China - Long-term issues and options (Vol. 7 of 7) : Annex F : transport

Report No. 5206-CHA China: Long-Term Issues and Options Annex F: Transporn. May 22,1985 Transportation Division 1 Projects Department East Asia and Pacific Regional Office FOR OFFICIAL USE ONLY Document of the World Bank This report has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS The Chinese currency is called Renminbi (RMB). It is denominated in Yuan (Y). Each Yuan is 1 Yuan = 10 jiao = 100 fen In early 1984 the official exchange rate of the Yuan to the US dollar was around Y 2 = $1. The internal settlement rate (ISR) of Y 2.8 - $1, however, was used in most merchandise transactions. The official exchtange rate is now about Y 2.8 = $1. On January 1, 1985, the Government abolished the ISR. FISCAL YEAR January 1 to December 31 WEIGHITS AND MEASURES tkm = ton-kilometer pkm = passenger-kilometer mu 0 u.0667 hectare (ha) kWh kilowatt hour (= 860.42 kcals) CTK converted ton-km, traffic unit ( 1 passenger-km I ton-km) mt = million tons mtpy million tons per year MW megawatt kcal = kilo-calorie kV kilovolt kVA = kilovolt-ampere ABBREVIATIONS AND ACRONYMS AIC - Average incremental cost CASS - Chinese Academy of Social Sciences CBEs - Commune and Brigade Enterprises DWT - Dead weight ton GDP - Gross Domestic Product GNP - Gross National Product GVIAO - Gross Value of Industrial and Agricultural Output GVIO - Gross Value of Industrial Output LRMC - Long run marginal cost LRVC - Long run variable cost MR - Ministry of Railways MOC - Ministry of Communications NMP - Net Material Product  FOR OFFICIL USE ONLY CHINA: LONG TERM ISSUES AND OPTIONS ANNEX F TRANSPORT Table of Contents Page No. SUMMARY .......................................................... vi-xiii PART I: TRANSPORT INDICATORS: CHINA AND THE INTERNATIONAL EXPERIENCE ............ 1 1. TRANSPORT IN THE ECONOMY ...................................... 1 2. FREIGHT TRAFFIC............................................... 4 A. Freight Growth ........ .................................... 4 B. Freight Intensity ......................................... 6 C. Average Transport Distance and Modal Split ................ 13 D. Port Traffic .............................................. 20 3. PASSENGER TRAFFIC ............................................. 23 A. Growth of Passenger Traffic ............................... 23 B. Travel Purpose ............................................ 27 C. Modal Spi-t .............. 29 4. CAPITAL INVESTMENT AND PRODUCTIVITY ........................... 30 A. Infrastructure ............................................ 30 B. Rolling Stock ............................................. 33 C. Investment ................................................ 33 PART II: FUTURE DEVELOPMENTS AND ISSUES ..... .................... 37 5. FUTURE TRANSPORT TRENDS ....................................... 37 A. Freight Transport Demand .... 37 B. Modal Split ..... 39 C. Railways ................................................. 41 D. The Role of Road Transport .... 42 Road Infrastructure ..... 43 Road Transport ..... 45 E. Domestic Water Transport .... 47 F.' Port Development ..... 49 G. Passenger Transport ..... 49 H. Modal Split ..... 50 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.  - ii - Page No. 6. TRANSPORT AND ENERGY .......................................... 53 A. Energy Transport Issues . . ................................. 53 Coal .................................................... 53 Petroleum ............................................... 59 B. Energy Consumption in the Transport Sector . . .............. 59 Railways ... 59 Road Transport ... 60 7. TRANSPORT AND INDUSTRIAL LOCATION .. 61 A. Natural Resources ....................................... . 61 B. Heavy Industry ............................................ 62 Cement .................................................. 62 Iron and Ste!l ... 64 C. Light Industry ............................................ 67 8. TRANSPORT AND AGRICULTURE ..................................... 68 A. Grain ..................................................... 68 B. Fertilizer ................................................ 72 C. Transport and Rural Development .. 73 D. Conclusion ................................................ 74 TABLES IN TEXT 1.1 Transport as Percentage of GDP ............. .. .............. 1 1.2 Freight and Passenger Transport Intensity . . 2 2.1 Freight Traffic Volume .. 4 2.2 Total Freight (Billion tkm) and GNP (US$ Billion): A Cross-Country Comparison .. 6 2.3 Freight Intensity, 1980 .. 8 2.4 Total Freight, 1980 .. 8 2.5 Percentage Distribution of CDP .. . 9 2.6 Freight Intensity, 1980-81 .. 10 2.7 Coal Preparation in Percentage of Raw Coal . . 10 2.8 Transport of Timber and Products .. 11 2.9 Freight Transport Intensity, Rail Network and Heavy/ Light Industry Share by Province . . 14 2.10 Modal Split of Freight Traffic in Z of Total tkm . . 15 2.11 Rail and Road Tonnage .. 15 2.12 Japan: Coastal Shipping .................................... 17 2.13 India: Intermodal Share by Commodity ....................... 19 2.14 US: Cement and Fertilizer Distribution by Mode, 1977 .. 19 2.15 Port Traffic in Selected Countries .. 21 2.16 Foreign Trade Component of GNP for Selected Countries .. 21  - iii - Page No. TABLES IN TEXT (cont'd) 3.1 Passenger Traffic V o luo e .................me .* ...... 23 3.2 Intercountry Comparison of Passenger Traffic Per Capita and GNP Per Capita ............ I........ 0... 25 3.3 US: Household Trips by Purpose, 1977 . . 27 3.4 Intercountry Comparison of Passenger Traffic Growth Rates and Income Elasticity ................*.......*........ . 28 3.5 Intercountry Comparison of Modal Split of Passenger Traffic 29 4.1 Transport Network, 1952-82 ....... ........ ..... ... 30 4.2 Transport Network Density in Selected Countries ... 31 4.3 Railway Traffic Density in Selected Countries ............... 31 4.4 Truck Productivity in Selected Countries . ............. 32 4.5 Transport Rolling Stock in Selected Countries ............ .. 34 4.6 India: Share of Public Sector Investment in Transport ...... 35 4.7 Annual Transport Sector Investment in Selected Countries .... 35 5.1 Elasticity of Freight Transport to GVIAO .................... 38 5.2 Freight Growth Scenarios .. 39 5.3 Scenarios for Future Freight Modal Split . . 40 5.4 Domestic Water Transport .. 48 5.5 Foreign and Domestic Trade through Major Coastal Ports ...... 49 5.6 Passenger Traffic Growth Scenarios .......................... 50 5.7 Rail and Road Passenger Traffic ..so.......... ... 50 5.8 Scenarios for Future Passenger Modal Split . . 51 6.1 1982 Interregional Rail Coal Transport ...................... 54 6.2 Net Interregional Coal Flows ................................ 54 6.3 Coal Production and Transport-Cost Comparison ............... 55 6.4 Regional Coal Production, 1980-2000 ........................ 55 6.5 Combined Rail-Water Routes for Coal ......................... 57 7.1 Patterns of Coal and Timber Transport by Rail ............... 61 7.2 Cement and Iron and Steal Transport by Rail ................. 62 7.3 Cement Price and Transport Tariffs ... 63 7.4 Major Pig Iron Producing Provinces ............................ 65 7.5 Iron and Steel Production and Rail Transport . . 66 8.1 Grain Transport . ... 68 8.2 Rail Transport of Grain, 1982 ............................... 69 8.3 Net Interregional Grain Flows .............................. . 69 8.4 Indicative Data on Grain Handling Facilities at Major Ports .... ...................................... 71 8.5 Fertilizer Rail Transport, 1982 .. . . .. . . . 72  1 v - Page No. FIGURES 1.1 Comparative Growth of Net Material Product, Output of Agriculture and Industry and Transport in Tons and Ton-Km .. 3 2.1 GVIAO and Total Freight Growth Trend, 1952-1982 ............. 5 2.2 Total Freight and GNP for Selected Countries, 1960-1981 ..... 7 2.3 China and Japan Coastlines: Comparative Distances Between Major Ports .. 18 3.1 Passenger Traffic, 1952-1982 .. 24 3.2 Intercountry Comparison of Passenger Traffic and GNP Per Capita ....................I.............................a............. 26 APPENDICES A. Coal Transport versus Electricity Transmission ............... 75 B. Statistical Tables .. 79 B.1 Intercountry Comparison of Freight Traffic Growth and GNP Elasticity ....................................................... B 0 B.2 Total Freight and GNP for Selected Countries, 1960-81 ... 83 B.3 Intermodal Freight Intensity by Province, 1981 . . 84 B.4 Intercountry Comparison of Average Freight Distance, Selected Years, 1940-82 ........................... 86 B.5 Short Distance Rail Transport in Large Urban Areas ...... 87 B.6 Short Haul Transport on Selected Trunk Lines . . 88 B.7 Freight Turnover by Mode of Transport .................. 89 B.8 US - Freight Turnover by Mode of Transport . . 89 B.9 USSR - Freight Turnover by Mode of Transport . . 90 B.10 Brazil - Freight Turnover by Mode of Transport b ....& 90 B.11 India - Freight Turnover by Mode of Transport . . 91 B.12 Japan - Freight Turnover by Mode of Transport . . 91 B.13 Japan -Water Transport by Commodity ... 92 B.14 Rail Freight by Commodity .. 93 B.15 USSR - Rail Freight by Commodity . . 94 B.16 India -Rail Freight by Commodity . . 95 B.17 Japan - Rail Freight by Commodity ....................... 96 B.18 Brazil - Rail Freight by Commodity .. 97 B.19 India - Commodity Flows and Intermodal Shares, 1978-79.. 9b B.20 Japan -Road Freight by Commodity . . 99 B.21 Intercity Passenger Traffic and Per Capita Income for Selected Countries, 1960-81 ............................. 100 B.22 US - Household Trips by Main Purpose, 1977 ............ .. 101 B.23 Passenger Traffic by Mode of Transport .................. 102 B.24 US - Passenger Traffic by Mode of Transport ............. 102 B.25 USSR - Passenger Traffic by Mode of Transport . ...... 103 B.26 Brazil - Passenger Traffic by Mode of Transport ......... 103  Page No. APPENDICES (cont'd) B.27 India - Passenger Traffic by Mode of Transport .......... 104 B.28 Japan - Passenger Traffic by Mode of Transport .104 B.29 Railway and Road Network - Density in Selected Countries 105 B.30 Rail Density by Province ..106 B.31 Investment in the Transport Sector .107 B.32 Freight Traffic, QUADRUPLE: Case 1 - Low Road Share 108 B.33 Freight Traffic, QUADRUPLE: Case 2 - Low Road Share 109 B.34 Freight Traffic, BALANCE: Case 3 - High Road Share .110 B.35 Freight Traffic, BALANCE: Case 4 - High Road Share .111 B.36 Road Network. 112 B.37 Passenger Traffic, Case 1 - High Road Share .113 B.38 Passenger Traffic, Case 1 - Low Road Share .114 B.39 Passenger Traffic, Case 2 - High Road Share .115 B.40 Passenger Traffic, Case 2 - Low Road Share .116 B.41 Coal Production, Transportation and Consumption Summary by Province, 1980 .117 B.42 Railway Energy Consumption Plan .118 B.43 Cement - Production and Rail Freight Destination, 1982 119 B.44 Iron and Steel Products - Production and Rail Freight Destination, 1982 ................. - - . - . 120 B.45 Grain - Production and Rail Freight Destination, 1982 121 B.46 Fertilizer - Rail Freight Destination, 1982 .122 M-APS IBRD A18284 Transport Intensity, Rail Intensity and Share of Heavy Industry IBRD 18224 Major Coal Flows, 1980 IBRD 18225 Coal Production by Region, 1980-2000 IBRD 18226 Planned Capacity Increase for Power Generation by Year 2000 IBRD 18227 Major Rail Flows of Petroleum and Products, 1980 IBRD 18228 Rail Electrification Projects - Under Construction and Near Term Plans IBRD 18229 Coastal Shipping - Port Traffic, 1979 Errata: Map 18226 Symbol for power generatioz in West Henan-Anhui should be mine-mouth and not load-center  -vi- SUMMARY 1. China's annual investment in the transport sector has been small in comparison to that of other countries. The new economic policies are generating rapid economic development that is accompanied by growing demand for transport of freight and passengers. Inadequate transport systems are becoming a bottleneck in, for examples, the movement of coal from mine to user, the transport of agricultural and light industrial products from rural to urban areas and the delivery of imports and exports. 2. Massive investment in the sector is necessary though some rationalization of freight operations is possible and should certainly be pursued. Preliminary processing near the point of origin (preparation of coal, milling of lumber, etc.) would reduce the tonnages that must be shipped. Bulk handling of products like cement and chemical fertilizer would reduce the waste that occurs when bags are shipped. Use of the truck fleet (especially own-account vehicles) could be improved. Further measures to rationalize transport could be identified through distribution studies for major commodities such as coal, iron and steel, cement, fertilizer and grain. 3. It is unrealistic to think, however, that such improvements will render existing facilities sufficient to accommodate all future traffic growth. Major investments are needed to increase transport capacity and avoid stifling economic growth. In past years, China's investment in transport has averaged about 1.1% of GNP, which is low compared to other countries. Future five-year plans should consider increasing the magnitude of annual transport investments to well over 2Z of CDP. 4. How could these investments be financed? Parts of the system, such as railways and ports, already generate substantial net revenues. In railways, the level of cash generation (including net operating revenues and depreciation exclusive of major repairs) amounts to over 4 billion yuan per year. For other transport facilities such as roads and ports, financing mechanisms need to be developed at all levels of Government. In particular, a review should be made of present arrangements to charge road users and consideration given to other financing alternatives such as a fuel tax. 5. The transport sector should also evolve toward more balanced allocation of traffic between modes. The responsibility system in agriculture is creating a rapid demand for road transport; water transport both coastal shipping and inland waterways, would also play a larger role to relieve the pressure on railways. The greater development of all transport modes will require inter-modal coordination. Infrastructure, as well as policies and regulations, need to be developed to promote inter-modal transfers, in cases when multi-modal services have an economical advantage. Freight Transport 6. Transport Intensity. A number of factors affect freight transport intensity: country size; location of natural resources; industry and population; composition of GNP; the level of processing of raw materials such  - vii - as ores, lumber, agricultural products; and the degree of vertical integration of industry. The composition of GNP is probably the major factor in explaining China's high freight transport intensity. Heavy industry is more transport intensive than light industry which in turn is more transport intensive than services. When the service sector, which is unusually small in China, is excluded from CNP, the freight intensity of China appears somewhat more in line with that of other countries. 7. The location of natural resources in relation to that of population and markets is another important factor as illustrated by the movements of coal and timber. These two resources together account for over 40% of the ton-kn in the freight traffic of the Chinese railways. While coal is found in many parts of China, timber is concentrated in the Northeast. As a result, the share of inter-regional transport of timber is twice that of coal and the average transport distance almost three times as long. 8. While coal is presently produced in most of China, the Largest and best reserves are in north China, particularly Shanxi and Nei Monggol. A question for the future is the optimal balance between coal development near major markets in northeast and east China and coal development in Shanxi. Analyses to date show that Shanxi coal can be competitive in coastal China despite transport distance of the order of 1,000 km. Deposits in Large seams are easily accessible and both investment and operating costs are lower than in other regions. In addition, coal is generally of higher quality and calorific value. Therefore further develop'ment of coal in north China appears to make good economic sense and the region is expected to account for almost 60% of the coal production increase by the year 2000. This will considerabLy increase inter-regional movement of coal and transport intensity per ton of output. 9. With greater concentration of coal production in north China the issue of transporting coal or electricity needs to be addressed. An analysis of this matter in Appendix A concludes that, in broad terms, rail transport would be more economical than electricity transport for coal above 5,000 kcal/kg. Since Shanxi coal is generally of high calorific value (6,000 kcal/kg or more), mine-mouth power generation and long distance transmission is likely to be justified only in the case of use of middlings, which are expected to become available in greater quantity with increased coal preparation before shipment. 10. By the year 2000, some 450-500 million tons of coal will need to be transported out of Shanxi, or over 4 times the planned volume for 1984. Increased use of coastal shipping from northern ports to southern destinations can help meet the demand, but coal will still need to be moved by rail to the ports. Therefore, substantial capacity needs to be added to the corridors leading from Shanxi to the east coast. In addition to increases in rail capacity, coal slurry pipelines are being studied, but the scarcity of water in Shanxi province may hinder such development. 11. Industrial location and plant size also affect transport patterns. The cement industry was developed in the Northeast and is only gradually becoming more evenly distributed. The raw material is usually fairly  - viii - abundant, and given the high incidence of transport cost on final product value, location of production close to markets is common in most countries. The drop in cement rail transport distance from 600 km in 1970 to slightly under 400 km now indicates that some progress has been made in dispersing production from its original concentration in the Northeast. Nevertheless, cement is transported in China over distances longer than those found economical in other countries and further rationalization appears possible. 12. Chemical fertilizer transport is an exampLe of the effect of plant size. As 13 large plants were put in operation after 1978, chemical ferti- lizer moved by rail more than doubled between 1977 and 1981 from 12 to 25 million tons and average transport distance by rail increased 30%. Further rationalization of the industry is likely to reinforce these trends. 13. Both cement and fertilizer are presently transported mainly in bags, and losses during transport are high, particularly in fercilizer. For both commodities, consideration should be given to grdduaLly increasing the proportion of bulk transport. As China is one of the largest fertilizer importers in the worLd, the shift to bulk handling in ports would result in immediate savings in shipping costs as well as in port capacity by reducing ship turnaround times. 14. Finally, the economic system also has a major impact on the freight intensity of the economy, as the case of iron and steel illustrates. Because of the low price of iron ore, pig iron is produced in virtually every province and 50 million tons of iron ore are transported by rail over an average distanice exceeding 300 km. As sc.aLting iron ore greatly reduces its weight, transporting 12-13 million tons of pig iron would be more economical than moving 50 million tons of ore. Iran and steel products are also traded heavily; many provinces both ship and receive iron and steel to and from almost all other provinces. Duplication of transport results from administra- tive rigidities and in particular the different spheres of responsibility for steel production at the national, provincial and Local levels. The existing technology, with little integration in steel making also requires much trans- port of intermediate products. Iron and steel transport is third after coal and timber in rail transport and, together with iron ore, account for 10% of rail traffic. Any rationalization of the distribution system would provide substantial benefits. 15. The heavy freight transport demand could be better managed. To find exactly where and how, some distribution studies for the major commodities being transported such as coal, timber, cement, fertilizer, iron and steel and grain should be made. These studies would necessarily be intersectoral involving producers, users, and transport agencies. The difficulty of carrying out such studies which cut across various administrative jurisdictions is recognized, but institutes under the State Planning Commission as well as universities, research institutes or consultants should be suited for this type of work. 16. Future Trends. The discussion of freight intensity shows that future de-'elopments are likely to increase transport demand in relation to output for some products and decrease it for others. For coal, average  - 1X - distances will certainly increase with reliance on concentrated mine location in North China. The same would apply to fertiLizer with further concentration of production in large plants. For cement and iron and steel, however, rationalization of the induscry and distribution should reduce transport demand. Overall, freight elasticity to gross value of industrial and agricul- tural output (CVIAO) is expected to remain near 1.0 to the year 2000. By then, domestic freight turnover could be about 3,000 billion tkm or about three times the level reached in 1983. 17. Modal Distribution. The dominance of rail in the transport sector is unique to China. Despite some reduction in share over the last 30 years rail still accounts for 66% of the traffic in ton-km, probably the highest share of any country in the world after the USSR. Covernment efforts have concentrated on developing the railways to serve the requirements of heavy industry for moving Large quantities of coal, oil, timber, mineral ores and construction materials. The slower growth of light industry and agriculture has contributed to a reLatively lower demand than in other countries for the transport of goods normally carried by road transport. Despite some possible rationalization in the transport of natural resources and industrial products mentioned above, the transport demand for these commodities and consequently rail traffic will continue to increase. It could well triple its 1980 level by the year 2000. 18. By the year 2000. however, China's transport system is likely to have evolved toward a more balanced modal allocation reflecting the more diverse needs of the economy. The new economic policy and in particular the emphasis on light industry and on the responsibility system in agriculture is creating a rapidly growing demand for road transport. With a growth rate of 10-11% p.a, the share of road transport could increase rapidly to 17-20t of domestic ton-km from 9% in 1980. 19. While it is widely recognized and agreed in China that road trans- port should and will play a larger role, there seems to be no concrete plan to bring this about. The complexity of the situation seems to have overwhelmed the system. It is difficult to determine which need to address first: (a) larger trucks would be more efficient than the present ubiquitous 4- 5 ton capacity model, but the roads and bridges are not built to take the corresponding axle loads; (b) energy efficiency of trucks should be improved but that would require better fuels and therefore changes in refineries; (c) the output of liquid fuel may not keep up with demand if road transport develops too fast. Transport infrastructure and operating agencies seem to have very little contact with vehicle manufacturers, and they in turn have little contact with producers of refined products. Given the importance and urgency of the matter, the Government should consider developing a coordinated plan for improving roads and road transport.  20. Development of road transport should cover: (a) infrastructure, including both major intercity highways and bridges, rural roads and terminals for forwarding road freight and for intermodal interfacing with other transport modes; (b) vehicles, including the technological improvement of engines, the diversification of output both in terms of small 0.5 to 2 ton trucks and of heavy trucks; (c) management and operation of the truck fleet, both under MOC and own account vehicles; and (d) reforms of the commodity circulation system to help reduce the need for transport and multiple handling. 21. Crowing road transport will greatly increase liquid fuel consump- cion. Supplies are already very tight and in some cases shortages constrain the use of vehicles. Even with improvement in fuel efficiency, consumption in the year 2000 could be 3 to 5 times higher than in 1980, i.e., 30 to 50 million tons. The increased provision of liquid fueL through reduction of exports or through imports may well be economically justified to avoid choking economic development. Consideration shouLd also be given to developing a national fuel distribution system with stations open to all those vehicles entitled to procure fuel. This would increase the flexibility of trucking and eliminate the need for vehicles to carry fuel in drums when they travel beyond distances they can cover with a full tank. 22. Domestic water transport could also play a larger role. Average route density of freight traffic is only 36 million tons on coastal lines and 14 million tons on the Yangtze River. By comparison a modern, newly built single track rail line can carry up to 20 million tons p.a. and a double track line up to 90-100 million tons p.a. Coastal shipping offers an economical alternative to the congested North-South rail lines. In the South, the dense river and canal system can perform the role of roads in other areas. Passenger Transport 23. Despite the rapid growth of passenger traffic in recent years, the mobility of people in China is still very low; it reached 270 pkm per capita in 1982 versus 785 in India, where GNP per capita is lower than in China. The international comparison shows that personal mobility is strongly related to income. In addition, elasticity relative to income is also higher at lower than at higher income levels. Assuming an income elasticity of 1.5-1.7, well below those of India and Brazil, would mean that, by the year 2000, mobility in China would be 5 times larger than now. 24. Recent developments in rural areas have generated a vastly increased demand for short distance travel which has brought about a shift toward the road mode as well as a fall in average travel distance. Road now carries 70Z of the inter-city passengers but scill accounts for only 31Z of the passenger- km. It is likely that as income levels grow, the demand for greater mobility  - xi - will be accompanied by a demand for private vehicles as it has been in most other countries, including centrally planned economies. In many developing countries this demand has been satisfied at first by two wheelers such as motorcycles, scooters and mopeds. Personal transport could be accommodated without great problems in China's rural areas. It is only in large cities that low road density and lack of parking space would make intensive use of personal transport a critical problem. Whatever level private motorized transport eventually reaches in China, the need to build a good public transport service is imperative. Each mode of transport can contribute advantages. Bus transport can best serve rural areas and their conneccions with cities; rail should cater to medium distance inter-city travel; and air transport will serve the longer distance inter-city routes. Investments 25. China's annual investment in the transport sector is small in com- parison to that of other countries. At about 1.1% of GNP, it is comparable to the level of India. The pe-centage is about 1.4 in USSR, goes up to 2.0 in Korea and 3.3 in Brazil. L_ is much higher in developed countries when private vehicle purchases are included. Has this low investment in the sector made transport a bottleneck to economic development? A number of examples can be cited to support the view that it has. Last year, some 10 million tons of coal were stored in Shi;i for lack of transport; part was destroyed by spontaneous combustion. Many industries report a shortage of railcars, or at least a great difficulty to obtain sufficient cars when they need them. Production outside the plan cannot be transported by rail. Rural areas are short of transport means both for agriculture and local commune and brigade enterprises. Ports have been congested. The lack of bulk handling facilities is preventing the import of bulk fertilizer, while for grain, ship waiting time has exceeded 15 days at some ports, resulting in large demurrage charges. 26. There is no doubt that in the forthcoming decades, a much larger investment effort will be necessary in the transport sector to catch up with the backlog of present needs as well as accommodate a sharp increase in demand. Failure to do so may well affect the growth rate of the economy. 27. Railways. For railways, investments should be concentrated on increasing capacity in congested corridors east of the Beijing-Guangzhou line rather than in greatly expanding the network. Multiple tracking and electrification are likely to be the most efficient ways to increase capacity. Present work on railway routes out of Shanxi province should be sufficient to allow the lines to handle traffic until the late 1980s. Construction of a new double track line from Datong to Qinhuangdao for heavy unit trains is scheduled to start in 1985 and should be adequate to accommodete traffic in that corridor in the 1990s. Other lines from central and south Shanxi may also become necessary in the 1990s. Besides the east- west corridors out of Shanxi, capacity increases will also be needed in the north-south corridors from the Northeast to Shanghai and Guangzhou through the Beijing-Tianjin area. Rolling stock production should be stepped up greatly both for freight cars and passenger coaches. Present production is grossly inadequate. In recent years, freight car production has only been able to meet half the combined need for replacement and new additions. Present plans  - xii - for increasing passenger coach production will fall far short of needs. Electric and diesel locomotive produccion will also have to be stepped up. To supply short term needs imports may have to be considered for both rolling stock and locomotives, until local production capacity can be increased. For all these items, investments of Y 7 to 9 billion per year will be needed. 28. Roads. In the road subsector, the pace of construction has decreased considerably in the last few years; from 1979 to 1982 only 10,000 km of new roads were built annually. Until recently, most roads in China were under provincial and local management and financing. It is now time to attend to a national network and its financing. Improving or building some 4,000 km of crucial links in this network would cost about Y 2 billion. The upgrading of roads with high traffic volumes would cost Y 20-30 billion while building some 20,000 km of rural roads per year, which would roughly double the present rural road network by the year 2000, would cost about Y 4 billion per year. For road infrastructure alone, investments of Y 6-7 billion per year appear justified. In recent years road investment in China has amounted to less than a billion yuan. (Social labor contributions may not be included in this figure.) Financing mechanisms need to be developed at all levels of government to achieve the necessary extension of the network as well as the maintenance, strengthening and improvement of all classes of road. Truck and bus production will also need to be increased substantially. Present truck production would just cover renewals of the fleet on the basis of a 10-year life of vehicles. Since traffic is increasing, vehicles are kept well past their economic life, resulting in high cost of road transport. 29. Ports. In recent years much effort has been put into port development, and much more is necessary to modernize the ports as well as increase their capacity. The emphasis should be on bulk handling facilities and containerization to greatly reduce the turnaround time of ships in port. Estimated necessary port investment would be Y 2.0 billion per year. A necessary complement to port development is the inland distribution system, including intermodal transfer points and inland freight terminals. Dalian- Shenyang, Tianjin-Beijing and Shanghai-Nanjing are major transport corridors originating at 3 major ports. Integrated transport studies in these corridors are necessary to rationalize transport investments in the ports as well as in the modes serving the ports. 30. Multi-modal Transport. The combined use of rail and water may be particularly suitable for coal transport. However, this may not be economical on all routes. Careful analysis on a case by case basis should be performed before assigning traffic to a particular route and modal combination. For instance it appears that coastal shipping from northern ports will be more attractive than the use of the Yangtze River to deliver Shanxi coal to the Shanghai area. Northern ports are being developed for the shipment of coal; however, development of receiving ports in southern areas seems less advanced. The problem of receiving ports may not be so serious if consideration is given to simplifying unloading facilities and using self- unloading ships and floating terminals. Large coal users such as steel mills and power plants on coastal locations usually develop their own terminals. Using self-unloading ships can save costly port infrastructure such as berths, shore equipment and dredging. Floating terminals anchored off-shore or in  - xiii - estuaries can transfer loads to smaller ships and barges for final distribution through rivers and canals. In addition to ports development, investments will be needed on waterways and also in ships for coastal shipping, inland navigation and foreign trade. 31. Aviation. For passenger transport, besides the development of rail and road, major investments will be needed in civil aviation including modernized airports with air traffic control and landing facilities and new, Larger and more efficient aircrafts. 32. Combining the partial estimates above for railways, roads and ports, the total investments in the transport sector should be of the order of Y 15- 17 billion per year between now ax.d the year 2000. This does not include road vehicles, ships or any investments in inland waterways and civil aviation. With those included, annual transport investment could well exceed 2X of GDP which would appear very reasonable in comparison with other countries, but would be double the current percentage rate. Transport Tariffs 33. As in any other sector, tariff-setting is important for efficient use of the transport system. A series of costing studies would be a necessary basis for rational pricing of transport services. A costing study has now started for the railways, and some studies are underway in selected ports. However, prices in the transport sector cannot be changed without changing other prices too. For instance, the low price of natural resources such as iron ore greatly contributes to the movement of ore rather than iron or steeL. The price of cement produced by centrally controlled factories is so Low relative to the price of cement from locally controlled factories that the centrally controlled cement can be sold competitively as far as 2,000 km from the factory, well over the economical transport distance. A change in transport tariffs would not by itself control such anomalies. Cost conscious- ness is not yet prevalent in management as other criteria like plan ful- fillment are more important than profit making. For many products, transport cost accounts for only a small portion of the product's final value. Despite the fact that unit transport cost savings are small, they can add up to large sums when large volumes are involved; however they can have an impact only if the profit motive becomes a more important management tool. In the medium- term, administrative guidance would have, in many cases, to substitute for prices in attempts to rationalize the transport markets.  PART I TRANSPORT INDICATORS: CHINA AND THE INTERNATIONAL EXPERIENCE i. Part I compares China's transport indicators with those of other countries and explores prospects for future development of transport in China. The first chapter outlines the share of transport in the economy. Chapters 2 and 3 detail freight and passenger transport, their growth over time and in relation to GNP, their intensity, average distance and modal split. The fourth chapter discusses investments in the sector. ii. International comparisons of transport indicators are difficult since transport is country specific. The size and shape of the country, the extent of its coast line, the navigability of its lakes and rivers, the location of natural resources and population, the nature and composition of the economy all affect transport, particularly the movement of freight. Therefore, we have limited our systematic analysis to large countries, namely US, USSR, India, Brazil and Japan, with occasional reference to other countries on particular topics. 1. TRANSPORT IN THE ECONOMY 1.01 The share of transport in the net material product of China (NMP) has been stable around 4X since 1952. This share in gross domestic product (GDP) in 1981 is estimated to have been no more than 4.1% even with adjustment of distorted prices. This is lower than the share in other countries used in our comparison: Table 1.1: TRANSPORT AS PERCENTAGE OF GDP 1950-60 1960-70 1970-77 India 4.9 5.G 5.3 Korea 4.1 5.9 6.5 Brazil 6.0 5.6 5.2 Us 6.6 6.4 6.4 Japan 8.1 7.4 6.3 Source: World Tables, The Second Edition (1980), World Bank 1980. From the above table it appears that the transport share of GDP increases in countries at low levels of income (India, Korea) and decreases when income exceeds a certain level (Brazil, Japan, US). It is surprising that the share of transport in NMP has remained constant in China since transport has been  growing faster than NMP (Figure 1.1) From 1952 to 1981, NMP at constant prices increased over five times while freight tonnage increased over seven times and freight ton km over 11 times. Passenger traffic also increased more than 10 times. This wouLd indicate that transport costs have decreased in relation to other costs, which may be explained by the high intensity with which the network and the rolling stock is used (see para. 4.02). 1.02 China's transport sector is characterized by an unusually high freight intensity, higher than in any country except the USSR. Passenger intensity is more in line with other countries, but low relative to India and Brazil. The table below relates freight and passenger traffic to GNP in 1980- 81 for the various countries of our sample: Table 1.2: FREIGHT AND PASSENGER TRANSPORT INTENSITY Traffic ynit/ Tkm Pkm GNP Tkm/$GNP Pkm/$GNP $GNP al bln bln bln US$ China (81) 852 250 275 3.10 0.91 4.01 USSR (80) 6,021 891 1,393 4.32 0.64 4.96 US (81) 4,766 2,572 2,635 1.81 0.98 2.79 India (81) 266 542 159 1.67 3.41 5.08 Brazil (81) 343 450 245 1.40 1.84 3.24 Korea (81) 29 53 61 0.47 0.87 1.34 Japan (80) 439 634 1,071 0.41 0.59 1.00 a/ Traffic unit = tkm + pkm. Source: Appendix Tables B.2 and B.21 The apparent contradiction between the high freight transport intensity of the Chinese economy and the low contribution of transport to GDP can be explained by the dominance of rail transport in China (see para. 2.16). Rail transport is much less labor intensive than road transport and therefore has a much lower value added per ton km than road transport. In China road freight is seven times more labor intensive than rail while the value added is about three times that of rail. In the US road is about five times more labor intensive. The two following chapters analyze in more detail freight and passenger transport in China and other countries of our sample.  Figure 1.1: COMPARATIVE GROWTH OF NET MATERIAL PRODUCT, OUTPUT OF AGRICULTURE AND INDUSTRY AND TRANSPORT IN TONS AND TONKM CINDEX 1962=100) ,2200- - LIGHT&HEAVY INDUSTRIAL OUTPUT 2,000 - TTKM -- T.ONS 1.800- - -- NET MATERIAL PRODUCT GROSS AGRICULTURAL OUTPUT 1 4100-/ X 1, 2893 400- .~~ ~ ~ ~ ~~~~~ / . l .- "'S-~~~~~~~~~~- - - 1 ........... 1950 1955 1Q60 1985 1970 1975 1980 SOURCES, 1. STATISTICAL YEARBOOK OF CHINA, 1963. 2. CHINA, RECENT ECONOMIC TRENDS AND POLICY DEVELOPMENT ANNEX 2, TABLE 2.3.  -4- 2. FREIGHT TRAFFIC A. Freight Growth 2.01 Domestic freight transported in 1984 reached almost 1,100 billion ton-km (tkm) a 15-fold increase since 1952, or an average annual growth rate of almost 9%. This makes China the third largest country in the world in terms of freight transport (tkm) after the USSR and the US. Table 2.1: FREIGHT TRAFFIC VOLUME (billion ton-km) Road Domestic Pipe- Civil Ocean Year Rail /a waterway /b lines aviation Total Shipping 1952 60.2 1.4 11.8 - - 73.4 2.8 Modal spLit (X) 82.0 2 16 - - 100 - 1977 456.8 25.1 102.1 38.7 0.1 622.8 174.1 1978 534.5 27.4 129.2 43.0 0.1 734.2 248.7 1979 559.8 74.5 139.0 47.6 0.1 821.0 317.4 1980 571.7 76.4 152.3 49.1 0.1 849.6 353.2 1981 571.2 78.0 150.7 49.9 0.2 850.0 364.3 1982 612.0 94.9 170.8 50.1 0.2 928.0 376.9 1983 664.6 108.4 181.1 52.4 0.2 1,006.1 397.7 1984 724.7 118/c 198/c 57.2 0.3 1,098.2/c 435/c Modal split (Z) 66 11 18 5 - 100 /a From 1979 includes all road transport not only that done by road trans- port departments. /b Excludes ocean going transport which is often included in Chinese statis- tics. In 1979, coastal shipping accounted for 85 billion tkm and inland waterways for 54 billion tkm. The figures for 1982 were respectively 106 and 65 billion tkm. /c Estimates. Note: These data exclude transport by traditional means which is certainly sizeable in terms of tonnage but mostly on short distance. 2.02 For the period 1952 to 1982 in China, the growth of freight trans- port has been somewhat higher than the growth in the gross value of industrial and agriculturaL output (CVIAO) (see Figure 2.1). Freight transport rose sharply between 1952 and 1960, slowed down from 1960 to 1975, and accelerated again after 1976. For the purpose of international comparison freight growth has been related to CNP in constant terms, a more commonly available measure than GVIAO. A regression analysis has been carried out with GNP in constant 1979-81 US$ as the independent variable and freight tkm as the dependent variable (Appendix Table B.1).  Figure 2. 1: GVIAO AND TOTAL FREIGHT GROWTH TREND 1952-1982 - - GVIAO (IN BIL. YUAN) _ ----TOTAL FREIGHT (IN BIL. TON-KM) - - 700- 7 / .D / H~~~~~~~ / w / 600- e- 20 X1 1000 1952 1os6 1960 1964 1968 1972 1976 1980  -6- 2.03 Expected freight elasticity is generally higher in the earlier stages of development; from 1960 to 1981, both Brazil and Korea had elasti- cities over 1, while the US and Japan were below I (Figure 2.2). The table below shows freight elasticities relative to GNP in various countries. The very high elasticity for the USSR reflects the emphasis placed on development of resources in the country's far eastern region which greatly increased average transport distances, at a time of slower overall economic growth. Table 2.2: TOTAL FREIGHT (Bil. tkm) AND GNP (US$ Bil.): A CROSS-COUNTRY COMPARISON Years Elasticity China 1965-1981 1.034 US 1960-1981 0.941 USSR 1965-1980 1.427 India 1960-1981 0.921 Brazil 1960-1981 1.109 Japan 1960-1980 0.756 Korea 1961-1981 1.218 Note: The multicountry regression of freight transport (tkm) and GNP ($) gives the following equation: log of Freight Transport = -0.45 + 1.085 log of GNP (r square = 0.795) (see Figure 2.2). Further details are given in Appendix Table B.1 together with analyses of sub-periods within the 20 year spans. 2.04 It is difficult to predict freight elasticities for China. The attempt to deemphasize heavy industry combined with the campaign for indus- trial energy savings point toward lower freight elasticity. On the other hand, rationalization of industry, which may mean larger plants in some sectors and less regional self-sufficiency, could lead to higher freight elasticities. B. Freight Intensity 2.05 At first sight China's economy appears highly intensive in freight transport. Expressed in ton-km of freight per dollar of GNP, China's economy is almost twice as intensive as the US, India and Brazil and eight times as intensive as Japan and Korea, both smaller size countries. High transport intensity appears in most socialist economies; the intensity in the USSR is about 40% higher than that of China. There are a number of factors affecting transport intensity: country size, location of resources and population, composition of GNP, the level of processing of raw materials such as ores, lumber, agricultural products, and the degree of vertical integration of industry. The effect of each factor is discussed below on the basis of the international comparison. An analysis of provincial transport intensity in China is also added at the end of this section.  Figure 2.2: TOTAL FREIGHT AND GNP FOR SELECTED COUNTRIES, 1960-1981 X ~ ~~~~~~~~~~~~~~~~USSR> /. ~USA 1000 slC 100 I~~~~~~~~~~~~~~BAI 1.0 / KOREA 1ie 1 00 1 000 GNP (BIL. US*)  -8- Table 2.3: FREIGHT INTENSITY, 1980 Country TKM/$CNP China 3.17 USSR 4.32 US 1.87 India 1.67 Brazil 1.40 Korea 0.47 Japan 0.41 Note: See Appendix Table B.2 for details, time series and sources. 2.06 Country Size and Resource Location. Country size is an obvious factor affecting transport intensity measured in ton-km. This can be illu- strated by comparing freight transport in Japan versus that in the US for 1980. Table 2.4: TOTAL FREIGHT, 1980 Country area Tons Ton-km Average distance (million km) (millions) (billions) (km) Japan 0.38 5,985 439 73 uS 9.36 5,501 4,827 877 China 9.60 5,457 850 156 While Japan moves a higher tonnage than the US, the average distance is Less than one-tenth, reflecting not only Japan's relatively small size but also the concentration of its population along a limited portion of the eastern sea- board. Size and resource location are certainly factors in the high intensity of Soviet transport. While average distance for total Soviet freight transport is not available, the average distance for rail freight has increased from 800 km in 1965 to some 925 km in 1980; reflecting the need to exploit more distant resources in Siberia and further east particularly for coal, ores and timber. 2.07 The implications for China are again somewhat balanced. While the country is as large as the US, the population is far more concentrated. Over 70X of the population lives east of a Beijing-Guangzhou line, while in the US both the east and west coasts are highly developed. For China, this wouLd imply generally moderate transport distances within its densely populated areas, with a smaller proportion of freight moving to and from the distant western parts of the country. Regarding resources, the most common, such as  -9- coal, are spread throughout the country. However, the conditions for coal exploitation in the Shanxi-Nei Monggol area are more favorable than in the Northeast and Southwest. Concentrating on these resources will increase transport intensity more than would the development of local production closer to demand areas or the development of industry close to primary resources. The trade-offs are between higher mining costs and lower transport costs of regional mines. The same applies to feed grains and meat production. The former are produced more efficiently in the Northeast; the latter is produced more intensively in the South. The trade-offs are between shipping grain from north to south versus shifting some of the meat production co the north. 2.08 Composition of GNP. This is probably the major factor affecting transport intensity. Heavy industry is more transport intensive than light industry which in turn is more intensive than services. Therefore it should be expected that, all other things being equal as an economy diversifies, transport intensity decreases. Such a decrease is very noticeable in Japan where freight intensity decreased from 0.62 tkm/$GNP in 1960 to 0.41 in 1980 while the service sector increased from 422 to 53% of CNP. The case is some- what less clear in the US where a long-term decrease from 1.91 in 1960 to 1.69 in 1977 was somewhat reversed in 1981 (1.81). This can be explained by the very small changes which occurred in the sectoral composition of the US econ- omy between 1960 and 1980. The table below gives the broad sectoral dis- tribution of GDP in the countries of our sample. Table 2.5: PERCENTAGE DISTRIBUTION OF CDP Agriculture Industry Services 1960 1981 1960 1981 1960 1981 China 47 35 33 46 20 20 India 50 37 20 26 30 37 Korea 37 17 20 39 43 44 Brazil 16 13 35 34 49 53 Japan 13 4 45 43 42 53 US 4 3 38 34 58 63 Source: World Development Report 1983, Table 3 p. 152-153. China has the smallest service sector and the largest industrial sector among the countries included in Table 2.5. This is certainly a major factor in explaining the high freight intensity. Assuming that the 3ervice sector does nct generate any freight transport, the freight intensity related to total GNP of a country where the service sector is 20% (China) would be double that of a country where the service sector is 60% (US) with the same intensity of freight for the other sectors (agriculture and industry). The freight intensity of China is thus smaller than that of the US when the service sector is excluded, but still higher than that of India. The following table shows freight intensity related to the nonservice sectors of the economy.  - 10 - Table 2.6: FREIGHT INTENSITY, 1980-81 Freight Non-service intensity Tkm/$GNP sectors/GDP excluding service (1) (2) sector (1:2) China 3.17 0.80 3.96 US 1.87 0.37 5.05 India 1.67 0.63 2.65 Brazil 1.40 0.47 2.98 Korea 0.47 0.56 0.84 Japan 0.41 0.47 0.87 Source: Tables 2.3 and 2.5. 2.09 Processing of Raw Materials. Any weight-reducing processing done at the source of raw materials will reduce transport intensity. Coal, which is being transported in huge quantities in China, is a prime candidate for pre- transport treatment. The following international comparison speaks for itself. Table 2.7: COAL PREPARATION IN PERCENTAGE OF RAW COAL China Japan France Britain Germany USSR 18.0 94.7 92.5 88.3 87.4 63.4 Source: Ways to Improve Economic Benefit of the Coal Industry Enterprises by Li Shaoxun and Ji Zhongshi in Research on the Economics and Management of Technology No. 4, December 31, 1982, p. 52-56. In the case of the USSR, it is said that further coal preparation would save 20-25 million tons of transport per year; (out of a total moved by rail of over 700 million tons) and that ex essive humidity and rock content mobilize 200,000 extra rail cars per year.- In China a reduction in the volume of coal transported by rail of 5 to 10 would save transporting 20-40 million tons. 1/ Herve Gicquiau, "Une Crise Durable des Transports Interieurs de l'URSS", Le Courrier des Pays de L'Est no. 251, Mai 1981.  - 11 - 2.10 Other examples from the USSR 21 indicate that, every year, 11 mil- lion tons of fertilizer are transported which contain only half the normaL nutrient level. Also, bc4ause scrap steel is not weLl pressed and packed, it is estimated that 250,000 extra rail cars are needed to move the total volume of 58 million tons. Regarding timber transport, it is estimated that raw logs require 35-40Z more rolling stock than would sawn timber. By contrast, the table below indicates that in the US large tonnages of primary forest and wood raw materials move only a very short distance (93 km) while smaller quantities of sawmill products and plywood move distances which, for rail transport, are comparable to those in China and the USSR for timber (which includes both logs and sawn timber). This indicates that in the US processing generally occurs near the production sites. From the US figures above, it can also be seen that the share of rail increases as the average transport distance for the product becomes longer (Modal split is discussed in more detail below). Table 2.8: TRANSPORT OF TIMBER AND PRODUCTS Average Tons Z tkm distance million rail million km US (1977) Primary forest and wood raw materials 296 27,638 93 of which by rail 45 15.2 5,235 116 Sawmill products 61 36,346 596 of which by rail 16 26.2 22,302 1,394 Millwork plywood and prefab 26 18,885 726 of which by rail 8.8 33.8 13,933 1,583 China (1981) timber by rail 40 - 50,000 1,250 USSR (1979) timber by rail 145 - 242,600 1,673 Source: 1977 Census of Transportation, Commodity Transportation Survey Summary, US Department of Commerce, Bureau of the Census, June 1981. 2.11 The 1982 Almanac of China's economy mentions iron ore and phosphate rock where dressing could save transport. However, since detailed information 2/ Herve Gicquiau, op. cit.  - 12 - on raw material processing for products other than coal is lacking, it is not possible to estimate how much total transport demand could be reduced in China by processing and preparing before transport. It is probably safe to say, however, that savings could be substantial. 2.12 Vertical Integration of Industry. It is not clear to what extent this factor affects transport intensity in China, as industries tend to be more integrated than in other countries, that is, they manufacture all components in-house. If that is true, further rationalization of production may imply more transport per unit of output. For instance, if a refrigerator factory decides to buy its compressors rather than make them, transport will increase. However this may be more economical if the refrigerator factory would otherwise make compressors in uneconomic quantities. In the USSR, it appears that fragmentation in certain industries is leajdng to excess transport of parts and unfinished heavy steel products.- Also the vertical integration within Ministries results in purchases of components without regard for transport distances, i.e. ignoring a source for the product which may be nearer but under a different Ministry. 2.13 Freight Intensity by Province. Freight intensity measured in tkm/yuan GVIAO is above the national average in the North, Northeast and three southern provinces of Hunan, Guizhou and Guangxi (Appendix Table B.3 and Map A18284). WhiLe the data on tkm by province include all modes, the road and water transport totals allocated to the provinces amount to only one-third their national totals reported in global figures. For roads, provincial fig- ures cover only the transport bureaus under HOC; for waterways, it is not clear what is covered, but the provincial distribution probably excludes coastal shipping. If all water and road transport were recorded by province, the transport intensity of provinces in the Yangtze river basin would cer- tainly increase and Jiangxi and Hubei could exceed the national average also. 2.14 Most provinces with a transport intensity above the national average also have above average rail network density, both in terms of population and area (Map IBRD A18284). The same provinces also tend to have a higher proportion of heavy industry than provinces with lower transport intensity. The table below shows the provinces with above and below average transport intensity along with their rail density and the share of heavy and light industry. Only in Shaanxi, Henan, Anhui and Guangxi is the share of heavy industry lower than average while transport intensity is higher. This can easily be explained for Henan by its location at the intersection of two very important rail lines; the north-south Beijing - Guangzhou and the east-west Longhai lines. The same explanation, although to a lesser extent also applies to Shaanxi, Anhui and Guangxi, where transit traffic is probably higher than average. A regression analysis shows that there is a significant relationship between transport intensity in tkm/Yuan of GVIAO and the share of heavy indus- trial output in a province. The equation is Y = -0.335 + 2.544 x in which Y is tkm/Yuan of CVIAO and x is the heavy industry output share in percent of total industrial output (GVIO). Details are shown in Appendix Table B.3. 3/ Herve Cicquiau, op. cit.  - 13 - C. Average Transport Distance and Modal Split 2.15 References to transport distances were made in the previous section about transport intensity. Appendix Table B.4 gives average freight distances for China, US and Japan, the only countries for which information is availa- ble. For other countries, average distances are only available for rail freight. In China, average transport distance increased substantially from 159 km in 1949 to 382 km in 1982. The increase has not been gradual but in brief spurts folLowed by many years without change. As the economy recovered from the war, it increased rapidly to 249 km in 1951 and then remained around this level until 1962. A new surge occurred and the average reached 282 km in 1965, stagnating at this new level until 1976, probably as a resuLt of the policy of increased regional self-sufficency. From then, it has been increas- ing very rapidly, adding 30X in 5 years. The present average transport dist- ance in China is still much lower than that of the US, a country of equal size. The location of population and resources probably explain this lower figure (para. 2.07). Thus, average transport distance may well continue to increase in China with rationalization of industrial production and greater reliance on specialization of areas endowed with specific production advantages (coal in the north, feed grains in the northeast). 2.16 The dominance of rail in the sector is unique to China. Despite a reduction in share from over 80% in 1950 to 66% in 1984, rail in China still has the largest share of freight traffic of any country in the world. Con- versely, the use of road transport in China is among the lowest in the world. The only country with a lower road share is the USSR mainly because of the very large share handled by pipelines (32%).  - 14 - Table 2.9: FREIGHT TRANSPORT INTENSITY, RAIL NETWORK AND HEAVY/LIGHT INDUSTRY SHARE BY PROVINCE Transport Rail Density per Heavy/light intensity per area population industry tkm/Y GVIAO km/sq km km/000 pop. ratio North and Northwest Hebei 2.14 16.8 0.6 52/48 Inner Mongolia 2.39 4.0 2.3 55/45 Shanxi 1.32 13.8 0.8 68/32 Shaanxi 1.00 9.8 0.6 46/54 Gansu 1.66 5.8 1.2 77/23 Ningxia 1.99 6.7 1.1 - 67/37 Northeast Liaoning 1.26 25.2 1.1 64/36 Jilin 1.28 19.3 1.6 57/43 Heilongjia*g 1.28 10.8 1.5 66/34 Central Henan 1.50 22.5 0.5 45/55 Anhui 1.01 10.0 0.3 45/55 Southwest Guizhou 1.17 8.1 0.5 57/43 Guangxi 1.07 8.9 0.6 37/63 Hunan 0.97 12.2 0.5 53/40 NATIONAL AVERAGE 0.85 5.6 0.5 49/51 Central and South Shandong 0.56 11.7 0.2 44/56 Jiangsu 0.44 7.1 0.1 39/61 Zhejiang 0.55 8.3 0.2 35/65 Fujian 0.61 8.6 0.4 37/63 Jiangxi 0.83 8.6 0.4 49/51 Hubei 0.74 8.9 0.3 48/52 Guangdong 0.36 5.2 0.2 35/65 Sichuan 0.48 5.2 0.3 49/51 Note: Further etails are given in Appendix Tables B.3 and B.30. Source: Statistical Yearbook of China 1981.  - 15 - Table 2.10: MODAL SPLIT OF FREIGHT TRAFFIC /a IN Z OF TOTAL TKM Rail Road Water Pipeline China (1982) 66 10 18 5 India (1981) 63 33 4 - USSR (1982) 57 7 4 32 US (1981) 31 18 31 19 Brazil (1981) 23 60 13 3 Japan (1980) 9 41 51 /a Rounded figures. Further details, as well as time series and sources are given in Appendix Tables B.5 to B.10. 2.17 Considering tons rather than ton km gives a higher share to the road mode in China, but this is also typical of all other countries. Table 2.11 RAIL AND ROAD TONNAGE tail Road Road/rail -- million tons -- ratio China (1980) a/ 1,113 3,855 3.5 USSR (1979) 3,687 23,000 6.2 Korea (1982) 47 356 7.6 Japan (1980) 74 5,318 72.0 US (1977 industrial products only) 548 1,891 3.5 /a The road figure for China includes 760 million tons handled by the Highway Transport Departments and 3,095 million tons by own account trucks, according to surveys conducted by MOC. The above statistics may not be entirely comparable in coverage. Some deal with intercity traffic only (Korea, US); others like Japan probably include trips within urban areas. In Japan, almost 38% of the read tonnage consists of sand, gravel and stone moving an average distance of only 16 km. The figure for the US covers only industrial products and thus falls far short of total road transport which includes large quantities of primary materials moving short distances. The figures above confirm that modern road transport in China is still very smali. However, the total tonnage moved on roads is probably much higher when one considers all the traditional modes, including  - 16 - agricultural tractors which represent most of the traffic on many roads in China. 2.18 The burden placed on the Chinese railway by short distance traffic is well known. Hauls shorter than 100 km accounted for 23% of total rail freight traffic and amounted to 250 million tons in 1980; hauls shorter than 50 km accounted for 14% of rail traffic. The same problem appears to affect the USSR with its railway transporting 380 million tons (13%) within less than 50 km and 700 million tons (24% of total) within less than 100 km. In India, however, tonnage moved less than 100 km accounts for only 6% of total rail traffic while 50% of the traffic moves over 700 km. In China much of the short haul rail traffic happens in and around large metropolitan areas such as Beijing, Tianjin, Shanghai, and Taiyuan. For instance, 49% of the 30 million tons loaded in the Beijing urban area are moving over distances Less than 50 km. Percentages are 33% and 25% respectively for Tianjin and Taiyuan. Major commodities in this metropolitan rail transport are construction materiaLs, coal, ores, oil and steel generally moving from sidings to sidings (Appendix Table B.5). On a sample of main lines, short haul traffic ranges only from 0.5% to 11% of total traffic, except for some short movements of ore and coal between a mine and a plant on two of the lines (Appendix Table B.6). Therefore, little capacity could be saved by removing short haul traffic On trunk lines; but the situation would need to be studied in large metropolitan areas. 2.19 In almost all countries, the rail share of traffic has decreased markedly over the last 20-30 years (Appendix Tables B.7 to B.12). The mo-t striking case is Japan where the share of rail dropped from 51% in 1955 to 9% in 1980. The exception to this trend is Brazil which, however, does not have a major railway infrastructure and where the share of rail increased from 19% in 1960 to 23% in 1981. Most of the shift from the railways has been toward road transport, except in USSR where pipelines account for 32% of total trans- port. As illustrated by the international experience, there is a large poten- tial for the development of road transport in China. In India, road transport has proven more economical than rail for most commodity movements up to 200 to 250 kilometers. For some commodities, particularly perishable fruits and vegetables and small machinery, road transport has a comparative advantage over rail for distances up to 450 km (see para. 2.22). The average distance for road transport of industrial products in the US is 235 km, within the economic range indicated for India. In comparison, average road transport distance in China is a mere 35 km, up from 10 km in the 1950s. The average is apparently only 18 km in the USSR. Increasing the share of road transport in China will face a number of problems, i.e., the need for major development efforts on the road network; the production of vehicles, particularly fuel efficient ones; and the demand for liquid fuel, which could easily amount to 10 times the present level (a more detailed discussion of the last two topics is in the "Energy Consumption in the Transportation Sector" paper). 2.20 In Japan, the shift from rail has brought traffic not only to roads but also to coastal shipping which has replaced rail as the dominant mode and accounts now for 51% of the traffic (the share of rail in 1955). In 1980, coastal shipping in Japan moved 450 million tons over an average distance of 470 km (Appendix Table B.13). In comparison, the two transport bureaus under  - 17 - the Ministry of Communication in China moved only 50 million tons in 1979 albeit over distances exceeding 1,000 km. The figure may somewhat understate the total amount of coastal shipping as some will be done by agencies not under central authority. Nevertheless, there seems to be scope for further increasing coastal shipping in China. The distances for coastal shipping in Japan and China are very comparable. The distance between the new industrial port of Tomakomai in Hokkaido to Nagasaki in Kyushu is equivalent to that between Qingdao and Cuangzhou (Figure 2.3). In Japan, five commodities account for some 80% of the coastal shipping traffic: Table 2.12: JAPAN: COASTAL SHIPPING (Major Commodities in Z of Total) 1965 1980 tkm Z of tkm Z of Rank billion total Rank billion total Coal 1 23.7 29.3 5 12.3 5.8 Iron & steel 2 14.2 17.6 3 31.5 15.0 Petroleum products 3 11.8 14.7 1 58.3 27.7 Cement 4 6.3 7.8 4 20.6 9.8 Mon metallic minerals 5 6.1 7.6 2 46.7 22.2 Total 62.1 77.0 169.4 80.5 Note: See further details in Appendix Table B.13. Traffic for these five commodities increased about three times in 15 years or at an average annual rate of 7% p.a. In the case of China, coal, petroleum products and grain are very good candidates for increased transport by coastal shipping. 2.21 Modal split varies greatly from commodity to commodity. Typically, bulk commodities are more economically carried by either rail or water trans- port than by road, even over short distances. Indeed, in most countries, including China, about three quarters of all rail traffic is comprised of at most 10 commodities. As shown in Appendix Tables B.14 to B.18 these commodities are the same for all countries and their ranking is fairly stable. They consist of coal, petroleum, ores, timber, iron and steel, cement, construction materials, grain and fertilizer. 2.22 Modal split by commodity requires special studies and surveys which are not readily available in many countries. A survey conducted in India for interregional freight traffic, dividing India into 289 regions, shows the following modal split for various commodities in 1978-79.  - 18 - QINHUANGDAO BEIJING O / ~~~~~AN TIANJIN QINGDAO [' ~~~~~~NANJINGG 0~~~~~~~~ SHANGHAI / ~~To'kyo 01 OsakaV XIAMEN Nagasa z = Figure 2.3: go a -CHINA AND JAPAN COASTLINES: o.GUANGZITCU Yz S COMPARATIVE DISTANCES BETWEEN MAJOR PORTS  - 19 - Table 2.13 INDIA: INTERMODAL SHARE BY COMMODITY Rail Road --- share in Z --- Minerals, coal, iron and steel, salt, cement, mineral oils, fertilizers, stones and products of agriculture >80 <20 Electrical equipment, leather goods, hides, edible oils, chemicals and drugs, non-ferrous metals, livestock, tyres and tubes, machinery 40-60 40-60 Automobile and parts, milk and products, textiles, tea and coffee, paints and dyes, cycles and parts, fruits and vegetables, manufactured articles 4-30 70-96 Source: Report of the National Transport Policy Committee, May 1980, p. 24, see also Annex Table 2.19 for detailed figures. 2.23 As a further example, the table below shows the modal split for cement and fertilizer distribution in the US. Cement is distributed over rather short distances. Because of the ubiquity of the basic raw material, the industry locates near markets and, therefore, the most common transport mode is road (83% of the total). For fertilizer, the average distribution distance is almost three times that for cement. As a result, much larger proportions of the product go by rail and water than for cement. Table 2.14: US: CEMENT AND FERTILIZER DISTRIBUTION BY MODE US 1977 Cement Fertilizer tons Z tkm Av. tons X tkm Av. million million dist million million dist TotaL 86.8 100 15,114 174 47.9 100 22,782 476 Rail 7.9 9 2,440 309 21.2 44 15,614 737 Road Common 42.7 49 6,626 155 9.2 19 2,037 221 Private 29.1 34 3,435 118 9.3 19 1,309 141 Water 6.8 8 2,613 384 5.3 11 3,760 709 Other .3 - - - 2.9 6 - - Source: US Census of Transportation, 1977. Op. cit.  - 20 - 2.24 By comparison, the average transport distance of cement by rail in China is 395 km in 1981, substantially higher than in the US despite the fact that a higher percentage of the production moves by rail in China (30% versus 9% in the US). This long distance is somewhat puzzling since plant size in China is smaller than in the US and indeed much of the production occurs in very small plants. One possible explanation is that there may be cross hauls of cement as a result of the structure of administration. For the USSR, the data is only available for construction materials as a whole, including cement. The average rail transport distance is a very high 480 km. A com- parable average for China, adding cement and construction materials, would be only 245 km. Construction materials consisting of sand, gravel and bricks would generally be expected to move only very short distances, and primarily by road. In Japan, sand, gravel and stone account for almost 30% of the tonnage transported by road (1,492 million tons in 1980), but move on average only 16 km (Appendix Table B.20). 2.25 For fertilizer, the average rail distance in China, 731 km, is very comparable to that in the US (737 km) and both are considerably lower than averages for the USSR (1,079 km) and India (1,015 km). In average rail distance, the Location of the plants and their average size have a major impact on a particular country's transport pattern. Distribution studies for specific commodities have been done in many countries. Coal, cement, ferti- lizer, and grain are aLl potential candidates for similar studies in China in view of their dominance in the transport system. D. Port Traffic 2.26 Port traffic in China is still low, mainly as a result of the self- sufficiency policy and the little development of coastal shipping mentioned above. It is increasing rapidly, however, having grown over 50% since 1976, from 142 million tons to 219 million tons in 1981 for the centrally admin- istered coastal ports. The Government has recognized the importance of having adequate ports to handle the traffic generated by its change in trade policies and major capacity investments are taking place in many ports. The present five-year plan envisages the construction of 132 new berths by 1990. Looking at port development in other countries, it can only be expected that further large port investments will be needed in China over the next two decades to keep up with expanding trade and increase the use of coastal shipping. 2.27 Recent figures on total port traffic in selected countries are as follows:  - 21 - Table 2.15: PORT TRAFFIC IN SELECTED COUNTRIES (million tons) Total Foreign trade Domestic China 1982 238a/ 82 156 Japan 1979 2,885 815 1,225 US 1980 1,950 n.a. n.a. India 1978-79 major ports 70 n.a. n.a. Korea 1980 79 n.a. n.a. a/ Coastal ports only. In 1981 Changjiang river ports handled 87.9 million tons. n.a.: Not available. Japan underwent the most explosive growth in port traffic in recent years. Traffic increased 10 times between 1955 and 1975, from 250 million tons (about China's present level) to 2,500 million tons. This is equivalent to an average annual growth rate of over 12%. Over a twenty year period between 1960 and 1980, port traffic in the US more than doubled from 942 miilion tons to 1,950 million tons. This corresponds to an average growth race of 4% p.a. It is not surprising that this figure is lower than for Japan as the US economy was already more developed in 1960 and port traffic was then four times that of Japan. Also Japan imports most of its raw materials and energy. 2.28 Port traffic is related in part to foreign trade, in part to the development of coastal shipping, both of which have increased rapidly in Japan over the last 20 years. The table below shows the relative share of foreign trade in the GNP of selected countries. Table 2.16: FOREIGN TRADE COMPONENT OF GNP FOR SELECTED COUNTRIES (in % of GNP) 1960 1980 Exports Imports Exports Imports China 2.37 2.48 6.46 6.91 Brazil 10.14 11.66 8.84 11.04 US 5.85 6.75 10.23 11.22 Japan 5.44 9.55 13.78 14.68 Korea 2.59 10.34 37.70 44.78 India Source: World Bank EPD data base.  - 22 - 2.29 While the figure above may be lower than other estimates, China's foreign trade is still low by international standards. However, a continua- tion of the present economic policies is likely to foster foreign trade increase at a rate higher than that of the overall economy. Coastal shipping may also grow faster than overall economic growth. Combined, these two factors will generate a growth in port traffic which may well be 50% higher than the growth in CNP, i.e., between 10 and 12% p.a. This would mean that port traffic in the year 2000 could be five times the present level. While some rationalization may be possible to reduce transshipments between seaward and inland waterway traffic, traffic demand would still imply the need for very large port investments to avoid bottlenecks that would stifle economic development.  - 23 - 3. PASSENCER TRAFFIC A. Growth of Passenger Traffic 3.01 Passenger traffic reached 274 billion passenger-km (pkm) in 1982, an 11-fold increase since 1952 and an average annual growth rate above 8%. Growth of passenger traffic has been more erratic than that of freight. For instance, the traffic levels reached in the early 1960s were not achieved again until the early 1970s. Since 1978 however, the growth has averaged an impressive 12% p.a., well over twice the overall rate of economic growth (Figure 3.1, Appendix Table B.21). This illustrates the potential demand for travel as income grows. It is likely that this growth would have been even higher had there not been limited capacity in passenger services. Table 3.1: PASSENGER TRAFFIC VOLUME (billion passenger-km) Year Rail Road Waterways Aviation Total 1952 20.1 2.3 2.5 - 24.9 Modal split Z 81 9 10 - 100 1977 102.3 44.8 9.8 1.8 158.7 Modal split X 64 28 7 1 - 1979 121.6 60.3 11.4 3.5 196.8 1980 138.3 72.9 12.9 4.0 228.1 1981 147.3 83.9 13.8 5.0 250.0 1982 157.5 96.4 14.5 6.0 274.4 1983 177.6 110.6 15.4 5.9 309.5 1984 204.6 129.4 15.2 8.4 357.6 Modal split Z 57 36 4 2 100 Growth rates Z p.a. 1952-77 6.7 12.6 5.6 - 7.7 1979-84 11.0 16.5 5.9 19.2 12.7 1983-84 15.2 17.0 - 42.4 15.5  Figure 3.1: PASSENGER TRAFFIC SELECTED YEARS, 1952-1982 CIN BMILION PASS-KM) 300e ~2e0 is- CD z w lee Ce, 1b9?. 1950 1983 1984 1938 1972 1978 1980  - 25 - 3.02 Despite the rapid growth in passenger traffic, the mobility of people in China is still very low; it reached 270 pkm per capita in 1982 versus 785 in India where CNP per capita is lower than in China. The table below shows the mobility and GNP per capita in the countries used as com- parators. Table 3.2: INTERCOUNTRY COMPARISON OF PASSENGER TRAFFIC PER CAPITA AND GNP PER CAPITA Year Pkm/capita GNP/capita CHINA 1981 252 278 India 1981 785 232 Korea 1981 1,368 1,576 USSR 1980 3,356 5,244 Brazil 1981 3,735 2,027 Japan 1980 5,416 9,173 US 1981 11,193 11,465 Note: The multicountry regression of mobility (pkm/capita) versus income (GNP/capita) gives the following equation: Mobility = -26.62 + 0.865 Income (r square = 0.91) (see Figure 3.2). (Although the best fit is for natural numbers, the graph is in log scale for presentation purposes.) Further details as well as time series and sources are given in Appendix Table B.21. The low mobility in China most likely results from demand restraint combined with inadequate supply of passenger transport capacity. Demand restraint is partially a consequence of the comparatively low share of CNP ',o:ng to house- hold consumption, perhaps 50% in China as compared with some 70X in India. The mobility in the USSR is also low; it is just Wlow that of Brazil where per capita income is only half that of the USSR. - The comparison of mobility across economies indicates that it is generally lower in centrally planned economies than in market economies (Figure 3.2). 3.03 As indicated above, personal mobility is strongly rel.ed to income levels. Elasticity relative to income is also higher at lower income levels than at higher ones. Countries at low levels of income show increasing elas- ticities as income grows (China, India, Brazil, Korea) while higher income countries have decreasing elasticity (US, Japan). Detailed elasticity figures derived through regression analysis are given in Table 3.4 for seven 4/ USSR statistics do not include travel by private car, which has been estimated to add about 10% to the total pkm (Soviet Economy in the 1980's: Problems and Prospects Part 1, Joint Economic Committee Congress of the United States, December 31, 1982).  Figure 3.2: INTERCOUNTRY COMPARISON OF PASSENGER TRAFFIC AND GNP PER CAPITA USA U. CL W a.