. At Kathmandu, there is an aeronautical fixed telecommunication network station with international circuits. The ATS direct speech circuits between Kathmandu Air Control Center (ACC) and Calcutta ACC and Varanasi ACC in India are not fully operational. Work is in progress to establish a satellite link with Bombay. Instrument approach and departure procedures are established at only two domestic airports besides Kathmandu. Most airports operate by visual flight rules. There are established airways between India and Nepal but no domestic network exists. Air Transport Services 43. RNAC's fleet of aircraft presently consists of two each of Boeing-757 and B-727, acquired on long term credit leases in the period 1987-89, and used exclusively for international services. For its domestic services, RNAC had in 1988: 3 British Aerospace Hawker Siddley 748 (HS-748) aircrafts; 10 DeHavilland Canada DHC-6s (Twin Otter); and 3 Pilatus Porter PC-6 type aircrafts. The domestic route network of RNAC is nationwide, serving all areas of the country. Most of the important towns are linked to Kathmandu with direct flights. Local flights to and from regional centers like Nepalgunj, Biratnagar, Bhairahawa and Dhangadi are also operated. The larger BS-748 aircraft (with 44 seats) are used on routes to Biratnagar, Meghauli, Nepalgunj, Bhairahawa and Pokhara. Twin Otter aircrafts (19 seats, but weight restricted to certain higher altitude airfields) are used to all other airfields except Manang, which is higher than 3000 m, beyond the range for which Twin Otter is certified to fly. PC-6 aircrafts are used for this route. As of 1990, six airfields namely Dhorpatan, Langtang, Syangboche, Palungtar, Chandragadi and Rolpa were not served by scheduled services. The first three airfields, being above 3000 m altitude, require a PC-6 type aircraft. Palungtar is very close to the Pokhara-Mugling road and demand has fallen, leaving little justification for a scheduled service. Only a charter service is now offered. 44. The number of domestic airports to which RNAC operated schedule services dropped to 32 in 1990. This seems to be in line with the recommendation of the 1988 ADB study. Charter services mainly transport food grains and construction materials to inaccessible areas. The 1988 study estimated this business to be varying between 10-20 percent of the total ton-kilometers offered in the 80s. Climate and other physical conditions strongly influence the scheduling of air services in Nepal. Because of high air turbulence and thick clouds that usually develop around mid-day in the valleys in the hill region, air services to many areas are only possible during the period 6:30-12:00 noon. During certain times of the year, extensive and dense early morning mist is experienced in Kathmandu, prohibiting flights from Kathmandu airport and thus further constraining the utilization of equipment. Table 5.11 provides details -77 - Annex V Page 21 of 26 of aircraft utilization in domestic service. In addition to climate and physical conditions, aircraft utilization is affected by the age of the existing fleet. Most of the aircraft in RNAC's domestic fleet are old and some of them are reaching the end of their commercially useful lives. The HS-748s in particular are known to have high maintenance costs. Table 5.11 also gives details of the age structure of RNAC fleet. TbLe 5.11 AIRCRAFT UTILIZATION AND FLEET AGE STRUCTURE Aircraft Aver ge Daily Utitization Hrs) A rcraft Age in 1987 (Years) 1981/82 1982/83 1983/84 1984/85 0-4 4-9 10-14 15-20 HS-748 2.74 2.67 2.50 2.66 - - 3 (3 Aircraft) DHC-6 2.67 3.05 3.02 2.83 2 3 3 2 (10 Aircraft) PC-6 0.39 0.48 0.34 0.46 - - 2 1 (3 Aircraft) I I Source: Transport Sector Profile, Asian Development Bank, Nanita, 1988 Air Traffic Growth and Proiections 45. Domestic passenger traffic growth over the ten-year period from 1976/77 to 1986/87 has averaged about 4 percent p.a. Numbers of domestic passengers grew from about 185,000 to about 275,000, while passenger-kilometers grew from 32 million to 48 million over the period. Traffic has however been fluctuating between strong growth in the late 1970s, equally steep declines in the early 1980s, followed by a recovery in the mid-1980s. It is understood that capacity constraints resulting from equipment shortages and maintenance problems accounted for the declines in traffic. The slightly higher growth rates recorded for passenger-kilometers compared with numbers of passengers, is due to the addition of 10 new destinations of which 8 were distant destinations. The overall upward trend in traffic, however, was not applicable universally to all airfields in the country. Of the 38 airfields receiving scheduled service in 1987, eight have experienced declining traffic, and in general these declines were related to the provision of road links in the zone of influence of these airfields. 46. The most comprehensive future growth forecasts, taking full account of the expanding and (planned) future road network were prepared by the 1988 ADB study. It gives future passenger volumes for the 38 stations where scheduled services were operative in 1987. Table 5.12 gives details of these forecasts and is based on the following approach: * No forecast was made for airports that have already shown sharp declines in traffic (often as a result of road link becoming available), except Tumlingtar, Bhairahawa and Meghauli. * For airports with all-weather road access committed, projections were made to the date of completion only, assuming future air traffic volumes will never be higher. * For other airports projections were made at an annual growth rate based upon the historical trend but adjusted upwards as airports with declining traffic have been excluded. -78- Annex V Page 22 of 26 The projections indicate that about 12 domestic airports will emerge as major traffic centers in the network generating over 10,000 passenger trips a year. Pokhara is expected to handle almost 150,000 passengers p.a. in the year 2010 based primarily on its attractiveness as a tourist destination. Nepalgunj and Biratnagar in the Terai will reach about 100,000 passengers p.a., from their current levels of about 60,000. Lukla and Meghauli, both tourist destinations, are expected to handle between 30,000 to 40,000 passengers, which is a very large increase over present traffic levels of just over 10,000 passengers. Both Lukla and Meghauli are now Category D airfields, and according to these projections will overtake all of the Category C airfields and even Bhairahawa which is a Category B airfield. Table 5.12 DOMESTIC AIRPORT TRAFFIC FLOWS - 1981/82 TO 1985/86 AND PROJECTIONS - 1990 TO 2010 ('000 PASSENGERS) Schedule Service Traffic Lows (a) ProJections Airport 1981/82 1985/86 2000 2010 Bhadrapur (b) 8.7 0.5 -(g) Biratnagar 44.9 46.9 70.1 95.3 TapLejung 0.3 1.3 awr(h) ar Tzntingtar 13.2 11.0 awr ar Bhojpur 5.3 7.9 ar awr Lamidande 9.2 10.6 swr aur Rumjatar 0.3 5.6 awr awr Phaplu 0.2 0.8 1.2 1.6 LukLa 8.9 10.6 24.0 42.9 Rajbiraj 0.7 0.3 - * Siara 14.6 11.5 17.4 23.4 Janakpur 6.4 5.4 nrlCi) nrt Ramechap 0.1 2.9 awr ar Jiri 0.8 0.2 - - Bharatpur 1.0 0.1 - Neghauli 9.2 8.1 18.3 32.8 Pokhare 23.3 36.4 82.3 147.4 BagLung(e) 3.5 10.0 aur owr Jamsam 3.7 6.0 9.1 12.2 Manang 0.0 0.5 0.8 1.1 Bhairahawa 14.2 10.9 10.9 10.9 Dang 7.7 1.4 - - NepaLgunij 24.1 46.4 78.7 106.8 Rolpa 0.1 0.1 our aur DoLpa 1.4 2.5 3.8 5.1 Chaurjhari(d) 5.2 8.3 12.6 16.9 Junla 2.6 8.3 12.6 16.9 Surkhet 9.3 11.6 ar our Tikapur 1.8 ar ar Dhangedhi 9.7 10.9 ur ar Sanfebagar 4.6 13.3 air aur Silgari (Doti) (e) 5.2 8.2 11.3 15.2 Bajura * 1.9 2.9 3.9 BaJhang 3.0 4.7 7.1 9.6 Simikot 0.4 1.6 2.4 3.3 Mahendranagar 3.6 7.2 sar awr Baitedi(f) 1.0 3.4 aur ar Darchute - 0.2 0.3 0.8 (a) Traffic flows are sum of arriving and departing passengers (b) Also known as Chandragadhi Cc) Also known as BaLewa (d) Also known as Rukum (e) Also known as Dipayal (f) Also known as Patan (g) Dash indicates declining traffic (h) aur denotes all weather road (M) nrL denotes new road link Source: ADB 1988a -79- Annex V Page 23 of 26 47. The 1989 JICA study F made projections for international passengers to the year 2010. According to these projections the number of international passengers is expected to increase from just under 600,000 in 1991 to almost 2 million by the year 2010. The mix of passengers between Nepalese and foreigners. would also change from around 70 percent foreign today to about 50 percent foreign in the year 2010. 48. Cargo traffic volumes were also projected by JICA for both domestic as well as international air cargo. In the case of domestic cargo, relatively moderate growth is forecast with traffic growing from about 2900 tons in 1990 to about 4200 tons in the year 2010, i.e., a growth rate well below 2 percent per year. International air cargo traffic, on the other hand, is projected to grow rapidly from around 24,000 tons in 1990 to 138,000 tons in 2010, which is an average annual growth rate of over 8 percent. This large increase in international air cargo traffic has been projected based on recent increases in the exports of ready-made garments and to a lesser degree, carpets. Export cargo is in fact expected to grow much faster, from about 14,000 tons in 1990 to 92,000 tons in 2010, which is an annual average growth rate of 9.5 percent. As a result, international air cargo traffic which is fairly balanced today is expected to become significantly imbalanced in the future with export cargo becoming about twice as large as import cargo, and this will mean that freight rates for export air cargo is likely to increase significantly over present levels. Air Transport Costs and Tariffs 49. Domestic air fares in Nepal are generally low. Table 5.13, based on findings in 1987, shows a comparison between fares and estimated operating costs for some selected routes. RNAC's operating cost for a Twin Otter aircraft was around NRs 11,500 an hour or about NRs 590 per passenger-hour at full capacity of 19 passengers. However, because of weight restrictions on many routes, these planes operate with 15 passengers, increasing the unit operating cost to NRs 740 per passenger-hour at effective full capacity. The estimated operating cost for a HS-748 aircraft was around NRs 25,180 per hour which corresponds to a unit operating cost of NRs 570 per passenger-hour at full capacity of 44 passengers. The load factor for RNAC domestic operation (average for 1984-85, 85-86 and 86- 87) has been 65 percent. Consequently, the average actual operating costs per passenger-hour for a Twin Otter was NRs 900 and for a HS-748, NRs 880 per passenger-hour . Among the routes shown in Table 5.13, cost coverages are the lowest on long direct Twin Otter flights out of Kathmandu (e.g., those to Bajhang, Dhangadi, Jumls and Tumlingtar), all in Hill regions. Cost coverages are comparatively better on the flights to Terai [ADB 1988b]. Thus, in general, it may be concluded that domestic air services were being operated at a loss; with services to Hill region destinations and other long haul flights by Twin Otters operated at a substantial lose. This was, and perhaps still is, a part of the policy to subsidize services to areas in the Mid and Far-Western regions. 50. The fares shown in Table 5.13 have been effective since 1985. Latest figures for the hourly operating costs in 1988 were NRs 28,200 and NRs 12,600 for HS-748 and Twin Otter aircrafts, respectively. The maximum subsidy has also come 2' "Development Study for Civil Aviation in Nepal", Japan International Corporation Agency, Tokyo, Japan, 1989 - 80 - Annex V Page 24 of 26 down to about 55 percent from 67 percent in 1987. No further information about the fare revision or subsidy situation is available. Despite the fact that the range of implicit subsidies on domestic services in Nepal is up to 55 percent, the overall average revenue-cost ratio on these operations was estimated to be around 82 percent for the year 1986-87. This indicates the large contribution to revenue made by non-Nepalese nationals paying much higher fares on tourist routes like Kathmandu-Meghauli (Royal Chitwan National Park), Kathmandu-Pokhara, Kathmandu-Jomsom, Pokhara-Jomsom and the Mountain flight. Moreover, RNAC makes profit on its international operations more than compensating for the losses on its domestic services. Overall, the Corporation expects to continue to show a modest surplus revenue over operating cost. Table 5.13 RNAC DOMESTIC AIRFARES AND ESTIMATED AIRCRAFT OPERATING COSTS 1986/87 (NRa) Fareb/ Estimated Estimated lapLicit Sectorg/ (NRs) Flying Time operating Subs dyc/ (Hours) Costb/ (NRs/Pax) (NRa/Pax) (% of Cost) Twin Other KTM-Bajhang 615 2.00 1,806 1,191 65.9 KTM-Dhangadi 650 2.00 1,806 1,191 65.9 KTM-Janakpur 320 0.58 524 204 38.9 KTM-JumLa 430 1.58 1,427 997 69.9 KTM-Simra 250 0.33 298 48 16.1 KTM-TuaLingtar 270 0.92 831 561 67.5 BIR-TunLingtar 225 0.42 379 154 40.4 KEP-Junta 215 0.67 605 390 64.5 KEP-Simikot 310 0.92 831 521 62.7 HS 748 KTM-Bhairahawa 400 0.67 590 190 32.2 KTM-Biratnagar 450 0.83 731 281 38.4 KTM- 560 1.33V/ 1,171 610 52.2 Nepalgunif/ P/ KTM: Kathmandu, SIR: SIratnagar, KEP: NepaLgunj 1/ Domestic Local fares since 27.4.85, confirmed 23.3.87 2/ Expressed in decimal fractions, not hours and minutes S/ Adjusted for average domestic network Load factor of 65 percent With average domestic toad factor 1/ Via Bhairahawa. If Nepelgunj was flown direct the operating cost would be NRa 1,027 and the subsidy MRs 467 or 45.5 percent of cost Source: ADS 1988a DCA Revenues and Expenditures 51. The Department of Civil Aviation charges fees for the use of airport facilities in the form of landing fees, parking fees, and housing charges. Other, though minor sources of revenue, are fees charged for navigational aid services provided to incoming and outgoing flights; room rentals; rent of land and hangars; income from oil sales; vehicle parking and entrance fees. The major source of income, however, is the embarkation tax paid by passengers which accounts for more than 60 percent of DCA revenues. Besides air port tax (embarkation tax) which is collected from passengers at the check-in counter, minor fees such as entrance and vehicle parking fees are collected in cash. All other charges are billed to clients and there is no assessment available as to how much of these invoices have been paid and how much are in arrears. Billing, accounting and auditing of these invoices, and payments against them, are done -81. - Annex V Page 25 of 26 at local levels. Reporting of invoices and subsequent payments is poor, and the DCA headquarters in Kathmandu does not have an up-to-date record of local revenue collections. Kathmandu's Tribhuvan airport generates the bulk of DCA revenues, generally accounting for about 95 percent of all revenues collected, which in the late 1980s totalled about NRa 45 million. 52. Recent data on DCA expenditures were not available, and the most recent information is for 1984/85. At that time regular (or recurrent) budget expenditures were at about NRs 20 million, while development budget expenditures were on the order of NRs 40 million per year. Based on this limited information it appears that DCA recurrent expenditures were considerably below the level of revenues collected (about NRs 45 million), and were grossly inadequate for proper operations and maintenance of civil aviation infrastructure. Investment Needs 53. To improve both international and domestic operations, the 1989 JICA study suggested projects for implementation on a priority basis. Table 5.14 summarizes these projects and their respective estimated costs in terms of 1991 prices. The original costs based on 1988 prices were revised to reflect the estimated domestic and international annual rate of inflation of 10 percent and 5 percent, respectively. The projects for Kathmandu and Pokhara airports were scheduled to be completed in two phases - Phase I was recommended to be completed by 1994 and Phase II by year 2000. These developments, if implemented, will enable the airports to handle the expected traffic by 2010. All other works were recommended to be completed by 1995. Clearly, the levels of investment proposed, quite apart from the economic viability of individual projects, is considerably beyond levels affordable for the aviation sector. 54. The strategy for the aviation sector calls for reducing the number of airports, and concentrating resources on those airports which are (i) essential to provision of basic services in remote areas, and (ii) potentially important to development of tourism in the country. Airports in areas with limited tourism potential, and where recent expansion of the road network has provided road access are slated for closure. Altogether some 12 to 13 airports have been identified as candidates for closure in the 1990s. 55. The airport maintenance budget in recent years has been of the order of NRa 3-4 million for domestic airports and NRa 7 million for TIA. The total amount is considered highly inadequate for the existing network, but if fewer airports (possibly 13 less) are in operation and some of the more problematic STOL airfields are paved, the need for maintenance will be less. However, under the current conditions the maintenance requirements are considerably higher than what is spent now and probably an allocation of US$3-4 million per year will be necessary for a minimum level of maintenance. -82 - Annex V Page 26 of 26 TabLe 5.14 SUMMARY OF PROPOSED FUTURE WORKS FOR AVIATION DEVELOPMENT 1992-2010 IN NEPAL (US$ MILLIONS) Conponent Investment Maintenance Total SoA.O 1. TIA Kathmandu (extension) 580 102.00 2. Pokhare Airport (construction at new site) 102* 1eco 3. Jomsom, Simikot, Lukta, Syangboche (extension, improvement) 2 8e= 4. Construction of New Airports (Mugu and 3 others) 1.0M 5. Lengthening and Surfacing of Runways 1.50 3.O 6. Navaids & Telecommunication seas b.oo 7. Maintenance of Airports 8. Aircraft Acquisition for RNAC (domestic operations) 02. a WU.m ,o 0.5 n0so C. Other Transport Modes 56. In addition to aviation and the road and trail networks there are minor transport services provided by railways and ropeways. Neither of these modes figure prominently in the overall economy of Nepal, though for some very localized areas they may have a significant impact on economic activity. For instance the 51 lam Janakpur Railway operates primarily as a passenger railway serving local traffic between Janakpur in Nepal and Jayanagar in India. The 43 km Kathmandu-Hetauda Ropeway is mainly used for transporting the output of the Hetauda Cement factory, and for some other bulk movements to the Kathmandu area. - 83 - Annex VI Page 1 of 12 WATER SUPPLY AND SANITATION 1. Adequate access to safe and reliable water supplies and proper sanitation facilities has been universally accepted as a basic civic amenity. Ensuring these facilities for every individual is an urgent goal for every nation. Actual coverage, however, varies considerably depending upon physical, financial and human resources and the priority assigned to this objective. Nepal, being one of the least developed countries in the world, faces an enormous task of providing water and sanitation facilities to a population widely scattered over its rugged terrain. By the end of the financial year 1991-92 only 42 percent of the country's population - 39 percent in rural areas and 67 percent in urban areas - is expected to have access to safe and reliable water supplies. The coverage in the field of sanitation facilities is even more limited with just 3 percent of rural and 34 percent of urban population (equivalent to 6 percent of the national population) being covered with existing or projected facilities in 1991-921. 2. That poor water supply and sanitation conditions are taking their toll is apparent from the health profile of the country which reveals a high incidence of illnesses related to water, sanitation and household/personal hygiene. Diarrhoea and dysentery are leading causes of morbidity (40 per 1000 cases) and child mortality (16.5 percent of total deaths) and typhoid, hepatitis and parasitic infections of the gut are very common. 3. Several government agencies in five ministries are involved in the water supply and sanitation sector (WSS) in Nepal. The Ministry of Housing and Physical Planning (MHPP), through its Department of Water Supply and Sanitation (DWSS) and the Nepal Water Supply Corporation (NWSC), plays the lead role. The Ministry of Local Development (MLD) executes a number of integrated rural development projects that include drinking water supply and sanitation components. Investment in the WSS sector during the 1980s increased from NRs 1,127 million in the Sixth Plan (1981-1985) period to NRs 2,302 million in the Seventh Plan (1986-1990) period, but maintained a steady 4 percent level in the national development budget over the period. Donor participation in the sector during the decade (1981-1990) has been of the order of 40 percent of total investment. 4. WSS in Nepal is facing various institutional, operational, financial and policy issues. Although, the government has stated its desire to decentralize effectively the activities in the sector, the organizational framework of lead agencies is still a constraint. Much remains to be done to realize the government's intention to get local communities actively involved and share the responsibility of providing water and sanitation services. Absence of effective legislation has also complicated matters in areas such as the priority For more detailed Information, refer to 'Nepal: Water Supply and Sanitation Sector leaues Paper. IBRD Report No.11479- NEP. 1993. - 84 - Annex VI Page 2 of 12 of water use, right to access and establishment of achievable national standards for the sector. 5. The preferred strategy for the future may be to shift emphasis to (a) educating the general populace as to the importance of safe drinking water and sanitation and their subsequent involvement in sector activities, (b) appropriate human resources development (HRD) within the concerned government agencies to better perform various tasks, (c) rehabilitation and reorganization of existing schemes to enable them to deliver their expected output and become financially independent, and finally (d) planning for future development with rational targets to ensure a balanced expansion of population coverage for both water supply and sanitation facilities. A. Existing Facilities and Coverage 6. Sources of drinking water supply in the rural areas of Nepal are piped water supply schemes, shallow handpump tube wells, private lined and/or unlined wells, natural streams and ponds, and so on. In urban areas piped water supply schemes are the main source of drinking water. As of 1990, there was a total of 2250 piped water supply systems in Nepal along with some 11,725 shallow tube .wells. The number of traditional (mostly private) wells is not known. About 70 percent of the piped systems and 90 percent of the shallow tubewells were provided by various public agencies. All 33 towns in of the country are served with piped water supply schemes. Coverage 7. Table 6.1 gives details of the water supply coverage as of 1991-92 by development zone and region. Table 6.2 shows the historical record of populations served with drinking water supply. There are some discrepancies between the overall figures in Table 6.1, which are based on district level estimates, and the historical figures in Table 6.2 which could not be reconciled. Existing water supply coverage in rural areas is estimated on the basis of the design population of completed water supply projects (piped water supply and shallow handpump tubewells installed by various governmental agencies and non- governmental organizations). In urban areas, the coverage is estimated according to urban water use, the rate for which varies from 45 liters per capita per day (lcd) to 180 lcd depending upon the type of connection. As of 1990, an overall 38 percent national coverage is estimated. This includes rural coverage of 35 percent (about 5.75 million people out of a total of 16.5 million) and 75 percent for urban (about 1.2 million out of a total of 1.6 million urban people). The figures in Table 6.2 indicate a steady deterioration in the urban water sector from nearly universal coverage in the early 1970s to 75 percent coverage at present. 8. There are two points worth mentioning. First, the coverage indicators employed to arrive at the above estimates can be questioned as safe water does not necessarily mean piped water or that from tubewells. Protected traditional wells or even some natural springs can provide a level of quantity and quality -85- Annex VI Page 3 of 12 sufficient to safeguard a community's health. Second, the estimated coverage figures may be somewhat higher than the actual as they include the service population of all completed schemes some of which are in need of rehabilitation and probably not serving all of the service population. 9. Estimation of actual sanitation coverage in Nepal is even more difficult as no systematic records are available. Except for about 25 percent of the population of the Kathmandu Valley towns - Kathmandu, Lalitpur, and Bhaktapur - a proper sewerage system is not in place. Many of the recent buildings in major towns have their own sanitation facilities (septic tanks). Sanitation coverage in the municipalities is still low. In rural areas sanitation coverage is virtually nonexistent. Most estimates of sanitation coverage are based on the number of facilities constructed directly through various projects, mainly for demonstration purposes. However, there are some privately constructed facilities as well. Sanitation coverage estimation is further confounded by the lack of a consistent definition of adequate sanitation. In 1990 the population using household latrines has been estimated to be 3 percent in rural areas and 34 percent in urban areas or 6 percent of the total national population. Table 6.3 gives details of sanitation coverage in Nepal. Table 6.1 ESTDIMAT ATR SUPPLY U OF POPULATION COID) 1991-92 Development eton_ Ecological Par-West Mid-West east Central East Total NMuntain Rural 44.0 38.0 76.0 23.0 30.0 40.9 Total 44.0 38.0 76.0 23.0 30.0 40.9 RiLl 44.0 42.0 43.0 32.0 30.0 37.9 Urban 12.0 98.0 78.0 86.0 85.0 85.0 Total 45.0 43.0 45.0 46.0 31.0 43.0 Teat Rural 68.0 72.0 16.0 50.0 26.0 40.5 Urban 16.0 57.0 50.6 51.0 51.0 47.0 Total 60.0 71.0 19.0 50.0 28.0 41.1 Total Rural 53.0 52.0 34.0 43.0 27.0 39.0 Urban 24.0 66.0 67.0 77.0 54.0 67.0 Total 51.0 53.0 35.0 48.0 29.0 42.0 Sources Aggregated from "Water Supply and Sanitation Coverages", DWSS, SHO, atbmandu, 1992 -86 - Annex VI Page 4 of 12 Table 6.2 DIHIE UATER AVAILABILITT AMD) WOPULATI(R SERVE Uater Availability 1970 1980 1990 aml Coverage Water Availability (OLtlion Itiday) __ DNSS: Rural 1,957 32,705 Urban -7, 930 852 NWSC: Urban - 5,000 42,000 populaMIon served (000'.) Total 647 1,576 6,949 Rural 186 862 5,753 Urban 461 714 1.196 c) Total 5.7 10.8 38.3 Rural 1.9 6.1 34.9 Urban 100.0 80.0 75.0 Source: *wNepal, National Report for UN Congres on Environment and Developmenta, New Era (Consultant), Eatbmandu, 1991 Water Supply and Sanitation Coverages"* Department of Water Supply and Sanitation (DWSS), EMN, Kathmandu, 1992 10. The above figures for water supply and sanitation coverage, in general, fall short of the original and revised targets under the Sixth (1980/81 - 1984/85) and Seventh Plan (1985/86 - 1989/90). A national coverage target was set as 68 percent for water supply by 1990, with 67 percent rural and 94 percent urban coverage. For sanitation the coverage targets were 13 percent and 21 percent for rural and urban population, respectively. In 1985, at the time of mid-decade review under the UN International Drinking Water Supply and Sanitation Decade (IDWSSD, 1981-1990), these targets were revised. The revised national target for water supply was 54 percent (52 percent in rural areas). Estimated coverages for water supply, as shown in Table 6.2 are well below these targets. B. Level of Service 11. The level of water supply services in urban areas of Nepal is generally inadequate and deteriorating as a result of a lack of maintenance and rapid population growth in urban areas. In most of the 33 municipalities where a water supply system is in place, the supply is intermittent with some areas receiving less than 6 hours of supply per day in two shifts. The level of service among the municipalities varies depending on the agency responsible for 87 Annex VI Page 5 of 12 the system and the type of connections. Thus, for the 19 municipalities under the jurisdiction of DWSS the designed consumption rates are 150 lcd for fully plumbed connections, 65 lcd for yard taps and 45 lcd for public standposts. The corresponding rates for the towns under the jurisdiction of NWSC are 180 lcd, 120 lcd and 45 lcd, respectively. No information on actual per capita consumption in Nepal is available. 12. Besides these quantitative shortages there are various problems related to quality of water as well. Groundwater generally contains relatively high proportions of iron and manganese in Terai towns and ammonia in Kathmandu valley. Due to the intermittent nature of supply, infiltration due to poor quality construction may cause contamination. Furthermore, only a few cities have facilities for proper treatment of supplies from surface and ground water sources, elsewhere treatment is limited to occasional chlorination only. Table 6.3 ESTDATED SANITATLGM CMERACX IW NIPAL.1990-91 I CW POPULATfW COERED) Development Region Ecological Region Par-West Mid-West West Central East Total mntaln Rural 0.9 0.8 12.6 6.7 4.7 3.8 Total 0.9 0.8 12.6 6.7 4.7 3.8 Hill Rural 0.6 1.6 3.4 9.1 2.8 4.1 Urban 30.8 26.9 33.9 57.9 8.6 51.4 Total 1.2 2.1 4.7 21.2 2.3 8.8 TeraL Rural 1.9 1.7 0.7 2.7 2.4 1.8 Urban 8.9 28.1 9.8 32.5 17.2 20.1 Total 3.1 3.9 1.4 5.6 3.2 3.8 Nepal Rural 1.2 1.5 2.5 5.4 2.8 3.0 Urban 14.9 27.8 47.9 49.4 59.6 34.0 Total 1.9 2.7 3.5 12.1 3.1 6.0 Source: Aggregated from OWater Supply and Sanitation Coverages*, DWSS, EMN, Kathmandu, 1992 13. The level of sanitation services is even lower and far more limited. Only about 25 percent people in the three cities in Kathmandu Valley have access to a proper sewerage system. The fact that just an access to a latrine is - 88 - Annex VI Page 6 of 12 considered sufficient for sanitation services indicates the expected level of service. In fact 94 percent of the country's population does not have even this facility. There is no operating sewage collection and treatment facility in Nepal and only Kathmandu has a solid waste collection and disposal system. The solid waste system was started with German assistance in 1980 and is still functioning. The disposal uses a landfill site. About 58,779 cubic meters (cbm) of waste was collected and disposed of in 1987-88 and some 62,460 cbm was collected and disposed of during the first 8 months of 1988-89. Elsewhere the municipalities are responsible for waste collection and disposal and perform these functions without proper planning or environmental impact consideration. Discharging untreated water from individual latrines, factories and municipal sewers into natural water bodies is common. Also common is the disposal of dead animals by water bodies contaminating these sources of drinking water. 14. Another related problem is the lack of storm water drainage. At present, except for a few storm water ditches along some urban roads, there is no arrangement for storm water drainage. This creates a number of environmental and other problems, especially in the Terai where the terrain is more or less flat and the amount of rainfall is high. Stagnant storm water is considered to be polluting groundwater sources and is a nuisance in urban areas creating problems for human and vehicular traffic flows. Installation of storm water drainage systems in some of the major cities is proposed in the future development plans (Eighth and Ninth Five Year Plans). Technologies Used 15. For water supply in the rural areas of Nepal either a system of gravity pipes or shallow tube wells or hand dug wells fitted with shallow well handpumps is used. Generally, the gravity piped system is used in mountainous and hill areas and wells are used in the Terai. Some piped system in the Terai and lower hills are pumped systems. Urban water supply is generally based on piped systems, some of them employing lift pumps. 16. For the piped systems the sources of water are mainly upland springs, streams -or rivers, and the intakes are usually located away from human habitation, 8-10 km in case of large urban systems. In rural areas some stream intakes normally designed with sedimentation tanks have posed problems during high flow-periods and floods. A piped system in Nepal generally consists of the source and its protecting/intake structure, transmission pipes, storage reservoir, distribution pipes and public taps or house connections in case of urban systems. A typical shallow tube-well is 10-15 meter deep with a suction type handpump and each well is expected to serve about 150 people. Indigenous pumps (mostly Nepal No. 6 type) are used in these shallow wells. However, about 20-25 percent of the Terai population lives in areas where groundwater levels are deeper requiring community lift pumps which have to be imported. 17. The DWSS piped systems, generally being of large size, are constructed by contractors with the work supervised by DWSS staff. The practice is followed for tube well construction. However, smaller piped systems for populations not exceeding 1500, a majority of which are UNICEF supported, are constructed through - 89 - Annex VI Page 7 of 12 directly hired labor, with varying degrees of voluntary labor and materials contributed by the community to be served. 18. Previously, there used to be a grant-in-aid program under which funds were provided by the government through village Panchayats for WSS related small scale projects. No professional or technical assistance was given by concerned agencies for such projects and as a result it is believed that almost all of such projects are no longer functional. No record for such projects is available and with the change of government in 1989, the program appears .to have been abolished. 19. Technologies employed in the sanitation subsector for household latrine construction, varied from direct pit with a wooden/bamboo platform to twin offset pits with masonry squatting platform, water seal pan and chain pipe. No municipal sever system is in place in rural areas, and a sizeable portion of the urban population as well is not covered. Some recently constructed buildings though have septic tanks. Water Ouality and Environmental Impact 20. For the rural piped systems, an effort is made to prevent or minimize the bacteriological pollution of water by preventing human and cattle access to the water source and as much of its catchment area as possible. Some mini treatment units have been tried for rural systems but no wider application is planned, at least currently, due to doubts about maintenance capability. An evaluation of the quality of water from wells in Terai area was sponsored by UNICEF in 1990 and the chemical and bacteriological quality of water was found to be generally acceptable except for high iron and manganese concentration in most cases. However, exceptionally high levels of fecal coliform contamination were found in most household stored water. 21. No comprehensive evaluation of the environmental impact of sewage disposal practices has been made, but the lack of proper sewage treatment facilities suggests a high probability of a potential environmental hazard. Some studies in urban areas have indicated high but localized concentration of pollutants and hazardous wastes in natural water bodies close to these areas. C. Sector Policies and Strategies 22. In 1986, the HKGN formulated and launched a Basic Needs Program which identified access to safe and adequate water supply and sanitation facilities as priority elements within six basic needs. The Government policy on health also emphasized the availability of safe and adequate water supply and sanitation facilities throughout the country. Although these policies call for a balanced and simultaneous improvement in water supply and sanitation facilities, the approach so far has primarily been oriented towards water supply. 23. Schemes for water supply projects are mostly identified by local communities and are included in the annual plan after being channelled through - 90 - Annex VI Page 8 of 12 the Village Development Boards (VDB) to District Development Boards (DDB) and then finally to DWSS/MHPP. The annual program is prepared taking into consideration the available financial resources, sector priorities and periodic plans and the annual budget ceilings. The DWSS programs are classified as the "Central Level" program if some foreign funding is involved or the project covers more than one district. All other programs are known as "District Level" programs. For a particular year the DDBs decide their annual programs based on the budget ceiling. There is a tendency, to add new projects every year despite a recommendation from the center to give priority to ongoing projects. Table 6.4 gives details of ongoing and completed water supply projects and it maybe seen that a large number of new projects are taken up each year compared to relatively few that are completed. 24. The previous strategy in MHPP/DWSS and NWSC has been to execute projects through contractors with agency supervision and then assume the responsibility for operation and maintenance (O&M), without much involvement of the local community. In February 1990, however, a directive was drafted to implement a decentralization policy mainly concerning DWSS activities. The directive provides guidelines for DWSS to actively promote involvement of communities in the planning, design, construction and O&M stages of all its projects, especially those in rural areas through User Committees. This policy recognizes the fact that the operation and maintenance of all such community schemes is the responsibility of the users including generation of the required financial resources. The directive, which is limited to water supply schemes, does not cover environmental sanitation or health education. The implementation of this policy so far has been only partial. For the urban areas, however, the NWSC is to continue its responsibilities. Table 6.4 as-coIW AND 00MKTED KAE XPPLY PJECTS DE ISS _agoln Completed Tear Central District Total mal Cumslatve Up to 79-80 84 80-81 49 - 49 14 98 81-82 75 - 75 28 126 82-83 106 - 106 24 150 83-84 117 - 117 33 183 84-85 162 - 162 17 200 85-86 6 161 167 27 227 86-87 45 170 215 13 240 87-88 60 208 268 29 269 88-89 68 743 811 18 287 89-90 90 895 985 128 415 Source: MHPP 1991 -91 - Annex VI Page 9 of 12 Sector Budget Allocation 25. The WSS sector has been receiving around 4 percent of the total national development budget over the past 15 years. Donor contribution has been on the order ,of 40 percent of total investment in the sector during 1980s. Table 6.5 summarizes details of the sector allocations and donor contributions during the Sixth and Seventh Plan periods. As can be seen from the Table, donor involvement has been concentrated in urban water supply, with no involvement in the sanitation sector, and only a small presence in rural water supply investment. 26. For operation and maintenance through DWSS, there are two sources of funds. The "regular budget" and a DWSS budget for "renewal and extension". The regular budget provides for salaries and wages of all staff, including the operation and maintenance staff at project sites. The total regular budget amounts for the sixth and seventh plan were NRs 31.3 million and NRs 77.1 million, respectively. The renewal and extension budget provides funds for repair cost of water systems, based on specific proposals and estimates from project engineers. The renewal and extension budgets have been NRs 31.7 million for the sixth plan and NRs 17.2 million for the first four years of the seventh plan. Figures for the operational and maintenance budget of NWSC could not be ascertained. Table 6.5 WSS SECTOR INVESTMENTS - 6TH & 7TH FIVE YEAR PLANS sixth Plan Investmens Seventh Plan Xavestmen Total Donor Shae of Total Door Shae Of Donor Dr (---R Millions-) (Z) (-Rs Millias---) (Z) Mater Supply Rural 661.00 139.7 21 1655.2 540.3 33 Urban 389.0 295.5 76 570.1 340.8 60 Total Water 1050.0 435.2 41 2225.3 881.1 40 Supply Sanitation Rural 2.8 - 0 3.8 - 0 Urban 73.9 - 0 73.1 - 0 Total 76.7 - 0 76.9 - 0 Sanitation Total USS 1126.7 435.2 39 2302.2 881.1 38 Source: *Nepal - Drinking Water Supply and Sanitation Sector Review and Development Plan (1991-2000)0, Ministry of Housing and Physical Planning, Kathmandu, 1991. 27. Operation and maintenance of most water systems in both rural and urban areas has been inadequate and is believed to be a factor in the -92 - Annex VI Page 10 of 12 continuously low rates of improvement in effective coverage. Most of the DWSS water supply schemes were completed with the intention of being eventually handed over to local communities. However, due to lack of financial and administrative capabilities at the local level, DWSS undertook the responsibility to operate and maintain the systems for 2-3 years as a temporary solution. In practice, however, this became an ongoing arrangement as expectations to involve local communities in operation and maintenance of the systems did not materialize. 28. Evaluations of the O&M of water supply systems by DWSS carried out with WHO financing in 1984 and 1988 indicated poor supervision, low level of support for maintenance workers, and inconsistent establishment of users committees, as major factors contributing to poor system condition and performance. A survey in 1990 to determine the O&M status of the 306 completed piped systems revealed that 268 rural and 14 urban systems were in need of repair/rehabilitation to varying degree. 29. Cost recovery from both urban and rural water supplies need to be increased. In the larger urban areas under the responsibility of NWSC, tariff rates for water supply were revised after 9 years in December 1990. It is believed that the prevailing rates still do not meet the actual 06M costs and NWSC needs a further tariff increase of about 75 percent to meet its current O&M costs alone. Future Plans and Targets 30. There are at least two, somewhat different, development plans for the WSS sector in the 1990s. One is the plan proposed jointly by the MHPP, World Health Organization (WHO) and the UNDP/World Bank Water and Sanitation Program, and the other is the WSS sector targets and investment projections as proposed by DWSS for the Eighth and Ninth development plans. The Joint MHPP, WHO and UNDP/World Bank plan has ambitious programs and targets for institutional and human resources development, improvement in sector management and greater community involvement. 31. Cost estimates in both plans are based on the stated government policy that all capital investment for household sanitation and O&M cost for water supply projects are to be borne directly by beneficiaries and hence do not include these items in their (cost) projections. If materialized, these plans would make available-a considerable amount of funds presently being spent for O&M activities. However, as stated earlier a large number of the completed water supply schemes are in need of rehabilitation before they can be handed over to local communities. Both plans give priority to the rehabilitation of these schemes. Some 268 piped systems in rural areas and 14 in urban areas were identified to be in need of rehabilitation. The DWSS has estimated the average cost of rehabilitation of a typical rural piped system to be NRs 0.5 million. A total of 200 such schemes is included in the Eighth Plan. The average cost for rehabilitation of rural piped water supply schemes for the Ninth Plan period (1995-2000), as estimated by DWSS, is NRs 0.4 million. A total of 500 schemes is planned to be rehabilitated during the Ninth Plan period. This brings the total for the rehabilitation of rural piped water supply schemes to 700 during the 1990s at an estimated cost of NRs 300 million (US$8.6 million) in 1990 prices. Along with these rural schemes, a large scale urban water supply rehabilitation program concentrating on Kathmandu Valley towns has also been proposed during the eighth plan with a budget of NRs 2000 million. However, - 93 - Annex VI Page 11 of 12 these projections and proposals were all prepared before the change in government in Nepal in 1989 and some modifications may be expected. 32. The population coverage target of HHGN for water supply by the year 2000 is 100 percent. A Sector Review conducted in 1986 under the UN International Drinking Water Supply and Sanitation Decade Program, however, suggested a somewhat different scenario keeping in view the financial, institutional and human resources constraints. According to this review, a target for coverage of 100 percent of urban and 81 percent of rural populations with water supply schemes by the year 2000 would be more realistic. This plan calls for the rehabilitation of completed schemes as discussed above, and then it proposes to complete the ongoing, some 900, projects which are expected to cover an additional 4.0 million people by 1995, raising the rural population coverage to 50 percent. Based on these coverage estimates and assuming that future investment will remain at their present level with an annual 10-15 percent increase to cover inflation, etc., the coverage targets were revised, as follows: 1995 - 75% urban, 50% rural, and 53% overall; 2000 - 90% urban, 75% rural, and 77% overall. 33. Lack of reliable data about existing coverage and the uncertainty associated with the future approach for sanitation programs to rely almost entirely on community and household financing make it difficult to project future sanitation coverage especially for the rural areas. Improved urban coverage to some extent is expected through sewerage schemes. In view of this and the experiences in other developing countries, a modest 25 percent rural and 75 percent urban sanitation coverage target is projected for the year 2000. The corresponding HMGN targets based on its previously declared policy of ensuring "provision of basic needs for all" is 100 percent by the year 2000. D. Financial Requirements 34. Details of incremental population that will need to be served over the decade (1990-2000) in order to achieve the above water supply coverage targets are given in Table VI.6. The estimated average per capita construction costs for different types of water supply schemes in 1990 prices is quoted by DWSS to be as follows: Large urban system - NRs 4,750 per capita Rural piped system - NRs 750 per capita Rural shallow tubewell - NRs 50 per capita Based on the targeted coverage figures and the above per capita costs, the construction cost for water supply systems (both urban and rural) is estimated at NRs 3,860 million for the period 1991-95 and NRs 6,270 million for 1996-2000 in 1990 prices. Total cost for various support programs including institutional strengthening, sanitation/hygiene education and human resources development for the corresponding periods can be estimated to be NRs 712 million and NRs 686 million, respectively for the first and second periods. Thus, the overall investment need for the sector has been estimated to be NRs 4,572 million for the period 1991-95 and NRs 6,956 million for 1996-2000, or NRs 11,528 million for the entire decade (1991-2000). This compares to NRs 6,273 million in the Eighth Plan (FY 93-97). - 94 - Annex VI Page 12 of 12 For achieving 25 percent rural and 75 percent urban sanitation coverage, an estimated 1.2 million latrines will have to be constructed by the year 2000. Based on an average cost of NRs 1000 per latrine (somewhat lower than prevailing rates, to be achieved through technology refinement and use of local material) a total private investment of NRs 1,200 million will be required during the decade for this purpose. Table 6.6 ROJXCTE C ANHD AMDDO1AL PULTIOW E oPjected Coverage (000* ) Additinal Populatin (000 s) 1990 1995 2000 1990-1995 1995-2000 Urban Mater Supply Population Served 1,196 1,716 2,629 520 913 Coverage (Z Total 661 75% 901 Population) Rural Water Supply Population Served 5,753 9,450 15,250 3,697 5,800 Piped Systems 1,910 2,435 Tube Wells 1,787 3,365 Coverage (Z Total 341 501 752 Population) 35. Donors contribution to the sector during the 1980s has been of the order of 40 percent of total investment in the sector2. Assuming a 50 percent share, the donor contribution can be estimated to be about US$20-25 million per year during the period 1991-2000. This amount appears to be well within the limits of donor capacity and interest. Therefore, it seems that institutional capacity issues rather than the financial resources constraints will drive the pace of WSS sector development in Nepal during the 1990 decade. However, besides financial assistance, donors' cooperation will also be required in technical assistance and institutional support including human resource development. 2 Nepal Water Supply and Sanitation Sector Issues Paper. IBRD Report No. 11479-NEP. 1993. - 95 - ANNEX VII Page 1 of 14 TELECOMMUNICATIONS 1. In 1991, Nepal had an average of only 0.32 telephones per 100 inhabitants, one of the lowest in the world. About 64 percent of the existing telephone lines in the country were concentrated in Kathmandu Valley. Out of 75 districts, only 56 had access to telephone services. It was expected that the percentage of population having access to telephone services would increase from 6 percent in 1984 to 12 percent by 1990. Present plans are to make telephone services available to all 75 district headquarters and some 400 plus rural community centers by 1995. Telegraph and telex services are extremely limited. The country has a satellite link facilitating international calls and an international telex exchange is functioning at Kathmandu. Nepal Telecommunications Corporation (NTC) under the Ministry of Communications is responsible for all public telecommunication services in Nepal. Established in 1975 as a government-owned statutory corporation, NTC functions in accordance with policy guidelines of the government and its budget and tariffs are subject to approval of the government. A. Existing Facilities 2. Telecommunication facilities in Nepal, although still very limited, have been expanding continuously since the 1960s. The number of telephone lines and exchanges has increased substantially, and outdated equipment is gradually being replaced, resulting in a considerable improvement in the quality of service. Table 7.1 gives details of the growth of telephone facilities over the period 1960-1991. Table 7.1 rar.comuWICATIOU PAC3.ITIES 1N WPAT. no. of Telephone LiaMS Wo. of Telephoms Type of Faclity year (talled Embans*s (Pacemage*) Capacity) Manual Antomatic 1969 4,900 3 9.0 91.0 1972 5,450 7 9.0 91.0 1977 10,770 13 16.0 84.0 1984 20,580 24 22.0 78.0 1989 55,060 0 0.0 0.0 1991 (projected) 77,450 44 1.3 98.7 Source: World Bank 1969, 1973, 1978, 1985, 1989, 19911 NTC 1990, 1991a -96 - ANNEX VII Page 2 of 14 3. Targets for 1991 aimed at a 14-fold increase in installed telephone lines over the period 1969-1990 with the rate of growth during the period 1984-91 being the highest at about 26.7 percent per year. The overall growth in the number of telephone lines for the period 1969-90 was about 14 percent per year. The 1991 data reveals a rise in the coverage to 0.33 telephones per 100 persons. At the completion of the IDA supported IVth Telecommunication Project in 1991, of the total of 42 telephone exchanges in Nepal, only 4 smaller ones, with a total installed capacity of 600 lines, would remain manual. The total installed capacity of modern digital exchanges will be 67,100 lines or about 87 percent of the total available capacity. The total capacity of the crossbar type exchanges is expected to be more than 9,000 lines. 4. In addition to the places with functional exchanges, long distance calling facilities are also available at about 50 remote locations through Public Call Offices (PCOs). The operator assisted long distance service is gradually being replaced with direct Subscriber Trunk Dialling (STD) system. Data about the STD coverage is not available, however, it is believed that these facilities exist in all major towns and some newly installed remote PCOs. In 1991, of the total 82 centers where telegrams were accepted (at least one in every district) 52 were equipped with teleprinter. The remaining 30 centers transmitted messages by morse or as phonograms. Until 1985 telex facilities were available only in Kathmandu and used exclusively for international service. By 1991 telex services were available to 599 subscribers in Kathmandu and other towns, through a 768 line exchange. However, telex is being fast replaced with fax service. 5. International telephone service has considerably improved with the installation of the satellite earth station at Kathmandu in 1982 largely replacing the old land lines or microwave system based service of the past. The satellite link has helped improve both the quality and availability of the international telephone and telex services. Sizeable increases in the number of international calls and revenues was noted in the first few months of the establishment of the satellite link, indicating potentials for future growth if adequate facilities are provided. In fact, it is believed that the tourism sector of the Nepalese economy has especially benefitted from the improved telecommunication facilities. 6. Two important projects undertaken with foreign assistance during the 1986-90 period are the Japan International Cooperation Agency (JICA) supported rural telecommunication project and IDA supported IVth Telecommunication Project. The rural telecommunication project was planned to be implemented in four phases over a six year period (1986-92) and was aimed at providing a digital radio transmission network mainly in the Trans-Himalayan and mid-hill regions of Nepal to serve 63 rural communities. Each community was planned to be provided with 5 telephone connections, four of which to individual administrative subscribers and one to serve as a long distance PCO. Other major features of the project were a) to provide 27 facsimile sets in 10 important towns (including 4 in Kathmandu); b) prefabricated transportable structures at all sites for housing equipment; and c) solar power at all sites which have no public power supply. - 97 - ANNEX VII Page 3 of 14 7. The salient features of the IVth Telecommunication Project (IVth TP) were as follows: - Providing automatic switching equipment for 35,000 local lines including associated cable distribution network and subscriber plants to connect about 31,000 new subscriber lines and replace 1,370 lines from manual exchanges. - Providing and installing about 2,500 additional terminations in national/international trunk exchange. - Installation of about 75 transmission systems in the long distance network. - Providing and installing equipment for expansion of existing (satellite link) earth station and telex exchange. - Installation of about 550 local and long distance public call offices (PCOs) in urban areas and about 200 long distance connections and PCOs in about 90 rural communities. An important feature of both the JICA and IDA supported projects is their emphasis on digital technology where feasible and full automation of domestic telephone and telex service. These provisions are expected to result in the least life cycle cost including initial investment and maintenance and operating costs. Though still ongoing, the project has made substantial progress and is expected to be completed by 1992. B. Access to Service 8. During the past three decades though the telecommunication network in Nepal has expanded considerably, accessibility to service is still severely limited. The density of telephone connections per 100 inhabitants in Nepal, at 0.29 in 1989, was one of the lowest in the world. Even this very low density does not actually reflect the real situation which varies substantially from one region to the other. Thus, in 1989, there were 3.29 telephones per 100 inhabitants in Kathmandu Valley and only 0.11 in the rest of the country. The estimated number of telephones in 1991 as per NTC's projection for the year is shown in Table 7.2. NTC's projection was made separately for Kathmandu and for rest of the country. These estimates were expected to raise the coverage per 100 inhabitants to 0.33 for the entire country. The coverage for Kathmandu Valley would increase to 3.53 per 100 inhabitants and for the rest of the country to 0.13. Table 7.3 provides details of actual coverage by development region and ecological zone. 9. As is evident from Table 7.2 about 61 percent of all telephones in the country in 1991 was expected to be concentrated in the Kathmandu Valley which accounts for just 5.5 percent of the population. This is lower than the 1989 - 98 - ANNEX VII Page 4 of 14 figure of 64 percent but higher than the 1984 figure of 57 percent. Even within the regions, the existing telephone facilities are mainly concentrated in relatively larger towns and district headquarters with virtually no access for the vast rural population. For the rural population telephone service is available only in the form of some 50 PCOs that provide long distance calling facilities., Although a total of 162 PCOs were planned to be installed in rural areas under the IDA supported IVth Telecommunication Project and the JICA supported Rural Telecommunication Project, more recent documents indicate that these targets could not be achieved and the number of PCOs in remote areas remains around 50. The earlier target of making telephone service available at all 75 district headquarters also could not be achieved as NTC's Vth Telecommunication Plan includes target to provide telephone service to 26 district headquarters which are still (in 1990) without such facilities. Table 7.2 Projected Number of Telephones Per 100 Inhabitants by the End of TVth Proiect (Planned by 1991) Ream Population Total Installed umber of Telephaus per (00018) CapeoLty Lnas 100 Pop. (in") Eastern 4,759 12,390 7,010 0.15 Central 5,012 8,700 7,355 0.15 Eathmandu Valley 1,064 43,860 37,576 3.53 Western 3,886 5,000 4,762 0.12 xLd-Wstern 2,473 4,000 2,323 0.09 Par-Western 1,746 3,450 2,622 0.15 Total 18,940 77,450 61,648 0.33 Source: NTC 1991& 10. Access to telephone service by non-subscribing population in urban areas is also through PCOs. In 1984 there was a total of 46 urban PCOs, 25 in Kathmandu and 21 in other towns. The IVth Telecommunication Project had a target of installing an additional 200 local PCOs in various towns and another 50 long distance PCOs, by 1990. The JICA Project has reportedly completed installation of all 33 planned long distance PCOs in rural areas. In fact, the 6 telephone exchanges now functioning in rural areas were installed to replace the PCOs (out of the above 33) where demand was excessively high. International calls can also be made, through assistance from Kathmandu operators, from these PC0s. 11. Estimates about the percentage of population having access to telephone facilities in 1984 suggested that overall only about 6 percent of the country's population had access to telephone service. It is understood that if desegregated on the basis of rural and urban areas, the percentage will be much lower for rural areas, somewhat higher for urban areas other than Kathmandu Valley, and considerably higher for Kathmandu Valley. A more recent estimate of population with access to telephone service after the completion of the IVth -99 - ANNEX VII Page 5 of 14 Telecommunication Project, as planned, is about 12 percent. There is evidence that while the planned targets for increase in the number of telephone exchanges and lines have largely been achieved, the number of PCOs could not be increased as planned. This may affect the population coverage targets, especially those pertaining to rural areas. Table 7.3 Actual Telephone Coyerage in Nepal by Develoment and Ecological Reaions - 1991 DEVLEPM REMI (Telephosms/100 Pop.) Par-West Kid-Weast Mst Central East Total Mountain 0.00 0.00 0.00 0.00 0.00 0.00 Bill 0.02 0.05 0.10 1.50 0.04 0.49 TeraL 0.47 0.21 0.17 0.19 0.23 0.22 Total 0.17 0.10 0.13 0.74 0.14 0.32 Source: World Bank 1992 C. Service Usage 12. Previous country-wide trends and recent information about the usage of PCOs in remote areas indicate that the predominant use of telephone service is for business purposes. In 1984, 69 percent of all telephone lines in Nepal were connected to business and government subscribers. Table 7.4 gives a profile of telephone subscribers in Nepal in 1984. The actual number of telephones Table 7.4 PROFILE OF TELEPONE SUBSCRIBERS IN NEPAL - 1984 Activity of Subscriber Psesn of Total LInes Comerce (retail and wholesale trade, banking, 34 insurance and farming) Residences 31 Government 27 Industry 5 Others (mainly tourists and transport) 3 Total 100 Source: W orld Bank 1985 -100 - ANNEX VII Page 6 of 14 recorded under personal names was higher, as in Nepal many professional, business and commercial units are registered under personal names rather than company names. The figure in Table 7.4 for the residential category was obtained by analyzing sample of telephones listed under personal names. After business and residential subscribers, the government is the third largest subscriber of the telephone service accounting for 27 percent of all telephone connections in Nepal. 13. No substantial changes are expected in the pattern of usage of local service. However, with the improvement in the inland long distance calling and international calling facilities, the share of tourist and tourism related activities is expected to increase. No updated data about more recent trends are available. A recent report on the JICA project indicates that besides the use for official (government) purposes, private use of PC0s was primarily for business, which accounted for 60 percent of all private calls. Personal calls, and those made by tourists, mountaineers and trekkers accounted for 20 percent each. 14. Up-to-date data about the number of telephone calls placed by subscribers in various categories are not available. A 1980 survey, however, revealed the calling pattern and frequency of use by subscribers in various categories. The survey indicated that on average 4,000 calls originated from each telephone line annually, but there were considerable variations among subscribers in different categories. The most intensive telephone use was made by subscribers in the tourism category with about 12,500 calls per year per line, followed by banking and government with 8,300 and 7,900 calls per year per line, respectively. Residential phones were used, on average, for 2,400 calls per year per line. Some of these residential phones were also used for work related purposes. A 1984 estimate suggests that about 80 percent of all telephone calls and revenue generated in that year were either related to business or government activities. D. Demand for Telephone Service 15. Both the estimate for current demand and forecast for future demand are constrained by various factors that cause the expressed demand to be somewhat depressed. These factors, among others, include lack of service in many areas of the country; discouragement of potential subscribers because of long delays; relatively poor quality of local and long distance service; no registration of demand from the government sector; and limiting the demand estimates only to areas where service is available. NTC's estimates for demand are limited to the number of applications for new telephone connections. Such applications are received only in areas where service is available and, moreover, as connections to government offices and agencies are given on priority basis, they are not included in the waiting lists for new connections. It appears therefore that demand estimates in Nepal are supply constrained and do not express real demand. The net result is an understatement of the demand growth rate. - 101 - ANNEX VII Page 7 of 14 16. Table 7.5 gives details of the expressed demand and the growth of telephone network in Nepal over the period 1980-1995. Expressed demand as used in the Table implies working lines plus registered applications for new connections. The Table gives details separately for Kathmandu and rest of the country. In 1989, there were a total of 26,785 working lines in Kathmandu against an unmet demand of 33,196 registered applications for new connections. For the rest of the country the unmet demand was 18,750 against a total of 19,133 working lines. Thus, in 1989, only 44.7 percent of expressed demand in Kathmandu and 44.9 percent of that in the rest of the country could be satisfied. NTC's forecast for 1991 is a total expressed demand of 52,000 against 34,800 working lines for Kathmandu and an expressed demand of 26,500 against 21,600 working lines for the rest of the country, indicating a total unmet demand of 22,100 or 28.2 percent of the total expressed demand. NTC's original Vth Project documents Table 7.5 TELEPMB DEUAD AD SUPPLY 1980-1989 ACTUAL 1980 1985 1989 Local Exchan-e Kathmandu 7,000 13,000 35,000 Capacity Rest 5,870 8,580 20,060 Total 12,870 21,580 55,060 Working Lines Kathmandu 6,347 11,834 26,785 Rest 4,524 7,223 15,306 Total 10,871 19,057 42,091 Telephones Isthmandu 8,250 14,792 33,481 Rest 5,429 8,667 19,133 Total 13,679 23,459 52,614 Registered Waiters Kathmandu 6,086 29,196 33,196 Rest 4,402 14,684 18,750 Total 10,488 43,880 51,946 Total Expressed Kathmandu 12,433 41,030 59,981 Demand Rest 8,926 21,907 34,056 Total 21p359 62,937 94,037 Expressed Demand Esthandu 51.0 28.8 44.7 Satisfied (M) Rest 50.7 33.0 44.9 Total 50.9 30.3 44.8 Working Lines Per KAIhmandu 0.85 1.39 2.63 100 Population Rest 0.03 0.05 0.09 Total 0.07 0.11 0.23 Telephones Per Kathmandu 1.10 1.73 3.29 100 Population Rest 0.04 0.05 0.11 Total 0.09 0.14 0.29 Source: NTC (1990) -102 - ANNEX VII Page 8 of 14 indicated that the gap between number of working lines and expressed demand would be narrowed to 13.2 percent by 1995, providing 111,200 working lines against a total expressed demand of 128,063 in that year. More recent details reveal that although the demand as well as the number of lines planned to be added by 1995 have been revised substantially upward, delays in the implementation of the project (now expected to be completed by 1997) would result in some unmet demand even after the completion of the project. In absence of any further expansion plans beyond 1995, the gap between demand and supply will again start increasing. E. NTC Finances 17. NTC's financial records for the period 1984-92 reveal a satisfactory financial performance in that it was able to show an increasing trend in net operating income throughout the period. Table 7.6 gives details of NTC's income statement from 1984 through 1992. As a result of the introduction of automatic long distance domestic service and establishment of the satellite link for international service, NTC experienced a substantial increase in revenue with annual revenue growth rates being about 26 percent and 30 percent, respectively for these services, over the period 1987-91. By 1987, revenues from telex services had surpassed the estimated amount for 1992, indicating a great demand for the service. Revenue from local telephone service, telegraph, and telex services have leveled off since 1990. - 103- ANNEX VII Page 9 of 14 Table 7.6 WTC RVEUES AD EXEESES (MRS MILLIO): 1984-92 1984 1985 1986 1987 1988 1989 1990 1991 1992 (Et.) I. REVEU Installation 19 8 Basic Rental 68 114 Extra Local Calls 30 29 38 65 87 124 141 80 80 Trunk Calls 15 17 20 42 68 119 159 181 202 International Calla 21 33 48 66 90 182 294 363 433 Total Telephone 66 79 106 173 244 426 593 712 837 Total Telex 20 30 45 59 62 65 65 69 62 Total Telegraph 7 8 8 8 8 8 7 6 6 Leased Private Circuit 3 6 7 5 8 8 8 8 9 Other 1 5 3 5 9 11 26 12 12 T0TAL REVENUE CC) 97 128 169 250 331 516 699 807 927 II.EPE Salaries 22 32 35 43 49 72 129 167 209 0 &M Expenses 9 10 12 23 24 26 35 36 44 Depreciation 19 22 33 56 111 122 186 200 240 Others 9 11 19 19 27 35 46 48 55 TOTAL ER. MSES 58 76 99 141 211 255 396 451 548 (D) Net Oper. Rev. (A-D) 39 52 70 109 120 262 304 356 378 Interest Payments (E) 23 30 31 33 57 87 109 152 193 Net Non-Oper. Expens. 4 2 8 14 22 3 32 52 22 Income Befr. Tax 12 21 31 63 42 172 163 152 164 (C-D-E-Non Op. Ez) Income Tax (F) 7 12 23 26 11 76 66 68 74 BTPT 5 9 8 37 30 96 97 84 90 (C-D-E-F-En Op. Em) Source: World Bank 1991 18. Wages, allowances, and pensions were the largest expense items until 1987, accounting for 38 percent of total operating expenses that year. However, since 1987, depreciation has become the largest expense item accounting for about 44 percent of total operating expenses in 1991. Maintenance expenses were expected to increase in the future as a result of ongoing and planned future development and expansion of NTC's network; however, as the emphasis has been on less maintenance intensive technology, increases in maintenance expenses would not be directly proportional to the network expansion. 19. NTC tariffs include a 20 percent increase, which became effective in July 1986. Another tariff increase was anticipated, and an IDA/NTC tariff study under the IVth Telecommunication Project, in 1990, recommended further increases in application/registration charges, installation charges and STD long distance * 104 - ANNEX VII Page 10 of 14 and international calling charges. Proposals have also been made to introduce peak/off peak price differentiation for both local and long distance calls, and it is expected that these tariff revisions would become effective in 1992. F. Telecommunications Development Expenditures 20. HMGN has been allocating modest amounts for communications development works in its five year development plans. The allocations accounted for 0.85 percent of total development expenditure, or about NRs 75.3 million during the Fifth Plan period (1975-80). The allocation for communication sector during the Sixth Plan period (1980-1985) increased substantially to NRs 339.1 million accounting for about 1.5 percent of the total development expenditure under the plan. The allocations under the Seventh Plan were NRs 290 million or 1 percent of total expenditure under that plan. Table 7.7 gives details of the total investment in the telecommunications sector in Nepal for the period 1982-90. The total investment given in the Table includes allocations from HMGN as well as from other sources. Table 7.7 plVESMOT ZW TELECOMMICATIONS iAR NTCM GDP GOP ENSTfr (MRa KUita) (M) (MRs Killions) 1982/3 70.2 33,761 0.2 1983/4 106.8 39,390 0.3 198415 126.1 44,417 0.3 1985/6 342.5 50,428 0.7 1986/7 716.1 59,248 1.2 1987/8 744.4 68,858 1.1 198819 134.4 78,259 0.2 198910 217.2 88,711 0.2 1990/1 280.0 100,628 0.3 Note: NTC's investment includes sources other than MV Development Expenditure. Source: World Bank 1992 21. Traditionally, there has been active donor involvement in the telecommunication sector in Nepal. Donors include, among others, IDA/World Bank, USA, UK, India, and Japan. Three major telecommunication projects were completed with IDA's assistance during the last two decades and a fourth one is under completion. NTC's ongoing IVth Telecommunication Project is also partly financed by IDA. NTC's contribution, after covering operating costs, debt servicing and income tax for its development program of 1986-90, was 23 percent. The remaining 77 percent of the requirements was planned to be met by IDA credit (13 percent), government equity contribution for the 1986-90 foreign exchange component of the JICA project (22 percent), and borrowing from other sources (42 percent). 22. It is estimated that during the period 1986-1992, a net fiscal surplus - 105 - ANNEX VII Page 11 of 14 of about US $45 million would accrue to HMGN from NTC's operations. NTC's generation of foreign exchange through its international operations was estimated to produce about US $0.5 million in 1985 and was expected to increase to about US $1.2 million by 1990. Thus, it is believed that NTC will be able to cover a substantial part of its future debt service on foreign loans for the sector from its own resources. Table 7.8 Telehon Demand sad SupplV (1980-1996) PRGJBCMB 1990 61992 1995 Local Bachans Xathmandu 42,500 46,500 78,500 Capacity Rest 34,040 33,700 47,400 Total 76,540 80,200 125,900 Working Lines Kathmndu 31,782 40,700 67,800 Rest 20,149 24,200 43,400 Total 51,931 64,900 111,200 Telephones Kathmandu 39,728 50,875 84,750 Rest 24,178 29,040 52,080 Total 63,906 79,915 136,830 Registered Kathmandu 46,304 16,900 10,965 Waiters Rest 15,882 6,800 5,898 Total 62,186 23,700 16,863 Total Expressed Kathmandu 78,086 57,600 78,765 Demand Rest 36,031 31,000 49,298 Total 114,117 88,600 128,063 Expressed Demand Kathmandu 40.7 70.7 86.1 Satisfied (1) Rest 55.9 78.1 88.0 Total 45.5 73.3 86.8 Working Lines Per Kathmandu 2.99 3.50 5.11 100 Population Rest 0.11 0.13 0.21 Total 0.27 0.33 0.51 Telephones Per Kathmandu 3.73 4.38 6.39 100 Population Rest 0.14 0.15 0.26 Total 0.34 0.40 0.63 a/ Estimated Source: NTC (1990) . 106. ANNEX VII Page 12 of 14 G. Future Plans Development Plan 23. Although details of allocations for the telecommunication sector under the Eighth Five Year Plan (1992-97) are not known, NTC has proposed a Vth Telecommunication Project, 1992-97. Salient features of the proposed project are given below. (i) Local Service - Increasing the installed capacity of telephone lines by 60,850 lines to a total of 138,550. Of the new lines, 42,640 will be for Kathmandu Valley and 17,940 for the rest of the country. - Establishing 28 new independent exchanges with a total capacity of 27,250 lines. - Extension, relocation and replacement of existing provisions with new improved ones involving some 23 exchanges and a total of more than 15,000 existing lines and 31,550 new lines. Thus, by the end of Vth project (i.e., 1997) over 138,550 telephone lines (86,500 lines in the Kathmandu Valley and remaining in the rest of the country) from 58 exchanges would be available. (ii) Long Distance Service - Installation of a new national and international trunk switch with 114 two-way 2 Mega-bytes/sec (Mb/s) trunk terminations at Patan in Kathmandu. - Installation of new main transmission links and a number of spur links to connect new exchange sites to the existing network. - Completion of digitalization of the main transmission links with the introduction of a new 68 Mb/s radio system linking Kathmandu to Pokhara, Bhairahawa, and Nepalgunj. The same system will have spurs to Bharatpur and Bhairahawa. This will make alternate link routes available for Pokhara and Bhairahawa. - Expansion of radio links established under the IIIrd and IVth TP to their full capacities. Biratnagar-Damak-Bhadrapur 34 Mb/s system will be converted to 68 Mb/s system and the Bharatpur-Tandi link to 8 Mb/s rate. - Introduction of about 90 route-kilometers of Optical Fiber Systems of capacities up to 140 Mb/s in Kathmandu valley to connect major exchanges, the earth station (satellite link) and Nepal television. (iii) International Service - Extension in international routes by adding circuits in the earth station. - Establishment of a new digital radio link between India and Nepal (Biratnagar-Farbeshgunj). 107 ANNEX VII Page 13 of 14 (iv) Rural Telecommunication Network - Installation of 25 multi-access rural radio telephone systems (MARTS), 6 VISTA type small earth stations and 12 point-to-point microwave systems to provide telephone service to the remaining 26 district headquarters and more than 400 rural community centers. (v) Data Communication Systems - Installation of a packet-switched network that can interact with the existing telex exchange with provision for service for district headquarters. (vi) Land Acquisition and Buildings - Land acquisition and building construction for new exchanges at Thimi and Butwal. - Construction of NTC Headquarters in Kathmandu and Patan Building Complex. Besides the above mentioned physical development works, the Vth Telecommunication Project also proposes activities related to institution building. These include further computerization of NTC operation, completion of the computerization of billing process, expansion of the telecommunication training center, and training and expatriate consultancy services. 24. Financial requirements to implement the above mentioned development plan as estimated by NTC are shown in Table 7.9. The Table shows component-wise fund requirements based on 1991 prices and an exchange rate of 1 US$ --NRs42. Total project cost, exclusive of the ongoing JICA Rural Telecommunication Project, is about US $127 million. Approximately US $19 million is expected to be covered out of the financial arrangements for the IVth project. The provisions of the new Vth project are therefore estimated to cost about US $108 million. NTC plans to make US $21 million (16.5 percent) available from its own internal resources, and expects to get a World Bank/IDA credit of US $21 million (26 percent) to finance the project. Other donors including the Danish International Development Agency (DANIDA), the Finland International Agency (FINNIDA) and JICA are expected to finance the remaining foreign components of- the cost. 25. NTC's operational and maintenance expenses for 1991-92 are estimated at about NRs 44 million (US$ 1.1 million). The projected 06M expenses for the year 2000 are of the order of NRs 139 million (US$ 3.3 million). The projected expenses are moderate owing to the fact that most of NTC's facilities are new and the technology chosen for them ensures relatively low maintenance requirements. It is believed that NTC is capable of providing necessary financial needs for O&M from its own resources. -108 - ANNEX VII Page 14 of 14 Table 7.9 1C0 S PUTURE FINABCIAL EZUErS (s MILLI ) IV 0a1cOI PROJECT V PROJECT TOTAL IV & V PROJECTS ITEM DESCRIPTION LOCAL FOREIGN TOTAL LOCAL FOREIGN TOTAL LOCAL FOREIGN TOTAL SWITCHING 0.000 0.000 0.000 0.996 23.107 24.103 0.996 23.107 24.103 CABLE NETWORK 0.500 8.407 8.907 5.081 26.256 31.337 5.581 34.663 40.244 SUBSCRIBER FACILITIES 0.000 0.000 0.000 0.022 2.200 2.222 0.022 2.200 2.222 TRANSMISSION & DATA EQUIPMENT 0.759 5.873 6.632 1.028 15.978 17.006 1.787 21.851 23.637 OFFICE AIDS & TRAINING EQUIPMENT 0.000 0.000 0.000 0.032 1.150 1.182 0.032 1.150 1.182 INSTITUTION BUILDING 0.000 0.000 0.000 0.020 3.100 3.120 0.020 3.100 3.120 (CONSULTANCY, TRAINING & COMPUTER FOR CORPORATE DATABASE TRAINING CENTER EMPANSION 0.000 0.000 0.000 0.100 0.000 0.100 0.100 0.000 0.100 VEHICLES 0.000 0.000 0.000 0.010 1.000 1.010 0.000 1.000 1.010 MISC. VEHICLESIPOWER/ AC/PC ETC. 0.041 3.140 3.181 0.000 0.000 0.000 0.041 3.140 3.181 LAND AND BUILDINGS 0.000 0.000 0.000 8.100 0.000 8.100 8.100 0.000 8.100 TOTAL 1.300 17.420 18.720 15.473 78.497 93.970 16.774 95.917 112.690 DUTIES 0.348 0.000 0.000 1.570 0.000 1.570 1.918 0.000 1.570 TOTAL BASE COST 1.649 17.420 18.720 17.043 78.497 95.540 18.692 95.917 114.260 PHYSICAL CONTINGENCIES 6 52 0.000 0.000 0.000 0.852 3.925 4.777 0.852 3.925 4.777 PRICE CONTINGENCIES @ 82 0.000 0.000 0.000 1.363 6.280 7.643 1.363 6.280 7.643 TOTAL PROJECT COST 1.649 17.420 18.720 19.259 88.701 107.960 20.908 106.121 126.680 Source: World Bank 1991 - 109 - ANNEX VIII Page 1 of 17 XNFRASTRUCTURE DEMAND AND RESOURCE MOBILIZATION 1. In planning the level and composition of infrastructure investment in Nepal, it is necessary to ensure that investment for various types of infrastructure is consistent both with the likely evolution of infrastructure demand, and with the capacity for mobilizing adequate levels of domestic resources to cover the costs of infrastructure provision. Aggregate demand for various types of infrastructure can generally be related to the overall levels of economic activity, and by utilizing appropriate sectoral Incremental Capital Output Ratios (ICORs) can provide rough approximations of desirable investment levels. Financing these investments, and paying for infrastructure services requires an assessment of various avenues for resource mobilization such as increased tax effort, user charges for public infrastructure, and/or private provision of specific types of services. This annex examines the above issues and also seeks to provide some guidance on the desirable composition of infrastructure investment both sectorally as well as spatially, i.e. among various parts of the country. A. Infrastructure Demand in the Economy 2. In order to project the demand for infrastructure into the future some relationships need to be established between the demand for infrastructure and the level of overall economic activity, since projections for the latter are fairly common place, are relatively robust, and there is a large degree of consensus around specific projections. The approach to assessing future infrastructure demand adopted in this report, is therefore based on projections for per capita GDP, and relationships between aggregate output in the infrastructure sectors and per capita GDP established through regression analysis of cross-sectional data from a number of countries at various levels of per capita income. Clearly, such an approach can only produce rough approximations of infrastructure needs since it does not account for numerous factors such as topographical features, climate, and differences in endowments such as the availability of clean water. However, our purpose in carrying out the analysis is a fairly modest one, which is to seek indicative assessments rather than very specific quantifications whose validity would in any case be questionable. Since the purpose of this analysis is to establish investment envelopes which would demarcate the boundaries beyond which there is a risk of considerable over investment, we have tended to err on the side of optimism so that the investment envelopes may be considered upper-bounds. For instance in projecting cer capita GDP growth, a 5% average annual growth rate is used whereas most observers of the Nepalese economy would select a growth rate for total GDP of about 5%, which when adjusted for population growth would result in a growth rate for per capita GDP of around 3% p.a. Aqqregate Demand for Transport and Communications 3. The relationship between per capita transport and communications expenditures and per capita GDP is shown in Figure 8.1. The plot using natural logs demonstrates a clear positive relationship between income levels and the demand for transport and communications among a cross-section of 36 developing - 110 - ANNX VIX Page 2 of 17 Figure 8.1 PCGDP and PCTransport & Communication Cross-Section of LDCs zc Transoor&Cornr-un;cctions-noturai logs e0 U 00 4.5 5 5.5 5 A.5 7 7.5 8 8.5 9 pcGOP - naturol logs daftefrom OESD. rngallagmer - 111. - ANNEX VIII Page 3 of 17 countries for which data were available. The following regression was estimated for this cross-section of 36 LDCs using 1985 data: in. (TC.) - - 3.03 + 1.02 in (GDP,) R - 0.67 (8.36) d.f. - 34 From the above regression, the aggregate demand for transport and communications can be projected to grow at the rate of per capita income times the income elasticity of demand for transport and communications (1.02 from the above regression) times the increase in population. 4. Table 8.1 presents projections of transport and communications aggregate demand to the year 2010. The three demand series correspond to the three income elasticities of 1.12 (one standard error above the estimate), 1.02 and 0.90 (one standard error below the estimate). Present estimates for the transport and communications sector indicate that aggregate demand is on the order of US$ 150 million which corresponds to the estimate computed using e - 1.02. Hence, the demand series produced using e - 1.12 is a reasonable upper bound especially in view of the high growth in per capita income (5% p.a.) assumed in producing the demand forecast. The high forecast for year 2010 of US$ 1.1 billion in aggregate transport and communications demand represents about 10% of GDP which would be up from the present level of around 5% of GDP. Table 8.1 PROJECTED GROWTH IN TRANSPORT & COMMUNICATIONS DEMAND Aggregate Transport & Camumication Demand Year GDPPC Population 1990 USS in Mill ns) (1990 USS) (MiLLions) e = 1.12 e = 1.02 e a 0.90 1985 16 16.5 208 128 72 1990 158 18.0 247 152 84 1995 202 20.0 361 216 116 2000 259 22.2 529 309 161 2005 332 24.6 776 442 224 2010 427 27.3 1142 634 313 Acrecate Demand for Water Supply & Sewerage 5. Unfortunately, the approach used above cannot be applied in the water supply and sewerage sectors, since comparable data are lacking. For the purposes of this analysis, therefore, it has been assumed that aggregate demand for water supply and sewerage is one tenth of that for transport and communications which appears reasonable in the light of recent experience in Nepal. Reqional Distribution of Infrastructure Demand 6. Future distributions of infrastructure demand will be determined both by the distribution of economic activity, as well as by the ability to pay for infrastructure services. The latter, which would largely depend on levels of - 112 - ANNEX VIII Page 4 of 17 discretionary income, will particularly influence the demand for consumption infrastructure services, e.g. drinking water supply. Those types of infrastructure that are primarily social overhead capital, i.e. needed to support the productive sectors of the economy, will be demanded in proportion to the levels of economic activity in various parts of the country. These two determinants of infrastructure demand can produce quite different assessments of the distribution of aggregate infrastructure demand. Table 8.2 presents data from the 1985 Multipurpose Household Budget Survey (MHBS) showing the variations in total income, cash income and discretionary income between rural and urban households located in the Mountains, Hills and Terai. As would be expected, there are very wide variations among these households, particularly with regard to cash income which increases by nearly a factor of four between rural Mountain households and urban Hill households, the latter being greatly influenced by the high cash incomes in the Kathmandu Valley. Table 8.2 AVERAGE ANNUAL HOUSEHOLD INCOME & EXPENDITURES (1985) (1985 NRs.) Incae, Expenditure & Savings Mountain Hil Teral Rural Rurat Urban Rural Urban Incomes 1. Cash Income 4.583 5,485 17,618 6,044 10,924 2. In-Kind Income 8.1 a02 7.676 9-3955,6 3. Total I ncomie 137*_ 1350 T55.2C3- __ 6___8 _ Expenditures 4. Non-Discretionary 9,428 9,956 17,208 10,977 11,677 5. Discretionary 3.405 326 073 3,701 A 6. TotaL Expenditures 72,ffi 13,211 r'W 14,678 148 7. Savings (3)-(6) 561 293 2,013 761 1,991 8. Discretionary Income (5)+(7) 3,966 3,548 8,086 4,462 5,212 9. Avg. Household (persons) 5.36 5.74 5.23 6.87 5.8 Source: Multipurpose Household Budget Survey 1988 7. In order to assess the future evolution of infrastructure demand spatially, the MHBS data was combined with population projections to the year 2010. These projections are presented in Table 8.3. It should be noted that the projections of aggregate cash incomes and discretionary incomes assume that the relative levels of discretionary income and cash income per capita do not change. Also, no growth in income is assumed. These assumptions, among various types of households are conservative in view of the objective of the Table, as will be seen below. - 113 - ANNEX VIII Page 5 of 17 Table 8.3 PROJECTED DISTRIBUTIONS OF DISCRETIONARY AND TOTAL CASH INCOMES Mountain Hill Teral Nepal Rurat Rural I ural I Urban Total --------------------Millions*-----------****) Paulatlon 1891 1.4 7.55 0.86 7.79 0.82 18.46 (7.8) (40.9) (4.7) (42.2) (4.4) (100) 2010 1.74 8.79 2.67 11.85 2.27 27.31 (6.4) (32.2) (9.8) (43.4) (8.3) (100) ----------Costent 1990 USS Mltofo *------- Discretionary Incoum 1991 40.0 175.0 49.9 189.8 27.6 482.2 2010 48.3 203.8 154.8 288.7 76.5 772.1 Increase 8.3 28.8 104.9 98.9 48.9 289.9 Distr. of Increase (%) 3 10 36 34 17 100 Total Cash Income 1991 45.9 270.5 108.6 257.0 57.9 739.9 2010 55.5 314.9 337.3 390.9 160.3 1258.9 Increase 9.6 44.4 228.7 133.9 102.4 519.0 Distr. of Increase (%) 2 9 44 26 20 100 8. Increases in discretionary income are computed in Table 8.3, as an indicator of the evolution of infrastructure demand, particularly for consumption infrastructure. The largest proportion of the total increase in discretionary income occurs in the urban Hills (36%), followed by the rural Terai (34%). Urban Terai accounts for another 17%. Together these three groups (urban Hills, and rural and urban Terai) while accounting for only 61.5% of the population in the year 2010, account for 87% of the increase in discretionary income. The remaining 38.5% of the population located in the rural areas of the Hill and Mountain zones, account for only 13% of the increase in discretionary incomes. 9. Cash incomes are a good indicator of overall economic activity in various parts of the country, and therefore the spatial distribution of cash income is an indicator of the distribution of infrastructure demand in general, and of the demand for social overhead capital in particular. The analysis of total cash incomes in Table 8.3 indicates that increases in cash income are even more strongly concentrated in the urban areas and the rural Terai. Rural Hill and Mountain populations account for only 11% of total cash income even though they constitute almost 40% of the population. Based on cash income, the urban sector accounts for 63% of the increase in infrastructure demand, even though they constitute only 18% of the population in the year 2010. Rural -v- Urban Infrastructure Demand 10. According to these indicators, the rural sectors should receive no more than 37% to 47% of total infrastructure investment, whereas the urban sector should receive between 53% and 63%. Although the rural population in the year 2010 may still constitute more than 80% of the population, their share of infrastructure demand would likely be below 50%. In fact, the simplifying assumptions underlying this analysis make these conclusions conservative. Growth in per capita incomes would increase the share of urban areas in total - 114 - ANNEX VIIX Page 6 of 17 infrastructure demand, and so would increasing rates of urbanization. Nepal is presently at a very low level of urbanization (9%), which is among the lowest in the world. The above analysis assumes that by the year 2010, urbanization will reach 18% which would still be among the lowest in the world. If in fact urbanization exceeds these projections, and this could very easily be the case, then the rural share of infrastructure demand would reduce correspondingly. B. Investment Levels 11. Estimation of an appropriate level of future investments in infrastructure is fraught with numerous difficulties. However, indicative levels can be projected using estimated ICORs and projections for aggregate demand for the various types of infrastructure. The previous section provided estimates of future aggregate demand levels based on cross-country comparisons of infrastructure demand and the level of per capita GDP. To derive the levels of investment needed to create the capacity to meet those demands requires adoption of sectoral ICORs appropriate for each infrastructure sector. However, sectoral ICORs can vary widely among countries, and temporally due to a number of reasons such as utilization levels of existing infrastructure stock, the efficiency of capital utilization, dependence of sectoral output on performance in other sectors, macroeconomic factors affecting overall economic performance, etc. 12. Some estimates of sectoral ICORs are available for Nepal!, and these are presented in Table 8.4. The Table also shows comparable ICORs for India. In general, Nepal and India appear to have similar levels of economic performance when viewed in relation to international comparatorsJ2. Over the period 1965- 81, both countries experienced GDP growth rates in the 2% to 4% range, and overall ICORs have ranged between 5 and 7. Over the same period, LDCs in general appear to have experienced GDP growth rates ranging from less than 2% to over 8%, while ICORs have ranged from less than 3 to ratios greater than 9. Hence, lOORB estimated for the Indian economy may be taken as indicative of the likely levels of ICORs for Nepal. V See "Approach Paper for the Eighth Plan*, National Planning Camission, 1990. See NSavings, Investment and Trade in the Third WorLd", Panchamukhi, V.R., et aL, New DeLhi, 1987. - 115 - ANNEX VIII Page 7 of 17 Table 8.4 SECTORAL ICORS FOR NEPAL AND INDIA Nepal India Sectors National Planning Comission a/ Share Panchamukhi Chitale 1975-90 1992-97 1977 b/ 1971-75 c/ 1961-81 d( ole Economy 4.5 4.1 4.0 4.2 6.0 Agriculture 2.7 2.7 2.3 2.5 3.2 Manufacturing 3.2 2.5 3.4 6.3 8.8 Construction 1.8 1.8 ) 1.4 1.4 RaiLways ) ) )8.7 7.4 18.6 Other Transport )17.8 )12.9 ) 5.3 6.4 Comunications ) ) ) 4.5 5.7 Other Services 14.4 A/ NApproach to the Eighth Plan 1992-97", HNG/ National Planning Comission, Kathmandu, 1991 b NepaLese PLan Performanceu, Sharma, G., NiraLa Publications, New DeLhi, 1989. S/ "Capital Formation and Output in the Third World", Panchamukhi, V.R., Radiant Publishers, New DeLhi, 1986 41 "Capital Output Ratios in the Indian Economy", Chitate, V.P., Radiant Publishersr New Delhi, 1986 13. The wide variations in ICORB presented in the Table 8.4 are caused by a number of factors, the two most important being estimation techniques and the period for which the ICORs were estimated. In general, most researchers have noticed an upward trend in the Indian and Nepalese ICORs since the early 1980s, and this could explain some of the differences in the Table. Estimation techniques can also affect ICORs substantially. Some estimates appear to be based on single period data, whereas others explicitly try to account for lags between capital investment and output increases. Chitale, for instance, presents ICORs estimated using a variety of investment lags and data smoothing using moving averages; his ICORs presented in Table 8.4 are the ones based on 5-year moving averages with a 3-year lag. Given these difficulties, any attempt at accurate prediction of future ICORs is bound to be problematic. For the purposes of this analysis we have adopted the following ICORs: Transport and Communications - 10; Other Infrastructure - 15; and we have assumed that these ICORs correspond to output increases occurring with a 5-year lag. The investment levels derived in Table 8.5 using these ICORs should therefore be treated as indicative best estimates, and should not be viewed as firm assessments of appropriate investment levels. - 116 - ANNEX VIII Page 8 of 17 Table 8.5 INVESTMENT NEEDS IMPLIED BY GROWTH IN DEMAND (millions of 1990 USS) Infrastructure Sectors 1991 - 1995 1996 - 2000 2001 - 2005 2006 - 2010 High Ned. High Ned. High Ned. High Ned. at Transport & Comunications Annual Increase in Demand 23 13 34 19 49 27 73 38 Annual Investment Level 230 130 340 190 490 270 730 380 Public Sector Investment k 115 65 170 95 245 135 265 190 5-Year Public Investment 575 325 850 475 1225 675 1325 950 Water SuoLy & Sewerage Annual Increase in Demand p/ 2 1 3 2 5 3 7 4 Aual Investment Level 35 20 51 29 74 41 110 57 Public Sector Investment k 32 18 46 26 67 37 99 51 5-Year Public Investment 160 90 230 130 335 185 495 255 N/ High and Nedium estimates correspond to High and Nedium demand forecast from Tables 8.1. Public investment as a percentage of total investment is taken as 50% for transport and commications and 90% for water supply and sewerage. c Demand for water supply and sewerage assumed to be 10% of demand in transport and communications. 14. Table 8.5 presents High and Medium estimates for investment needs into the future. As was stated earlier, the Medium scenario corresponds to present levels of demand in the economy. In the near term, therefore, the Medium scenario would be applicable. In the longer term, if the Nepalese economy were to transform itself into a more infrastructure intensive economy, the higher projections would apply. These projections, however, do not take into account any serious lag effects between investment levels and output increases. If a 5- year lag were to be assumed, then the investment level shown in Table 8.3 for a given 5-year period could be increased to correspond with the level shown for the following period. 15. Table 8.6 compares estimated investment needs with the investments in the Seventh Five-Year Plan (7FYP, 1986-90), and planned investments in the Eighth Plan (8FYP, 1993-1997), for water supply and sanitation and transport and communications sectors. Clearly, in the case of water supply and sanitation sectors, the levels of planned investments are considerably lower than needed to meet the estimated demand. For the transport and communications sector, investment levels appear to roughly correspond with those estimated based on the "High" demand scenario. This may be a slightly higher level of investment than desirablel however, as was noted earlier, the precision of the estimates are not sufficient to categorically assert that the Eighth Plan investments proposed for transport and communications is excessive. - 117 - ANNEX VIII Page 9 of 17 Table 8.6 ESTIMATED INVESTMENT LEVELS AND FIVE YEAR PLANS (Constant 1990 USS Millions) Scenarios Mater Supply & Transport & Sanitation Comunications Seventh Five Year Plan (1986-90) 54 540 Eight Five Year Plan (1992-97) 165 a/ 850 Estimated Five-Year Investment Needs 1991-1995 High 160 575 Medium 90 320 1996-2000 High 230 850 Mediu 130 475 Estimated Investment Needs 5-Years Lao 1991-1995 High 230 850 Medium 130 475 1996-2000 High 335 1225 Mediun 185 675 p/ Based on the estimates in paragraph 34. Source: Five Year Plan investments from "Approach to the Eight Plan 1992-1997", NMG Planning Coamission, Kathmandu, 1991. C. Trends in Infrastructure Spendin9 16. This section briefly discusses the overall spending on infrastructure by government over the recent past. It examines central government and local level infrastructure expenditures and compares both of these with other countries as well as with expenditures undertaken by the private sector. Central Government Expenditures 17. It is, of course, impossible to categorically declare that a certain proportion of central government spending should be devoted to infrastructure. Certainly in Nepal, the share of infrastructure spending in the development budget is high, approximately 50%. And, the development budget is also relatively large compared to overall government spending, about two-thirds. Overall, infrastructure spending represents one-third of all central government spending. The obverse is that recurrent spending on infrastructure is rather low at only 1.6% of total government spending. This points up the need to ascertain the adequacy of spending on operations and maintenance (O&M) rather than new capital spending. 18. In a survey of 44 of the low income countries of the world, data for infrastructure spending by central governments was available for 18 countries for the year 1985. For these 18 countries their central governments have devoted an average 3.2% of GDP to infrastructure spending j, while for Nepal the central government spent about 5.2% of GDP on infrastructure. The only countries of this group that spent a greater share of their GDP on infrastructure were Malawi (7.8%), Ethiopia (5.8%), Vanuatu (5.5%) and Pakistan (5.4%). It is not clear that these countries should be spending more or less of their respective GDPs than does Nepal, nor is it clear that Nepal should be devoting a greater share of GDP to infrastructure than do the other 13 countries in the sample. What is Actually, these data include a limited amownt of spending on non-Infrastructure items such as tourism affairs and trade promotion. - 118 - ANNEX VIII Page 10 of 17 clear, however, is that it does not seem that the share that Nepal does spend on infrastructure is particularly odd based on this inter-country comparison. Local Government 19. Town and local development expenditures are relatively small in Nepal. For instance, as a percentage of HMG's total development spending over 1979/80 - 1984/85 these expenditures ranged only between 6.2% and 1.7%. As a percentage of GDP they ranged-only between 0.99% and 0.17%, over the same period. Total municipality investment is compared to private and total public sector investment in Table 8.7. Clearly, development spending by municipalities is dwarfed by HMG1s and the private sector's development spending. Table 8.7 SOURCES OF INVESTMENT SPENDING (1984/85 Constant Prices, milLions of NRs) Sectors Private Total Development Public Municipalities Sector Expenditures Sectoral Agriculture, Irrigation & Forestry 8,380 550 8,350 17,280 Industry, Mining and Power 7,040 30 3,770 10,840 Transport and Comfcations 5,130 600 1,530 7,260 Social Services 6,450 820 7,760 15,030 Total 27.000 2r000 21,410 50.410 I/ PubLtc Sector excludes Municipalities Source: HMG's NPC June 1985 D. Resource Mobilization 20. The Government of Nepal is extremely dependent on foreign grants and loans. Foreign grants amount to about one quarter of the domestic revenues raised by the Nepalese Government, which is similar to the levels in low income Sub-Saharan Africa. Foreign loans come to an additional five-eights of domestic revenues. Overall, foreign loans and grants combined are only about 15% less than total domestic revenues generated by HMG. Hence, slight declines in official development assistance or other foreign financing available to Nepal can result in rather large budgetary setbacks. It is imperative that if Nepal is to follow a sustainable development strategy then the government must lessen its dependence on external financing. This can be done in one of two ways: either the amount of external assistance accepted by the Government is reduced and greater reliance is placed on private sector investment, including private sector investment in social overhead capital, or government undertakes efforts to raise domestic resources. Central Government Tax Effort 21. Domestic resource mobilization by the central government in Nepal has been at relatively low levels, although rising slightly during the 1980s. For instance, central government revenues during the 1980s averaged only about 9% of GDP, whereas in 1985 the average ratio of current central government revenues to - 119 - ANNEX VIII Page 11 of 17 GDP for low income countries of the world came to about 15.4%.!/ This ratio has risen in Nepal, from 8.2% in 1982 to 10.1% in 1989. This is a reasonable trend to expect given that over the same period real per capita income in Nepal rose by about 12%. This low level of revenues to GDP is indicative of a relatively weak "tax effort." On the surface, it appears that there is great potential for the central government to increase its revenue take considerably, although this requires both political will and administrative capacity. A systematic comparison of Nepal's effort to raise revenues with that in other countries indicates, as shown below, that HMG could raise an additional 4% to 8% of GDP without imposing taxes higher than international norms. 22. A number of factors are said to contribute to a government's ability to generate revenues. Primary among these is the level of per capita incomes and the corresponding higher levels of urbanization in an economy. It is expected that at higher levels of per capita incomes and urbanization governments find it easier to raise higher relative (i.e. revenues as percent of GDP) revenues. Tanzil/ expects that urbanization leads to ever greater demands for public services, which are accompanied by greater generations of government revenues to fund these services, and that urbanization also facilitates tax collection. In Tanzi's words, "It is a fact of life that in the majority of countries a large proportion of total domestic taxes originate in the capital city or in large urban centers, whereas much of the public expenditure also takes place there". Assuming that urbanization is highly correlated with per capita incomes, Tanzi estimated the following equation using ordinary least squares regression on data from 86 countries. (Total Tax Revenue)/(GDP) * - 4.86 + 3.38 log (GDP,:) R2 - 0.26 (1.19) (5.61) d.f. - 84 23. Using somewhat more up-to-date data and a more limited number of countries (Eastern bloc countries were not included nor was the Republic of South Africa, nor were upper incomes countries) similar regressions were estimated. However, we were concerned not only with taxation but also non-tax revenues. There is often little difference between non-tax and tax revenues. For instance, in Egypt a large share of the government's revenues are royalties on petroleum exports, whereas in other countries this is treated as a tax on exports. The results of these regressions are presented below: (a) Revenue/GDP - 17.16 + 0.005004 GDPw R2 - 0.28 (4.179) d.f. - 45 (b) Revenue/GDP = 17.87 + 0.004482 GDP, R = 0.14 (2.8772) d.f. - 51 (c) ln (Revenue/GDP) - 1.25 + 0.2652 ln (GDP,.) R2 - 0.20 (3.62) d.f. - 51 See TabLe 2.2 in World Devetopment Report, WorLd Bank, 1988. Tanzi, Vito mQuantitative Characteristics of The Tax System of Developing Countriesu, in "The Theory of Taxation for DeveLoping Countries", Oxford University. Eds. Newberry and Stem, 1987. - 120 - ANNEX VIII Page 12 of 17 Regression (a) in all senses provides the greatest "goodness of fit". Regression (b) is very similar to regression (a) but does not eliminate severe outliers. Including these outliers, however, does not significantly alter the conclusions about the relationship between per capita income and revenue generation, though it reduces the level of statistical confidence. Regression (c) was run merely to test if the relationship might be non-linear. Regression (c) also does not alter our conclusions about the expected relationships. Based upon expected values generated from regression (a), Nepal should have been able to generate current revenues of about 17% of GDP. Based upon regression (c) the expected revenue level would have been about 13% of GDP. Instead, in 1986 (the data are from 1986) the Government of Nepal only raised current revenues of about 9% of GDP. Hence, greater revenue efforts by the Government of Nepal could raise an additional 4% to 8% of GDP without exceeding international norms. 24. In HMG/N's revenue structure, taxes represent about 82% of the government's revenues, while non-tax revenues supply the rest. The greatest source of tax revenues is indirect taxation, with somewhat greater reliance on revenues generated by taxing domestic goods and services compared to import duties, and with export taxes representing only a very tiny contribution. In all, indirect taxes during the 1980s ranged between 35.6% to 44.5% of total government revenues. Direct taxes during the 1980s ranged from 12.9% to 17.2% of total government revenues, with the greatest source being the tax on personal incomes. Tax on immovable property (such as real estate) amounted to only a very small share of central government revenues, ranging from 1.0% to 3.5%. Local Government Resources 25. Local government bodies (district and lower level) are dependent on HMG for about 90% of their revenues. On a per capita basis, own source revenues for local governments (not including municipalities) ranged between Ru. 0.22 and Re. 1.906/ ($0.01 to $0.10) in the mid-1980s. Even in a country as poor as Nepal, this amount of locally generated revenues is extremely low. About ten times this amount is allocated to local government by HMG. Much of the allocation of funds to local governments is based on population size and district resources. There is apparently no aspect of matching grants or other specific incentives for district level mobilization of resources. Among the resources available to districts from HMG are administrative, developmental, and special programs grants. The administrative grants are to assist in covering certain recurrent costs, particularly to pay for the salary of employees, and tend to be based on a progressive redistribution as evidenced by the fact that higher per capita allocations are made to the poorer districts in the mountains. 26. There are 33 municipal governments in Nepal, each of which generates revenues for local uses from the octroi (a local duty on the movement of goods), property taxes and various fees. The octroi is by far the major source of revenues for these municipalities, averaging around 80% of all revenues and grants. In addition, municipalities receive specific grants from the central government, but these grants only average about 10% of all financial resources available to the municipalities. As a share of Nepal's GDP, municipal revenues are rather insignificant, ranging from 0.09% of GDP to 0.23% of GDP during 1984- 89. As a share of urban incomes municipalities total revenues and grants See report to the Asian Development Bank: "Oecentralization in Nepatw, Bienen, N., Kapur, D. et at, Princeton University, 1988. - 121 - ANNEX VIII Page 13 of 17 averaged only about 1.2% during 1985-89. Clearly the magnitudes are not great. Data from a selected number of countries indicate that property taxes range from 5 to 25% of recurrent receipts of local governments (see World Development Report, 1988). In Nepal, property taxes come to well below 1% of municipal taxes. Property taxation has a number of advantages over other types of taxation, especially compared to octroi, and is an untapped source of local government financing. The opportunities for increasing reliance on this source of financing is discussed below. E. Options for Domestic Resource Mobilization 27. To the extent possible costs and benefits of public services should be directly linked through user charges. When specific services are provided to specific users relevant charges should be levied for those services. When broad services are delivered to specific geographic locations specific fees, charges and taxes should be imposed upon the residents in the area. User charges can promote greater economic efficiency by (a) allocating services according to the marginal benefit/marginal cost principle and (b) by causing users to demand higher quality of service. When users are not required to pay for such services through user charges they are less likely to effectively lobby for high quality service. At the same time, since the users are not required to pay charges they will tend to "over-consume" the service in question, lobby for services that are not economically justified or create inefficient congestion in public services. 28. User charges are of two types: consumption and benefit related. A consumption charge is levied for the use of a public service and directly relates consumption of the service to payment for the service. A major benefit related charge is the local development tax, usually collected to pay for a specific local development. These are often lump sum taxes. In most countries such development taxes are collected primarily from land holders. In much of the developing world user charges are relied upon greatly at the local level. In Nepal, however, such charges amount to only a very small share of municipal government revenues. An expansion of user charges by municipal governments should not be very difficult to manage. Such expansion of user charges would have numerous benefits, specifically: this would better relate local revenues to benefits, would be economically more efficient by reducing distortions, and could improve equity considerations, since many of the beneficiaries of government services (such as sewerage, trash collection, city water, etc.) are the wealthier members of society. Pricing of Infrastructure Services 29. There are many administered prices in Nepal. In addition to user charges collected by various government agencies, there are a number of services provided by public enterprises whose prices are regulated by government. The reasons for such regulation include: regulation of monopolies; generation of appropriate rates of return on public capital; social goals that require subsidization to promote "merit goods"; and to make services available to the poor. -122 - ANNEX VIII Page 14 of 17 30. Water Surolv prices in Nepal are extremely low, and coat recovery for both urban and rural water supply schemes will require substantial increases in water tariffs. In the larger urban areas under the responsibility of the Nepal Water Supply Corporation, tariffs for water supply were revised in 1990 after 9 years without any increases, and were again increased 100 percent in January 1993 to NRe 5.00/1000 litres. In the case of the Kathmandu Valley, where existing groundwater sources are being rapidly depleted, the cost of schemes to bring in water from outside the Valley are estimated to require a 300 percent real increase in prevailing tariffs over the next 10 years. 31. Irrigation plays an important role in the agricultural development of the country. At present, the total irrigated area of the country is limited, at about 13% of all cultivated land. Over 80% of the total irrigation facilities are concentrated in the Central Terai. Only very few areas in the Hills enjoy irrigation facilities. Moreover, the western part of Nepal, where rainfall is relatively low, has very limited irrigation facilities. According to numerous reports there is a radical difference in private and public irrigation schemes in the Terai. Public irrigation schemes have the potential to supply water reliably to about 5% of the country's cultivated land, yet the actual area supplied is considerably less. This poor performance is due to a combination of inadequate design, lack of a tertiary distribution networks, and woefully inadequate spending on operations and maintenance. One study found that it costs the government NRs. 3,096 to deliver only NRs. 678 worth of water to farmers. The study concludes that most government irrigation projects are not functional at all, mainly due to poor maintenance which is largely due to the government allocating almost nothing for maintenance of these schemes. In contrast, privately run irrigation schemes, which account for about four-fifths of all areas under irrigation, cost about one-fifteenth the per hectare cost of public schemes. In addition, the private schemes are adequately maintained and cover operating costs on their own. One disadvantage of the private schemes, however, is that because of the constant need for reconstructing weirs and parts of the system using local wood they contribute to deforestation, which is an important ecological problem. 32. Water charges on publicly provided services are very low. The annual rate of NRs. 200 per hectare for surface irrigation and NRs. 400 per hectare for deep tubewells are insufficient to meet operations and maintenance expenditures. At the same time, the quality of irrigation services from government schemes is very low, with water supply not being reliable and the amount supplied often insufficient to be useful. These problems with the quality of irrigation services from public schemes has made it extremely difficult to get farmers to pay the water fees they are obligated to pay. Hence, it is inadequate to merely raise the prices of water to meet O&M costs, the quality of service must also be improved. Additionally, in view of the high costs of public schemes vis-a-vis private irrigation schemes, the point may be moot if it is instead suggested that Nepal move to ever greater reliance upon private irrigation schemes. 33. Transport facilities in Nepal are generally operated at very low utilization levels, and hence the opportunities for imposing user charges in the form of tolls are extremely limited. With the exception of some urban streets in Kathmandu, road use is well below capacity even in the industrial "ribbon zones where the highest traffic levels are only on the order of 800 vehicles per day. Funding of roads may therefore be best accomplished using a variety of taxes, such as excise taxes on fuel, etc., and dedicating such tax revenues for the road sector. - 123 - ANNEX VIII Page 15 of 17 34. Creating a Road Maintenance Fund financed with a dedicated share of road user charges for road maintenance is being considered. The Fund would assign specific revenues, usually from sources such as fees or percentages of certain types of tax revenues, to specific road expenditure categories. The attractiveness of a Road Maintenance Fund as a concept is that by linking revenues to a share of specific expenditures (to be complemented with Central Government resources) make it possible to ensure that payments to contractors are timely and the expenditure program is adequately funded. Hence, tying excise taxes on fuel to road maintenance can ensure that as long as fuel is bought and roads are travelled there will be adequate funding for maintenance of roads. Additionally, the concept of road user charges is an extension of the benefit principle of taxation, where beneficiaries of government services not only pay for those services but where those fees and taxes that are paid are directed back to the provision of those very services. The Road Maintenance Fund has at least two other potential benefits. First, it can provide greater stability and continuity in the funding of specific services, hence ensuring timely and adequate operations, maintenance and expansion. Second, by linking fees and taxation to ear-marking for specific purposes it may make such revenue generation more palatable to tax payers, and hence more politically feasible. 35. However, clear and transparent guidelines are neded to avoid possible problems with the Road Maintenence Fund. For instance, lack of performance budgeting and improved accountability may lead to a misallocation of funds, where programs of lesser economic and social worth may receive greater funding than other programs, especially in light of fiscal austerity. By the same token, the Road Maintenance Fund may reduce flexibility of budgeting and infringe on the power and discretion of the executive branches of government. Not all the issues considered to be problems will be universally seen as such. Given the nature of the executive branches of governments it may make good policy to impose restrictions on their powers and discretion. In the case of Nepal, however, given recent developments of the past few years in terms of political liberalization, it may impede the political and administrative development of the country if ear-marking were to occur on a grand scale. 36. Generally, World Bank ex erience points to the advantages of Road Maintenance Funds. A recent paperY provides adequate evidence of instances where dedicated shares of road user charges should be considered. Among the instances where it might be useful with regard to Nepal, are: (a) Where the extension of the benefit principle is most clearly served by the Road Maintenance Fund. That is, where the beneficiaries of spending can clearly be linked to a revenue source. For instance, a portion of the excise taxes on fuel can be dedicated for road maintenance; and (b) A Road Maintenance Fund is necessary because there is reason to believe that important programs will otherwise not be adequately funded. This seems to be the case in Nepal, where funding for operations and maintenance is well below optimum. McCleary, W., "The Earmarking of Government Revenues: A Review of Some World Bank Experience", The World Bank Research Observer, Vol. 6, No. 1, January 1991. - 124 - ANNEX VIII Page 16 of 17 Yet, this mechanism would not solve the issue of why operations and maintenance have consistently been under-funded in Nepal, and there are very definite budgeting issues involved that need to be addressed in other fora. Imrovinq Local Revenue Systems 37. Efforts to decentralize (government responsibilities, especially to the municipalities may yield considerable benefits in terms of raising accountability and making government more responsive to the electorate. This is not, however, a sine qua non. It should also be kept in mind that in terms of resource mobilization and allocation there has been almost no progress at the local levels, and perhaps some backsliding to date. In addition, although there have been efforts to improve administrative capacity of municipal governments, such as the Management Support for Urban Development (MSUD) project, administrative and analytic capacities at local levels remain weak. 38. A report on the finances of local municipalities-' describes the resource mobilization efforts as almost exclusively relying on the octroi, with low level of central government grants, very little reliance on direct taxes (e.g. professional and property taxes), and none or negligible reliance on user- based fees. The municipal revenue system can best be characterized as inefficient and difficult to administer, mainly due to the heavy reliance on the octroi.VI While the MSUD team clearly feels the need to replace the entire octroi with more direct taxes, they also realize that this cannot be done quickly and easily. Because these municipalities are so dependant on the octroi, removing the octroi is difficult since these municipalities could wind up with very limited revenues. In fact it may be argued that no effort should be made to dismantle the octroi while local government resources remain so severely constrained. Rather, efforts should concentrate on developing new revenue sources, thereby reducing the reliance on octroi. 39. There is a great deal of scope for improving the ways revenues are generated at the local level in Nepal. It has already been mentioned that greater reliance on marginal cost pricing could be used to enhance both the allocation of government provided services as well as to mobilize greater domestic resources. At the same time, there are some very fundamental improvements that could be made to the overall revenue structure of municipal financing in Nepal. It is suggested that such a fundamental program would comprise three aspects: (a) Greatly reduce reliance on the octrai; (b) increase reliance on user charges; and (c) increase drastically the amount of revenues from property taxation. OResource Nobilization for Urban Development: Financial Points of VieW* TuLadhur, Raju, UNDP/WorLd Bank Project (NEP/88/046). A nanber of reports discuss the inefficiency and primitiveness of the octrol. See Stout, Don et at, "Octrol Tax Study.0 NSUD Project Study Report. 1990: OA Study of Local Taxation in Nepal*, NSUD. 1990); TuLadhar, Raju OThe Octrol: Its Structure, Role and Issues In Municipal Finance, MSUD, 1990. - 125 - ANNEX VIII Page 17 of 17 The property tax has several advantages over other types of taxes. First and foremost, all municipalities contain properties, hence there is a base. Second, the property tax, combined with good government can be self-liquidating. That is, where good government provides high quality, desired public services, the value of property rises. Hence, if we can ensure that the property tax will be imposed jointly with improved services, then we can ensure that rising property taxes can be offset by increased land values. Property taxes are primarily imposed on the wealthier segments of society and with some specific exemptions can be fine-tuned to ensure their progressivity. Finally, the tax on land is a tax on an economic rent and hence is non-distortionary. Thus, property taxes are efficient relative to almost any other tax. Poverty Concerns 40. Poverty concerns are of great importance in Nepal; with a per capita income of only $154 in 1991, only ten countries of the world were poorer. The recommendations for tax restructuring and marginal cost pricing made here are generally progressive in income. For instance, the octroi imposes an internal trade tax, which tends to be borne more heavily by poorer segments of society, while a property tax is, in the first instance, imposed on the wealthier (i.e. land owning) segments. Marginal cost pricing generally tends to exempt the poorest of the poor from paying for services that they cannot at any rate take advantage of. These basic tenets are, of course, generalizations and greater study and targeting of tax incidence and expenditure are necessary before such sweeping generalizations can be asserted with any degree of certainty.