Bolivia national transport study (Vol. 1 of 5) : Existing conditions and forecasts

61812 V1 , BOLIVIA NATIdNAL TRANSPORT STUDY 'FINAL REPORT VOLUME ONe EXISTING CONDITIONS AND FORECASTS PREPARED FOR t:;OVERNMENT OF THE REPUBLIC OF BOLIVIA tl"JHh~!S,l'~':lV OF TRAhtSPORT ANO COMMUNICATIONS) nl,¥T'E'1i;;:'~'\.I,r.;";'rg:"11",.B:'.l.<. ~~.~,Ii\H< ~~!~")~ ReC·ONS'T'~UC;nON AI'\ID CEVELOPMENT eJN!TEO NAT'!(.U\.IS OE\lr:;LO~MeNT PROGRAMME PREPAREOBV ~""A ~~.t~9",~ "' i;-:-~~:JL~'\llr~ '11."0'18'  BOLIVIA NATIONAL TRANSPORT STUDY FINAL REPORT VOLUME ONE EXISTING CONDITIONS AND FORECASTS PREPARED FOR GOVERNMENT OF THE REPUBLIC OF BOLIVIA (MINISTRV OF TRANSPORT ANO COMMUNICATIONS) INTERNATIONAL BANK FOR RECONSTRUCTION ANO OEVELOPMENT UNITEO NATIONS OEVELOPMENT PROGRAMME PREPAREDBV WILBUR SMITH AND ASSOCIATES IN ASSOCIATION WITH CONSA SRL & ECOVIANA SRL LA PAZ. BOLIVIA ..JULV 18B1  BANKERS TRUST TOWER CABLE WILSMITH P. O. BOX 92 TELEX 57-3439 COLUMBIA, S. C. 29202 July, 1981 PHONE (803) 771·8844 Senor Cnl DEMA Rene Guzman Fortun Mr_ James B. Newman Minister of Transport and Ports, Railways and Aviation Division Communications Latin America and Caribbean Regional La Paz Office Bolivia The World Bank Washington, D.C. U.S.A. Gentlemen: Wilbur Smith and Associates is pleased to submit this Final Report of Phase I of the Bolivia National Transport Study. The Report is in five volumes. Volumes I and II review the existing situation and present the main analyses, findings and recommendations. Volumes III and IV contain supplementary appendices and the fifth volume is a summary of the Report. The Report represents the conclusion of over two years effort by a professional team made up of consultant's staff and Bolivian nationals. During the course of the Study, several interim reports and technical papers were produced, culminating in a draft Final Report in November, 1980. The first review period was concluded in December with the final meeting of the Study Steering Committee, when all initial comments on the draft report were pre- sented by Bolivian entities and by the World Bank. A more lengthy review and discussion period followed in which additional valuable contributions were made to the Final Report. The Report presents and supports a recommended National Transport Plan set in the con- text of the expected development of the Bolivian economy over the next ten years, and with consideration for possible developments in the following ten years. It is not intended that the recommendations be taken as a fixed plan; forecasting the future is a hazardous task, and it is impossible to foresee all possible developments. Therefore, it is strongly recommended that the Government set up a permanent transport planning group, utilizing the Bolivian counterpart team involved in this Study,to monitor developments and adapt the Plan to changing circumstances. The first steps have been taken in this direction by selecting a counterpart team for Phase II of this Study. Hopefully, this will form a solid basis for the continuing transport planning process, and for a permanent National Transport Center_ The cooperation and assistance we received throughout the Study from officers of the Government and others concerned with transport in Bolivia, is gratefully acknowledged. We also wish to recognize the contributions made by our colleagues on the national counter- part team, and by our local associate consultants, CONSA SRL and ECOVIANA SRL. We appreci- ate the opportunity to have participated in the study effort, and look forward to the imp- lementation of the recommendations. Respectfully submitted WILBUR SMITH AND ASSOCIATES ~r~ ~~~~ma~' tf Smi ALLIANCE, OH - AMMAN, JORDAN - BRISBANE - CAMDEN, NJ- CHARLESTON, SC- COLUMBIA, SC- FALLS CHURCH, VA - HONG KONG - HOUSTON KUALA LUMPUR, MALA YSIA - KNOXVILLE - LA PAZ, BOLIVIA - LEXINGTON, KY - LOS ANGELES - MELBOURNE - MIAMI· NEW HAVEN NEW YORK· PITTSBURGH - RALEIGH, NC - RICHMOND - .SAN FRANCISCO - SANTA CRUZ, BOLIVIA - SINGAPORE - TORONTO - WASHINGTON, DC  TABLE OF CONTENTS Chapter 1 INTRODUCTION 1 Background of Study 1 Purpose and Scope of Study 2 Working Papers and Earlier Reports 4 Structure of Final Report 5 2 OBJECTIVES OF TRANSPORT POLICY 7 Use of Existing Infrastructure 8 Regulation of Operators 10 Improvement of Transport Infrastructure 11 3 DATA COLLECTION AND STUDY METHODOLOGY 13 Data Collection 15 Transport Model System 19 Project Evaluation 27 HDM Model and Highway Analysis 31 Budget Constraints 33 4 EXISTING TRANSPORT SYSTEM 35 Railroads 35 Roads and Road Transport 39 River and Lake Transport 44 Transport by Pipeline 47 Air Transport 48 Overall View 50 5 TRANSPORT COSTS AND TARIFFS 53 Treatment of Costs 54 Railroad Costs and Tariffs 57 Road Transport Costs and Tariffs 64 River Costs and Tariffs 72 Air Transport Costs and Tariffs 77 Pipeline Costs 84 6 THE FINANCIAL FRAMEWORK OF THE TRANSPORT MODES 87 The Roads 87 The Railroads 91 River Transport 93 Air Transport 97 -i-  TABLE OF CONTENTS (Cont. ) Chapter 7 THE BOLIVIAN ECONOMY 99 Socio-economic Factors Affecting Transport Demand 99 The Bolivian Economy in 1980 100 The Future of the Economy 114 Alternative Development Patterns 121 Quantification of Development Projects 126 Socio-economic Characteristics in 1989 and 1999 126 8 RESOURCES AVAILABLE FOR TRANSPORT INVESTMENT 133 Discussion of the Problem 133 Past Investments 135 Projection 136 Conclusion 142 9 EXISTING AND FUTURE TRANSPORT DEMANDS 143 Freight Transport Demand 143 Freight Transport within Bolivia 144 Freight Transport between Bolivia and Neighboring Countries 151 International Transit through Bolivia 155 Passenger Travel 157 Transport Demand in Major Corridors 160 10 TRANSPORT PROJECTS SELECTED FOR EVALUATION 163 Deficiencies of the Transport System 163 Committed Projects 168 Candidate Projects 170 The Computer Transport Networks 179 -ii- ..  CONTENTS OF VOLUME TWO Chapter 11' Improvement and Maintenance of Existing Highways 12 New Penetration Roads and Other Transport Links in Remote Areas 13 Rail Transport 14 Air Transport 15 Improvement to International Trans- port Links 16 Organizational Aspects of Transport 17 Future Tariff Policy and Financial Prospects 18 Investment Program 19 Future Transport Planning 20 Conclusions and Recommendations TABLES Number 4-1 Length of Roads by Districts and by Type of Road Surface 40 4-2 River Navigation System in Bolivia - Amazon River Basin 45 5-1 Total Railroad Costs, 1977 58 5-2 Costs of Passenger Transport by Train, 1977 59 5-3 Costs of Goods Transport by Rail, 1977 60 5-4 Impact of Rehabilitation Policies on Train Operating Costs and Track Maintenance Costs 63 5-5 Costs of Transport by Road 66 5-6 Vehicle Parameters Adopted for Cost Equations, 1977 68 5-7 Vehicle Cost Parameters, 1977, 1989 and 1999 70 5-8 Unit Costs of Road Maintenance Operations 71 5-9 Tariffs for Trucks and Buses, 1977 73 5-10 Cost of Transport on Ichilo-Mamore River System 75 5-11 Costs of Scheduled Air Transport, 1978 78 5-12 Costs of Non-scheduled Air Transport, 1978 80 5-13 Airport Costs by Region and Function, 1979 82 5-14 Financial Transport Costs by Pipeline, 1977 85 -iii-  ,", TABLES (continued) '", Number 6-1 Expenditure and Income for Road construction and Maintenance, 1976-79 88 6-2 Road User Charges, 1976-79 89 6-3 Expenditure and Income of National Railroad Company, 1976-79 92 6-4 Expenditure and Income of Lloyd Aereo Boliviano, 1978-79 95 6-5 Expenditure and Income of AASANA 97 7-1 Trend of Gross Domestic Product, 1970-80 104 7-2 GOP by Expenditure, 1970, 1977 and 1978 105 7-3 GOP by Industrial Sector: Averages for 1970-76 107 7-4 Agricultural Production, 1978 110 ,", 7-5 Value of Exports, 1970, 1977 and 1978 112 7-6 Production of Ten Principal Minerals 113 7-7 Production of Hydrocarbons, 1970, 1977 and 1978 115 7-8 Industrial Production, 1970, 1977 and 1978 116 7-9 Summary of Socio-economic Variables - 1977 and Projections for 1989 and 1999 128 7-10 Supply and Demand of Commodity Groups - 1977 and Projections for 1989 and 1999 129 7-11 Production, Consumption, Imports and Exports: 1977 and Projections for 1989 and 1999 131 8-1 Transport Investment, 1973 - 1979 137 8-2 Transport Investment and National Resources, 1973-1979 138 8-3 Transport Investment and National Resources, 1980-2000,Development Pattern C 140 8-4 Transport Investment and National Resources, 1980-2000,Development Patterns A and 0 141 9-1 Existing and Predicted Internal Freight Tonnages 145 9-2 Internal Ton-Kilometers by Mode 146 9-3 Internal Co~modity Movements by Road and Rail 147 9-4 Observed and Predicted Import Tonnages by Major Mode 152 9-5 Observed and Projected Export Tonnages by Major Mode 153 9-6 International Transit Traffic on Bolivian Rail System, 1979 156 9-7 Passenger Travel, 1977 158 10-1 Candidate Rail Projects 172 10-2 Candidate Road Projects 174 10-3 Candidate River Projects 176 10-4 Miscellaneous Candidate Projects 178 -iv-  FIGURES Number Following Page 3-1 Roadside Interview Stations 15 3-2 Analysis Zones 20 3-3 Freight Transport Models 21 3-4 .Hodal Allocation Diversion Curves 24 3-5 Highway Transport Cost System 32 4-1 Existing Railroad Network 36 4-2 Trends in Rail Transport 37 4-3 Existing Network of Primary and Secondary Roads 38 4-4 Navigable Waterways 44 4-5 Existing Pipelines 47 4-6 1980 Domestic Air Passenger Services 49 4-7 Accessibility for Ground Transport 50 5-1 Price Indices at Current Rates: 1960 to 1979 55 5-2 Railroad Costs Related to Traffic Density 61 5-3 Normal Functions Relating Inter- provincial Truck Tariffs to Distance 72 7-1 Main Population Centers 102 7-2 The Mining Bel t 111 7-3 Oil and Gas Fields 113 7-4 1989 Development Pattern 122 7-5 1999 Development Pattern A 123 7-6 1999 Development Patterns C and D 125 9-1 1977 Daily Road Vehicle Movements 160 9-2 1989 Daily Road vehicle Movements 160 9-3 1999 Daily Road Vehicle Movements 161 9-4 1977 Annual Tonnage Flows by Rail 161 9-5 1989 Annual Tonnage Flows by Rail 161 9-6 1999 Annual Tonnage Flows by Rail 162 10-1 Committed Transport Projects 169 10-2 Candidate Rail PrOJects 171 10-3 Candidate Road Projects 173 10-4 Candidate Airport Projects 175 10-5 Candidate River Projects 175 10-6 Miscellaneous Candidate Projects 177 10-7 The 1989 Test Network: Candidate Projects 180 10-8 The 1999 Test Network: Additions to the 1989 Test Network 181 -v-  ,.., APPENDICES Volume 3 lA Available Working Papers 3A Roadside Interview Surveys 3B Collection of Other Transport Data 3C Socio-Economic Data Collection •• 3D Definition of Analysis Zones 3E Definition of Commodity Classi- fications 3F The Transport Model 3G Case Study of Project Evaluation 3H Evaluation of New Modes on Existing Links 4A Existing Characteristics of Rail- road Transport 4B Principal Results of Roadside Origin-Destination Survey 4C Existing Characteristics of Waterborne Transport 4D Existing Pipeline Characteristics 4E Existing Characteristics of Air Transport 5A Costs of Railroad Transport 5B Costs Related to Road Transport 5C Costs of River Transport 5D Costs of Air Transport 5E Other Aspects of Transport Costs 6A Road User Charges 6B Financial Analysis of AASANA 6C Financial Analysis of Rail Transport 4 7A Overall View of the Bolivian Economy 7B Definition of Alternative Development Patterns 7C Forecast and Distribution of Socio-Economic Variables 7D Socio-Economic Projections by Zone 9A Internal Traffic Projections 9B Forecasts of Foreign Trade Movements by r-1ode 9C Potential Transit Traffic Through Bolivia -vi-  APPENDICES (continued) Volume 4 IDA List of Candiate Projects llA Vehicle Axle Load Measurement Program lIB Costs of Betterment of Unpaved Roads llC Highway Design Standards lID Highway Capacity and Congestion lIE Summary of Evaluations of Unpaved Roads llF Bridge Analysis llG '" Seco-Desaguadero Highway The R~o l2A Analysis of New Penetration and Feeder Roads l2B Evaluation of the Santa Barbara-Bella Vista Highway l2C Evaluation of the Highway Eteramazama- San Ignacio de Moxos l2D Evaluation of the Ichilo-Mamore River System Cleaning Project l2E Evaluation of the Railroad from Rio Grande to Trinidad l3A Assumptions on the Costs of Railroad Rehabilitation Projects l3B Evaluation of Railroad Rehabilitation Projects l6A Analysis of Regulations on Road Transport l6B Organizational Structure and Regulation of Water Transport 17A Projections of ENFE's Freight Operating Costs and Revenues leA Committed Highway Investments -vii-  CHAPTER 1 INTRODUCTION CHAPTER 1 INTRODUCTION In Bolivia's vast territory, problems of access and communication have always been particularly acute. Large, potentially productive regions remain unexploited, partly because access to them is impossible or very difficult. Where ways of communication do exist, they have often been built to a low standard with the result that transport costs are relatively high. There is no lack of ideas on how to improve the coun- try's transport system, and many projects have been pro- posed, some of which have been the subject of detailed feasibility studies. Many of them, however, have never become a reality, usually because there were no funds available. Sometimes, the scarce funds were used on projects of doubtful benefit, thus preventing more deserving projects from being built. Thus, the problem is not only to conceive "good" transport projects, but to establish the "right" priorities for them. This Study aims to resolve the problem by identifying those transport investments which are likely to be of greatest benefit to the Bolivian economy. It also examines the organizational and financial aspects of transport. The resulting'proposals, together with the suggested investment program, offer a new transport strategy recommended for the 1980's. The government, for reasons beyond the scope of this technical study, may opt for certain solutions not recommended in this Report. Should this be the case, however, the Study's analyses will still be useful in quantifying the economic costs of those solutions. Background of Study The last major transport study in Bolivia was completed in the late 1960's and gave valuable guidance on transport investment and organization (1). There have been many changes since then, both with regard to economic development and transport characteristics, and it was therefore decided to conduct another comprehensive analysis. (1) Estudio Inte ra1 de los Trans ortes en Bolivia (Bolivia Natl.ona1 Transport Study , prepared by Daniel, Mann, Johnson and Mendenhall; Stanford Research Institute; and Alan M. Voorhees and Associates for the Bolivian Government, the World Bank and UNDP, 1968. -1-  In October 1974, an agreement was signed between the Bolivian Government and the United Nations Development Programme to carry out such a study. Two phases were specified. The first was to recommend a transport policy and, among other things, to identify investment projects suitable for international funding agencies such as the World Bank. The second phase was to consist of technical assistance to the Bolivian Government in implementing the recommendations of Phase I. The World Bank (International Bank for Reconstruction and Development - IBRD) acted as Executing Agency for this Project. In October 1978, it contracted Wilbur Smith and Associates to carry out Phase I, and the Study commenced in November 1978. The analyses and findings are described in this Report. Purpose and Scope of Study The work to be covered by the Study was detailed in the Terms of Reference prepared by the World Bank in consultation with the Bolivian Government. The basic objectives were defined as follows: "The objectives of the transport survey (Phase I) are to provide the Government of Bolivia with a National Transport Master Plan for the development of a coordinated transportation system; to define the appropriate governmental and investment policies for each mode of transport, taking into account all ,~ existing modes in use in the country as well as those that might become feasible during the planning period; and to help strenghten DPC under MTCCA (2). The Master Plan shall take into account the country's future economic and financial capability. To achieve these objectives, the survey shall be carried out under these Terms of Reference and shall consist of: (i) a review of recent studies; (ii) the collection of all such data and any other information deemed necessary to prepare a master plan for the development of a coordi- nated transportation system, and the identification of major areas where data are insufficient; (iii) reco~mendations for the strengthening of the DPC under the MTCCA, which is responsible for =====----- (2) Directorate of Planning and Coordination under the Ministry of Transport, Communications and Civil Aeronautics. -2-  the execution of intermodal transport studies and the preparation of transport sector input to the five-year plani including recommendations on its permanent staffing; (iv) the identification and analysis of principal problem areas in the transport sector. The presentation of recommendations for procedures and methodologies for solving the problems, giving consideration to the technical, operational, planning, economic, coordination, organizational, administrative, managerial, financial, and training aspects of such problemsi (v) a detailed highway maintenance program for 1980-1989 (3); (vi) a detailed transport investment and financial program for 1980-1989 (3); (vii) an indicative transport investment and financial program for the period 1990-1999 (3); (viii) reco~mendations for the improvement, if required, of the operations, planning, organization, administration and management of each transport mode; and (ix) recommendations for the improvement, if required, of overall transport policies (including regulatory measures and transport pricing) and coordination among modes." The third of the above items refers to future transport planning. Special emphasis was to be placed on the creation of a group which would not only implement the Study's recommendations in Phase II but also update its analyses and give advice to government on transport matters in the years thereafter. With this in mind, particular care had been exercised in selecting a qualified and experienced counterpart team of Bolivian professionals, and these were closely integrated with the Consultants' team to participate in the analyses and their presentation. Through this direct contribution to the analyses under the Consultants' direction, it is felt that an effective way was found to transfer the transport planning technology to the counterpart team. (3) During the course of the Study, these periods were modified to 1981-1990 and 1991-2000, respectively. -3- Working Papers and Earlier Reports Apart from this Final Report, the Terms of Reference stipulated the following four reports to be produced by the Study: • Inception Report • Progress Report • First Interim Report • Second Interim Report. The Inception Report, published in December 1978, presented a detailed work program for the Study, including the basic definition of the transport model and of the various surveys designed to provide the necessary data. This was complemented in September 1979 by the Progress Report, which elaborated the economic methods to be used in the Study. It also reported on the results of a Regional Planning Workshop held by the Study in June 1979 under the auspices of the Ministry of Planning and Coordi- nation and the United Nations Development Programme. The purpose of the workshop was to identify the possible directions that economic development might take in Bolivia, as a basis for the prediction of future transport demands. The First Interim Report was published in December 1979. It described the various surveys conducted by the Study and the socio-economic projections which form the basis of the traffic forecasts for 1989 and 1999. It also discussed the existing characteristics of each transport mode. The Second Interim Report, published in July 1980, dealt primarily with the development and initial application of the transport model, but also discussed the program of project evaluation. Working Papers - In a study of this Size, numerous indivIdual analyses are made which contribute to the ultimate findings. The calculations and results of these analyses normally remain in files, inaccessible to persons outside the Study team who wish to investigate them. In an attempt to document the many individual Study analyses, approximately 60 working papers were published, containing descriptions, calculations, tables and, in some instances, computer printouts. Depending on the author, these papers were written in either English or Spanish, but most of the English texts were then translated into Spanish. -4-  LA PAZ - COTAPATA (LA CUMBRE) ZONGO ROAD LA PAZ - COTAPATA (LA CUMBRE) ROADS IN MOUNTAINOUS TERRAIN  The extent of these documents far exceeds what could be covered in the Final Report, even in the Appendices which have been taken mainly from the working papers. Appendix lA lists those working papers that were published before the completion date of the Final Report. Structure of Final Report The Final Report aims to give a concise description of the Study's criteria, evaluations and conclusions. The appendices contain the most important background data and a fairly detailed description of the computer transport model. Throughout the Final Report, reference is made to the working papers which contain the detailed background data and analyses underlying the conclusions of the Study. The Report commences with a discussion of the basic objectives of transport policy, followed by a chapter on the Study methodology. The next chapters examine various aspects of the transport system and the Bolivian economy. Chapters 10 through 15 describe the evaluations made of candidate projects, and these are followed by analyses of the organizational and financial aspects of providing trans- port services. Chapter 18 presents the transport invest- ment program for the 1980s and the outline program for the 1990s. This is followed by the last two chapters which con- tain a discussion of future transport planning in Bolivia and a summary of the Study's conclusions and recommendations. -5-  CHAPTER 2 OBJECTIVES OF TRANSPORT POLICY CHAPTER 2 OBJECTIVIES OF TRANSPORT POLICY Transport serves seven basic functions, which are briefly discussed as follows: (1) Exploitation of Geographical Differences - Transport enables the benefits of particular geographical situations to be shared by people living in other situations. Few countries have such a variety of geographical situations as has Bolivia. Almost every type of crop can be grown, almost every kind of mineral found, but transport is needed to exploit these rich possibilities. Much of the natural wealth of the country remains untouched and untouchable for lack of adequate transport. (2) Specialization - One of the keys to economic development is specialization, which is only possible if there is a wide enough market. This often requires transport. In particular, the advance from subsistence to commercial far~ing can only be made if there is transport to a suitable market. Similarly, the successful establishment of industry in La Paz, Cochabarnba and Santa Cruz will be much easier if the three cities form effectively a single market by virtue of good transport connections. (3) Other Economies of Scale - Specialization is only one way in which economies of scale can be achieved. Cheaper transport widens the market for both outputs and inputs and thus makes possible larger-scale units of production in both industry and agriculture, with resulting economies of scale. This is of great importance for the development of Bolivian agriculture and processing industries. (4) Location of POEulation - An efficient and desirable distribution of population is not possible without transport. Bolivia is a large country with a small population, but the majority of its people live on the Altiplano with little opportunity for agricultural expansion. Enormous areas of land are totally unused and unusable for lack of access. A large part of the population lives on poor land while elsewhere better land lies idle, partly for lack of transport. -7-  ". (5) Social Purposes - The development of social activities and, in particular, the spread of ideas depend on transport of people between different parts of the country, and to and from other countries. Passenger transport between different departments of Bolivia is relatively difficult or impossible, except by plane, and is therefore ,largely confined to a small minority because of the high cost involved. (6) Culture and Recreation - The higher levels of culture and recreation depend on transport to bring people together. There are great opportunities for recreational activities and hence tourism in Bolivia, notably in the Andes, but they are largely untouched, partly for lack of transport. (7) Political and Military Purposes - Government control and national integration require a certain minimum of transport facilities for passengers and goods. Such facilities are lacking in some parts of the country, especially in Pando, Beni and the eastern part of Santa Cruz department. These are the basic ways in which transport can enable the community to use its territory productively and organize itself effectively. The aim of transport policy should be to develop and improve the transport system so that it can fulfill these functions more efficiently. However, the priorities among the objectives must be established in the country's political process. This report focusses therefore on quantifying aspects of transport economics and regional development, with reference being made to other decision criteria where appro- priate. Use of Existing Infrastructure The first task of transport policy is to use existing transport facilities to best advantage. This implies: An Economic Maintenance Policy - The method and frequency of maintenance should be calculated to minimize the overall cost of transport, i.e. the sum of vehicle operating, track maintenance and investment costs, on each link of the network. On both roads and railroads in Bolivia, maintenance policies and in particular the balance between maintenance expenditures and investment in new infrastructure have not been decided very scientifically and could be improved. Optimal Pricing Policies - The volume of traffic and the choice of mode are influenced by tariffs. High tariffs may impede economic development. In Bolivia, where distances -8- are long and transport is inevitably costly, tariff policy could help or hinder the development of the country and, by attracting traffic to the modes with lowest marginal costs, could help to reduce the average costs of transport to the community. Current tariffs are largely the product of a policy of setting prices equal to average costs. Theoretically there are advantages in setting them equal to marginal costs, although in practice there may be difficulties in identifying the marginal costs and in finding an acceptable method of charging them. Despite the difficulties, however, it is especially important in Bolivia, where there is ample spare capacity at all times on practically the entire transport infrastructure, not to suppress traffic and the activity that it reflects by tariffs needlessly higher than the marginal costs attributable to that traffic. Since tariff policy influences the volume of traffic on individual transport links, it affects the benefits froIT link improvements and hence the selection of investments. It is an important part of transport policy and has been treated as such in this Study. Regulation of Vehicles and Drivers - It is universally accepted that questions of safety require regulation of vehicle design and maintenance, and also of driving standards. It should be considered whether Bolivian regulations are adequate and effectively enforced, particularly in view of the difficult and dangerous nature of many of the roads. Regulation of vehicle size and weight is also generally accepted as desirable in order to prevent undue damage to road surfaces and obstruction by oversize vehicles. It should again be considered whether these regulations are well conceived in economic terms, are sensibly applied to different classes of road and are effectively enforced. The environmental effects of motor vehicles can in theory be controlled by either economic disincentives or regulations. Governments have generally chosen the latter. This problem is less serious in Bolivia than in many other countries but it may nevertheless require some attention. Social Subsidies - It may be considered desirable, for non-economic reasons, to relieve isolated and remote communities from some of the costs of transport. This may be done by remission of taxes or by payment of grants towards the provision of cheap and frequent services by bus, train or plane. -9- Regulation of Operators The second main aspect of transport policy concerns the efficient provision of transport services on the available infrastructure, subject to the various charges and regulations just discussed. The basic objective is to obtain, as cheaply as possible, services that match consumers' demand in terms of frequency, convenience and reliability. This is an area in which private enterprise can work well, subject to certain conditions. The most important condition is the avoidance of monopoly and, with it, price-fixing. Where traffic volumes are reasonably large, this is normally not a problem; monopolies are easily broken by allowing free entry into the industry. But where traffic volumes are small, monopoly may flourish, and this seems to be a- problem in Bolivia, particularly in the trucking industry. It may be desirable in such cases to control rates and to take action against restrictive practices by operators, as discussed in Chapter 16. Even where traffic volumes are large and real competition among operators can be achieved, there is unending debate about the need for regulation. One must consider the case for regulating (a) rates and fares, (b) entry into the industry, (c) services offered, (d) service quality and (e) coordination between different operators. Other countries vary from complete freedom to comprehensive regulation, but the general trend today is towards deregulation on the roads and in the air, for both passenqer and goods services. On the railroads monopoly prevails amost everywhere, for technological reasons, and necessitates a greater degree of control, which is usually obtained by the simple expedient of public ownership. But even there, in face of financial losses and bureaucratic incompetence, there has been a tendency to decentralize management and give greater freedom to act commercially, i.e. to maximize profits. In Bolivia, as in many developing countries, there is a tendency to think that transport services can best be provided by public corporations or private monopolies under public control. Established operators tend to ask for protection against new competition, employing arguments about "excessive" or "unfair" competition for this purpose. One of the tasks of the Study is to offer advice on these important policy problems. There is no way in which a particular policy on competition can be proved the best. One can only put forward the arguments for and against, in the Bolivian situation, and point to experience elsewhere. This is an area where experience and judgement are necessary. -10- Improvement of Transport Infrastructure Only when the existing infrastructure is well managed and the services operating upon it are efficiently organized, is one well placed to consider the possibilities of improving and expanding it. For this purpose the use of cost-benefit analysis is well established, though not without difficulties. The justification of an infrastructure project, however, must be influenced by the way in which the infrastructure is used or misused. That is why it is logically desirable to determine transport policies towards use and operation first before turning to infrastructure policy. Clearly, much improvement in the infrastructure is possible and desirable. The road network is rudimentary as a network and generally of a very low standard. Huge areas in the north and east are devoid of roads. There is no practical limit to the amount of road construction and improvement that one would like to see in Bolivia. On the railroads there is the conspicuous gap between the Andean and the Eastern systems, both of which are in a state where re- habilitation is desirable. River transport is relatively important in providing access to remote regions of the country, but no port infrastructure has yet been provided. Lacking direct access to the sea, Bolivia has two international river outlets through the Paraguay and the Madera Rivers, but both are very poorly connected with the rest of the transport system and both are little used for the transport of Bolivian goods. Air transport is better developed than the other modes. The network of airports reaches into many parts of Pando, Beni and Santa Cruz departments that are not served, or only very badly, by other modes. Even between cities relatively well connected by road or rail, e.g. La Paz-Cochabamba-Santa Cruz, air transport is the dominant means of carrying passengers, because of the combination of terrain, distance and conditions of the existing infrastructure. More remote cities such as Tarija, Sucre and Trinidad are highly dependent on air transport. Nevertheless, many airports need better runways and navigation aids, and many of the aircraft in ser- vice are old and dangerous in the light of current inter- national flight safety standards. The air services themselves, while giving extensive coverage of the country, are often of low frequency, with only two or three flights a week. Thus a great deal of improvement is desirable, for both economic and social objectives, but the cost would far exceed the resources that could conceivably be obtained for the purpose. The problem of choosing the best projects is not easy_ As far as economic objectives are concerned, cost- -11- benefit analysis is able to give sensible conclusions. The main policy difficulties arise over non-economic objectives because they are more difficult to evaluate. Inevitably there arise conflicts between the use of limited resources for economic or non-economic objectives. In Bolivia a major objective is national integration, which unfortunately tends to conflict with the needs of the economy. The basic problem is to place values on non-economic objectives, e.g. the connection of isolated communities or the assistance to underprivileged sections of society. Projects with such objectives may yield low economic returns; the problem is to value their non-economic benefits. This problem is no different from that of giving operating sub- sidies for similar purposes. The best approach to the problem is to show (a) the benefits expected in physical terms, e.g. the numbers of people affected and the amount of use they are likely to make of the facilities providedi and (b) the opportunity cost, i.e. the alternative benefits to be sacrificed to make this project possible. It is also important to (c) ensure that cheaper ways of producing the same or similar benefits are not neglected. The final assessment must be a political one, but often a clear statement of the facts will point irresistibly towards the right conclusion. -12-  YAPACANI SANTA CRUZ· TRINIDAD (IN CONSTRUCTION) NEARMACHACAMARCA RECENT BRIDGES  CHAPTER 3 DATA COLLECTION AND STUDY METHODOLOGY CHAPTER 3 DATA COLLECTION AND STUDY METHODOLOGY The Study is divided into three logical and conventional phases: first, an analysis of the current transport situation; secondly, a study of the ways in which transport needs and problems will probably develop in the next 20 years, and of the options for improving the transport system to meet the challenge; and thirdly, an evaluation of those options, subject primarily to economic constraints, leading to recommendations for structural and organizational improvements to the system. In order to acquire a good knowledge of the transport situation, inventories were built up of the infrastructure roads, railroads, airports, ports and pipelines -- and surveys were made of transport services and traffic on all modes. Information was collected about the organization, regulation, tariffs, costs and financial position of the state transport authorities, and of private operators of trucks, buses, aircraft and river boats. A computer model of traffic movements was constructed from the data describing transport facilities, tariffs and traffic, together with socio-economic data of population, income, production, consumption, urbanization and foreign trade, all distributed among 77 zones. The data concerning goods movement, i.e. production, consumption and foreign trade, were broken down into 39 categories of goods, selected according to their transport characteristics. The purpose of the model was to estimate the complex effects on traffic movements that would result from future changes in transport facilities and tariffs, or in the aforementioned socio- economic characteristics of the country. A national transport study is faced with an extremely large and intricate set of variables which, through a complicated process of competition and complementarity, produce the traffic pattern at a given time. In order to estimate how this pattern would respond to changes in either the transport system or the social and economic situation, or both, a good model is invaluable; it handles complexities beyond the capacity of any human brain and makes massive calculations with far greater speed and accuracy than any statistician. It permits the different modes in every part of the country to be planned as a system and ensures that traffic predictions on the system are logical and consistent with the socio-economic activities that give rise to the traffic. . -13-  The second phase considered the likely development of transport needs up to 1989 and 1999. As discussed in Chapter 7, forecasts were made of national population, income, pro- duction, consumption and foreign trade, in the detail required for the model. The possibilities and uncertainties of Bolivia make forecasting a hazardous exercise, particularly as far ahead as 1999. For this reason, three alternative develop- ment patterns were produced for 1999. The ~odel was then used to predict the resulting traffic on a minimum network in 1989, including only transport improvements already committed. Thus were revealed the deficiencies of the transport system, as they would emerge in the future if only minimal improvements were made. In order to plan an optimal program of improvements and extensions, a large nu~ber of proposals were compiled from allover the country and, after an initial process of selection, were reduced to a list of projects meriting more detailed costing and evaluation. Projections were made of the budget likely to be available for such projects. The third phase of the Study evaluated the candidate projects and policy proposals and arrived at conclusions. The model was used to test an arrbitious network in 1989 with t'\110 widely different sets of tariffs and in 1999 on a network with further improvements. In the latter test the central, most probable development pattern was used. The results contributed towards the analysis of tariff policy and the evaluation of projects. At the same time, investigation of the organization and regulation of the transport industries led to recommendations for reform. The evaluation of projects was thus made in the light of proposals for a new tariff policy and institutional regime. All the projects were evaluated first on their economic merit, mostly by the conventional cost-saving method whereby benefits to both existing and generated traffic are estimated directly from reductions in transport costs. Most of the evaluations were based on model results but some projects, notably those affecting international links, were not included in the model and were examined manually. A separate air model was used to generate air passenger flows. Most road projects were evaluated by the Highway Design and Maintenance Standards (HDM) model, using traffic predictions from the main model. In all cases, however, the evaluation depended on traffic predictions and estimates of the costs of vehicle operation and infrastructure maintenance, with and without the project. The HDM model was also used to evaluate road maintenance policies. Some projects entailed new extensions to the transport network, notably penetration roads, on which.all or most of -14- the traffic would be newly generated. The cost-saving method was not applicable to such projects and a producer surplus method, which attempts to place a net value directly on the economic activity reflected by the traffic, was used instead. Following economic evaluation, questions of regional dis- tribution and social policy were considered before an invest- ment program was drawn up. The final output of the Study, as described in this report, thus consists of conclusions and re- commendations on transport investment and maintenance, regula- tion, organization and tariff policy. But, because of the close relation between transport and economic activity, the conclusions also imply a number of non-transport recommenda- tions concerning land use, migration, farming methods and in- dustrial development. The remainder of this chapter discusses the collection of transport flow data, computer modelling and project evalua- tion procedures. Data Collection The data on existing transport and factors affecting transport which were collected by the Study, fall into four main groups. These are transport flows, infrastructure, costs and tariffs, and socio-economic data. Information on these is given in Chapters 9, 4, 5 and 7 respectively of this report. This section deals with the procedures used to collect the data for each of the transport modes and also the socio-economic data. More detailed information on data collecting is available in the Study's First Interim Report published in December 1979. Road Transport - Data on road transport flows were collec- ted in an extensive roadside interview survey. The object of the survey was to provide as complete a picture as possible of passenger and freight movements. To achieve this the in- terview sites were chosen so as to intercept all traffic leav- ing or entering the departmental capitals, which generate most of the long-distance traffic in the country, together with sites at important intermediate junctions. The interview sites selected are illustrated in Figure 3-1. The survey was conducted in two parts, from February 20 to March 27, 1979 during the wet season, and from June 5 to 25 (w~th the exception of Beni department which was surveyed later) during the dry season. Usually the interviews covered all traffic from 6 a.m. to 10 p.m. for three days in each season, although at some sites the interviewing was continued for the whole 24-hour period. During the dry season the night- time traffic was counted as most sites. The survey and data coding procedures are detailed in Working Papers 1 and 2. -15-  The information collected in each interview, where applica- • able, was as follows: • Place, direction, date and hour of interview • Vehicle type • First number 6r letter of registration plate (indicating use of vehicle) • Origin and destination of trip • Purpose of trip • Service type of vehicle • Number of seats and passengers transported • Cargo capacity and quantity of cargo carried by type. These data were then coded and punched onto computer cards. Each interview was then subjected to a set of checks to ensure valid codes, e.g. commodity code between 1 and 69, and logic, e.g. origin and destination compatible with inter- view site and direction. Where errors we~e revealed, they were, if possible, corrected for coding and punching mistakes or, if necessary, rejected. At the end of this process 134,823 valid interviews had been obtained. . Subsequently the interviews were factored to annual volumes of 1977, the Study's base year. This process is described in Working Paper 15, and the factors are summarized in Appendix 3A. After the application of these factors, the data were used to produce many matrices and tabulations used in the development and calibration of the transport models. Additional data on road transport flows were obtained from counts carried out by the National Road Service (SNC). This organization has been conducting counts for many years and produces annually a flow map showing movements over the major road system. Their current program involves monthly manual classified counts at 132 locations using personnel from SNC sub-district offices (residencias). This program will be greatly increased shortly when SNC will have available a large number of automatic traffic counters to supplement the manual counts. Largely to obtain information of transport operations, extensive interviews were undertaken and discussions held with trucking and intercity bus operators and with the agen- cies responsible for the regulation of road transport ser- vices. The data collection efforts concerned the following items: • Truck and bus operations, organization and regulation • Function of transit check points • Vehicle standards • Vehicle import controls • Transpor~ tariffs and costs. -16- t!ill: SurveYI near Trinidad only f or dry lealon . LEGEND All other stationl refer to both wet and INTERNATIONAL BOUNDARY dry season survey campaions . In addition, EXISTING ROAD dry season interviewl were conducted SYSTEM at R iberalta and Guayaramerl n . ~ . ROADSIDE ~ INTERVIEW STATION ORURO Cruce Mochocomorco ,I I ,, I I I / r-- --------1 " / ... ..; ROADSIDE INTERVIEW STATIONS U)f,.IIu't fIlmil!. and .Aoaocialea FIGURE 3-1  As well as road flow data, it was also necessary to col- lect and examine data on existing road infrastructure. The basis for this was a road inventory conducted by SNC which at the start of the Study existed only for the districts of La Paz, Cochabamba and Santa Cruz, but has since been completed for the whole country. This inventory was used as the basis for much of the input required by both the main model system and the HDM model, described later in this chapter. To check and complement this work, two other surveys were conducted by the Study. The first was a check on the accuracy of the SNC inventory on a sample basis, the results of which are described in Working Paper 5. This was done by members of the Study team visiting the SNC survey teams while the invento- ry was being carried outi also, checks were made in the field for those areas for which the inventory was complete. The second survey was a windshield survey primarily to establish travel times required for the highway network, but also used to check several of the items in the SNC inventory. Rail Transport - The collection of rail flow data did not require the extensive effort needed for road transport • . This was because the National Railroad Company (ENFE) records an increasing amount of its flow information on computer tapes. Discussed in Working Paper 17 and summarized in Appendix 3B, these tapes gave full 1977 freight movements for the Andean system, freight movements for the Eastern system and passenger movements for the Andean system for 1978, which with other data provided by ENFE could be factored back to 1977. Eastern system passenger flows were taken from manual records since this infor- mation had not then been introduced to the computer system. River Transport - Data on river transport freight move- ments were obtained from records kept at the major river ports by the Bolivian Navy (FNB). These records indicate the ori- gin, destination and commodity class for freight movements. In a similar manner to the processing of the road data, the information was coded, punched on cards and edited before matrices and tabulations were produced. In order to obtain data on boat operations on the Ichilo- Mamore river system, a questionnaire survey, combined with a field inspection, was carried out by the Study during March and June 1979. The data represent a summary of boat operations and conditions including: • Boat characteristics • Financing conditions • r.1aintenance • Taxes • Salaries • Origin/Destination patterns. -17-  Air Transport - The majority of domestic air passengers are carried by the national airlines LAB and TAM, and data on their flows were made available to the Study for analysis. The majority of domestic freight movements are carried by the non-regular air services. Detailed flight records . ' were obtained from the Ministry of Transport, and these were processed through the computer in a similar way to the road data. Pipelines - The pipeline system of Bolivia is operated by the National Hydrocarbons Company (YPFB), and all data on the origins, destinations and quantities of the pipeline flows were obtained directly from that company. Foreign Trade - The foreign trade flows were treated separately both in the data collection and the subsequent analyses. The source for these data was a set of computer tapes, provided by the National Statistics Institute (INE). These tapes gave information, on both exports and .imports, about the origin and destination inside and outside Bolivia and about the nature of the commodity and the weight of the movement. The flows were aggregated to the Study's zone system and commodity classes and tabulated for use in the analyses. Socio-economic Data - Existing socio-economic profiles and future projections constitute one of the principal inputs into the transport model. Thus the collection of these items was one of the key tasks of the Study. The principal data sources were reports and other documents provided by various Government ministries and agencies, but these were supplemen- ted by various surveys and analyses described in this section and, in more detail, in Appendix 3C. The regional development corporations which are situated in the departmental capitals formed one major source. The method of data collection was a questionnaire survey followed up by visits to each of the corporations by a survey team. The items collected for each province within the department were: • Agricultural production • Livestock production • Mining and oil production • Manufacturing production • Future development plans. A further set of questionnaires, followed by interviews, were sent to 22 major public and private enterprises. These -18-  THE RIO GRANDE RAil BRIDGE AT PUERTO PAlLAS NEAR SANTA CRUZ included such bodies as the National Mining Company (COMIBOL), the National Hydrocarbons Company (YPFB) and the Secretariat of Integration. From these enterprises much information was obtained on transport as well as socio-economic characteristics. To obtain data on private manufacturing, a sample of 88 companies covering each industrial branch was visited by an interviewer from the Study team. The questions asked included the type and amount of production and inputs, the distribution of goods, and information on transport modes used. As over half of the total consumption of goods in Boli- via consisted of agricultural products, it was obviously vital to make special efforts in this area. In the face of this situation, the assistance of the Bolivian Institute of Agricul- tural Technology (IBTA) and of the Ministry of Agriculture was requested. The latter promised the cooperation of the provin- cial extension agents in carrying out a survey. The survey forms consisted of an eight-page questionnaire complemented by a brochure explaining the survey procedures as documented in Working Paper 11. In order to check the accuracy of the replies, they were compared against the data previously re- ceived from the regional corporations and data provided by the Ministry of Agriculture at a departmental level. From the available data , a consistent set was selected and used for the development of agricultural production and consumption by analysis zone. Another type of data required for input to the transport models was population and employment for each analysis zone. The former of these had been published by lNE from the results of the 1976 national census. For the latter, although the information had been collected in the census, it had not been processed or published. However, the computer tapes contain- ing the data were made available to the Study to derive these figures independently, which is described in detail in Work- ing Paper 16. Transport Model System The object of the transport model system is to predict, in a logical and consistent manner, how transport flows will be affected by changes in population, income, production, con- sumption, transport infrastructure and transport tariffs. To this end, a complex series of computer models have been de- signed, built and calibrated for use in Bolivia. This section gives a brief overview of the model system which is described in greater detail in Working Paper 52 and Appendix 3F of this report. -19-  The transport model system is composed of a freight model and a passenger model, which operate separately on the various input parameters, but which together give a total picture of flows on the transport facilities of the country. It was designed to represent movements which take place wholly within Bolivia. The analysis of foreign trade move- ments was undertaken separately and is described in Chapters 9 and 15. Analysis Zones - An important simplification in the model system is the concept of the analysis zone. As illus- trated in Figure 3-2 and defined in detail in Appendix 3D, the country was divided into 77 zones for which all parameters are aggregated. Thus for each zone, population, income, pro- duction and consumption are defined and all transport activity associated with that zone is assumed to begin or end at one point within the zone. The model is limited to the prediction of traffic between these zones, rather than between each possi- ble origin and destination in the country. In defining the analysis zones, the following five criteria were observed: (i) The zones should be relatively homogeneous in respect to the current production patterns and future development prospects. (ii) They should contain sufficient population and productive capacity to generate and attract significant volumes of traffic in goods and persons. (iii) Zone boundaries should follow the borders of the administrative units on the basis of which statistics of population, employment and pro- duction are maintained. (iv) Zone boundaries should conform to natural local trading, agricultural or industrial areas. (v) The total zone system should reasonably reflect the significant variations in the configuration of the country's transport system and actual traffic flows. -20-  • ZONE CENT ROID 47 ZONE NUMBER \ ~~~ <' < l_~ 70 • dl o () :-­ 1) \) , c L V () ~ \) " V - ~ V> :.~-.( (> '"' \ ~ ~ () r- o T I ;"J A ANALYSIS ZONES %11"" [jImil/' and Jiuociale6 FIGURE 3-2  Freight Model System - The object of these models is to predict freight flows by road, rail, river and pipeline, both in terms of tons and vehicles. The freight movements are de- rived separately for each of 39 commodity groups, which are defined in Appendix 3E and further discussed in Chapter 7. As shown in Figure 3~3, the sequence of the model system is: • Freight network model • Freight distribution model • Freight modal allocation model • Vehicle model • Assignment model. These are described below. Freight Network Model - This model has two basic inputs which are a transport network description and a set of trans- port tariffs and cost data. The network defines, in numerical terms, the transport infrastructure to be considered. Thus it will include the major roads, railroads, rivers and pipe- lines of the country, with a specification of their relevant physical characteristics, e.g. length, surface type, width and alignment for roads. The transport costs are then assigned to each part of the network, in general, as a function of physical characteristics. The most important part of this model is the creation of paths or routes through the network from each zone to every other zone. Several sets of paths are built corresponding to the different modes available. These paths have two useSj firstly in the assignment model described later, and secondly in the calculation of zone-to-zone costs and times which are used in the subsequent three models. Freight Distribution Model - The object of this model is to link the production and consumption of goods, specified for each zone and commodity class, into zone-to-zone tonnage movements. The manner of the linkage is a function of the zone-to-zone costs produced by the network model and the par- ticular commodity class. The following three submodels are used to accomplish this process: • Cost minimization submodeli • Cost independent submodel~ and, • Gravity submodel. -21-  The choice of which submodel to use for any commodity , class was made on the basis of which was found to best represent the observed flows. The cost minimization submodel is best suited for commodities containing one type of rela- tively uniform product, such as wheat, cement or fertilizers. It minimizes transport cost involved in distributing the commodity, i.e. the matrix T .. is chosen such that 1.) n n r i=l r: j=l T .. 1.) d .. is a minimum and 1.) n n L i=l T .. = C. 1.) ) and r::: j=l T .. = P. 1.) 1. Where: T .. = Tonnage flow from zone i to zone j 1.) d .. = Cost from zone i to zone j 1.) C. = Consumption in zone j ) P. = Production in zone i 1. n = Number of zones. The cost independent submodel distributes flows in such a manner that each destination zone receives, accordinq to the level of consumption, its proportional share of the production in the origin zone. This model is suitable for the commodity group encompassing vehicles and transport equipment, which consists of a number of discrete products. Each of these is produced in only one zone, and the relative proportion of their consumption is reasonably constant in all zones. The mathematical formulation is: P.1. x C. ) T .. = n 1.) r Pk k=l Where P k = Production in zone k, and the other variables are as shown before. In the gravity submodel, tonnage flows are distributed :-r:: between zones according to a probability function of cost which decreases as cost increases. Thus if the cost between -22-  NETWORK I TRANSPORT MODEL TARIFF AND COST DATA I FREIGHT NETWORK PRODUCTION AND CONSUMPTION DATA I '.j , ROUTE CHOICE t MODEL RAIL FAVORED t t ROAD PATHS PATHS , , ,-l/ r RAIL PARAMETER I t ROAD PARAMETER 'V MATRICES MATRICES , "- I /' "- r t MINIMUM TAR IFF ICOST MATRICES , '" . DIST.RIBUTION .. . . . , ... ......... ................ ........ .. " ......... ~ MODELS ~ ·······t···········, . DISTRIBUTION MODEL H- DEPOT CONVERSIONI ! IF REQUIRED I ,l- t FLOW MATRICES BY COMMODITY '-v .. ............................. ............. :MOD AL ALLOCATION MODEL ~ · . . . ·. ·. . ········ . . ··t·· .......................................... EXCESS TIME I , .- MODAL ALLOCATION MODEL I COSTS I RAIL TONNAGE FLOW MATRICES + + ROAD TONNAGE FLOW MATRICES I / "- .................................................... ····t·· . ·. . ·. ··· ·. ··· . ·. ·. ·t· . ·. . ·. · · · . . . . . . : VEHI CLE MODELS , ,~ .. RAIL WAGON RUCK / ... ......+.... I MOVEMENTS MOVEMENTS ..... . ......... : ASSIGNMENT MODEL ". .... ..... . .. . ... .. ... VEHICLE t EMPTY TRUCK ASSIGNMENTS ~ MOVEMENTS FREIGHT TRANSPORT MODELS FIGURE 3-3 a zone pair is decreased by a transport improvement, the probable movement size increases. The formulation is: C. x F(d .. ) ) 1) T .. P. 1) = 1 Where: T .. , d .. , C., P., n are as before 1J 1J 1 1 F = Distribution probability function. The gravity submodel in effect supposes that a commodity may sometimes be transported extreme distances, from a pro- duction unit to the furthest consumption zone, but that cheaper (shorter) journeys are more likely. With this sub- model, the output of a production zone is mostly transported to neighboring zones, with diminishing (though non-zero) quantities being distributed to zones further away. The value of this submodel is chiefly that is represents a com- promise between the two extreme situations covered by the two submodels described earlier. It applies to such mixed commodity groups as fruit, beverages, textiles and machinery. Freight Modal Allocation Model - The object of this model is to allocate the zone-to-zone flows produced by the distri- bution model between the modes available to carry that flow. The basis of the allocation are the relative costs for each mode as derived from the network model. The method used is first to calculate the cost ratio be- tween the two modes being split and subsequently allocate a percentage to each mode based on diversion curves derived from observed data. The cost ratio is defined as: Cl Ratio = C + 0 + D + E + F 2 Where: = Cost of zone-to-zone trip by first mode (e.g. road) = Cost of zone-to-zone trip by second mode (e.g. rail) = Cost to reach second mode in origin zone (e.g. a road trip to railhead) D = Cost from second mode to destination zone (e.g. a road trip from railhead) E = Cost of excess time of second mode over first mode F = Factor applied to average curves to give commodity specific curves. -23-  . The diversion curves for the allocation of agricultural and non-agricultural products between road and rail are shown in Figure 3-4. Their derivation, which is based on the results of the roadside interview survey and ENFE's computerized records of 1977 flows, is discussed in Appendix 3F. Vehicle Model - The model converts the mode-specific zone-to-zone tonnage flows into vehicular flows. Thus for road flows, the tonnages are converted into movements of loaded trucks of three diffe~ent sizes, the proportions de- pending on the zone-to-zone journey times derived from the network model. Empty truck volumes are estimated based on loaded truck movements to give total truck flows. Rail tonnages are converted directly into wagon loads using commodity specific factors. Thus, there are the following three submodels: • Loaded truck submodeli • Empty truck submodel; and, • Rail wagon submodel. The loaded truck submodel converts tonnages of zone-to- zone road flows into zone-to-zone truck flows. The basis of , ''Il the method is a set of curves, derived from the results of the roadside interview survey, which relate trip length to the number of small, medium and large trucks required to carry the tonnage. Factors are also applied to the curves for each individual commodity group, as detailed further in Appendix 3F. The empty truck submodel adds empty truck movements to the full truck movements. The method assumes that total truck movements between any zone pair are symmetrical, and so the first part of this submodel adds the required number of empty trucks to the direction of zone-to-zone flows which is lower. The second part complements this by a number of additional empty trucks, according to a formula derived from the roadside survey. The formulation is: (FAB x FBA}q Where: = Number of additional empty truck movements from zone A to zone B of a particular truck size = Number of loaded truck movements from zone A to zone B of that truck size '. = Travel time from zone A to zone B 1< p, q, r = Constants determined by regression analysis. -24- ;,  I""'" I ~r- ._- f-- PEf CE~ T R "'IL I""," ._-.-- ~---~ ...Q[HER "_"_ ~. i l7 GRI~1I1 TIJRF i I J V --~. - i - - ,- "'" 7 1- , ....... --~.--- --.---- I I AI'! V Ir --- V I I I v 1)--- - J .r. 1-- ! ------ ----~\-- I v v j..o' ~ ~ I ~ ~ .1 .2 .3 .4 .5 .6 .7.8.9 I 2 3 4 5 10 RO A 0 / R A I Leo ST RAT 10 MOOAL ALLOCATION DIVERSION CURVES Ci)it'l~Ui flmilh and .A66oct'ale6 FIGURE 3-4  The ~ail wagon submodel converts rail tonnages into rail wagon flows, by applying commodity specific load factors to each tonnage flow. These factors were derived from 1977 ENFE statistics reporting wagon movements and tonnage flows by commodity. Assignment Model - This model takes the zone-to-zone vehicle flows produced by the vehicle model and assigns them to the routes estimated by the network model. The accumula- tion of all movements using a particular road or rail sec- tion gives the total freight vehicle flow on that section. Alternatively, the tonnage movements output by the modal allocation model may be input, which would result in tonnage flows on each section of infrastructure included in the network. Passenger Model System - The modelling concepts used for passengers are very similar to those described above for freight. There are three sets of models in this system which are as follows: • Surface public transport (bus and rail); • Private vehicles; and, • Air. Each has its own network model with representations of the available infrastructure, services and costs. Upon this, a model was derived for each by regression analysis, which predicts zone-to-zone passenger flows as a function of zonal income and trip costs. In the case of surface public transport, these flows are then allocated to either bus or rail services. The subsequent stages of the passenger model are an assignment of passenger flows, an estimation of vehicle movements and an assignment of these. The passenger vehicle flows (buses and private vehicles) are then added to the freight vehicle flows to give total traffic on each network section to be used in the evaluation. The surface public transport model predicts volumes of bus passengers, buses and rail passengers. The total passenger flow between each zone pair is determined by the following basic equation: FL = lOa x A + b x B + C x C + d Where: FL = Annual passenger flow between a zone pair A = Logarithm of minimum travel cost between the zones (either road or rail) B = Logarithm of total GDP in the origin zone C = Logarithm of total GDP in the destination zone a,b,c,d = Constants determined by regression analysis. The resulting annual passenger flow is then split into a rail flow and a bus flow in a similar manner to freight flows, i.e. by using a diversion curve which gives, as a function of the modal cost ratios, the percentage using each mode. -25-  The mathematical expression to estimate private vehicle flows has a similar structure as the previous equation. How- ever, travel cost is replaced by travel time and GDP by urban population, as these parameters were found to give better results in representing base year flows. A two-phase model was developed for air passenger trans- port. The first part estimates passenger flows based on a regression analysis, while the second part adjusts flows according to the flight frequency between zone pairs. The e- quation is expressed as follows: FL = lOaxA + bXB + c-C + dXD + e Where: FL = Annual air passenger flow between a zone pair A = Logarithm of trip cost B = Logarithm of travel time difference between surface public transport and air C = Logarithm of total GDP in origin zone D = Logarithm of total GDP in destination zone a, b, c, d, e = Constants determined by regression analysis. The second part of the model adjusts the number of passengers on each link as deduced from this formula. It was observed that on high frequency routes the regression model underestimated traffic for the base year, while overestimating flows for low frequency routes. Thus, a curve was drawn up linking the ratio of actual to regressed trips against fre- quency, and applied to the results of the first part of the model. This is discussed further in Appendix 3F, together with other aspects of the passenger model system. Use of the Models - The models are used in two dif- ferent phases, firstly, the calibration phase, and secondly, the predictive phase. ,, The calibration phase is applied to the Study's base year, for which estimates of production, consumption, in- come and transport tariffs are known. Also known are the transport flows resulting from these data. Thus in this phase the various model parameters are adjusted until an adequate representation of the observed flows is reproduced. In the predictive phase, the model is applied using the socio-economic and cost parameters estimated for the forecast years. Moreover, several network alternatives can be "tested", i.e. the model calculates the transport flows that would re- sult from the implementation of several assumed network ." improvements. The flows estimated by the model for 1989 and 1999 are presented in Chapter 9. -26-  Model Calibration - As appropriate, model functions were calibrated to represent as closely as possible current trans- port characteristics. In the gravity distribution model, for instance, the probability function F shown in the previously described formula had to be calculated. Subsequently, several components of the model system were validated against data observed for the base year. In the case of the distribution model, the total tonnages surveyed to enter and leave the seven southern department capitals were compared against the values estimated by the model. Overall, the model results were within 4 percent of observed flows, but showed a certain overestimation in the cities with low tonnages. The results of the modal allocation model were validated against the detailed rail transport statistics available by commodity group. The model estimated an overall rail tonnage which was one percent above actual flows. With regard to ton-kilometers, the model results were four percent high. The final validation of the freight model system consisted in the comparison of the results from the vehicle model against observed truck and rail wagon movements. Except for certain discrepancies encountered at two low- volume locations and at Santa Cruz where the dense road traffic in the immediate environs cannot be well represented by a national transport model, the truck model results were within one percent of observed flows. With regard to rail wagon movements, the model estimate was within 1.3 percent of ENFE's records. The conclusion is that the freight model system represents very well current transport characteristics. The validation of the passenger model yielded somewhat inferior results, with an overall underestimation of bus movements by 13 percent and of private vehicle movements by 18 percent. For the purposes of the this Study, however, the passenger model is deemed adequate, considering that freight flows are the principal factor in justifying potential transport improvements. A more detailed description of the model calibration is contained in Appendix 3F. Project Evaluation An evaluation is an attempt to predict the net value to the community of following one course of action rather than another. It recognizes the existence of alternatives, one of which is always the possibility of "doing nothing". -27- The purpose is to discover which of the alternatives is most likely to produce the greatest net benefit to the community as a whole. Although there can be difficulties in deciding exactly how to value (or weight) different kinds of benefit accruing to different classes of people, the general aim is clear, and in this Study the predominant aim is to estimate the economic benefits and costs affecting society as a whole. Questions of political expediency lie outside the scope of the Study. Technically, an evaluation is a comparison of two alter- native courses of action. There may be more than two alter- natives, in which case they can be evaluated in pairs until they can be placed on a scale of net benefit relative to one which is chosen as a benchmark. The evaluation compares the alternative courses over a long period of time, as long as may be relevant, but this is usually done by examining the situation in two or more single years and interpolating and/ or extrapolating for the other years. Hence one carries out single-year evaluations as part of an overall evaluation. Costs and Benefits - Four items affecting cost or benefit enter into every evaluation: (i) Capital costs, most commonly costs of construction. (ii) Differences in unit transport costs of all affected movements by all available modes and vehicle types. (iii) Traffic volumes on all affected links, with and with- out the project, by mode and vehicle type. , .~ (iv) Annual maintenance costs on the affected parts of the infrastructure, with and without project. These costs depend partly on the volumes of traffic. These four items normally include all the costs and be- nefi ts that need to be :rreasured. They provide an indirect, but logically good way of measuring most of the net costs and benefits proceeding from a transport project.--rlt should be noted that many benefits are exactly offset by costs else- where and therefore do not need to be measured). Other net costs and benefits, however, can arise in particular cases, e.g. environmental costs, changes in accidents, and effects on the national balance of payments. Where these are impor- tant, they are taken into account in the final judgement of the project. There are also, of course, non-economic impacts, e.g. political and social effects, which may be important and are assessed after the economic evaluation. " The Cost-saving Method of Evaluation - Except for evalua- tions where the differences in unit transport costs are re- latively large, the conventional cost-saving method is used. The impact of a transport investment takes place over a long period of years and must be evaluated accordingly. In prac- tice this normally means that the impact has to be examined first at two or more points in time by calculating single- year rates of return; then a way must be found to take account of the whole life of the project, resulting in an internal rate of return (IRR) and a net present value (NPV). -28- .>  LA PAZ - CARANAVI (NEAR COROICO) FEEDER ROAD NEAR SAPECHO END OF THE GUABIRA - CHANE ROAD WOODEN BRIDGES  The first step is to evaluate the impact of the project in a single year. This is done by the following simple for- mula in which two alternative situations, X and Y, are com- pared, Y requiring more investment than X: Bn - M ARn = _-:-:-_-11_ K where AR = annual rate of return, in year n, on capital cost K, n M = cost of maintaining infrastructure in situation Y, n less that in situation X, in year n, K = capital cost of Y less capital cost of X, compared to year n, B n B = annual net traffic benefit, Y over X, year n, to nm vehicle class m. = 1 2 (C nm - C' nm ) (Q nm + Q' nm ) C = unit cost of transport, vehicle class m, in situa- nrn tion X, in year n nm= tion Y, in year n C' unit cost of transport, vehicle class m, in situa- Qnm= volume of traffic, vehicle class m, in situation X, year n Q' = volume of traffic, vehicle class m, in situation Y, nm year n ARn can thus be calculated from the four sets of data described above. The overall rate of return on the investment is estimated over its expected life or 25 years from commencement of con- ~ction, whichever is less, by means of the IRR and the NPV. These two conventional measures are defined as follows: IRR is the value of r that satisfies the following equa- tion: -29- where n is the year measured from 1980, and N is the total number of years considered. ''Ii NPV equals: N L n=l where i = marginal rate of return. Values of i were con- sidered between six and twelve percent. A case study of pro- ject evaluation following the cost-saving method is given in Appendix 3G. The method of evaluating a new transport mode to be added in a corridor already served by transport is .., described in Appendix 3H. The Producer Sur Ius Method of Evaluatic" , Khere the " difference in unit transport costs Cnm - C nm' is h~gh, or infinite in some cases, the cost-saving method is not valid. These cases occur when a penetration road or so~c other new transport facility gives access to an area formerly extremely inaccessible. Before the project there is no traffic, or very little, and therefore no production except for local con- sumption, if indeed there are any inhabitants. After the project there is traffic, which of course is generated traf- fic and reflects production generated by the o~~~=tu~~ty to transport it to the market. The producer surplus method at- tempts to estimate the net gain (or profit) frorr. such generated production after deducting all relevant costs including other investment costs and transport costs to the market. This estimate of producer surplus is in effect ~ direct estimate of the value represented by generated traffic which, in the cost-saving method, is estimated indirectl7 f~~~ the reduction in transport costs. In principle the benefits thus estimated as producer surplus can be inserted as B in the evaluation formula. n Project Specification - Each candidate project, i.e. a project selected for evaluation, was systematically ~ ,ccified on a project specification sheet giving the following informa- tion: brief description, location specifying nodes and dis- tances, traffic speeds and link lengths before and after pro- ject, traffic volumes in 1977, construction time, construc- tion costs (by year if construction takes more than ~ne year), infrastructural maintenance cost before and after pr~ject, and unit transport operating costs before and after project in 1977, 1989 and 1999. More details on project specifications are given in Working Paper 93. -30- Ii'  Specification of Evaluations - Since an evaluation is a comparison of two, and only two, alternative situations, the final evaluation of a project may require several indi- vidual evaluations if there are more than two alternatives. Also, after an evaluation, it may be desirable to modify the project in some way, or to change the land-use assumptions, and make another evaluation. Each evaluation must therefore be clearly defined as to the two situations being compared. HDM Model and Highway Analysis The highway analysis was concerned with three classes of roads: (i) New roads, where no link previously existed (ii) Existing feeder roads (iii) Other roads. The new roads defined for evaluation were mostly in the underdeveloped regions of the country. Although such links were included in the transport model networks, the predicted volumes using these links were, as expected, very low. The key to the evaluation of these links was the producer sur- plus analysis described earlier. There are many thousands of kilometers of existing feeder roads in the country ranging from scarcely used tracks to im- portant access roads to agricultural areas. With the help of the National Road Service, the main feeder roads in each region were identified and an assessment made of their importance, and of the costs of maintenance and improvement. This was largely an exercise based on judgement and was not carried out to the same level of detail as the other road analyses. The analysis of the 12,460 km of the basic road system, i.e. roads coded in the computer transport network, was car- ried out using the World Bank's Highway Design and Maintenance Standards Model (HDM) (1) based on traffic forecasts estimated by the main transport model. The function of the HDM model is outlined below, followed by a description of the work pro- gram using the model. (1) Highway Design and Maintenance Standards Model (HDM), Transportation, Water and Telecommunications Department, World Bank, Washington, USA, June 1979, revised January 1980. -31-  The HDM Model - The HDM model simulates the life of a road, taking into account the growth in traffic, the deterio- ration of the road due to environment and traffic, and the maintenance operations performed on the road. The impact of these factors on vehicle riding characteristics is then assessed and hence the effect on vehicle operating costs. Other costs, such as construction costs, can also be input. Year by year costs· of construction, maintenance and vehicle operations are calculated by the model and these are depen- dent on such specified inputs as traffic mix, road charac- teristics and maintenance policies. Alternative situations can be compared in the model. Hence it is possible to speci- fy a series of policies for such things as maintenance, road construction or vehicle size, and compare the total cost of highway transport in each case. Work Program Using the HDM Model (Where Appropriate) - The relationships between road design, road maintenance and the costs of vehicle operations are described in Working Paper 109 and summarized in Figure 3-5. Based on these relationships, the following program of work was defined: (i) Highway Characteristics - Existing highways were categorized by geometric alignment characteris- tics. Also, new road designs were analyzed to identify the relationship between design standards and the resulting geometric characteristics. Ter- rain type was an important consideration. (ii) Axle Load and Pavement Designs - Alternative pave- ment design policies were analyzed to identify the optimum. The effects of high axle loads on pave- ment deterioration were also analyzed in order to formulate policies on permissible vehicle types and vehicle taxation. (iii) Highway Design Standards - An analysis was made of the effect of highway design levels on vehicle operating costs. The effects of geometric align- ment and paving were considered and the results were strongly affected by initial traffic volumes and terrain. (iv) Road Policies - Policies for road improvement were considered with alternative design levels and al- ternative timings of improvement. Optimum policies were identified but these were influenced by bud- get limits. ", -32- Ii  TERRAl N CHARACTERISTICS rl STRUCTURES r REGIONAL CHARACTERISTICS , .. HIGHWAY DESIGN STANDARDS ~I GEOMEIRICS TOTAL HIGHWAY DESIGN --- HIGHWAY CONSTRUCTION COS TS - TOTAL COST OF MAX I MUM r+ HIGHWAY I- PROVISION PERMI SSI BL E AU E LOADS L.I SURFACE I 1r"o HIGHWAY MAINTENANCE COSTS - i ~ TOTAL POll CY FOR r+ COST OF HIGHWAY - ... r TRANSPORT VOLUME SIZE, TYPE, SUR~~~EO~~~~WAL - SOClO- ECONOMIC OF r-t' AND NUMBER I- I~AIIHEr;AI;CE SAFETY FACTORS CARGO OF TRUCKS rt' AND r- , SURFACE COMfORT CONDITION TOTAL ... I;UMBER AllO '-to COST Of . HIGHWAY TYPE OF USE - NUMBER !!EAVY VEH I CLES VEHICLE USER COSTS r- r+ Of BUSES I- VOLUME AND TRANSPORT I""- COSTS ... VOLUME OF PASSENGERS ---+ COMPOSITION OF TRAFFIC NUMBER OF LIGHT ~ VEHICLES HIGHWAY TRANSPORT COST SYSTEM %'IluiI fl'milli and ..;!(/4q.()Ciale~ FIGURE 3-5  (v) Bridges and Delays - Potential bridge projects were identified together with an estimate of cur- rent delays due to lack of bridges. Delays were oosted and a feasible program of bridge construc- tion established (independent of the HDM Model). (vi) Maintenance Policies - Alternative maintenance policies were evaluated for each type of surface and by traffic volume class. (vii) Investment Program - Road improvement (including bridges) and maintenance policies were evaluated to select an optimum balance between these two types of highway expenditure within probable bud- get limits. Initial appraisals were made for high- ways alone, but the final program included all transport investments. Budget Constraints In defining an investment program it is not sufficient simply to list all projects which have been found to be economically worthwhile. In general the costs of these pro- jects will exceed the available budget, if not over the whole period under consideration then at least in some years dur- ing that period. Therefore it is necessary to have a logi- cal, systematic method of maximizing some form of economic benefit while staying within budgetary constraints. For the purposes of this Study it was decided that net present value (NPV) should be maximized (2); the method is based on a dynamic programming algorithm. The basic concept underlying the approach used is termed the investment unit. An investment unit consists of a set of projects which are mutually exclusive, i.e. at most one may be implemented. For example, for a particular stretch of highway the investment unit could consist of: gravel the road, pave to a low standard, pave to a higher standard, each with three different starting dates, reSUlting in nine different projects. The investment unit may be applied to road, rail, air or any other mode, or to maintenance pro- jects. For each project there must be specified the NPV of (2) 'Ihawat Watanatada and Clell G. Harral, Dete:c:nination of Economical- ly Balanced Highway Expenditure PF1Eeams under Budget Constraints: A Practical Approach, presented a~ l\brld Conference on Trans- port Research, wndon, April, 1980. -33- that project, discounted to a common year for all projects, and the costs of the project by year. As long as a project can be defined in this way it can be subjected to the same treatment. In practical terms the method may be employed in several different ways. For instance, if a highway investment budget is known independently, then the best selection of highway pro- jects may be made. Alternatively all transport projects for the country may be analyzed together. An additional output of the method is the global NPV obtained at varying budget levels, thus providing knowledge of the sensitivity of the global NPV to changes in budget level in any or all of the investment periods under consideration. '. -34-  CHAPTER 4 EXISTING TRANSPORT SYSTEM CHAPTER 4 EXISTING TRANSPORT· SYSTEM The history of Bolivia's economic development is closely related to the provision of transport infrastructure. A century ago, the growth of the mining industry went hand in hand with the construction of railroad lines. In the 1950s, the economic rise of the Santa Cruz area followed the building of a road connection with the rest of the country; this was aided by the expansion of the hydrocarbon industry which itself called for the development of a pipeline network. More recently, penetration and feeder roads have been built to spur agricultural production, and the exploitation of the rich Mutun iron ores is seen in conjunction with improved access to foreign markets via the Paraguay River. This chapter describes briefly the principal charac- teristics of today's transport system. Each mode is first examined separately, followed by a general overview of the access provided to the various regions and the country as a whole. Railroads The rail network is comparatively extensive, being more than 2.5 times as long as all paved highways together. It is divided into two separate systems; a connection be- tween them by rail is possible, but would require a long detour via the Belgrano Railroad in Argentina. Most of the Andean System in western Bolivia was built more than 60 years ago, primarily to serve the mining industry on the Altiplano. In contrast, the Eastern system has been in operation for less than 20 years; its traffic volumes have reflected the economic growth of the eastern lowlands and now account for 39 percent of the country's ton-kilometers by rail (it makes up 38 percent of the national rail network). The following paragraphs summarize the characteristics of the railroads; further details are contained in Appendix 4A and Working Paper 27. -35-  .... Rail Network - Figure 4-1 shows the existing railroad lines in and near Bolivia. Rail construction beoan in 1872 with the railroad from Mejillones (near Antofagasta) towards the Altiplano in territory which is now administered by Chilei in 1892 this line reached Oruro. Nearly all the other lines in the Andean rail system were built between 1900 and 1920. The last extension to this system was in the 1950s when the now abandoned railroad from La Paz into the Yungas was extended to Chuspipatai the further extension to Rurrenabaque never went beyond the planning stage. " In the eastern lowlands, the Santa Cruz-Corumb4 railroad was completed in 1958 with Brazilian assistance. The Santa Cruz-Yacuiba railroad, built with Argentinian credits, has been in operation by the Bolivian railroad since 1967. Two sections of the planned railroad to Trinidad (also built with Argentinian aid), Santa Cruz-Santa Rosa and Santa Rosa-Yapacan!, were completed in 1976 and 1978, respectivelYi construction of the third section to R!o Grande commenced in September 1980. Today, the railroad network comprises 3,643 km of one- meter gauge, single-track lines. Of these, 2,257 km are in the Andean system and 1,386 km in the Eastern system. The alignment is characterized by its sharp curvature and steep grades in the mountainous areas with minimum radii of less than 100 meters. On the Altiplano and in the lowlands, .• there are generally long tangents with a few large-radius curves. In general, the track is stable, with the exception of sections between Oruro and Cochabamba, and between Villaz6n and Atocha, which often suffer interruptions during the rainy season. Also, in the exceptionally severe rainy season of 1978/79, a long section of the Santa Cruz-Corumb~ railroad was washed away near Robor~, interrupting operation for nearly half a year. The Bolivian network connects with the railroads in Argentina, Chile and Brazil, which in general are also single track with one-meter gauge, and thereby with several ports on the Pacific and Atlantic Oceans. Ferries on Lake Titicaca give access to the Peruvian railroads which, however, have standard-gauge (1.435 m) track. Transport Flows - In the year 1979, 807,000 passengers and 751,000 tons ot goods were transported on the Andean system. The Eastern system carried 405,000 passengers and 445,000 tons of goods. More than three quarters of ENFEts -36-  LEGEND +t-+++­ NARROW GAUGE (1 .00m .) ~ STANDARD GAUGE (I..S5m .) c. cl .... , ... ........ ,..... , .. . ... '''''(' ~ / o () 11\ ~ 2 o ) c } > I V '\ '" ~ () --r.j ~ ~, ~' - ~ >. \.i () ~ ~ " " o ''"'~ 1 :\J A EXISTING RAILROAD NETWORK IJ)fillu" fIlm;I" and .A()oocialeo FIGURE 4-1 freight traffic are foreign trade movements, which are heavily imbalanced towards imports (73 percent of foreign traae rrove:rrents by rail in 1978). By far the most important route, both for imports and exports, is the line to Antofagasta, accounting for 43 percent of ENFE's foreign trade tonnage (1978). During the 15-year period 1965 through 1979, the number of passenger-kilometers nearly doubled, primarily because of the growing number of movements on the Eastern system, and because of longer average trip lengths. This is shown in Figure 4-2, which also indicates that the same situation occurred with the transport of freight and for the same reasons. The number of ton-kilometers more than doubled, with the most significant growth in the period 1974 to 1977. Organization of the National Railroad Corn an - ENFE (Empresa Naciona e Ferrocarri es manages a 1 t e railroads of the country with the exception of the 105-km mining railroad from Machacamarca to Unc!a, which is operated by the Bolivian Mining Corporation. Technical administration is provided by the Ministry of Transport and Communications, while financial matters are under the Ministry of Finance. The Board of Directors is chaired by the Minister of Transport, with representatives of the Presidency of the Republic, the Finance, Mining and Industry Ministries and the Railroad Directorate of the Ministry of Transport; the General Manager of ENFE is also a board member. The Board, General Management and administrative and commercial departments are located in La Paz, while operations management is located in Oruro, with operating offices in the main railroad centers. The management of the Eastern network is located in Santa Cruz, together with all its operating offices. Operations - In June 1980. ENFE had 7,201 employees of whom 6,405 worked in operations and 79~ in rehabilitation. Of the total, 6,412 were permanent employees and 789 temporary. The total number of permanent and temporary employees amounted to 6,401 in 1975 and 6,047 in 1976 (excluding the approximately 400 persons working on the Guaqui line, and prior to large-scale activities on the track rehabilitation program). The rail locomotive fleet includes 9 diesel-hydraulic locomotives manufactured in 1968 and 54 diesel-electric locomotives, of which 20 are of 1968 vintage and 34 are 1977 and 1978 models. In addition, 14 ferrobuses are in operation, composed of one powered car and one or two towed cars; 10 are 1967 models and 4 were purchased in 1978. -37-  At the end of 1979. the passenger car equipment consisted of 187 cars, 89 of which were purchased prior to 1950. Of the remaining, 91 were purchased and 7 were primarily built by ENFE within the last three years. The company maintains 140 rail freight cars for its own use, and another 1,927 are available for commercial use. Most of these are relatively new, 1,145 having been purchased after 1973. ENFE's 1979 statistics also list 192 own-account cars, mostly belonging to the National Hydrocarbons Company (YPFB). The availability of the locomotive equipment reaches 75 percent, with about 25 percent out of service due to repairs. Repairs and maintenance of locomotive equipment are performed in the workshops located in Viacha, Oruro, La Paz and Santa Cruz. Repairs and maintenance of non- powered passenger equipment and railcars take place in Uyuni and Robor~. , 1\ ENFE uses the following communications systems for its operations: • A security system for the movement of trains, and for traffic regulation from central command posts. • A selective circuit for calls in the Andean network for the use of the operative departments. • A telephone system for communications between the busier centers. • Additional telephone and telegraph services in the I , Eastern system. o A block signal system on the Oruro-Cochabamba branch. Current Railroad Improvement Projects - Almost invariably, the railroad lines were built with the minimum possible invest~~nt, using a relatively light rail and wooden sleepers placed directly on the soil, without ballast. This, combined with a certain amount of deferred maintenance, has resulted in poor operating conditions, causing low speeds, high operating costs and risk of derailment. With the help of several loans from the World Bank, a iJ ~, track rehabilitation program was initiated in 1973. By the end of 1979, 433 kilometers of line, or about 12 percent of the rail network, had been brought up to a standard permitting efficient operation. Discussions are currently underway be- tween ENFE and the World Bank to decide a further phase of the track rehabilitation program, the desirability of which is examined later in this report. -38-  400 L I j til c .£ 350 PASSENGERS :;; 300 .. ... til .... E 250 0 ~ A ............. 200 .... .. Ol c til til 150 0 0.. 100 SYS EM 50 O~----~--+---~--~----r---~---+~--~--+----+--~~--T---~--~-- 1965 1966 1967 1966 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1______ 600 ! _. l +- L _ - 1·- .. - r L i l 550 : FREIGHT i '-~- 1-'_-- ! ! V 500 .. _- f-. / / til 450 i / ___~ V c I 0 I i ! - 400 I I - .- :;; : . : l /~NF -T-~~ ~~ .1 ,-~ r- '- .. ... til 350 - I---~J ~ND¥ I /l' I -- -1 . .,---- " 'E 300 --1---- 0 ~ YSTEM // 250 ~-I-- .....,1' . ,-- ----i-- -~ c ..... _....." I : I , j I - ........V .,.t. ~.- .... : T--T""T----r-- i .~-.-., 0 r- -- c--- ~.' I 200 f-- ..... _ ... I --~ -EA~TER1S~~~!!~./" 150 I 100 '~--T-- . I i / .'. 1 !......" """"r 50 r-'-' r-.-}:::-r-j-ir---- I i , I I 0 1965 1966 1967 1966 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 TRENDS IN RAIL TRANSPORT (J)fi"u", flmillt and .;t/(o(>odale6 FIGURE 4-2  PAVED I N GRAVEL SURFACE EARTH SURFACE :Ii ,.J~.--\ -- CO," , ~_ t'\ \, ~ \, d ./'-'1 ­ r I j . .... \ :.­ d ~ \ .... / (' ".... .. , ­ EXISTING NETWORK OF PRIMARY AND SECONDARY ROADS FIGURE 4-3  The only current construction of new railroads is the extension of the track from Yapacani to Rio Grande, part of the Santa Cruz-Trinidad rail line. However, several design studies have been completed or are being actively considered to investigate the following lines: • Santa Cruz-Mataral-Aiquile (Rail Interconnection Santa Cruz-Cochabamba); • Mataral-Tarabuco (Part of rail interconnection Santa Cruz-Sucre); • Motacucito-Mutun~Puerto Busch (Paraguay River); and, • Guaqui-Oesaguadero (Peruvian Border) • The last of these would form part of a link between the Peruvian and Bolivian railroad networks, which are currently connected only by the Lake Titicaca ferries; some construction is taking place in the Peruvian section Puno-Oesaguadero. All the rail connections mentioned above are among the rail projects evaluated by this Study. Roads a£d Road Transport Although the importance of road transport has grown considerably in the past two decades, the system of modern highways is still quite rudimentary. At the end of 1978, only 13 percent of the primary and secondary roac network was paved. The first major paved highway to be built in Bolivia was from Cochabamba to Santa Cruz, playing a most important role in national integration and in the economic development of the eastern lowlands. This 495-km highway, which for the most part is now in need of rehabilitation, was completed in 1958. The construction of paved highways then proceeded gradually, with only 244 km opened to traffic in the 1960s, and a further 550 km in the 1970s. The interurban road system is used predominantly by trucks and other commercial vehicles. Because of the virtual absence of private passenger cars, traffic volumes are generally very low and exceed a daily flow of 1,000 vehicles at only a few locations. This compares with the capacity of a good two-lane road of between 10,000 and 20,000 vehicles per day. Road Network - The National Road Service (Servicio Nacional de Caminos - SNC) classifies the road network into three types of routes: primary, secondary and local. At the end of 1978, there were 5,375 km of primary roads, 4,270 km of secondary roads and 29,183 km of local roads, totalling 38,828 km. Of this total, 1,289 km (3 percent) . were paved, 6,798 km (18 percent) gravel and 30,741 km (79 percent) earth roads; Table 4-1 gives a breakdown of the road system by surface type and district. As shown in Figure 4-3, the paved roads are concentrated in the districts of La Paz, Cochabamba, santa Cruz and Oruro (1,229 km, 95 percent). -39-  Table 4-1 LENGTH OF ROADS BY DISTRICTS AND BY TYPE OF ROAD SURFACE KII.DMETERS OF ROAD BY DISTRICT TYPE OF ROAD La Paz Chuq. Tarija Cboa. S.Cruz Oruro Potost Tupiza Beni Pando Totals PRIMARY NE'IWORK Paved 185 6 54 373 426 159 1,203 Gravel 492 834 371 190 250 441 511 90 30 3,209 Earth 86 191 35~ 284 48 963 Subtotal 763 840 616 563 1,030 600 511 374 78 5,375 SEXX>NOARY NEIWJRK , Paved 21 7 28 "', 0 " Gravel 164 97 356 62 164 172 115 176 1,306 Earth 393 24 136 126 562 926 322 318 79 50 2,936 Subtotal 578 121 492 195 726 1,098 437 494 79 50 4,270 LOCAL ROADS Paved 3 55 58 Gravel 1,201 711 371 2,283 Earth 3,651 3,480 1,882 1,578 4,230 3,299 4,858 3,433 362 69 26,842 Subtotal 4,852 3,480 1,882 2,292 4,656 3,299 4,858 3,433 362 69 29,183 '!'OrAL NE."ImRK Paved 206 6 54 383 481 159 1,289 Gravel 1,857 931 727 963 785 613 626 176 90 30 6,798 Earth 4,130 3,504 2,209 1,704 5,146 4,225 5,180 3,751 725 167 30,741 TOTAL 6,193 4,441 2,990 3,050 6,412 4,997 5,806 3,927 815 197 38,828 SOURCE: Estad!stica Vial 1978 (Road Statistics for 1978), Servicio Nacional de Caminos, Departamento de Planificacion, La Paz. *' oiI .. ;& a ,• • • • • • it " cl(I '"  Without doubt the most important road axis connects La Paz with Cochabamba and Santa Cruz. After completion of the Quillacollo-Confital and Confital-Caihuasi sections, which are currently or will soon be under construction, it will constitute a paved road of about 870 km. Between Cochabamba and Santa Cruz, the construction of a second paved road is planned to commence in 1981 (Chimor~-Yapacan1), which will run to the north of the existing road, mainly through flat terrain. These are committed projects and were therefore not included in the Study's formal evalua- tions. Another important axis runs from La Paz to the south, passing through Oruro, Potos1, Tarija and Bermejo. This axis, apart from the short paved Oruro-Machacamarca and Tomatas-Padcaya sections, is a gravel road. Two major roads branch off from this axis to connect with the Argentinian road system at Villazon and Yacuiba. Other international road connections exist at Fortin Villazon, Desaguadero and Tambo Quemado. These pOints are linked with the remainder of the Bolivian road network by either gravel or earth roads. Several international agreements have defined a number of transcontinental highways, some of which pass through Bolivia. The Pan American Highway enters Bolivia at Desa- guadero and continues via La Paz, Oruro and Potosl. It splits further south and connects with the Argentinian road system alternatively via Villazon and Tarija-Bermejo. Another branch south of Oruro goes via Sucre and Fortln Villaz6n into Paraguayan territory. Another international road (Carretera Marginal de la Selva) is planned to cross the Peruvian border near Puerto Heath and continue via R10 Yucumo and Chimore to Montero, where it would split into two sections: the one would provide an eastward connection with the Brazilian road system at San Matfasi the other continues in a southerly direction towards Argentina, via Santa Cruz and Yacuiba. The Organization of American States defined two further international roads passing through Bolivia. The Arica-Santos highway follows existing roads from Tambo Quemado to Santa Cruz, from where it would con- tinue eastwards to Corumba in Brazil. The LaPaz - Caracas highway would consist largely of roads that are yet to be built, either via Cobija or via Riberalta and Guayaramer{n. Traffic Volumes - In general, traffic volumes on the interurban road system are low. The latest system-wide data published by SNC (for 1977) indicate that only 300 km of road carry an average daily volume of more than 1,000 vehicles (1). Half of these represent four main roads leading into Montero, north of Santa Cruz. The other sec- tions in this category are in the immediate vicinity of Cocha- bamba and La Paz. (I) Estad{stica Vial 1978 (Road Statistics for 1978), Servicio Nacional de Caminos, Departamento de Planificacion, La Paz, December 1978. -41-  Even in the next category of 500 to 1,000 vehicles per day, there are only 357 km of roads throughout the country, which again are mostly clustered around the major cities. The LaPaz-Oruro highway is the only interurban connection carrying more than 500 vehicles per day throughout its • length. • The roadside interview surveys conducted in 1979 around the departmental capitals and at some major roa~ jupctions provided detailed information on the traffic using the road system; this is discussed extensively in Working Papers 14 and l4a and summarized in Appendix 4B. About 67 percent of all observed vehicles were trucks and vans, 18 percent buses and 15 percent private passenger vehicles. Of the observed trucks, 55 percent carried no freight (although about a fifth of these were used for passenger transport), and the remainder were on average loaded to SO percent of capacity. On the other hand, buses were operating with high load factors, averaging about 93 percent. SNC Organization - The National Road Service (SNC) is an autonomous agency under the Ministry of Transport and Communications. It is charged with planning, constructing, maintaining and administering highways in Bolivia. Overall, it employs 5,211 persons of whom 194 are engineers, 501 technicians, 4,329 manual workers and 187 others (1979). The SNC was established through Supreme Decree No. 6684 of February 21, 1964 and by Legal Decree No. 7390 of February 15, 1965. Its origin was the Bolivian-American Cooperative Road Service, which was charged with road maintenance in 1955, replacing the Direcci6n General de Vialidad. The head office in La Paz includes nine technical and administrative departments. Road maintenance and improve- ments are carried out at the district level. At present, there are ten districts, each of which has an independent admini- stration under the direction of a district chief. Six construction projects are currently underway. Two of these, La Paz-Cotapata and Quillacollo-Confital, are being built by contractors. The others (2) are under the direct administration of SNC, supervised by project chiefs who have a similar rank as the district chiefs. (2) Project Sapecho, i.e. Bella Vista-Rio Quiquibey, Rio Yucumo-San Borja and Rio Yucumo-Puerto Salines roads; Project San Ignacio, i.e. certain road improve- ments in the Beni; San RamOn-concepcion road; and San RamOn-Trinidad road. -42- SANTA CRUZ - TRINIDAD SANTA CRUZ - TRINIDAD (PROBLEMS WITH DUST) SUCRE - EPIZANA (OFFICIALLY GRAVEL) EARTH ROADS  Organization of Road Transport - Public road transport is largely provided by drivers operating their own vehicle. They must belong to a syndicate, of which there are 86 for non-urban transport: these are grouped into ten federations, all belonging to the Bolivian Confederation of Drivers. Transport cooperatives and private transport companies are also recognized by the State, but their activities are limited to the interdepartmental transport of passengers and the international transport of passengers and cargo. The Ministry of Transport and Communications fixes transport tariffs and has the legal power to regulate road transport. The interrelationship between Government and the road transport industry is examined in Chapter 16 of this Report. The drivers' federations, in accordance with agreements signed internally, act as the collectors and distributors of all goods movements inside Bolivia. The truckers must comply with the requirements imposed by the federations to be assign"ed cargo for transport. Freight transport is divided into two classes: • Long-distance transport • Interprovincial transport. Long-distance transport is characterized by travel distances greater than 200 km. The average capacity of the trucks is 12 tons and the average distance travelled by each truck is estimated at 40,000 km per year. Interprovincial cargo transport is used for trips less than 200 km in length and with trucks with an average capacity of 8 tons. This type of transport is closely linked to seasonal production of various agricultural sectors. It is estimated that the average distance travelled by vehicles in this service is 30,000 km per year. International cargo transport is provided by eight cooperatives and five private companies operating between Bolivia and Arica (Chile), Matarani and Lima (Peru), and lately and sporadically Rosario and Buenos Aires (Argentina). Interprovincial passenger services are provided by 74 syndicates with passenger and cargo vehicles. Buses that operate over mountainous roads generally have a capacity of 25 to 30 seats, while those which operate primarily on flat roads usually have a capacity of 32 to 40 seats. Passengers are also carried by truck; it was observed in the roadside interview survey that trucks carried an average of six passengers, often in addition to goods. -43-  Interdepartmental passenger services are furnished by 40 companies with an overall fleet of over 200 buses, averaging 36 seats per vehicle. They generally connect the departmental capitals, and the average annual distance travelled by each bus is about 75,000 km. Siver and Lake Transport The Madera River system in northeastern Bolivia, the Paraguay River near the southeastern boundary, and Lake Titicaca are all important waterways. Their geographical locations are shown in Figure 4-4. The following presents a sununary of their characteristics and usage, based on detailed data compiled during the course of the Study and described in Appendix 4C and in Working Papers 23 and 24. River Transport in Northeastern Bolivia - The rivers Beni - Madre de Dios - Orthon, Ichilo - Mamore, and Itenez form the Madera River which flows into the Amazon. River navigation is limited to about half the 10,000 km of principal rivers and major tributaries within Bolivia. They provide a natural means of communication and transport for passengers and goods to the country's northern region, which is nearly devoid of land access. Table 4-2 sununarizes the Bolivian waterways belonging to the Amazon river basin. It presents for each river system the principal navigable tributaries and the length of each. The total navigable waterways approximate 5,570 km. However, this length includes several sections, such as Puerto Linares - Rurrenabaque, where the channel depth only permits the passage of shallow-draft barges of limited capacity (15 tons). Moreover, the traffic of barges is considerably reduced during the dry season in the winter. The Ichilo - Mamore river system is the most important for transport; it is estimated that the total tonnage transported in 1978 was approximately 200,000 tons. This was mostly carried by 207 privately operated boats registered by the Bolivian Navy. The transport capacity of the 150 largest boats averages 75 tons per boat, considerably more than the boat capacities prevailing on the other rivers in northeastern Bolivia. Along the Ichilo - Mamore system, there are three main ports: (a) Puerto Villarroel, which is connected with Cochabamba by a 240 km long road; (b) Puerto Varador and Puerto Almacen near Trinidad; and (c) Puerto Guayaramer!n. About 82 percent of the recorded goods movements on the Ichilo - Mamore are between these ports; the remaining 18 percent are almost· all between the three major ports and other intermediate river ports. -44-  LEGEND 0 Department Capitols • Ports 0 Other Towns Navigable Rivers ~ ~ Other Rivers -f ~\m\\\I,\~1 Rapids National Boundary DLA PAZ < , OSANTA CRUZ \ o \, ORURO \ " '" '- <: POTOS~ 0 SUCRE f---- _-------.... Plo.SlIGrel; (Laguna QicefllS)J ",", ( MuNn/ . ~ronIlI l', / / ' ' ..... Plo.Bu.eh ~ ~( ( " OTARIJA / I \ l \ / I"'.. r -----....;' ,------, , I / \_-_/j \/ A R G E N T N A o I NAVIGABLE WATERWAYS FIGURE 4-4  Table 4-2 RIVER NAVIGATICN SYSTEM IN BOLIVIA - AMAZCN RIVER BASIN RIVER SYSTEMS AND NAVIGABLE TRIBUTARIES TERMINAL POINTS LENGTH IN KM 'I'Ol'AL KM 1. Ichilo-Mamore Rivers Ichilo River Puerto Grether-Puerto Villarroel (1) 143 Ichilo River Puerto Villarroel- R:Lo Grande 276 Mamore River Rio Grande-Guayaramer1n 1,041 1,460 T. Isiboro-Secure Rivers Chipiriri River-Mamore River 284 T. Iba.re River Puerto Alrnaren-Mamore River 46 T. Yacuma River Rio Bio-Santa Ana de Yacuma 40 (2) T. Yacuma River Santa Ana de Yacurna-Mamore River 43 T. Rro Grande-Yapacani River Yapacani Railroad Bridge- Mamon;~' River 150 (2) 563 2. Beni-Madre de Dios-Orthon Rivers Beni River Puerto Linares-Rurrenabaque (1) 145 (2) .t>o U1 Beni River Rurrenabaque-cachuela Esperanza 883 1,028 I 120 (2) 120 T. Madidi River Esrreralda River-Beni River Madre de Dios P~ver Puerto Heath-Beni River 483 483 Orthon River Puerto Rico-Beni River 315 315 T. Tahuamanu River Porvenir-Orthon River 332 T. Manuripi River Manurime River-Orthon River 200 (2) 532 Abum River Rapirran-Fortaleza del 'Abuna 210 210 3. Itenez River Piso Firroe-Mamore River 631 631 T. Machupo River San Raron-La Horquilla 110 (2) T. !>'1..achupo River La Horqui11a-Itenez River 15 T. Itonamas River Magdalena-La Horquilla 100 (2) 225 '!OrAL 5,567 T. Tributaries. (1) Navigable with restrictions. (2) Estimated distances . . sot1RCE: First Interim Fep?rt, Bolivia National Transport Study, D?cerrber 1979.  .. The Beni - Madre de Dios - Orthon river system converges on Riberalta, making this city an important port and distribution center. On this system, the number of registered boats in 1978 totalled 230, with capacities varying from 1 to 70 tons. Navigation along the Abun! River is limited to the settlements located alongside the river, extending from its confluence with the Rapirran River to the town of Fortaleza del Abun!. The Itenez River is potentially important for transport, with navigation possible from Piso Firme to its confluence with the Mamore River over a distance of 631 km. It flows through virtually uninhabited territory, and no information is available regarding the characteristics of the vessels using it, nor the transport of passengers and freight. However, since the minimum channel depth during the low water season is 0.9 m, the Itenez clearly provides better navigation conditions than the Rivers Beni, Madre de Dios and Orthon. Transport on the Paraguay River - This river, belonging to the River Plate System and connecting with the port of Buenos Aires (2,771 km from Corumb!), forms the border between Bolivia and Brazil for a distance of 48 km near Puerto Busch. It does not otherwise pass through Bolivian territory, although waterborne access would be possible from the Caceres Lagoon at Puerto Su!rez. Many years ago, ocean-going vessels did in fact anchor at this Bolivian town, but over the past decades the lagoon has become shallow even during the wet season and sometimes dries up completely. Consequently, Puerto Su!rez is no longer used as a port. At the port of Cor~.b~ in Brazil, Bolivia currently uses a wharf with berthing space for two 1,000-ton capacity barges. This wharf is utilized for the transport of iron ore originating at Mutan, of which about 100,000 tons were planned to be exported in 1980 via the Paraguay River. ~~out 2,000 tons of general cargo to and from Bolivia (1979) also use the port facilities at Corumb!i there are currently no co~mercial port activities at Puerto Busch. Transport on Lake Titicaca - About 40 percent of Lake Titicaca belongs to Bolivia and the remainder to Peru. Both countries have signed agreements controlling transport and general use of the lake. These agreements are reflected in the policy governing the Joint Permanent Coordinating Commission of Bolivia and Peru. -46-  Lake transport is an essential part of the railroad system connecting Bolivia with the Peruvian ports on the Pacific Ocean, linking the meter gauge Viacha-Guaqui railroad in Bolivia with the standard gauge Puno-Matarani railroad in Peru. The other major navigation route on the lake is be- tween Puno and Chaguaya, a Bolivian port through which the minerals from the nearby Matilde mine are exported. Both import and export volumes via the Lake have been generally declining during the 1970s. This is particularly acute on the Chaguaya-Puno route where the annual volume decreased from 67,000 tons in 1973 to 37,000 tons in 1978, due to the falling production at the Matilde mine. Transport by Pipeline Commercial oil production in Bolivia commenced in 1924, and by 1949 the level of production had risen to some 1,500 barrels per day. During that quarter century, refining of the crude oil had taken place in small refineries located in or close to the oil fields. In 1949, with the development of crude oil production in the vicinity of Camiri, the Bolivian National Hydrocarbons'Company (Yacimientos Petroliferos Fiscales Bolivianos - YPFB) constructed Bolivia's first transport pipeline from Camiri to Lamboyo where the line branched; one branch going to the refinery at Cochabamba, the other to the refinery at Sucre. Numerous other pipelines were built thereafter and the system, shown in Figure 4-5, now has an overall length of about 4,700 kID within Bolivian territory. About 20 percent of this consists of natural gas pipelines, all built within the last 10 years. Another 20 percent has been built to carry refined petroleum products; the remainder have been classified as crude oil pipelines. The current charac- teristics of pipeline transport are described in Working Paper 22 and summarized in Appendix 4D. Crude Oil Pipelines - There are nearly 2,500 km ot crude oil pipelines in Bolivia. They range from 2 to 12 inches in diameter, and were constructed between 1949 and 1978. Due to a lack of new oil finds, the production of crude petroleum has been declining over the past several years. Production fell from 2.74 million cubic meters in 1973 to 1.88 million in 1978, a decrease of 31 percent. -47-  On the other hand, the Bolivian crude oil is increasingly being refined within the country. In 1978, 1.56 million cubic meters of crude oil were processed at Bolivian refineries, compared with 0.88 million in 1973. As a consequence, the proportion of Bolivia's oil production exported in crude form declined from 68 to 23 percent during the same period, and some of the pipelines originally built for crude oil are now also used for refined products. ... . Gas Pipelines - Three gas pipelines are in operation. By far the largest one, the 24-inch pipeline from Santa Cruz to Yacuiba was built between 1968 and 1971 to export natural gas from the fields near Santa Cruz to Argentina. While natural gas and crude oil pipelines are generally considered as integral part of the petroleum industry and therefore somewhat apart from the transport community, they have, in terms of ton-kilometers, rendered a substantial part of the transportation service performed in Bolivia during the past three decades. In addition, the capital investment for pipelines during the past 30 years has constituted a significant part of total industrial and transportation investment in the country. Product Pipelines - The first pipeline constructed in Bolivia to transport refined petroleum products was completed in 1956 to transport products from the Cochabamba refinery to the cities of Oruro and La Paz. This 6-inch pipeline was paralleled by a second pipeline when the first approached capacity. A further 3-inch product pipeline was added in 1975 from the Sucre refinery to Potos!. A third pipeline between Santa Cruz and Villamontes, which was formerly used for crude oil, was changed to a petroleum product line in June 1979. The need for other product pipelines may arise with the development of the Bolivian economy in order to reduce the costs of distributing petroleum products. Product pipelines are in direct competition with other modes of transport. Therefore, a detailed analysis has been made of the economics of product pipelines, as discussed in Chapter 5. Air Transport Air transport has grown significantly during the last ten years. This has helped to alleviate the deficiencies of other transport modes, especially in the movement of passengers over long distances.and in providing access to isolated parts of the country. Air transport is vital for -48-  LEGEND - C rude Oil Pipelines Gos Pipelines Product Pipelines 6, Oi I Refineries A Arico SANTA Tite 0 "- "-......,,- I / POTOSI ( ",- ) C' , I ( / \ >..---? i Boyuibe I / / ( I \~ \ \ T;";"f 11 t \ 6, Villomonfes I I TARIJA I I I I / I 0 50 ~L~~_ _ _ _~!~~~~ 100 Approximate Scale ISO 200 km o I I 1 I"~ I I I ' ----- I I \..., \, \ r- ~/ ''---- I ,-- , 0-1-Yoc_u_i be __ "\ , / ! " '------ --,\ _/,J -- \ \ \ \Vi I I i I ',j \ ,, \ CHI L E / ARGE N TIN A \ / / / EXISTING PIPELINES FIGURE 4-5 the transport of several products to their markets. Moreover, it is important for social and economic integration by linking together the various parts of the country. It also provides necessary international communications. Aeronautical Infrastructure - The airport register of the Directorate of Civil Aeronautics lists 626 airports and airfields throughout Bolivia. The vast majority of these are small landing strips used by light airplanes, especially in regions which do not have easy access by land. Practically all these minor airfields have earth or turf surfaces, follow the natural terrain and have no navigational or landing aids. The National Airport Administration (Administracion de Aeropuertos y Servicios Auxiliares a la Navegacion Aerea - AASANA, operates 33 airports and airfields (see Appendix 4E). Six of these (La Paz, Cochabarnba, Santa Cruz, Trinidad, Tarija and Sucre) have paved runways which are regularly used by modern jets. The lengths of the paved runways range from 2,650 meters at Cochabarnba, to 4,000 meters at La Paz. The runways of the other AASANA airports have earth or gravel surfaces, and their length is generally in the range from 1,000 to 2,000 meters. Regular Air Services - The principal air carrier is the national airline Lloyd A~reo Boliviano (LAB); its domestic routes are shown in Figure 4-6. In addition, there are air services provided by the Bolivian Air Force, as well as a few private companies. Lloyd Aereo Boliviano which is owned primarily by the Bolivian government, provides domestic and international air transport services for passengers, cargo and mail. Its fleet includes seven jetliners (three Boeing 727-200, three 727-100 and one 707) for domestic trunk routes and inter- national services, and five turboprops (four Fairchild F-27, and one Aerocommander) for other domestic services_ It also provides international services to several Latin American countries and the United States. Transportes Aereos Militares (TAM) is an autonomous unit of the Bolivian Air Force and provides scheduled and non- scheduled passenger, cargo and mail service within Bolivia. With its fleet of two piston aircraft (C-47) and seven turb9props (six Fokker F-27 and one Lockheed Electra), TAM serves primarily population centers which are not otherwise served by air, and also Armed Forces units. Transporte AereO Boliviano (TAB) is another air transport enterprise of the Bolivian Air Force; with two turboprop aircraft (Hercules C-130), it offers a non-scheduled service for domestic and international high-volume cargo. -49-  • .. TWo other small airlines, L!nea A~rea Imperial (LAI) and L!neas A~reas Canedo (LAC), operate scheduled services between La Paz, Cochabarnba and PotosI. Eight foreign airlines serve La Paz and/or Santa Cruz, providing direct connections with other South American countries, the United States and Europe. Non-regular Air Services - TWenty-five private air transport companies offer non-regular cargo services within the country; 19 of these are based in La Paz, four in Cocha- bamba and two in Santa Cruz. These companies have a total fleet of about 50 planes, mainly piston aircraft of 1940s vintage such as the OC-3, OC-4, OC-6, C-54, Curtiss C-46, Convair 240 and 440, and Boeing B-17. The main function of these companies is to transport meat from the Beni, but they also carry other freight and some passengers between the major cities and areas with poor accessibility by land transport. There are also 112 air taxi operators, with 163 aircraft of different types, including propeller and turboprops. They are based primarily in the eastern lowlands (Santa Cruz and Beni, having 49 and 52 operators, respectively, but also in La Paz, Cochabamba and Tarija) and render an important low- volume passenger and freight transport service to remote points of the country_ TWenty-one private operators using specially equipped ,. small aircraft, ranging from helicopters to small jets, provide services of crop spraying, fumigation, aerial photograrnmetry, and mining and petroleum prospecting. The Bolivian National Hydrocarbons Company (YPFB) operates its own fleet of four small turboprop aircraft. Passenger and Freight Volumes - The number of domestic air passengers has shown a rapid growth in recent years, increasing from nearly 480,000 persons in 1975 to about 930,000 persons in 1978. LAB carried about 92 percent of the 1978 traffic, with TAM accounting for most of the remainder. International passengers using LAB increased at a slightly higher rate, from 100,000 persons in 1975 to nearly 200,000 persons in 1978. About 49,000 tons of goods were transported in 1978 by domestic air transport, with the non-regular air cargo services accounting for almost 80 percent of this tonnage. They carry primarily meat from the lowland cattle areas of the Beni and northern La Paz Departments to the consuming II" highland cities, chiefly La Paz. Overall View One notable feature of Bolivia's transport system is the absence of year-round access to large parts of the -50-  1 ~ LEGEND 50 0 100 200Km . REGULAR SERVICE BY; I I I LAB IBooinQ 727) LAB IF 27) TAM LAI LAC o () , o , <: ~ \\ " ~ \. <j.. 'i. ~ ~ ~ 0 :­ ~ () I,~ l' \ a .\ fI (j E XT I :II ,\

1980 DOMESTIC AIR PASSENGER SERVICES FIGURE 4-6  } NO LAND CONNECTIONS ALTIPLANO WITH DURING ENTIRE YEAR 50 0 ~ 100 200Km . I ! I ~ ttl ------ "d e = 1:" (') »: 11 ) o o o REP L ACCESSIBILITY FOR GROUND TRANSPORT FIGURE 4-7  TEMPORARY REPAIRS FLOOD DAMAGE (1979-1980 WET SEASON) ) SANTA CRUZ-CORUMBA RAILROAD TAPERAS ROBORE SECTION country. As shown in Figure 4-7, 36 percent of the total land area has no land connections at all with the rest of the country and therefore depends heavily on river and air transport. A substantial intermediate zone, accounting for 31 percent, finds its few access roads and paths impassable for up to six months during and after the rainy season. East of Santa Cruz, however, the railroad to Corumb~ generally operates during the rainy season and thus enjoys a virtual monopoly in that corridor. To sum up, only one fourth of the country (with, however, 86 percent of the population) can be considered to have land access during the entire year. But even here, railroads and roads are often blocked by landslides, and river crossing are often impassable after heavy rains so that land connections are interrupted in the rainy season for several days at a time. Bolivia's International Connections - As a country without its own seaports, Bolivia depends heavily on efficient land connections through neighboring nations. There are several railroad lines linking Bolivia with seaports in Chile, Peru (via Lake Titicaca), Brazil and Argentina. Road connections are much more rudimentary: a very long route to the Argentinian ports, an adequate link to the Peruvian port and a relatively poor connection to one of the Chilean ports. Most of these roads are unpaved, and with Brazil there is no road connection at all. As a consequence of the much better rail infrastructure available for international transport, most of Bolivia's foreign trade uses this mode. Not including the hydrocarbon exports by pipeline, 88 percent of all 1977 imports and exports were carried by rail, 11 percent by road and slightly less than 1 percent by air. This subject is further discussed in Chapter 9. Transport to and from Bolivia via the Paraguay River is still quite undeveloped and consists mostly of iron ore exported from Mutun. However, the relatively low cost of navigation makes this mode particularly attractive to Bolivia which, because of its remoteness from the seacoast and the associated high transport costs, has a competitive disadvantage in selling many of its potential exports on the world market. Use of Existing Infrastructure - It was pointed out that even the most heavily travelled roads are used to less than 10 percent of their capacity, except at a few locations on the fringe of the major cities. The general observation can be made that the use of existing transport infrastructure is low for the other modes, too. -51-  ,. • The freight tonnage per line carried by rail averages 165,000 tons per year over ENFE's network, ranging from a low of 9,000 tons to 500,000 tons on the most heavily used section (R!o Mulato - Uyuni). The capacity of a single railroad track depends on several factors, such as charac- teristics of passing tracks, signalling, length of trains, imbalance of transport demand and importance of passenger traffic. It is estimated that between 3 and 5 million tons of freight could be carried over a single-track line in Bolivia, which again vastly exceeds current transport demands. With regard to air traffic, the busiest airport in terms of aircraft movements is Trinidad with about 38,000 takeoffs and landings per year, followed by Santa Cruz (33,000) and La Paz (28,000). In contrast, the operational capacity of a runway easily exceeds 200,000 takeoffs and landings per year. Even in the case of pipelines, some excess capacity exists; because of the falling crude oil production, the pipelines built for the export of this commodity are not in use continually. Conclusions - The foregoing demonstrates that it is basic access and the lowering of transport costs rather than additional capacity, that is mostly required to improve the country's transport conditions. This may imply the construction of new roads into currently inaccessible zones, but these must be seen in the light of the migration that is needed to settle them. Much more important is the improvement of the existing infrastructure to reduce maintenance and operating costs. ,. -52- II' ,1 ,  CHAPTER 5 TRANSPORT COSTS AND TARIFFS CHAPTER 5 TRANSPORT COSTS AND TARIFFS It was shown in Chapter 4 that transport infrastructure in Bolivia, where it exists, is not lacking in capacity. Everywhere outside the cities, on all four modes, the levels of traffic are extremely low and impose no real pressure on the capacity of the system. The genuine weakness of the system lies in the poor quality of service and the complete absence of facilities in some parts of the country. Improvements to the system should therefore be directed almost exclusively towards higher quality of service, including the provision of transport where there is none at present. In economic terms this means reducing unit transport costs in the widest sense, i.e. reducing journey times, vehicle operating costs, accidents, losses, unrelia- bility, inconvenience, and the maintenance costs of the infrastructure. Better road surfaces, railroad rehabili- tation, all-weather runways, modern port facilities, all these and other improvements are desirable to ease the burden of transport, i.e. lower its economic cost, by making it faster, easier, cheaper and more dependable. The evaluation of proposed improvements, particularly where large investments are required, must therefore pay great attention to their likely impact on transport costs. For this purpose lengthy studies have been made of the costs, both infrastructural and operational, of all modes. The results are summarized in this chapter. The word "cost" will be used here, unless specified otherwise, to mean the money value of resources consumed. This meaning may be emphasized by the terIl" "financial cost" in contrast to "economic cost" which refers to the opportunity cost of resources consumed. To the cOIl"munity it is econoIl"ic costs that matter, because they express the real value of resources, whereas to individuals and companies it is obviously financial costs that matter. The latter are normally passed on by transport producers to transport consumers, at least in part, in the form of prices, rates or taxes, which may be summed up in the word "tariffs". It is tariffs, rather than costs, that influence the decisions of transport users. The relationship between tariffs and costs is fundamental to the economic efficiency of the transport system. The needs of tariff policy therefore provide a second major reason for the following study of costs and tariffs. -53-  • Finally, as the economy develops, and with it the transport system, the financial position of transport authorities and operators will be affected in many ways. A study of costs is necessary for an analysis of the financial prospects of the industry. Treatment of Costs The following information was required about costs, for each mode: (i) Total financial costs, broken down into attributable and unattributable costs; i.e. costs are attributed as far as possible to specific links and types of services, but there are some joint costs that cannot be thus attributed except arbitrarily. Given the traffic volumes, the attributable costs can be expressed as average costs per unit of I traffic by type of link. (ii) The effect of different infrastructure standards • on the costs of maintaining the infrastructure and on traffic operating costs, both financial and economic. (iii} The impact of changes in traffic volume on costs, both financial and economic. (iv) Construction costs: in order to estimate the costs of a large number of improvements, it is desirable to have general cost data that can be applied in similar situations . .tv} Prediction of future changes in transport costs due solely to changes in productivity and real factor prices, i.e. the prices at constant peso values of labor, fuel, equipment and other factors used for transport. Shadow Prices - Shadow prices are theoretical prices used for goods and services where there is no market price or where market prices are distorted. Shadow prices are usually intended to measure the marginal costs of the goods and services in question. They are used where market prices are seriously different from marginal cost. The cost elements subject to shadow pricing in the context of this study were fuel and interest on capital expenditures. Fuel is almost entirely produced in'Bolivia, and its domestic market price is sub- stantially below the international levels that the Nat- -54- ional Hydrocarbons Company (YPFB) could command if it were exported. Taking into consideration the costs of domestic distribution, an average 1977 shadow price of $b 2.50 per liter was calculated, or about 50 percent more than the domestic retail price. With regard to interest, a rate of 12 percent (in real terms, i.e. discounting for inflation) was applied. For all other cost components, the esti- mated market prices were used in the analysis. Costs in 1989 and 1999 - Throughout most of this report, money values are expressed in 1977 pesos. This means that, whatever the year in question, the indicated amounts of money represent the value of the peso in 1977. Thus the effect of inflation on the purchasing power of the peso is discounted by the use of "constant pesos". To avoid misunderstanding, two points should be noted. First, constant pesos do not mean constant prices: the general purchasing power of the peso remains constant but individual prices may change. Of course, if some prices go up, others must come down, to maintain the constant peso value. This is important because, when the economy develops, there must be changes in relative prices; in particular a rising GNP per head means necessarily a rise in labor costs relative to other costs. And, at the present time, it would be foolish to assume that fuel prices will remain, relative to other prices, at their 1977 level. Secondly, the discounting of inflation by the use of constant pesos does not mean that inflation can be entirely ignored in planning transport. It has important implications for financial planning and, when foreign loans are involved, can affect the economic evaluation of projects. These questions will arise later in the report. In Working Paper 40, a broad study was made of past trends in relative prices, as summarized in Appendix SE. The factors of production entering into transport were con- veniently grouped in five categories: labor, machinery, fuel, other materials and construction. Figure 5-1 shows that all five groups of costs had been rising relative to the general price level; in other words, transport equip- ment and infrastructure were becoming relatively more cost- ly. However, because of increasing productivity in their use, their higher cost did not necessarily lead to a sim- ilar rise in the costs of transport services. -55- .  • It was considered that some of these past trends-- supposing the base data to be reliable -- were unlikely to continue unchanged for the next ten or twenty years. With a little judgement, the following changes from 1977 prices were adopted for the years 1989 and 1999: 1989 1999 Labor + 10 % + 20 % Machinery + 10 % + 20 % ·iIfJ Fuel +100 % +200 % Other materials - 10 % The large increase postulated in fuel prices includes the increases already seen since 1977. The above factors were applied to all transport cost components in 1989 and 1999. Naturally the proportions of the five groups of inputs vary so that the final effect of the price changes varies from one transport mode to another. In road transport, for example, fuel accounts for an average of about 10 percent of the total cost to provide the service, or twice the proportion prevail- ing for rail. Applying the projected increases of the individual cost components, road transport in 1989 would be 19 percent more costly in real terms than in 1977, and rail transport 15 percent. With the proposed road improvements and the regulatory measures recommended in Chapter 16, it is expected that the utilization of trucks will increase substantially. It has been estimated that the average annual kilometrage performed by the various truck classes will increase by about 20 percent from 1977 to 1989. This increase in efficiency signifies that road transport costs have been calculated to grow by only 16 percent, rather than the 19 percent quoted above. No similar change in operating efficiency was projected for rail transport, so that the relative competitiveness (in terms of economic-costs) of road and rail transport estimated for 1989 remains virtually unchanged from the base year. Thus, there would have been no substantial change in the Study's analyses and conclusions, if a general price index had been used instead of breaking down cost increases into their individual components. -56-  1.5 1.0 ----------­- - - ­ __ en I&J (.) o z o.~ o --~--------------_+----------------+_--------------~------~ 1965 19701975 ..- LEG EN D COST OF LIVING INDEX OTHER MATERIALS .. --~. MACHINERY FUEL LABOR PRICE INDICES AT CURRENT RATES: 1960 TO 1979 FIGURE 5~ 1 Railroad costs and Tariffs Table 5-1 summarizes the financial costs of the rail system in 1977. The Andean system cost $b 486 million (excluding the Viacha-Guaqui line which links ~a ~az with Lake Titicaca) and the Eastern system $b 271 m~ll~on. It is estimated that passenger traffic cost $b 246 million, i.e. $b 0.627 per passenger-km, and goods traffic cost $b 511 million, i.e. $b 0.930 per ton-km; but these estimates depend on the somewhat arbitrary allocation of common costs (the table shows these in parentheses). Average costs - The financial accounting data for 1977 were broken down and analyzed to give the sort of informa- tion required by the Study (see Appendix SA and Working Paper 32 ). Separate data were available for 18 sections into which the railroad system was divided, and on each sec- tion the variable costs attributable to passengers by ferro- bus, ordinary passenger train and autocarril, and to goods by small load and wagon load were separated. The remaining, fixed costs are mostly common costs and cannot be separated between services except arbitrarily. For what it is worth, however, this has been done by ENFE, based on the criteria laid down by the International Union of Railways (UIC) as shown in Tables 5-2 and 5-3. The tables show that the average variable cost per passenger-km was 0.21 pesos by ordinary train, 0.15 pesos by ferrobus and 0.47 pesos by autocarril. It is of interest that the inferior service, the ordinary train, cost more to operate than the ferrobusi this is partly because it stops more frequently. The average fixed cost on the whole network was 0.43 pesos per passenger-km, i.e. 69 percent of the total cost. This high percentage of fixed cost is partly due to ENFE 1 s definition of fixed cost, which is taken from the International Union of Railways, but also to the very low volume of traffic. The average variable costs of goods traffic per ton-km were 0.98 pesos for small loads and 0.27 pesos by wagon load, but over 97 percent of the traffic went by wagon load. Fixed costs averaged 0.65 pesos, or 70 percent of the total cost, again because of very low traffic volumes. On average, over the whole network of 3,643 km, less than two goods trains and two passenger trains pass each day. The unit fixed costs are inevitably high. The tables also show great differences between lines. The unit fixed costs are much higher on the mountainous lines, like Potos!-Sucre and Oruro-Cochabamba-Aiquile, and of course they tend to be high on lines with relatively low traffic volumes such as Viacha-Guaqui and, again, Cochabamba- Aiquile and Potos!-Sucre. Variable costs also show big -57-  • Table 5-1 • TOTAL RAILROAD COSTS, 1977 ('lb:>usands of pesos) ANIEAN SYS'ln1 EASTERN a:>sT ITEM Passengers Goods Total SYSm·!(l) Traffic Handlin:] of baggage and freight Labor 14,031 1,649 2,252 6,813 8,579 14,031 8,462 10,831 4,520 4,174 5,163 18,551 12,636 15,994 .. Servicin:] of trains Interchange and rent of rolling stock 5,910 400 5,345 15,632 11,255 16,032 6,567 12,072 17,822 28,104 • Buildin:Js and services, etc. 5,460 7,605 13,065 6,347 19,412 Traffic total 29,702 43,974 73,676 38,843 112,519 Traction lDcarotives, labor and fuel 12,840 40,634 53,474 14,838 b8,312 Buildin:] and services 1,809 6,953 8,762 3,806 12,568 Traction total 14,649 47,587 62,236 18,644 80,880 Polling Stock lDcarotives, rraintenance 11,022 18,145 29,167 12,414 41,581 Coaches, maintenance 3,285 3,285 2,092 5,377 ''II Wagons, rraintenance 2,037 2,037 6,352 8,389 Coach building 515 2,190 2,705 717 3,422 Depreciation 17,628 25,787 43,415 22,698 66,113 Buildin:Js and services 772 1,200 1,972 2,124 4,096 ,4 Rolling stock total 33,222 49,359 82,581 46,397 128,978 Track Inspection (1,181) (4,099) 5,280 628 5,908 w::>rks (7,723) (26,804) 34,527 29,268 63,795 Depreciation (12,878) (49,780) 62,658 34,039 96,697 Buildin:Js and services (3,172) (11,011) 14,183 10,901 25,084 Track total (24,954) (91,694) 116,648 74,836 191,484 Telecamunications Signals and telecarmunications (291) (1,013) 1,304 790 2,094 Buildings and services (1,306) (4,531) 5,837 1,608 7,445 Telecamu.mications total (1,597) (5,544) 7,141 2,398 9,539 .Administration (32,736) (72,434) 105,170 51,897 157,067 Other Depreciation 3,051 7,556 10,607 6,649 17,256 Interest (8,520) (19,363) 27,883 31,016 58,899 148,431 337,511 485,942 270,680 756,622 (1) Breakdown between goods and passengers not available for the Eastern 51'staT'. NOl'E: These figures exclude the Viacha-Guaqui line, mich cost Sb 32.4 milliCll, and the Machaearnarea-Unc!a line which is not run by ENFE. Values shown in parentheses refer to camon costs, azbitrarily allocated to passenger and goods services. SOURCE: Castos del Transporte Ferroviario (Costs of Railroad Transport), Working Paper 32, Bolivia National Transport study. .' -58-  Table 5-2 COSTS OF PASSENGER TRANSPORT BY TRAIN, 1977 OFDTImRY TRAINS FERROBUSES Atl'IOCARRIIFS Annual Av.Cost/pass-kIn Armual Av.Cost/pass-km Armual Av.Cost/pass-kIn SOCTION OF RAIL SYSTEM lENGTH Pass-kIn Variable Pixed Pass-km Variable Fixed Pass-kIn Variable Fixed (kIn) (OOOs) ($b) ($h) (OOOs) ($b) ($b) (OOOs) ($b) ($b) I.a Paz-El Alto 17 1,908 0.47 0.98 1,378 0.29 0.60 52 1.98 7.02 El Al to-Viacha 25 2,907 0.41 0.76 2,027 0.24 0.46 77 1.53 5.45 Viacha-Oruro 204 19,225 0.29 0.46 16,546 0.15 0.22 Oruro-Machacamarca 24 6,677 0.23 0.50 417 0.24 0.58 Machacamarca-R:L.o Mulato 184 40,572 0.16 0.25 3,197 0.14 0.24 R:L.o Mulato-Uyuni 105 25,870 0.20 0.31 Uyuni-Atocha 93 19,313 0.19 0.31 Atocha-Villaz6n 196 27,654 0.25 0.42 I U1 Viacha-charafia 207 8,608 0.18 0.39 1.0 Uyuni-lNaroa 175 3,362 0.27 0.58 I Oruro-Cochabarnba. 211 19,418 0.17 0.47 26,980 0.15 0.23 Cochabamba.-Aiquile 216 1,514 0.84 2.71 1,770 0.24 0.47 Rio Mulato-Potosi 174 6,899 0.29 0.63 2,785 0.16 0.31 Potosi-Sucre 175 3,601 0.20 0.84 2,346 0.27 0.62 1,998 0.85 1.06 Viacha-Guaqui 65 200 1.46 5.20 Andean System 2,071 187,528 0.22 0.42 55,676 0.16 0.27 4,097 0.65 1.16 San ta Cruz-Corurnl::>& 643 45,316 0.20 0.59 36,610 0.13 0.33 1,241 0.37 0.86 Santa Cruz-Yacuiba 539 42,858 0.18 0.54 20,336 0.13 0.33 2,126 0.17 0.40 Santa Cruz-Santa Rosa 104 306 0.42 6.77 Eastern System 1,286 88,480 0.19 0.59 56,946 0.13 0.33 3,367 0.25 0.57 G:E~.ND 'IOTAL 3,357(1) 276,008 0.21 0.47 112,622 0.15 0.30 7,464 0.47 0.90 (1) Excludes Sucre-Tarabuco, Uncfa-Machacarnarca, and Santa Fbsa-Yapacanf, which explains difference with Chapter 4. SOt.JRCE: Costos del Trans,[?9rte Ferroviario (Costs of Railroad Transr:ort), v.brking Paper 32, Bolivia National Transport Study.  Table 5-3 COSTS OF GOODS TRANSPORT BY RAIL, 1977 Stft..ALL LOADS \'lAGON LOADS Annual Av. Cost/Ton-km Annual Av. Cost/Ton-km SECTION OF RAIL SYSTEM LENGTH Ton-km VariaEle Fixed Ton-km VariaEle Fixed (km) (OOOs) ($b) ($b) (OO('ls) ($b) ($b) La Paz-El Alto 17 23 2.04 3.22 1,378 0.34 0.85 El Alto-Viacha 25 81 1. 57 2.30 4,886 0.26 0.50 Viacha-Oruro 204 582 1.74 2.15 34,810 0.37 0.65 Oruro-Machacamarca 24 165 1.33 1. 66 9,855 0.25 0.38 Machacamarca-Rfo Mulato 184 1,263 1.45 1.78 75,722 0.31 0.46 Rio Mulato-Uyuni 105 861 1.18 1.47 51,542 0.23 0.35 Uyuni-Atocha 93 319 1.61 1. 92 19,095 0.29 0.51 Atocha-Villaz6n 196 407 1.94 2.43 24,378 0.35 0.68 Viacha-Charaiia 207 300 2.31 2.87 17,929 0.41 0.92 Uyuni-Avaroa 175 1,183 0.99 1. 30 70,848 0.21 0.35 I O'l Oruro-Cochabamba 211 550 1.45 2.50 32,961 0.29 0.82 o Cochabarnba-Aiquile 216 19 4.53 16.42 1,109 0.93 9.91 I Rio Mulato-Potos1 174 244 1.49 2.20 14,603 0.31 0.77 potosf-Sucre 175 30 6.13 11.10 1,759 1.22 6.76 Viacha-Guaqui 65 5 6.40 10.00 3,252 3.47 6.07 Andean System 2,071 6,032 1.47 1. 98 364,127 0.32 0.64 Santa Cruz-Corurnba 643 6,644 0.43 1.25 117,088 0.17 0.52 Santa Cruz-Yacuiba 539 1,466 1.50 4.54 85,477 0.18 0.58 Santa Cruz-Santa Rosa 104 55 0.62 12.57 1,512 0.15 3.82 Eastern System 1,286 8,165 0.62 1.91 204,077 0.17 0.57 GRAND TOTAL 3,357(1) 14,197 0.98 1.94 568,204 0.27 0.61 (1) Excludes Sucre-Tarabuco, Uncfa-Machacamarca and Santa Rosa-Yapacanf, which explains difference with Chapter 4. SOURCE: Costos del Transporte Ferroviario (Costs of Railroad Transport), Working Paper 32, Bolivia National Transport Study. 4i " ;.)i. " "'.,l# . 4...~ 150. .,... it< " ., '" ill" .... II, • •• differ~nces between lines. The highest costs tend to occur, not surprisingly, on mountainous lines, like Cochabaroba-Aiquile, and on lines with very little traffic, like Viacha-Guaqui. The Eastern system shows much lower costs than the Andean system for goods, but not for passengers. Some of the differences between lines are difficult to explain, however, and may be simply due to the statistical process: the figures were coropiled from aggregated accounts and contain numerous estimates. Marginal Costs - Oespite the deficiencies in the data, the comparison of costs by line showed a clear relationship between average variable costs per ton-km and the number of ton-kro, and similarly with passenger-km. This reflects the very low traffic volumes on the system. At very low volumes, not only are the track and stations grossly under-utilized, giving rise to high fixed costs, but also the rolling stock and train crews, giving rise to high variable costs, too. As traffic volumes increase, the under-utilization of rolling stock and crews soon disappears and variable costs per ton-km or passenger-km become constant. Given this situation in which a growth of traffic would reduce average variable costs, it was of interest from the point of view of tariff policy to calculate marginal variable costs. The following results were obtained: passengers: 1-1 = (V - O. 14) (0. 7 1 - 2. 6 00 ) + O. 14 freight: M = (V - 0.16) (0.28 + 0.560) + 0.16 where M = marginal cost per passenger-km or ton-km, V = variable cost ($b per passenger-km or ton-km), and o = traffic density in millions of passengers or tons. The equations are applicable to each line for the relevant values of V, but may be generalized as follows: passengers: M = 0.067 e -2.600 0- 0 .29 (0.71 - 2.600) + 0.14 O 560 72 freight: M = 0.036 e • 0- 0 . (0.28 + 0.560) + 0.16 These curves are illustrated on Figure 5-2. Cost Functions - The operating costs suw~arized above are sufficient for the prediction of future traffic on existing lines and also for the evaluation of new lines, but they do not permit an evaluation of projects for the rehabilitation of existing lines. For this a knowledge is required of the relationships between the quality of the track and the costs of operating on it and maintaining it. Cost functions of this -61-  • • sort are ~for the roads as a result of extensive research • in other countries, but for the railroads, in Bolivian conditions, there is little such information. Consequently, • cost functions had to be produced by a combination of theo- retical deduction and judgement based on experience. Thus, any conclusions derived from them should be treated . cautiously: better data could be produced by ENFE, and this is an area where useful work can still be done in the future. The cost functions adopted were specifically related to three kinds of rehabilitation policy, as follows: Policy A: a low-cost improvement of track formation, with junction renewals, to permit average train speeds of 50 km/hour for passenger trains and 35 km/hour for goods trains in flat terrain, or 35 km/hour and 23 km/ hour respectively in mountainous terrain; Policy B: a more costly but longer lasting operation, employing stone ballast, to achieve the same running standards as Policy Ai Policy C: the same improvement as in Policy B plus track realignment, to per~it average train speeds of 60 km/hour for passenger trains and 40 km/hour for goods trains in flat terrain, or 45 and 35 km/hour respectively in ~ountainous terrain. These train speeds compare with present speeds averaging about 40 km/hour for passenger trains and 23 km/hour for goods trains in flat terrain, or 25 and 20 km/hour respective- ly in mountainous terrain. The cost reductions expected from these policies are given in Table 5-4. These reduction factors were applied separately to each line. Tariffs - Rail tariffs charged in 1977 were obtained on tape from ENFE for passengers and for a large variety of commodities, for station-to-station movements. Freight tariffs are based on a tariff book, which contains a large number of separate tariffs for different co~modities and different journeys, but many deviations are made from the tariff book. Analysis of the data enabled commodities to be placed in six groups, for each of which an equation was produced relating the tariff to distance. The equations did not give a very close fit to the tariff data and were used only for commodity-links where no actual tariff was available. The equations were as follows: -62- ft  I I I a .9 8 ,7 E x w ,6 '" " a. "- -- .0 ;; E x c " I- "- ,~ I- '" 0 u FREIGHT VARIABLE FREIGHT MARGINAL PASSENGER VARIABLE PASSENGER MARGINAL --+---r-----.----.----.-'------,r-----.--- - I "---,----,-- .5 ,6 .7 8 .9 1.0 TRAFFIC DENSITY (M''',on Tons/Year) Or (Million Pass/Year) RAILROAD COSTS RElATED TO TRAFFIC DENSITY Ci)tillut !:Imil" and .;J~a(Jce'ale~ FIGURE 5-2  Table 5-4 IMPACT OF REHABILITATION POLICIES ON TRAIN OPERATING COSTS AND TRACK MAINTENANCE COSTS (ca:'culated percentage changes) REHABILITATION POLICy(1) Flat Terrain Mountainous COST ITEM ABC ABC Depreciation of rolling stock passenger trains: -16 -16 -26 -22 -22 -35 goods locomotives: -22 -22 -30 -16 -16 -26 goods wagons: - 9 - 9 -12 - 6 - 6 -10 Maintenance of rolling stock: - 2 - 2 - 4 - 2 - 2 - 4 Crew costs passenger trains: -18 -18 -29 -25 -25 -39 goods trains: -25 -25 -33 -17 -17 -29 Fuel costs: - 8 -14 Track maintenance costs: -18 -40 -47 -20 -34 -48 (1) See text for definition. -63-  .., NO. COMMODITIES WAGON TYPE TARIFF ($b per ton) 1 Minerals Wagon 93 + 0.45 d 2 Refined petroleum products Tanker 290 + 0.55 d 3 Wood, wood products, primary and fabricated metals, vehicles and machinery Flat bed 21 + 0.71 d 4 Cereals Box car 1 66 + 0.38 d 5 Live animals Box car 2 385 + 0.35 d 6 All others Box car 3 62 + 0.49 d where d = distance in km. Passenger fares are related to distance but each line is weighted to reflect its individual cost characteristics. Track maintenance is more costly in mountainous than in flat country, and unit operating costs are higher on the lines with lowest traffic volumes. A weighting factor is determined for each line and fares are calculated as a standard rate per km multiplied by the weighting factor. The weighting factors, which are also used for freight tariffs, are as follows: La Paz - E1 Alto 1. 471 Uyuni - Avaroa 0.886 E1 Alto - Viacha 1.160 Oruro - Cochabamba 1.242 Viacha - Oruro 0.951 Cochabamba - Aiqui1e 1. 394 Oruro - Machacamarca 0.917 R!o Mu1ato - Potos! 1.425 Machacamarca - R!o Mu1ato 0.924 Potos! - Sucre 1.469 R!o Mu1ato - Uyuni 0.952 Viacha - Guaqui 1.108 Uyuni - Atocha 1.151 Santa Cruz - Quijarro 1. 000 Atocha - Vi11az6n 1. 398 Santa Cruz - Yacuiba 1.000 Viacha - Charana 1.048 Santa Cruz - Santa Rosa 1. 000 In the model tests for 1989 and 1999, except for the 1989 tariff test and the final 1989 test of the recommended system, it was decided to set tariffs equal to marginal costs. (The rail system is so little used that there is little difference between short and long-run marginal costs). As shown above, marginal costs are much lower than average costs or actual tariffs, generally less than one third, which led to relatively large volumes of rail traffic in the model results. In the 1989 tariff test, in order to demonstrate the sensitivity of rail traffic to price, tariffs were maintained at their 1977 level in real terms, subject to an increase of 9 percent on account of expected changes in relative factor costs. In the test of the recom- mended system, an intermediate tariff was assumed, lying between marginal costs and 1977 tariffs. Road Transport Costs and Tariffs " Whereas railroad cost data are concentrated in the accounts of FNFE, the same type of information for road transport is dispersed among a multitude of operators. -64- Thus, much effort was spent on obtaining the relevant data, as is described in a number of working papers. The National Road Service (SNC) was the principal source of road maintenance costs, also discussed in the following. The basic relation- ships of road construction costs are described in Chapter 11. Average Operating Costs - Six types of vehicles were selected to represent traffic on Bolivian roads. Private passenger cars as such were not included (but jeeps were), since they represent less than 7 percent of total traffic flow, except in the immediate vicinity of major towns. The selected vehicle classes were: • Toyota Jeep, 5 seats, gasoline, 128 h.p. • Medium Bus, Chevrolet 60, 26 seats, gasoline, 153 h.p. • Large Bus, Mercedes 0365, 36 seats, diesel, 145 h.p. • Small Truck, Dodge D-400/600, 4.4 tons capacity, gasoline, 176 h.p. • Nedium Truck, Toyota F-IOO, 6.4 tons capacity, gasoline, 138 h.p. • Large Truck, Volvo NIO or Faa, 11.6 tons capacity, diesel, 243 h.p. Cost data were obtained from suppliers and users, as detailed in Appendix 5B and Working Paper 33. Operators of buses and trucks are organized into syndicates, which eased somewhat the collection of data. These syndicates are grouped, for both buses and trucks, into interprovincial and interdepartmental (1) operators, whose costs differ in some respects on account of the different distances that they cover. In general, interdepartmental journeys are over 200 km, interprovincial journeys less than 200 km. This leads to big differences in annual kilometrage per ve- hicle, which has an important effect on transport costs. Between departments, buses average over 80,000 km a year, trucks about 40,000 km, whereas between provinces the corresponding averages are about 40,000 and 30,000 km. The diverse characteristics of the country's roads and regional differences between supply of and demand for trucking services lead to big variations in annual vehicle utilization and in other factors affecting costs, from one route to another. These differences are hidden in the averages used for the analysis. Table 5-5 gives details of the financial costs of the large buses and large trucks. Including the effect of medium buses, the average bus costs per seat-kilometer were $b 0.18, $b 0.23 and $b 0.28 for paved, gravel and earth roads respectively. This results in an average for all roads of $b 0.21 per seat-km which, (1) Interdepartmental passenger services are provided by bus companies rather than syndicates. -65-  Table 5-5 COSTS OF TRANSPORT BY ROAD (Pesos per 1,000 km, 1977) LARGE BUS lARGE TRUCK COST ITEM Paved Gravel Earth Paved Gravel Earth Fuel 756 790 807 692 723 739 Oil 53 106 106 53 106 106 Tires 518 763 1,096 572 843 1,211 I Spare Parts 1,279 1,422 1,5.47 1,480 1,645 1,790 0'\ 0'\ I Maintenance Labor 350 353 384 499 503 547 Crew 762 865 1,104 430 488 623 Depreciacion 1,022 1,685 2,436 1,714 3,626 5,796 Interest 609 818 946 865 1,495 1,912 Standing 1,070 1,360 1,685 1,261 1,886 2,545 Total 6,419 8,162 10,111 7,566 11,315 15,269 Cost per Passenger-km ($b) 0.19 0.24 0.30 Cost per ton-km ($b) 0.90 1.39 1.88 -.. .. "" ...ii ,", ~ ..J '" ;II .." .... l ~ . . .. Ii.. __ .. . ... .. .... .,. - ., M ... 111 .. jJ -., "" .. . when the observed load factor of 94 percent is applied, converts to $b 0.22 per passenger-kilometer. (The tariffs charged by truck and bus operators are discussed later in this chapter.) For trucks, the average costs per ton-kilometer of- fered are $b 0.57, $b 0.88 and $b 1.18 for the three types of road surface, with an average of $b 0.77 for each ton- kilometer offered. The observed average occupancy for large trucks, after the exclusion of very short trips, was 63 percent, which results in an average cost per ton- kilometer of $b 1.22. Vehicle Cost Parameters - As described in Working Paper 33a, operating costs for the six types of vehicle under different road conditions were calculated from the equations in the HDM model (2), which were based on research in Kenya (3). Modifications were made to allow for the effects of high altitude in Bolivia and to reflect better the costs of spare parts and labor for maintenance. The equations require a s~all number of parameters describing each type of vehicle. These include eight prices: the prices of the vehicle, a set of tires, a liter of fuel and of oil, the wage rate for the crew and the cost of a mechanic, the average value of time for passengers, and the rate of interest. Other parameters describe the vehicle: by type, fuel, horsepower, weight, standard axle equivalents, number of passengers, annual driving hours, annual kilometers on paved, gravel and earth roads, vehicle life on paved, gravel and earth roads, and the percentage of standing costs to total costs. The cost parameters are given separately in financial and economic terms and identifying the foreign exchange component. The parameters adopted for the base year, 1977, are shown in Table 5-6. The annual kilometrage indicates how much could be done entirely on paved roads, or gravel, or earth and a current average, in fact, of course, most vehicles drive on all three kinds of roads, in varying proportions. The same applies to vehicle life. The difference between financial and economic cost consists of taxation or, in the case of fuel and oil, subsidy; and in the case of the rate of interest it represents the discount for future inflation. (2) HDM Model, Model Description and Users Manual, Transpor- tation, Water and Telecommunications Department, World Bank, Washington, June 1979, revised January 1980. (3) H. Hide, S. W. Abaynayaka, I. Sayer and R. J. Wyatt, The Ken a Road Trans ort Cost Stud : Research on Vehicle Operating Costs, Transport and Road Researc Laboratory, Report 672, Crowthorne, England, 1975. -67-  1It • 0. Table 5-6 " " VEHICLE PARAMETERS ADOPTED FOR COST EQUATIONS,1977 " VEHICLE TYPE '. ,. 26-Seat 36-Seat Medium .FUel PARAMETER Jeee gasoline gasoline Bus Bus diesel Small Truck gasoline Truck gasoline large Truck diesel ',. BHP 128 153 145 176 138 243 Weight (tons) 2.3 8.9 11.7 4.4 6.4 11.6 ".«i Axle equivalent 0 0.40 1.47 0.11 0.97 4.70 No. of passengers 3.8 24.0 35.0 8.0 7.0 6.0 111\ Annual hours driven 420 1,454 1,655 790 950 1,200 Annual kms: - paved roads 25,000 88,500 89,000 37,800 41,800 50,400 - gravel roads 21,500 63,500 63,600 21,000 23,200 28,000 'It - earth roads 20,500 53,500 55,000 16,400 18,200 21,900 22,000 80,000 28,500 31,500 38,000 .1 - average 82,000 Vehicle life (years): - paved roads 13.0 8.5 7.5 12.0 12.0 11.0 0" - gravel roads 12.0 7.5 6.7 11.5 11.0 10.0 - earth roads 11.0 6.8 6.0 10.0 10.0 8.0 Vehicle cost:financial 223.8 529.2 1194.1 271.9 402.3 1012.1 ($b 000) economic foreign exchange 154.3 104.3 436.5 330.5 1071.5 511.9 235.5 164.3 354.0 236.9 902.4 754.3 " Cost of tires:financial 1,464 3,090 4,170 1,464 3,090 4,791 ($b ) eoonomic 1,330 2,812 3,789 1,330 2,818 4,352 ,'\ foreign exchange 935 1,890 2,470 935 1,890 2,700 .a Cost of mechanic:financial 130 120 116 130 120 116 ($b /hour) econanic 125 115 llQ 125 115 110 Cost of crew: financial 10 25 39 25 25 25 ~ -, ($b /hour) econcmic 10 24 37 24 24 24 1 0/ Cost of fuel: financial 1.67 1.67 1.33 1.67 1.67 1.33 ($b IUter) econanic 2.50 2.50 2.50 2.50 2.50 2.50 '! foreign exhange Cost of oil: financial 12.50 12.50 12.50 12.50 12.50 12.50 .., ($b lliter) econanic 13.20 13.20 13.20 13.20 13.20 13.20 foreign exhange 2.10 2.10 2.10 2.10 2.10 2.10 Value of t:irre ($b /hour) 12 5 5 4 4 4 Standing costs %: financial 15 21 21 21 21 21 econanic 14 20 20 20 20 20 ;~ Fate of interest % : financial 20 20 20 20 20 20 econanic 10 10 10 10 10 10 foreign exch. 12 12 12 12 12 12 SOURCE: Costos del Transporte Automotor (Costs of Road Transport)f Working P~per 33 f Bolivia National Transport Study. -68-  Projections to 1989 and 1999 - The general treatment of future price changes was discussed above, but specific technological changes can also be expected to affect costs and prices. It is expected that, with rising wages, the employment of assistants on trucks and buses will decline and, with rising vehicle ownership, the proportion of private vehicles with paid drivers will also decline. Similarly, the annual kilometrage will change for various reasons: that of light vehicles will diminish as ownership rises; that of buses will rise slightly as roads improve; and that of trucks will rise substantially as utilization improves. The life of light vehicles is expected to rise and that of other vehicles to fall, reflecting the trends in annual kilometrage. The foreign exchange components will be greatly affected by the likelihood that diesel fuel will have to be imported and by the development of a Bolivian motor industry~ the latter is also expected to lower the relative costs of small buses and trucks. Table 5-7 shows the projection of future costs and compares them to those of the base year. Road Maintenance Costs - Data on the costs of road maintenance were availabre-from the World Bank Highway Maintenance Project in Bolivia, 1978, and from studies by L. F. Baquero in 1975 and the SNC in 1978. Other data came from recent estimates by SNC for work on the Cochabamba-Santa Cruz and La Paz-Oruro roads. A detailed analysis of the data is described in Working Paper 38B arid principal results are given in Appendix 5B. Costs were first established for specific maintenance operations. On paved roads, six operations were costed: routine maintenance, patching, sealing, surface dressing, overlaying and rehabilitation. On gravel roads there were four operations: routine maintenance, grading, spot regravelling and gravel surfacing; and the first three of these are applicable to earth roads too. The costs adopted as typical for these operations are given in Table 5-8 for 1977 and projected to 1985. To allow for consistent differences found in different parts of the country, partly caused by availability of materials, regional factors were introduced which raise costs in Santa Cruz, Beni and Pando, and lower them in La Paz and Oruro. The 1985 projections were needed for the HDM Model; they took into account some improvements in organization and changes of technique recommended by the World Bank as well as changes in the relative prices of input factors. Road Tariffs - Surveys were made of truck and bus tariffs on a sample of routes. The data were analyzed to establish equations relating tariffs to distance, quality of road, and department. The analysis is described in Appendix 5B and Working Paper 33C. -69-  ,. Table 5-7 • VEHICLE COST PARAMETERS, 1977, 1989 AND 1999 • VEHICIE TYPE • 26-Seat 36-Seat Srrall .Mediun Large 11 PARAMETER ~ Jeep Bus Bus Truck Truck Truck . Financial Costs Cost of Vehicle 1977 223.8 529.2 1194.1 271.9 402.3 1012.1 ,. ($b 000) 1989 1999 246.0 269.0 450.0 400.0 1313.0 1433.0 200.0 180.0 442.0 483.0 1113.0 1214.0 • Cost of Tires 1977 1,464 3,090 4,170 1,464 3,090 4,791 ($b ) 1989 1,464 3,090 4,170 1,464 3,090 4,791 11 1999 1,318 2,781 3,378 1,318 2,781 4,312 II/ Cost of Mechanic 1977 130 120 116 130 120 116 ($b /hour) 1989 1999 132 144 132 144 132 144 132 144 132 144 132 144 ,. Cost of Crew 1977 10 25 39 25 25 25 41 ($b /hour) 1989 5 28 43 28 28 28 1999 3 30 34 17 30 30 Cost of Fuel ($b /liter) 1977 1989 1.67 5.0 1.67 5.0 1.33 5.0 1.67 5.0 1.67 5.0 1.33 5.0 "I Cost of Oil ($b /liter) 1999 1977 1989 10.0 12.5 17.5 10.0 12.5 17.5 10.0 12.5 17.5 10.0 12.5 17.5 10.0 12.5 17.5 10.0 12.5 17.5 .,. 1999 22.5 22.5 22.5 22.5 22.5 22.5 Eoonanic Costs " "~ Cost of Vehicle 1977 154.3 436.5 1071.5 235.5 354.0 902.4 ,iI ($b 000) 1989 170.0 400.0 1179.0 170.0 389.0 993.0 1999 185.0 350.0 1286.0 150.0 475.0 1083.0 Cost of Tires 1977 1,330 2,812 3,789 1,330 2,818 4,352 't ($b ) 1989 1,330 2,812 3,789 1,330 2,818 4,352 1999 1,197 2,530 3,410 1,197 2,536 3,917 Cost of Mechanic 1977 125 115 110 125 115 110 ($b /hour) 1989 132 132 132 132 132 132 " 1999 144 144 144 144 144 144 ,,. Cost of Crew 1977 10 24 37 24 24 24 ($b /hour) 1989 5 27 41 26 27 27 1999 3 29 32 16 29 29 " Cost of Fuel 1977 2.5 2.5 2.5 2.5 2.5 2.5 4' ($b /liter) 1989 5.0 5.0 5.0 5.0 5.0 5.0 1999 10.0 10.0 10.0 10.0 10.0 10.0 r"t Cost of Oil 1977 13.2 13.2 13.2 13.2 13.2 13.2 ($b /liter) 1989 17.5 17.5 17.5 17.5 17.5 17.5 d 1999 22.5 22.5 22.5 22.5 22.5 22.5 i 'i Foreis::: Exchange Costs Cost of Vehicle 1977 104.3 330.5 511.9 164.3 236.9 754.3 ($b ')00) 1989 114.7 200.0 563.0 80.0 260.5 830.0 1999 100.0 100.0 580.0 50.0 200.0 800.0 ' ." Cost of Tires 1977 93S 1,890 2,470 935 1,890 2,700 . ,j ($b) 1989 935 1,890 2,470 935 1,890 2,700 1999 840 1,700 2,200 840 1,700 2,430 11\ Cost of Fuel 1977 0 0 0 0 0 0 ($b /liter) 1989 0 0 5.0 0 5.0 5.0 , ·,1 1999 0 10.0 10.0 10.0 10.0 10.0 Cost of Oil 1977 2.1 2.1 2.1 2.1 2.1 2.1 i ~t, ($b Iliter) 1989 3.0 3.0 3.0 3.0 3.0 3.0 1999 4.0 4.0 4.0 4.0 4.0 4.0 sotJR:E: Costes del Transpc?rte lIlltarotor (Costs of Road Transport), Working Paper 33, t ~!\ Bolivia National Transport Sttrly. I ~I -70- fill ,·,1 GRAVEL ROAD GRADING NEAR PUERTO VILLARROEL  Table 5-8 UNIT COSTS OF ROAD MAINTENANCE OPERATIONS (1977 pesos) 1977 C(13TS 1985 <XlSTS REGICNAL FAC'roRS Foreign Foreign La Paz/ Sucre/Potos1/ OPERATICN UNIT Financial Econanic Exchange Financial EconaTlic Exchange Cbba. Oruro Santa Cruz Tarija Beni Panek> ----- Paved Roads 1.10 1.00 1.10 1.10 Routine kIn 7,000 6,700 1,420 7,900 6,700 1,420 1.00 0.90 2 1.10 1.00 1.10 1.10 Patch.in.g m 101 100 50 93 92 45 1.00 0.95 2 13 13 8 12 13 7 1.00 0.85 1.00 1.40 1.40 Sealing m 1. 00/1. 20 (2) 2 52 52 33 48 51 31 1.00 0.85 1.00 1.40 1.40 Surface Dressing m 1. 00/1. 20 (2) m 3 2,500 2,600 1,590 2,300 2,400 1,470 1.00 0.85 1.00/1.20 (2) 1.00 1.40 1.40 Overlaying Pehabilitation (1) I -.J -Surface Treat:m!:mt kIn 489,500 497,080 303,400 440,550 447,372 281,250 1.00 0.85 1. 00/1. 20 (2) 1.00 1.40 1.40 f-' I -Asphaltic Concrete kIn 1,120,000 1,140,000 700,000 1,008,000 1,030,000 650,000 1.00 0.85 1. 00/1. 20 (2) 1.00 1.40 1.40 Gravel Roads Routine kIn 5,250 5,000 1,060 5,250 5,000 1,060 1.00 0.90 1.10 1.00 1.10 1.10 Grading kIn 850 870 500 1,607(3) 1,644(3) 945 (3) 1. 00 0.95 1.05 1.00 1.05 1.05 3 106 105 39 180 (3) 179(3) 66(3) 1.00 0.90 1.50 1.00 3.50 1.60 Spot Gravelling m 3 76 79 46 180(3) 185 (3) 109(3) 1.00 0.90 1.50 1.00 3.50 1.60 Gravel Resurfacing m Earth Roads kIn 3,150 3,000 634 3,150 3,000 634 1.00 0.90 1.10 1.00 1.10 1.10 Routine kIn 680 700 400 643 660 380 1.00 0.95 1.05 1.00 1.05 1.05 Grading 3 106 105 39 180 179 66 1.00 0.90 1.50 1.00 3.50 1.60 Spot Gravelling m (1) Assuming 6 meter wide highway. (2l The tw::> values apply to near santa Cruz City and distant fran Santa Cruz respectively. (3) Increase mainly due to intrcduction of watering and canpacting techniques. SOURCE: Costs of Road Maintenance, Working Paper 38b, Bolivia National Transport Sttrly.  The quality of roads varies greatly, with substantial effect on vehicle operating costs. In the context of the road tariff analysis, the sample of routes was therefore divided into paved and unpaved, with a third class "mixed", formed by routes partly paved and partly unpaved. These three classes were further divided according to their terrain: flat, undulating or mountainous; and flat routes were divided between the lowlands and the Altiplano. The interdepartmental services of both trucks and buses were analyzed separately from the interprovincial services. Naturally it is more difficult to relate the longer-distance tariffs to department and terrain, since they relate to services passing through several different regions. The analysis of interprovincial truck tariffs was made separately for agricultural and non-agricultural commodities. With both classes of tariff the standing costs, i.e. the costs of waiting to be unloaded and loaded at each end of the trip, were found to be significant. This cost is unrelated to the distance of the trip or the quality of the road. It was established at 2.4 pesos per hundredweight (46 kg -- the usual tariff unit) for agricultural products and 0.83 pesos per hundredweight for non-agricultural products. The analysis of interprovincial truck tariffs showed a clear relationship with the quality of road. For agricultural products the effects of road surface, terrain and department could be detected, but for non-agricultural products only the road surface showed up as significant. For interprovincial bus tariffs the differences between departments were clear but could not be isolated from differences of terrain. The tariff equations adopted for the 1977 model are shown in Figure 5-3 and in Table 5-9. In the 1989 and 1999 model tests, bus and truck tariffs were adjusted slightly upwards to allow for rising component costs. The adjustments were calculated by applying the component cost increases to a breakdown of bus and truck operating costs. River Costs and Tariffs There are no port facilities on Bolivian rivers, that is to say, no quays, piers, cranes or storage buildings. River ports are simply places where the river bank permits boats to stop and load or unload and where there is road access. There are therefore no port costs of any significance-apart from the laborers who do the loading and unloading. River costs consist almost entirely of boat and barge operating costs, an analysis of which is fully reported in Working Paper 34 and summarized in Appendix SC. -72-  LEGEND y Agricultural Products 28 1 I. 2. Unpaved mountainous Unpaved undulating 3. Unpaved flat I 26 4. Mixed mountainous 2 24 5. Mi xed undulating 3 6. Mixed flat and paved mountainous 4 5 +- .<= 22 7. Paved undulating ., 0' 8. Paved flat 6 ~ ., 20 Non- Agriculturol Products 7 - "0 c: 18 9. Unpo v ed 10. Paved 8 '" :I: "- 16 .ci =- ~ ~9 14 10 .... 12 c t- 10 8 6 4 2 -+--------------~--------------r_------------~--------------_.--------------~ 0 50 100 150 200 __ 250 Ix Di stance - km NORMAL FUNCTIONS RELATING INTERPROVINCIAL TRUCK TARIFFS TO DISTANCE U)fell.u~ :Jlme'llt and Aj:>oceide:. FIGURE 5-3  Table 5-9 TARIFFS FOR TRUCKS AND BUSES,1977 TARIFF EQUATION(l) TRUCKS Interdepartmental: y = 0.036 x Interprovincial,agricultural products, normal equations: paved, flat y = 2.4 + 0.064 x undulating y = 2.4 + 0.070 x mountainous y = 2.4 + 0.076 x mixed, flat y = 2.4 + 0.076 x undulating y = 2.4 + 0.081 x mountainous y = 2.4 + 0.086 x unpaved, flat Y = 2.4 + 0.086 x undulating Y = 2.4 + 0.092 x mountainous y = 2.4 + 0.096 x These equations should be factored as follows by region: Altiplano 0.75 Yungas 0.80 Valleys 0.80 Santa Cruz 1.25 Pando 1.50 Interprovincial, non-agricultural products: paved y = 0.83 + 0.048 x unpaved y = 0.83 + 0.057 x BUSES Ihterdepartmental: y = 0.225 x Interprovincial: La Paz (Altiplano) y = 0.137 x La Paz (Yungas) Y = 0.256 x Oruro y = 0.140 x Santa Cruz y = 0.181 x Potosi y = 0.186 x Chuquisaca y = 0.184 x Cochabamba y = 0.180 x Tarija y = 0.230 x (1) y=tariff per hundredweight, x = distance in km SOURCE: Tariffs of Road Transport, Working Paper 33c, Bolivia National Transport Study. -73-  Average Costs - A sample survey of boats on the river Mamor~, the most important waterway within the country, was conducted in the first half of 1979. Thirteen usable replies were obtained. The boats used are made locally to a tradi- tional design; trains usually consist of a motor vessel with two non-powered barges attached to its sides. Two typical sizes were selected as representative although there are a variety of sizes in use. They were as follows: BOAT CHARACTERISTICS Class A Class B Length (m) 15.3 20.5 Width (m) 3.4 4.5 Draft (rn) 1.06 1.30 Horsepower 57 77 Number of barges Length of barges (m) 2 20.8 2 22.5 • j Width of barges (m) 4.5 5.0 Draft of barges (m) 1.2 1.4 Capacity of boat (tons) 25 50 Capacity per barge (tons) 51 92 Total capacity (tons) 127 234 The average operating costs found for these two classes of boats are given in Table 5-10. They are divided into fixed costs (per year) and variable costs related to the number of hours in operation. The fixed costs came to about $b 400,000 and $b 569,000 for Classes A and B respectively, and the variable costs were $b 56 and $b 72 per hour. It is difficult to generalize about the costs per ton-km because it varies enormously according to boat utilization (i.e. percentage of hours spent operating), load factors (i.e. percentage of carrying capacity used) and the speeds possible on the river. The costs per ton-km given in Table 5-10 are estimates for the conditions existing in 1979. They show that costs per km tend to be less over long dis- tances, because a smaller proportion of time is spent waiting in port. Costs are considerably higher upstream than downstream because of the difference in speed. For example, the total time required to navigate upstream from Guayaramer1n to Puerto Villarroel is 220 hours, compared to 120 hours in the opposite direction. It can therefore be deduced that • the variable costs per boat-kilometer are about 80 percent higher for upstream than downstream operation on the Ichilo- Mamor~ River. Because of the lower speeds, however, river boats can operate more hours per day when moving upstream. The total elapsed time for the Guayaramer1n to Puerto Villarroel trip is usually about 12 days, i.e. only 20 percent above the 10 days required for the downstrea~ trip. However, the allocation of fixed costs to either upstream or downstream transport is somewhat academic in view of the current underutilization of the river fleet -74-  Table 5-10 COST OF TRANSPORT ON ICHILO-MAMORt:: RIVER SYSTEM TYPE OF BOAT (INC.BARGES) Class A Class B Annual Costs per Boat (with Barges) ($b) ($b) Crew 243,360 341,050 Maintenance 43,250 77,900 Depreciation on boat 38,400 41,700 Interest on boat (10 percent) 18,400 31,250 Depreciation and interest on barges 19,600 21,750 Insurance 35,000 51,000 Registration 2,350 4,000 400,360 568,650 Operating Costs per Operating Hour Diesel Fuel 36.0 48.0 Oil 6.0 7.0 Oil change 2.7 3.3 Oil filter 1.2 1.4 Grease 4.0 5.0 Fees 6.5 7.5 56.4 72.2 Estimated Cost per ton-km Downstream Puerto Villarroel-Trinidad, 470 km 0.87 0.60 Trinidad-Guayaramerin, 860 km 0.62 0.42 Puerto Villarroel-Guayaramer1n, 1317 0.48 0.32 km upstream Trinidad-Puerto Villarroel 0.94 0.65 Guayaramer1n-Trinidad 0.68 0.46 Guayaramerln-Puerto Villarroel 0.55 0.36 (l)Bypassing access to Trinidad port. SOURCE: River Transport Operating Costs and Tariffs, Working Paper 34, Bolivia National Transport Study. -75-  It was found that boat utilization on the Mamore was very low. Boats operating between Puerto Villarroel and Trinidad were spending more than 75 percent of their time in port, and boats between Puerto Villarroel and Guayaramer!n about 65 percent. Load factors recorded in official port records were low, averaging about 40 percent, but probably some cargo is not recordedi a considerable amount of undocu- mented- imports from Brazil is known to travel up the Mamore. It can be calculated that if time in port were reduced to • eight days per trip (a quite normal time in Bolivia) and load factor raised to 60 percent, transport costs per ton-km would be reduced to about one third of the estimates in Table 5-10. Given the same utilization rate and load factor, the larger boat gives transport costs about one third less than the smaller boat, but the question is whether the former can indeed achieve the same utilization rate and load factor. There is evidence from tariff rates that the Mamore is cheaper than the other main navigable rivers. If the rates reflect costs, the costs on the Mamore are less than half those on the Beni, Madre de Dios and Orthon. The reasons for this lie mainly in the fact that the boats using the latter rivers are much smaller than those on the Mamore. Other Cost Data - Since river transport in Bolivia incurs virtually no infrastructure costs, there should be no difference between long-run marginal and average costs of a boat operating with reasonable utilization rate and load factor. Projects to improve rivers and ports are designed to save boats from delays and sometimes accidents, on which there is no cost information. The impact of time saving on operating costs is easily calculated from the data of fixed • • and variable costs. More complex cost functions, as in the case of road and rail improvements, are not required. River Tariffs - River tariffs are published by the Bolivian Navy for the rivers Beni, Madre de Dios and Orthon. Information about the river Mamore was obtained separately. Their analysis showed that there was no simple, general relationship between tariffs and distance that could be used for modelling. Rivers run at different speeds and have • different seasonal patterns. The speed of a river varies from one stretch to another. Some rivers contain hazards • • -76- • and some can take larger craft than can others. Hence tariffs are specific to each stretch of river and do not conform to any simple formula. In addition, tariffs are generally higher for upstream than for downstream transport. Consequently matrices were prepared, with different tariffs in opposite directions, for the zones or sub-zones directly linked by river transport. As there is no significant traffic between the four important rivers, separate matrices were defined for each as shown in Appendix 5C. The published tariffs, though government controlled, may be regarded as commercial rates which, by and large, cover the costs of boat operation. Since there is no capital investment in river infrastructure, and virtually no expenditure on roaintaining the rivers, the tariffs should reflect the economic costs of river transport quite well. Subject to changes in co~ponent costs, the same tariffs were used in all the 1989 and 1999 tests. Air Transport Costs and Tariffs The costs of air transport are relatively straightforward to analyze because only a small part concerns infrastructure, and the operating costs of a given aircraft are practically independent of the quality of the infrastructure. In fact, air transport is the only mode for which operating costs can be studied without reference to infrastructure. Of course, bigger and better runways permit larger aircraft to use an airport, with important repercussions on operating costs, but they do not affect significantly the costs of a given type of aircraft. When runways are put out of action by floods, which often happens in the Beni, there could be a fall in aircraft utilization and a consequent rise in unit operating costs, but this does not necessarily follow. Scheduled Services - Lloyd Aereo Boliviano (LAB - the National Airline) furnished data on the four principal types of aircraft engaged on scheduled services, both domestic and international. They are the F-27, which is used on secondary routes with flight times generally between 30 and 60 minutes, the B-727-100 and B-727-200, used on international and primary domestic routes, with flight times of about 30 minutes internally and 2.0 to 4.0 hours externally, and the B-707, which is employed only for freight. The detailed analysis of the data is described in Working Paper 35 and summarized in Appendix 5D. The average operating costs per flying hour, as observed for these four types of aircraft, are shown in Table 5-11. -77-  " • Table 5-11 COSTS OF SCHEDULED AIR TRANSPORT, 1978 (pesos per flying hour) • AIRCRAFT TYPE F27 B727-100 B727-200 B707 • Depreciation and interest 3,924 9,276 16,280 11,602 Major repair and overhaul 1,082 2,510 4,108 5,182 Maintenance 5,130 6,708 6,940 8,252 'I Fuel: domestic 2,358 9,042 6,944 8,124 foreign 3,470 7,618 8,088 Crew 1,594 3,516 4,300 4,020 Insurance 324 1,062 1,498 846 Passenger services 312 2,008 4,646 Landing fees 398 4,124 4,906 5,436 Administration 4,874 14,426 14,788 15,712 Total 19,996 56,142 72,028 67,262 Foreign exchange component 52.3% 51.7% 71.5% 70.0% Number of seats 36 125 165 Number of tons 40 Average flying speed(km /hour) 360 650 730 770 Seat-km offered per flying hour 12,960 81,250 120,450 Ton-km offered per flying hour 30,800 Cost per seat-km offered ($b) 1.54 0.69 0.60 Cost per ton-km offered ($b) 2.18 SOURCE: Costos del Trans orte A~reo Re ular (Costs of Scheduled Air Transport , Working Paper , Bolivia National Trans- port Study. -78- • • The costs per seat-km show how much more economical the larger planes are, given enough traffic to fill them. The data are easily divided into fixed and variable costs if one treats crew costs as being all variable. The equations are as follows: F27 C = 10.5 + 0.0109 t B727-100 C = 52.0 + 0.0314 t B727-200 C = 89.6 + 0.0395 t B707 C = 80.3 + 0.0391 t where: C = annual cost in million pesos, and t = annual number of flying hours. Non-scheduled Services - A large variety of aircraft are used for non-scheduled services, mostly freight transport, and their utilization varies greatly, from almost nothing to 100 hours a month. More than a hundred routes are flown, 90 percent of them based on La Paz or Cochabamba, and most flights take between 60 and 90 minutes. Clearly the costs of operation vary widely and are difficult to uncover. A questionnaire w~s submitted to a number of principal operators and was followed up by informal interviews. Attention was limited to the five aircraft types in most coromon use, which are all of elderly vintage. The analysis is given in Working Paper 35a. The results shown in Table 5-12 are based on the observed average route length of 1.15 hours and utilization of 600 flying hours a year. The costs per ton-km are complicated by the effect of airport altitudes; take-off load capacity at La Paz and other high airports is much lower than in the lowlands. Fortunately the greater part of air freight is meat IPoving from the lowlands to the Altiplano and there is no lack of capacity in the return direction. Air Transport Tariffs - For air freight, tariffs were not used in the modelling work. The passenger fares used in the 1977 analysis were the actual fares charged by LAB and TAM (Transportes Aereos Militares). For 1989 the following equation was used based on 1977 tariffs: y = 73 + 0.872 x where, y = fare in pesos, and x = distance in km. -79-  • • • Table 5-12 COSTS OF NON-SCHEDULED AIR TRANSPORT, 1978 (pesos) • AIRCRAFT TYPE Douglas Douglas Curtiss convair Martin C-54 DC-6 C-46 440 404 Capital cost ($b 000) 2,574 2,760 2,600 2,380 2,410 'Ibtal cost per flying hour: 8,300 11,400 7,200 7,350 7,000 Cost per ton-hour offered: IDwland 1,100 1,100 1,300 1,550 1,550 Andes 1,750 1,300 1,550 1,850 1,900 NOTE: Route length 1.15 hours, annual utilization 600 hours flyirq. SOURCE: Costos del Trans orte A~reo No-re ular (Costs of Non- sc e u e Air Transport , Working Paper 35a, Bolivia National Transport Study. •.. ,. -80-  For 1999 the equation was increased slightly to allow for increases in component costs, especially fuel, taking into account an expected improvement in aircraft utilization. Airport Costs - Data o~ 1979 air?~rt c~sts were received . froID AASANA, the National A~rportAdmin~strat~on. 'Ihese costs were estimates, given separately for the four regions of AASANA and broken down into six functional divisions: • Runway operations and maintenance • Air navigation • Aircraft services • Passenger services • Freight services • Administration. It is not certain how reliable this functional breakdown is; the figures are estimates and could perhaps be improved, but they were the only usable data available. Table 5-13 gives the summarized data. An attempt was made to find statistical relationships between these costs and aircraft IDovements during 1979, for which good statistics are available. Aircraft were divided by type into Classes 1 to 4 according to whether they normally used only Class 1 airports, or Classes 1 and 2, or Classes 2 and 3, or all four classes. The cost data were then related to statistics of landings, passenger movements and freight-tons moved. The analysis is described in Working Paper 35B and summarized in Appendix 50. The results are subject to considerable possibility of error and the work could usefully be refined in the future. With this proviso, the average airport costs associated with each flight are as follows (only Bolivian airport costs are included): CLASS OF AIRCRAFT 1 2 3 4 ($b per flight) Runway costs 1,575 262 147 110 Air navigation costs 3,750 2,250 400 237 Aircraft service costs 3,348 695 144 67 8,673 3,207 691 414 -81-  Table 5-13 AIRPORT COSTS BY REGION AND FUNCTION, 1979 (million pesos) REGION Santa Cocha- FUNCTION La Paz Cruz bamba Trinidad Total Runway operations and maintenance 6.5 2.9 2.8 4.1 16.3 Air navigation 18.5 11.6 10.3 8.9 49.3 Aircraft services 19.1(1) 1.9 1.0 0.8 22.8 Passenger services 6.7 2.3 1.8 0.8(2) 11. 6 Freight services 4.4 2.8 2.2 2.3(2) 11.7 Administration 10.2 5.2 6.0 4.5 25.9 65.4 26.7 24.1 21.4 137.6 (1) A large amount of aircraft repair and maintenance is concen- trated at La Paz; the cOpt covers much more than the inci- dential needsof aircraft on routine flights through La Paz. (2) An adjustment was made between passenger and freight services at Trinidad. • SOURCE: Airport Costs and Tarrifs, Working Paper 35b, Bolivia " National Transport Study. • -82- • • •  plus, Cost per international passenger: $b 36 Cost per domestic passenger: $b 7 Cost per ton of freight: $b 458. Thus a B-707, with 150 passengers, flying into Bolivia from abroad would be allocated airport costs of $b 9,736 for the Bolivian end of the flight. A B-727, with 100 passengers, flying within Bolivia would be allocated $b 3,907, and an F-27 carrying 2 tons of freight within Bolivia would be allocated $b 1,330. Airport Tariffs - Charges for transport services at airports are of four kinds: e Landing fees $ Air navigation charges • Aircraft parking • Passenger airport taxes. Landing fees, shown in Appendix 5D, are related to the category of airport and maximum gross weight of the aircraft. International flights pay more than domestic. International charter flights pay a surcharge of 50 percent on the standard rates. There are separate charges for domestic meat carriers and air taxis. A surcharge of 30 percent is applied to operations between 6 p.m. and 6 a.m. and on Sundays and public holidays. Flights using Bolivian airports pay air navigation charges according to the following formulae: International Flights: $b 0.255 n d..JW National Flights: $b 30VW where: n = number of radio aids available on route, d = distance flown over national territory, in nautical miles, W = maximum gross weight in tons. The landing fee entitles an aircraft to stay up to six hours. Afterwards there is a surcharge for aircraft parking of 40 percent for each 24 hours or part thereof. International passengers pay an additional airport tax of $b 250 when boardingi domestic passengers pay $b 18. -83-  All the above charges, except passenger airport taxes, are automatically included in aircraft costs and passed on in fares and freight charges. In the model, therefore, only passenger airport taxes were included. Pipeline Costs Cost data were estimated for refined product pipelines, as shown in Working Paper 22 and in Appendix SE. Capit.al and operating costs are both·affected by mountainous tei- rain. tn.order to produce typical figures for Bolivia capital and maintenance costs have been estimated for' mount- ainous terrain of medium difficulty in construction. Operat- ing costs have been calculated for pipelines in flat terrain assuming a pipeline-use factor of 15 percent. The data given in Table 5-14 are a guideline and must be modified if the pipeline terminates at a higher elevation than its origin. The pump power required to perform the lift is estimated at 4 h.p. per 100,000 cubic meters per year for each 100 meters lift. The capital cost of pumping facilities needed to perform the lift is estimated at 15,000 pesos per horse power. The transport costs in the table would rise by 0.02 pesos per barrel or 0.12 pesos per cubic meter for every 100 meters lift. If a pipeline falls in elevation the costs will be reduced in the same order as just described. • • • • • -84-  Table 5-14 FINANCIAL TRANSPORT COSTS BY PIPELINE, 1977 TRANSPORT COST PER 100 KM. QUANTITY Total Cost Unrt Cost BPD{l) Cu.m./year Fixed variafile variaOIe- - Total ($b OOO/year) ($b /eu.m./100 km) 4-ineh diameter 5,730 332,500 8,940 4,120 12.4 39.3 4,500 261,100 8,580 3,760 14.4 47.3 3,000 174,000 8,340 3,540 20.3 68.3 I 00 2,000 116,000 8,200 3,480 30.0 100.7 U1 I 6-ineh diameter 16,660 966,700 13,420 5,400 5.6 19.5 13,000 754,400 12,900 4,400 5.8 22.9 9,500 551,300 12,380 3,800 6.9 29.4 5,000 290,200: 11,940 3,480 12.0 53.1 8-ineh diameter 32,600 1,891,700 31,100 8,740 4.6 21·1 25,000 1,450,700 22,000 7,140 4.9 20.1 18,000 1,044,500 19,300 5,960 5.7 24.2 12,000 696,300 18,620 5,280 7.6 34.3 6,000 348,200 18,400 5,140 14.8 67.6 (1) Barrels per day. SOURCE: Existing Pipeline Charagteristies_, Working Paper 22, Bolivia National Transport Study.  CHAPTER 6 THE FINANCIAL FRAMEWORK OF THE TRANSPORT MODES CHAPTER 6 THE FINANCIAL FRAMEWORK OF THE TRANSPORT MODES The costs of transport facilities are borne initially by the agencies, companies and individuals who provide them. The costs must be recovered, else the service will break down, and tariffs are the principal means whereby this is done. But tariffs have a big influence in determining the volume of traffic and choice of service. Ideally -- and this will be the subject of Chapter 17 -- tariffs should be designed to induce the optimal use of the transport system, maximizing the surplus of benefit over cost; but these tariffs may fail to recover costs fully, or they may yield an excessive profit. The careful application of grants, subsidies or taxes can help to reconcile the aims of tariff policy with the need of each financially independent unit to balance its books. A source of finance must also be found for capital investment. In short, a financial frame- work is needed to enable the transport industry to follow efficient policies of pricing and investment. Th;s chapter examines the present financial framework. The Roads Financially, the roads are the most complex part of the transport system because there is no generally accepted way of charging for their use. The National Road Service (SNC) is the agency responsible for all the roads covered by this Study, and its annual expenditure is summarized in Table 6-1. At this stage in the development of road transport in Bolivia, it is not surprising that about 71 percent of SNC expenditure between 1976 and 1979 was on construction, 9 per- cent on improvement and only 20 percent on maintenance. In 1976-79, 63 percent of SNC revenue came from the Treasury and 31 percent from lending agencies. Little more than 1 percent came directly from road users. It is quite reasonable, in view of the high expenditure on investment, that much of this should be financed by borrowing. The di- rect contribution from road users comes largely from transit tolls but the revenue of $b 9.83 million in 1979 only just exceeded the cost of running the transit posts. Indirectly, road users pay far more, through fuel tax and customs duties 'on vehicles and vehicle parts. These payments do not go directly to SNC but are certainly user charges. Table 6-2 shows these user charges. Vehicle owners also pay annual registration fees to municipal authorities but these are used for local roads and are therefore not included in the table. -87-  Table 6-1 .. • EXPENDITURE AND INCOME FOR ROAD . • CONSTRUCTION AND MAINTENANCE, 1976-79 EXPENDITURES :1976 MILLIONS OF CURRENT PESOS 1977 1978 1979 .. • : Maintenance .. - Permanent - 'n:Ircfx::lrary 139.26 23.26 199.91 25.09 224.96 25.85 263.07 33.83 • - Special 6.03 6.02 231.02 19.61 270.42 34.42 331.32 • Subtotal' 168.55 • .. - Studies - Local improvements: i) Regular program 0.09 29.69 0.29 34.56 0.32 34.25 32.36 0.34 .• ii) With third parties 5.87 4.73 7.28 4.54 iii) Special program 29.82 55.93 78.04 113.86 Subtotal 65.47 95.51 119.89 151.10 .dI Construction - S'bldies 16.64 9.05 22.65 33.81 .. - Major improvements 178.92 236.25 223.72 235.47 - New roads 232.61 592.51 896.78 807.06 . Subtotal 428.17 837.81 1,143.15 1,076.34 GRAND 'IOTAL 662.19 1,164.34 1,533.46 1,558.76 m::x:ME • ... National Treasury COntributions 535.56 Special COntributions 864.23 890.21 827.19 . - Fram domestic institutions 1.63 4.33 124.86 22.38 - Fram foreign institutions 0.95 10.21 2.68 '30.15 Credits from IDB and similar • • institutions Specific taxes 133.88 297.43 505.83 580.83 . - 'lblls 8.33 9.04 9.63 9.83 - Road Service If=vy (1) - La Paz agricultural custans - Chba.agricultural custc:ms 2.52 1.46 2.05 2.67 1.51 2.42 2.33 0.42 2.76 1.90 1.34 3.92 .. Other incorce 2.78 12.14 8.99 6.12 'lbtal 689.16 1,203.98 1,547.71 1,483.66 Danestic Source 554.33 896.34 1,039.20 872.68 Foreign Source 134.83 307.64 508.51 610.98 (1) 'lax levied on residents in lieu of three days of road work required .. annually by law. sa:JRCE: Annual Balances of the National Road Service (SN:). -88- .. .  Table 6-2 ROAD USER CHARGES, 1976-79 (excluding municipal charges) MILLIONS OF CURRENT PESOS 1976 1977 1978 1979 Gasoline tax 7.0.7 75.5 81.4 102.0 Diesel tax 8.3 10.2 11.6 18.9 Customs duty, vehicles 281.1 201. 7 274.6 442.4 Customs duty, tires and 5.8 4.6 7.7 8.1(1) tubes Customs duty, spare parts 2.0 1.6 1.8 2.1 (1) Direct payments to SNC 14.4 15.6 15.1 17.0 Total 382.3 309.2 392.2 590.5 (l) Estimates SOURCE: Diez Anos de Estad!stica Petrolera, 1970-1979 (Ten Years of Petroleum Statistics), YPFB; and import Perrni ts. -89-  Road user charges increased dramatically in 1979 and will .• rise sharply again in 1980. In the three years 1976-78 they amounted to $b 1,084 million which comfortably covered the $b 951 million spent on the maintenance and local improvements, (such as straightening of curves or building of small bridges). ... The largest source of this revenue is customs duty for vehicles, of which 72 percent was paid for passenger vehicles and 28 per- cent for goods vehicles. This subject is further discussed in • Chapter 17, together with the formulation of a recommended policy for road user charges. More information on current road use- charges is presented in Appendix 6A. • Truck and Bus Operators - All trucks and most buses operat- ing a public service within Bolivia, are privately owned by small • companies or individuals organized in syndicates. The syndicates are strong and obviously endeavor to maintain tariffs at a pro- • fitable level. On the other hand, tariffs and fares are con- trolled by central government authorities. It is difficult to discover the true financial position of these private operators. There are no published financial balance sheets and an investigation depends upon personal contacts with ... operators, who are often reluctant to discuss the matter. For many of the smaller operators, it must be doubted whether they themselves have a coherent view of their own financial position. • Some figures were obtained, and are presented below (for 1977): SERVICE ROUTE INCOME COSTS DEFICIT .. . (thousands of 1977 pesos) Large Bus La Paz-Oruro 466 590 144 La Paz-Potosi 459 665 206 La Paz-Tarija 543 609 66 • • Large Trucks (General) 364 458 94 • Medium Trucks (General) 189 265 76 . • That is to say, poor results were obtained by all types of operators of inter-urban transport. However, for both passenger and freight transport, income shown in this table is an underestimate. Additional income is '•. obtained for trucks for transport between intermediate points ,. (12 percent of the income shown in the table) and for passenger transport, from the transport of small packages (2 to 5 kg) and documents. Costs were also reduced by postponement of vehicle servicing (only really possible with new vehicles) and amorti- .. • zation and avoidance of custom duties. Since 1977, two changes have occurred~ a rise in the price of • ., fuel in December 1979 and a further rise in the price of fuel, and also lubricants, in July 1980. These are summarized as follows: • -90- • tilt  Year -------.;;;..=--_._-- ITEM 1977 1979 1980 Price of fuel ($b liter) - diesel 1.5 4.0 6.0 - gasoline 1.5 2.5 6.0 - lubricants 12.0 12.0 lS.~ Exchange Rate in pesos per US$ 20.0 25.0 25.0 These price rises were accompanied by raises in te..rL"'fs! as follows: ITEM~ ____________________________ YEAR 1979 1980 Increase in tariffs - interdepartmental + 45 percent + 25 percent - interprovincial + 35 percent + 20 percent Following these two modifications, the financial situation of the road transport operators is much improved with a balance between costs and income. The Railroads The annual expenditures and income of ENFE for 1976-79 are given in Table 6-3, excluding new investment. Operating costs in this period comprised about 64 percent of total expenditure, the remainder being accounted for by depreciation, interest and taxation. The depreciation costs included here are higher than those normally calculated by ENFE, but have been accepted as being more realistic. Interest charges leaped from $b 24 million in 1976 to $b 166 million in 1979, i.e., from 4 percent to 20 percent of earned income. Taxation took another 8 percent. Table 6-3 also shows that freight provides 88.5 percent of earnings, passengers only 11.5 percent. ENFE is supposed to cover its own costs out of income. During 1976-79 there was a deficit rising from $b 88 million (12.1 percent] in 1976 to $b 309 million (27.1 percent) in 1979. Of total earnings, 46.0 percent came from import traffic and 19.2 percent from export traffic. Domestic freight contri- buted 16.7 percent, other freight-related services 6.6 percent and passengers 11.5 percent. The importance of international freight traffic to the railroads should be emphasized; without it they could not continue to operate. It is probable that much of this international traffic would be relatively insensitive to rises in tariffs. -91-  Table 6-3 EXPENDITURE AND INCOME OF NATIONAL RAILROAD COMPANY, 1976-79 MILLIONS OF CURRENT PESOS • EXPENDITURE 1976 1977 1978 1979 • Operations: • Administration Traffic 66.7 121.7 86.1 134.9 128.4 143.3 134.2 163.1 • Workshops 97.8 99.9 77.3 93.2 • Signalling 76.3 75.9 95.5 95.0 Track 111.9 109.8 177.7 193.2 Signals and Telecommunica- 9.7 11.1 15.3 16.9 tions ,. .•'. Subtotal 484.1 517.7 637.5 695.6 Depreciation Track and installations 86.9 99.7 86.8 106.6 Rolling stock 82.2 80.3 86.6 110.5 .- Interest 24.1 58.9 85.1 166.5 .. Taxes on traffic 41.8 47.2 62.1 52.2 Income tax TOTAL EXPENDITURE 7.1 726.2 8.3 812.1 9.0 967.1 9.3 1140.7 .• INCOHE .• Passengers (1) Freight (1) 78.4 509.4 87.9 587.5 92.4 665.9 88.2 659.6 ..• ... Baggage and parcels 13.2 15.5 20.8 25.1 Freight services 37.3 52.0 36.8 58.9 TOTAL INCOME 638.3 742.9 815.9 831. 8 (DEFICIT) (87.9) (69.2) (151.2) (308.9) - as percent of total expenditure (12.1%) (8.5%) (15.6%) (27.1%) i. (1) Including Tax SOURCE: Planning Department, ENFE; and Social Security Fund of Railroad Employees .. ... -92-  Of passenger earnings, two thirds came from ferrobus and first-ciass (pullman) services. With growing competition from road transport and aircraft, the prospects for this class of traffic cannot be good unless services are improved. New investment and capital replacement have been fin- anced by borrowing since the beginning of the rehabilitation plan in 1972 agreed with the International Development Association. Loans have been received of SUS 71.7 million from the World Bank, SUS 15.4 million from Japanese suppliers and SUS 10.8 million from various other suppliers. Credits of SUS 29.3 million have been given by the Treasury and Central Bank. These loans amount to SUS 127.2 million or $b 3,180 ~illion. Repayment is due to extend over the next 40 years, with a peak of $b 288 million due in 1982, assuming that no further loans are contracted in the mean- time. Clearly ENFE is in a position of acute financial difficulty. Traffic receipts in 1979 were disappointing and there can be no reason for optimism in the next few years. There must soon be either a big change in tariffs, which might not remove the deficit, or a change in ENFE'S financial obligations. Chapter 17 describes the study's recommendations for a future tariff policy. More infor- mation on current rail finances are presented in Appendix 6C. River Transport River craft are privately owned and their tariffs are controlled by the Ministry of Transport and Communications. It is extremely difficult to discover the true financial pos- ition of the operators. Their costs are dominated by the cap- ital costs of boats and barges, which cannot be established with any accuracy. Incomes almost certainly include items that do not enter into the accounts because of the likely occurrence of unrecorded trade. The situation on the Mamor~ may be indicative. Tariffs were fixed in 1974 and remained unchanged until 1980 when, after a petition by boat owners, they were substantially raised. Between 1974 and 1980 the Bolivian retail price index rose by 80 to 90 percent, and it is difficult to see how boat operators survived unless they were making large profits in 1974 or else receiving tariffs above the official rate. Even at the end of the period, just before the tariffs were raised, there appeared to be excess capacity on the Mamore and no sign that operators were going out of business. In the circumstances it would be difficult without an in-depth study of all relevant factors to corne to a view about the true financial position of this sector. -93-  .. Air Transport As described in Chapter 4, domestic transport services are provided by several airlines, the most important of which is Lloyd Aereo Boliviano (LAB). The operation of the major airports is in the hands of another Government agency, AASANA. The following paragraphs discuss the current financial situation of LAB and AASANA. • • Lloyd Aereo Boliviano (LAB) - Expenditure by LAB in 1978 and 1979 is summarized ~n Table 6-4. Over 80 percent of expenditure is related to passenger traffic. No further details of operating expenditure are published by LAB. A large additional item is the servicing of debt, which jurr@ed from $b 104 million in 1978 to $b 225 million in .. liP 1979, Le. from 11.6 percent to 18.1 percent of operating costs. The data on income given in the same table show a big increase in passenger revenue, from $b 694 million in 1978 to $b 954 million in 1979, while freight revenue declined from $b 237 million to $b 205 million. Total earned income exceeded operating costs by 12 and 8 percent in 1978 and . • 1979, respectively. However, because of the large debt payments, earned income fell 9 percent below total expenditure in 1979. International flights contributed 73 and 74 percent of income in 1978 and 1979, respectively. .• The deficit was partly covered by taxes on tickets and • freight. International tickets pay a tax of 13 percent, • domestic tickets 4.2 percent. Freight pays 4 percent if international, 4.2 percent if domestic. For the National Airline alone, these taxes produced $b 34.3 million in 1978 and $b 51.4 million in 1979, which LAB was allowed to retain. .. • A number of routes, particularly in the north of the country, are designated "unremunerative routes" and are • eligible for subsidies from the government. In 1978 and 1979 LAB estimated the loss on these routes at $b 80 million of which 29 percent was accounted for by Riberalta, 24 percent • ,. by Guayaramerin and 10 percent Oruro, but this estimate has not been agreed to by the Government. LAB's indebtedness really dated back to the inauguration of the international routes to Miami and Buenos Aires. Be- tween 1975 and 1979, some $b 1,310 million has been borrowed, 90 percent of it from abroad, principally for the purchase of . new aircraft. The outstanding debt at the beginning of 1980 was $b 923 million. .. • -94- .• • •  UNLOADING ON THE ROAD NEAR TRINIDAD IN THE HIGH-WATER WET SEASON PUERTO ALMACEN IN THE DRY SEASON THE DRY SEASON PORT OF PUERTO VARADOR (BOLIVIAN NAVY PETROL BARGES BEING UNLOADED) RIVER PORTS AT TRINIDAD  Table 6-4 EXPENDITURE AND INCOME OF LLOYD AEREO BOLIVIAN~ 1978-79 MILLIONS OF CURRENT PESOS EXPENDITURE 1978 1979 Passengers, including servicing of debt 815.1 1,198.9 Freight, including servicing of debt 189.4 212.7 Other services 0.3 0.3 Less servicing of debt - 104.4 - 224.7 Operating costs 900.4 1,187.2 Servicing of debt 104.4 224.7 Total expenditure 1,004.8 -1,411.9 INCOME Passengers 693.8 953.9 Excess baggage 27.5 72.7 Mail 3.2 3.9 Freight and parcels 236.7 205.2 Other 50.7 50.4 Earned income 1,011.9 1,286.1 Retained taxes 34.3 51.4 Total 1,046.2 1,337.5 PROFIT (DEFICIT) 41.4 (74.4) - as percent of total expenditure 4.1% (5.3%) SOURCE: Planning and Economic Analysis Division and Cost and Analysis Division, Lloyd A~reo Boliviano. -95-  II In order to correct the poor financial situation, domestic fares for LAB were increased in early 1981 approximately 40 percent (but there were regional deviations). This change came too late for detailed analysis by the Study, but it is thought that the increase has corrected LAB's financial position. • National Airport Administration (AASANA) - Expenditure and income-Xor AASANA are shown in Table 6-5 for 1978. The figures differ from those published by AASANA for deprecia- tion and interest, but are based upon detailed calculations carried out by the Study. It is clear from the table that AASANA was in a very grave financial position if true . account is taken of depreciation and interest payments which • approximate $b 48 million, 30 percent of the total. Indeed on the income figures provided by AASANA, the institution was unable to cover even its operating costs, yielding a 17 percent deficit on these alone. The table also reveals a distortion in the make-up of payments to AASANA related to costs: the income from passengers and landing fees far .. • more than covers the cost of providing these services--both categories serve to subsidize the low tariff charge for flight services, which cover only 15 percent of their own costs. . • AASANA is supposed to cover all its costs, including those arising from new investment, out of income. This it has found impossible to do in view of the high cost of .. • providing adequate aeronautical facilities on the relatively low volumes of traffic using the airport system. However . • the picture has somewhat improved recently: in December 1979 a 65 percent increase in tariffs was approved. This, on the basis of the half-yearly figures available for 1980, should allOW AASANA to easily cover its operating costs, .. .. and come close to meeting its financial charges. However, if total costs are to be covered (as defined on an adequate • basis), further increases may be required. • More information on current AASANA finances is presented in Appendix 6B. • . • .. - .. • .. -96- .  Table 6-5 EXPENDITURE AND INCOME OF AASANA ($b million) COSTS 1978 Operating Costs: Runways 15.69 Flight Services 80.78 Passenger-related Services 16.32 Total Operating Costs 112.79 Depreciation and Interest 48.00 Total Expenditure 160.79 INCOME Runways 58.14 Flight Services 12.21 Passenger-related services 25.83 Total Income 96.18 (DEFICIT) (64.61) - as percent of total expenditure (40.2%) SOURCES: AASANA and National Transport Study. -97-  CHAPTER 7 THE BOLIVIAN ECONOMY CHAPTER 7 THE BOLIVIAN ECONOMY The purpose of the transport system is to satisfy econo- mic and social needs for accessibility, and it is the econo- mic needs that form the main object of this Study. No one could deny that the needs of the economy for cheap, efficient transport are being met inadequately and in some places not at all. But the places in greatest need of improvement, from the economic pOint of view, are not always obvious. An essen- tial part of planning improvements to the transport system is to analyze the national economy, and in particular those as- pects giving rise to a demand for transport, and to make a forecast -- or if that is too difficult, a set of alternative scenarios -- of its long-term development. Socio-Economic Factors Affecting Transport Demand The demand for long-distance transport, as distinct from local transport, arises from people's wish to travel or send goods between different parts of the country, or abroad, and their ability to pay for it. It thus arises principally from the geographical distribution of population and economic ac- tivities, and from income levels. The leading factors may be defined as: (i) The size and distribution of population. (ii) The size and distribution of the urban population, because urban areas tend to generate much more traf- fic per head than do rural areas, even when incomes are equal. (iii) The cash incomes of the population, urban and rural, by area. (iv) Distribution of industry -- primary, secondary and tertiary .. - by area; in Bolivia, where non-agricul tu- ral development is small and scattered, it is de- sirable to know the specific location of the principal factories, oil wells, mines, mills and other plants. (v) Demand for industrial inputs; an important part of national freight traffic consists of raw materials and semi-finished products required by industry, i.e. traffic moving between different industries; this is the most difficult demand factor to measure and analyze. (vi) Imports and exports. -99-  , \ In order to analyze and forecast these socio-economic factors, and to introduce them quantitatively into the na- tional transport model, it is convenient first to study and forecast the national economy as a whole, with regard to broad regional trends, and then to distribute methodically the resulting national or regional figures between the 77 analysis zones (defined in Chapter 3). This work has been presented at length in eight working papers, and some of the principal find- ings are included in Appendices 7A, 7B, 7C and 7D of this Report. Working Paper 46 gives a comprehensive appraisal of the Bolivian economy today. Working Paper 45 considers its prospects over the next two decades and -- drawing on the Study Workshop described in Working Paper 44 -- identifies major development opportunities, which are quantified in Working Paper 47. The steel industry is treated separately in Working Paper 48. Because of the great uncertainties ahead, however, three alternative development patterns (or scenarios) are for- mulated in Working Paper 46a, drawing on the various development opportunities in different ways. Then Working Paper 41 presents the socio-economic statistics, at the national level, for 1977 with forecasts for 1989 and 1999, and these are developed at the zonal level in Working Paper 49 for analytical use. The Bolivian Economy in 1980 Bolivia is rich in natural resources. The size of France and Spain together (but with only six percent of their popula- tion) , the country contains a wide variety of climates, soils, vegetation and geological structure. Some 87 percent of the land is suitable for agricultural production; 1 percent is used for farming and about 40 percent for extensive cattle raising. Virtually every type of crop, from tropical to temperate, can be grown. There are vast areas suitable for livestock of every kind. Enormous forests contain large supplies of valuable tim- ber and, although the country lacks sea coast, it possesses large rivers and lakes well stocked with fish. " Despite this abundance of agricultural potential, Bolivia is better known for her wealth in minerals. Practically every valuable mineral is produced, except coal and bauxite, and even the latter is thought to be available. The one important mineral that is lacking is coal, but this is compensated for by substan- tial reserves of oil and natural gas. These generous resources are the heritage of a compara- tively tiny population -- about 5 million -- and lack of man- power in some regions is doubtless one reason why the natural potential of the country remains largely untouched. In 68 percent of the country, this being mostly in the lowlands, the population density is less than 1 per square kilometer, and in _ 85 percent of the country it is less than 5 per square kilometer. .e -100- Agricultural output is a small fraction of the physical poten- tial, and mineral production, apart from tin and antimony, could also be very much larger. Known oil reserves have been substantially exhausted but much exploration for new reserves remains to be done. Meanwhile the exploitation of natural gas is at an early stage. A small amount of light industry to sup- ply the domestic market has been established in La Paz, Santa Cruz and Cochabamba. Shortage of manpower, however, is not the main reason for lack of development. Indeed nearly a million Bolivians appear to have emigrated since 1950, for lack of opportunity in their home country. The main obstacles to development are geographical and historical. The terrain makes transport ex- tremely difficult and costly, both in the mountains, for ob- vious reasons, and in the flat lowlands where large, shifting rivers have to be crossed and huge areas are frequently flooded. But even if the produce of the country could be moved easily, there are few markets for them. Bolivia is surrounded by similarly vast and empty areas whose few inhabitants can pro- duce much the same agricultural commodities themselves and, until recently, had little use for Bolivia's minerals. Her position in the heart of South America, thousands of miles from the big markets for agricultural and mineral products, has not encouraged development. The course of history has led to numerous handicaps of which four may be mentioned. For various reasons, among which the desire to avoid tropical diseases was important, the popu- lation concentrated in that corner of the country which was among the least fertile, so that today, from an economic view- point, there is a highly inefficient distribution of population. Secondly, through most of its history, the country has been suffering from political instability. Thirdly, a long series of wars and quarrels with her five neighbors have not contributed to the growth of transport and trading links nor any other form of economic cooperation, although there has been progress of this sort in recent years. Fourthly, as a sad consequence of her border disputes, Bolivia has lost direct access to the sea, which has not encouraged the development of an efficient export corridor. Thus, for many formidable reasons, Bolivia is -- in the correct meaning of the term an exceedingly undeveloped country; this does not mean that she is very backward or poor, but rather that she has not gone far towards the high pros- perity that her natural and human resources should make possible. -101-  Population - The last two censuses showed a growth of population from 2,704,000 in 1950 to 4,648,000 in 1976, an annual growth rate of 2.1 percent, but these figures omit a number of forest dwellers and may also suffer an appreciable • 'i\ percentage of undercounting. The natural rate of increase is believed to be about 2.8 percent per annum, the difference o~ 0.7 percent per annum being attributable to net emigra- t~on, mostly to Sao Paulo, Buenos Aires, Lima, Arica and other cities in the neighboring states. But these figures are unreliable, since the statistical coverage of migration, residence, births and deaths, is far from complete. It is no secret that many Bolivians are living without official re- cognition in Argentina, Brazil and Chile. Both censuses showed the highly uneven distribution of the population, with the great majority living on the alti- plano or nearby in the high valleys, but some significant changes between 1950 and 1976 were revealed. The remainder of the country, measuring about 75 percent of the area, showed a strongly increasing trend in its population which ,+ rose from 395,000 to 1,052,000, or from 14.6 percent to 22.6 percent of the total. This was accounted for mainly by the development of cotton, sugar and oil near the city of Santa Cruz and, to a much smaller extent, the growth of cattle farming in the Beni. A second trend, to be observed in almost all countries, was the shift of people into the cities (Shown in Figure 7-1): the nine departmental capitals grew by 162 percent, raising their share of the population from 20.4 to 31.2 percent; other urban areas grew by 125 per- cent, raising their share from 14.7 to 19.2 percent, while rural areas grew by only 31 percent, diminishing from 64.9 to 49.6 percent of the population. But the biggest shift was abroad. This flood of nearly one million emigrants, which is an enormous number for a country with less than 3 million people at the beginning of the period, came mainly from the Departments of potosf, La Paz, Cochabamba and Chuqui- saca, in that order, and as a result, 14 provinces (in potos!, I' ,0., Cochabamba and Oruro departments) actually declined in popula- tion. The National Product - It is certainly no criticism of the statisticians to say that Bolivian economic statistics should be treated with caution. The' country is bounded by 11 several thousand kilometers of uncontrolled frontier offering all kinds of opportunity for unrecorded trade. A '1' ' -102-  NUMBER OF RESIDENTS (1976) II More than 100,000 • A 20,000· 50,000 e • 5,000 10,000 .. ~ , (' \ \ I / ATRINIDAD '-~-"" '---------- 1Iinef" ~\ \ cO, ~, BAMBA' abo Soo... Por'oc""~ ....IIIoIlIcrO -...... . Q.~ia(; P"".lo IlSANTA CRUZ ORURoII ••I(!Vi ~XX Llallaqll4iA.llIftCio TARIJA rUpilOe ..... I I a 50 100 150 200 km. , \j~ J 1 , \ \ \ MAIN POPULATION CENTERS (j)fiI6u't fl'millt and J/aj)liCialed FIGURE 7-1 large part of the population is engaged, partly or wholly, in production outside the market, which does not appear in the gross domestic product (GOP). In recent years there has been rapid, if not exactly galloping, inflation which makes ex- tremely difficult the estimation of gross national product and ll other "real indicators of economic performance. The Central Bank's estimates of GOP for the period 1970 to 1980 are reproduced in Table 7-1. They suggest that, in real terms, the national product rose by just 6 percent per annum between 1970 and 1975 and by 4.3 percent per annum be- tween 1975 and 1978. Allowing for a 2 percent per annum in- crease in population, the increase in GOP per head becomes 4 percent and 2.2 percent per annum during the two periods, which is not much for a country in an early stage of develop- ment. Allowing for the fact that the urban population, which accounts for most of the recorded GOP since the output of goods and services by the rural population is largely unrecor- ded, was rising at about 3.5 percent per annum, the statistical growth of GOP may partly amount to a conversion of unrecorded rural poverty into recorded urban poverty. The five basic categories of consumption and expenditure that absorb the GOP are shown in Table 7-2 for the years 1970, 1977 and 1978. The figures appear to show clear progress, with increases per head of 25 .,Percent'in private consumption, 67 per- cent in government consumption and 29 percent in investment: but one must bear in mind that the growth of private consump- tion and, to a lesser extent, investment is exaggerated by the shift of population from subsistence to the cash economy_ Also, the reported private consumption of $b 9,710 ($US 476) per head in 1978, which is clearly a very low figure, underestimates the real standard of living insofar as it excludes the non- cash economy. Such progress as was made, however, was made in an un- satisfactory way. Total consumption (private plus govern- ment) expanded its share of GOP from 81 percent to 86 percent, while investment raised its share from 19 percent to 20 per- cent. This expansion was achieved by maintaining the percen- tage of imports at about 24 percent but allowing the percen- tage of exports to fall dangerously from 24 percent to 18 percent. In short, the growth of GOP has been achieved by a 50 percent increase in the volume of imports and only a 12 per- cent increase in the volume of exports. -103-  Table 7-1 , i TREND OF GROSS DOMESTIC PRODUCT, 1970-80 (Million pesos, at 1977 prices) PERCENT YEAR GOP INCREASE 1970 46,856 1971 48,913 4.4 1972 51,810 5.9 1973 55,570 7.3 1974 59,408 6.9 " 5.0 1975 62,398 1976 66,306 6.3 68,842 3.8 1977 70,771 2.8 1978 1979(1) 73,662 4.1 1980(1) 77,068 4.6 '1 (1) Projections. SOURCE: Document No.8, prepared by the Central Bank for the Workshop on Alternative Strategies for Regional Development, La Paz, June 1979. i, i~ . -104-  Table 7-2 GOP BY EXPENDITURE, 1970, 1977 AND 1978 (1977 prices) MILLION PESOS ANNUAL GROWTH PERCENT PERCENT TOTAL EXPENDITURES 1970 ·19/1 1978 1970-78 1970 1978 (%) Private. consumption 33,351 47,031 51,415 5.6 71.2 72.6 Government consumption 4,561 9,019 9,405 9.5 9.7 13.3 Investment 8,982 14,696 14,298 6.0 19.2 20.2 I 24.0 I-' Exports 11,277 14,512 12,659 1.5 17.9 o Imports (-) 11,315 16,416 17,006 5.2 -24.1 -24.0 Ul I GOP 46,856 68,842 70,771 5.3 100.0 100.0 Balance of Trade -38 -1,904 -4,347 PESOS GROWTH PJ\TE EXPENDITURES PER HEAD 1970 1~77 19'8 1970-78 Private consumption 7,789 9,122 9,710 2.8 Government consumption 1,065 1,750 1,777 6.6 Investment 2,098 2,850 2,700 3.2 Exports 2,634 2,815 2,391 -1.2 Imports (-) 2,643 3,184 3,212 2.5 GOP 10,943 13,353 13,366 . 2.5 SOURCE: Derived from Document No.8, prepared by the Central Bank for the Workshop on Alternative Strategies for Regional Development, La Paz, June 1979.  i. "" By 1978 the balance of trade had fallen heavily into de- ficit and was financed by foreign loans. By the end of 1978 the external debt had risen to SUS 3,011 million (or $b 61,424 ,I million at the then exchange rate) although not all this amount had been spent. The servicing of the debt was absorbing 23 percent of export earnings and thus increasing the difficulty of balancing the external account without resort to yet more loans. But the level of the debt was already so high that new loans were becoming harder to find. Bolivia's credit-worthiness was almost exhausted. The National Product by Industry - The industrial sec- tors that make up the GDP are shown in Table 7-3 together with the recorded percentage of employment, and public and private investment, in each sector. The figures are averages for the seven years 1970-76. They are purely descriptive and do not intentionally carry any argument. The employment percentages are distorted by the difficulty of estimating the percentage in agriculture; equally the GDP contribution of agriculture is grossly underestimated. The figures indicate the relative amounts of labor in each sector but not the corresponding amounts of capital. Apart from agriculture, however, the figures do show the relative importance of the main sectors of the economy. Min- ing, as would be expected, made an important contribution. The largest item in the table, however, is commerce, which may partly be a reflection of inefficient retailing and high retail margins. The part contributed by transport appears large when one considers how little transport there is in Bolivia; this reflects both the high cost of transport and the low level of other economic activities. The annual growth rates refer to the years 1970-75. The agricultural growth rate could be quite misleading, reflecting little more than a shift from subsistence to commercial agri- culture. The other figures, however, are certainly significant. The negligible growth in mineral output indicates a state of stagnation in the nation's principal export industry. The ra- IP! pid development of oil extraction is evident. There appears to have been solid growth in the important sectors of manufac- turing, construction, electricity and transport. One may also note the high growth in government expenditure. I. ,~ " H ; -106-  Table 7-3 GDP BY INDUSTRIAL SECTOR: AVERAGES FOR 1970-76 (Million pesos, at 1977 prices) CONTRIBUTION PERCENTAGE OF: ANNUAL (1) INVESTMENT SECTOR TO GDP GDP Employment GROWTH Public Private Total (%) -- Agriculture 9,801 17.6 50.0 5.4 275.3 544.0 819.3 Mining and r.!etallurgy 4,667 8.4 2.8 0.6 1,310.5 401.3 1,711. 8 Hydrocarbons 1,113 2.0 0.4 20.2 2,806.3 737.2 (2) 3,543.5 I I--' Manufacturing 8,110 14.5 7.0 6.8 1,140.6 314.4 1,455.0 0 .....,J I Construction 2,192 3.9 3.2 5.6 347.9 347.9 Commerce 10,306 18.5 5.3 5.6 325.6 325.6 Government 4,768 B.6 9.4 8.5 1,599.9 1,599.9 Housing 4,525 8.1 9.4 4.1 470.5 470.5 Electricity, Gas, Water 788 1.4 0.2 7.2 399.1 111.8 510.9 Transport and Communications 4,608 8.3 2.9 9.9 1,751.3 478.2 2,229.5 Other Services 4,873 8.7 9.4 4.6 159.6 159.6 55,751 100.0 100.0 5.9 9,283.0 3,890.5 13,173.5 (1) J:970-1975. (2) F.stimate. SOURCE: La Econom!a Boliviana (The Bolivian Economy), Working Paper 46, Bolivia National Transport Study.  · 1 The sectoral growth rates may be compared with the in- vestment figures. The large investment in hydrocarbons was clearly associated with the rapid development of that industry. Agriculture, as the country's principal industry, received little investment, but heavy investment was made in transport, mining and government. ,.~ , J The fact that 70 percent of investment was in the public sector says something about the relative strength of the '" public and and private sectors. Mining, hydrocarbons, elec- tricity, gas and water, transport and communications, even manufacturing -- all are dominated by the public sector. The public sector is also involved in agriculture. Contruction, housing and commerce are the only big industrical categories left free from state involvement. The following pages will provide some more detail about the four main productive branches of the economy: the infra- structure, agriculture, mining including hydrocarbons and industry. Infrastructure - Before the 1950s, Bolivia was little more than a mining outpost. The population lived almost ex- clusively on the Altiplano and in the high valleys of Potosi, Chuquisaca and Cochabarnba, and there was little infrastruc- ture other than that required to serve the mines. Nearly all economic activity was in private hands; the government inter- vened as little as possible in economic affairs. After the revolution in 1952, attempts were made to draw up and implement a long-term development plan, and in 1956 a 10-year plan was produced with United Nations assistance. As I~ already discussed in Chapter 4, railroads were built from Santa Cruz to Corumba in the far east and to Yacuiba in the far south; in addition, a paved road was built from Cochabamba to Santa Cruz. Thus Santa Cruz became a transport center and the surrounding area was colonized for the production of cotton, at first, then sugar and rice. Long-term planning was continually frustrated by political upheavals. Nevertheless, by the early 1970s a Ministry of Plan- ning and Coordination was in operation, with regional develop- ment committees in each of the nine departments, all under the !j , supervision of a powerful Council of Economics and Planning. Despite the earlier lack of planning, considerable progress was made. Electrification spread rapidly. The road from La Paz to Oruro was paved and industrial infrastructure was provided to form the beginnings of an industrial axis from La Paz to Santa Cruz via Cochabarnba. Airports were built and expanded. Roads were built to support colonization schemes in northern La Paz, Alto Beni, Chapare and Yapacan{, producing tea, coffee, cocoa, citrus fruits, chickens and timber. -108- /'1 ,  MINOR ROAD NEAR LA PAZ FORD ON ORU RO- POTOSI HIGHWAY FORD NEAR BOUNDARY OF LA PAZ AND BENI DEPARTMENTS RIVER CROSSINGS Private enterprise moved into the Beni to develop cattle farming on the basis of private air transport to La Paz and Santa Cruz. With road improvements to Tarija, private enter- prise began to develop vineyards and grow fruit. The auto- pista to El Alto was built and the road to the Beni started. Despite unusual difficulties, a great deal of basic in- frastructure has been put in place in 25 years, and the face of the country has been changed a lot. It is no longer just a mining outpost. Agriculture - While mining was the dominant interest in Bolivia, agriculture was neglected. Society was divided into the ~ini~g community, which had suffic·ient income to import most of ~ts needs as contra-traffic to the exports of minerals via Arica and Antofagasta, and the campesinos who engaged in subsistance agriculture mainly on the Altiplano and in the valleys. A certain amount of food was sold to the small ur- ban communities but none was exported. The country could af- ford to import food that was not produced locally. With the rapid growth of the cities and the relative decline of mining, this situation could not last. Much lar9- er output from the agricultural community was needed to sup- ply the rest of the population. With an improving transport system there was also a growing case for developing major agricultural exports. But the industry had not yet achieved these ends. Agrarian reform after 1952 led to the disintegra- tion of large farms into small holdings, normally of less than 7 hectares. Productivity was low, quality poor, and supplies unreliable. There was little modern technology and little possibility of obtaining credit for improvements. Agricultural production, after growing steadily during the first half of the 1970s, appears to have stagnated in the last few years. The rate of growth between 1970 and 1978 is given as 3.5 percent per annum at constant prices, which is just about equal to the growth in the urban population. Table 7-4 shows details of 1978 production. Potatoes were the lead- ing crop in terms of value, followed by sugar cane, bananas, coffee, coca, yucca and corn. It is a big step from subsistence farming and selling small surpluses in the local market, to a highly organized and competitive export industry. Given the infrastructure, the possibility is there, but before it becomes a reality many other improvements must be made: improvements to species, quality control, packing, hygiene and marketing. The industry is a long way from achieving these needs. -109-  Table 7-4 AGRICULTURAL PRODUCTION, 1978 PRODUCT PRODUCTION VALUE (tons) lS6 million) Corn (ma!z blando) 132,400 422.3 Rice, unpeeled 88,600 291. 9 Wheat 59,900 248.5 Barley in grain 74,800 233.2 Other cereals 15,300 55.6 Cereals 371,000 1,251.5 , -1. Potatoes 793,000 2,949.2 Yucca 306,200 569.6 Other root crops 121,800 250.8 Root crops 1,221,000 3,769.6 Tomatoes 36,800 208.6 Corn (choclo) 105,600 207.0 Beans 58,000 180.1 Onions 48,800 173.2 Other vegetables 76,600 424.8 Vegetables 325,800 1,193.7 Bananas 399,000 840.6 Peaches 28,800 201. 7 Grapes 22,000 193.5 Oranges 80,000 186.7 Other fruit 102,600 405.8 Fruit 632,400 1,828.3 Sugar cane 3,477,500 991.1 Cotton in fiber 17,400 424.0 Other industrial crops 85,200 319.9 Industrial crops 3,580,100 1,735.0 Coffee beans 22,200 727.9 Coca leaves 19,500 571. 0 Cacao seeds 3,200 150.9 ,. ' Tea 1,500 7.3 Stimulants 46,400 1,457.1 Fodder 746,800 1,048.7 Total crops 6,923,500 12,283.9 THOUSANDS OF HEADS Cattle 490 1,513.6 Pigs 822 446.5 Sheep 1,719 367.9 Fowl 5,579 306.8 Goats, llamas, alpacas 342 106.4 Total livestock 2,741.2 TOTAL AGRICULTURE 15,025.1 SOURCE: Department of Statistics, Ministry of Agriculture (MACA). Oivision of Economic Studies and Aaricultura1 Marketing, MACA. Plan Operativo Aqropecuario 1979, MACA, Office of Sectoral planning. -110-  Nevertheless, some progress has been made. Table 7-5 shows exports of all commodities by group in 1970, 1977 and 1978. At constant prices, agricultural exports rose from $b 280 million to $b 783 million and processed goods from agricultural materials rose from $b 136 million to $b 479 million. Nevertheless, together they only comprised 10 per- cent of the value of exports. Meanwhile deficiencies have opened up in wheat, barley, cotton, soya and groundnuts, all of which can be grown in Bolivia. In 1978 only 60,000 tons of wheat were produced; a further 142,000 tons had to be im- ported, together with 121,000 tons of flour. Mining and H*drocarbons - Seventeen minerals are pro- duced, of which t e ten most important are shown in Taple 7-6. Tin is predominant and, in value terms, accounts for 72 percent of ' the minerals listed in the table. Bolivia is one of the world's largest producers of both tin and antimony. As shown in Figure 7-2, all mining is situated in the Altiplano apart from a small amount of iron ore extraction recently started at Mutun, but the Brazilian Shield in the east of the country is another large area believed to contain mineral deposits. Most of the mines are nationalized. The expansion of the industry in the last ten years has been : severely restricted by several factors. The world demand for tin has grown very slowly. The mines have been heavily taxed and this has reduced the incentive to exploration. World consumption of tin is actually less now than during World War 2, and Bolivian output reached its maximum in 1929. The poor progress of mining has been offset by the in- creased exploitation of oil. The National Hydrocarbons Company (Yacimientos Petrol!feros Fiscales Bolivianos - YPFB) was set up in 1936. After 1955 foreign companies were allowed to search for oil, and the Bolivian Gulf Company was success- ful in discovering the fields at Colpa, Caranda and R!o Grande, indicated on Figure 7-3. In 1969 the company was nationalized but other private companies have continued to explore in various parts of the country. Oil and gas have been found and are being produced at another six fields in Santa Cruz, three in Chuquisaca and six in Tarija. Production of oil peaked in 1973, at 17.3 million barrels, and has been falling since, to 11.8 million barrels in 1978. Known resources of oil have also fallen, from 205 million barrels in 1971 to 135 million in 1978. At first most of the oil produced was for exporti now domestic consumption takes the greater part and exports have declined. Since the oil tends to be mostly light, there is a growling scarcity of heavy oils for domes- tic consumption and it is expected that imports will soon be necessary to meet the demand. -111-  Table 7-5 VALUE OF EXPORTS, 1970, 1977 AND 1978 (1977 prices) VALUE IN MILLION PESOS PRODUCT 1970 1977 1978 - i· " Agriculture: Industrial crops 221 728 561 l' :! Livestock products 90 129 Forestry products 59 132 93 Subtotal 280 950 783 Processed agricultural goods: Refined sugar 31 458 254 Meat 74 Wood products 31 240 225 Subtotal 136 698 479 Manufactured goods 214 284 I' Minerals 9,388 9,810 9,197 .' Crude oil 701 1,348 755 I, 11 Natural gas and LPG (1) 1,348 1,425 GRAND TOTAL 10,505 14,368 12,923 f' (1) Liquid petroleum gas. SOURCE: Various. " -112-  LEGEND: Ag SILVER Au GOLD 8; BISMUTH ell COPPER Pb LEAD Sn TIN WOLFRAM Zn ZINC Unificodo Sn,Ag 5uer pt,r( -0 Sn o 50 100 150 200km. ~L~~~~__~~~~~__~' " \ , 1 i \j THE MINING BElT FIGURE 7-2  'l'able 7-6 PRODUCTION OF TEN PRINCIPAL MINERALS 1978 PRODUCT PRODUCTION VALUE (tons) ($b million) Tin 30,880 7,761.0 Antimony 12,672 798.2 Silver 200 759.8 Zinc 59,322 722.9 Lead 18,041 239.2 Wolfram 3,170 201. 7 Gold 0.77 102.7 Copper 3,266 88.5 Bismuth 482 45.3 Cadmium 196 20.1 TOTAL 128,230 10,739.4 SOURCE: Data from Ministry of Mining and Metallurgy quoted in Plan Operativo 1979, Ministry of Planning and Coordination. -113-  Production of gas started in 1960 and reached 151,246 million cubic feet in 1973, since when the level of output has remained fairly stable. Nearly all the gas and most of the oil comes from Santa Cruz. Production statistics for hydrocarbons are shown in Table 7-7. Industry - Bolivia clearly possesses the raw materials and the power necessary for manufacturing industry but she lacks what is more important, a large enough domestic market and a large enough labor force to develop the specialization and economies of scale necessary to compete with the world's industrial powers. The vast supporting structure of an indus- trial society is almost wholly lacking. However, there are prospects of Bolivia becoming an exporter of processed agri- cultural produce. With regard to other industrial goods, the membership in the Andean Pact and similar trade associations may contribute to the strengthening of this sector. There is also a potential for the further development of certain light industries, particularly those where trans- port costs penalize the imported product. Some 2,285 esta- blishments were listed by the Ministry of Industry, Commerce and Tourism in 1978, and 650 new industrial companies were registered between the beginning of 1974 and May 1979, employ- ing nearly 10,000 people. About half the industrial firms appear to be in La Paz and Santa Cruz; Oruro and Cochabamba are next in order of importance. About half the industrial firms are engaged in the food and drink industry; next in importance are metal work- if ing and textiles. Table 7-8 shows the production of industrial goods in 1970, 1977 and 1978. The annual growth rate in this sector was 6.4 percent which compares with an average growth rate of 4 percent in the population of the four industrial cities between 1950 and 1976. The industrial sector contributed 14.5 percent towards GDP in 1970-76 and 6 percent of exports, mostly sugar and timber. I' The Future of the Economy II -. Transport planning requires the best possible view of socio-economic development during the next twenty years. There is no official economic plan and it is not possible to fore- I, ' cast so far ahead with any confidence, especially in Bolivia. Rather than become committed to a single, fixed forecast, it is better to r~ze the possibility of alternative paths or IJ patterns of development and to seek a transport plan which is least vulnerable to a forecasting failure. The "least risk" transport plan is that which, while taking due account of the "most likely" forecast, also takes account of "less likely" forecasts, so that no disastrous mistakes will have been made, . , should one of the latter prove correct. -114-  LEGEND A Oil ! Oil and Gos m Gos ~ Reserves \ ) "\ r \ , \ \ i , ; oTRtNtDAD oLA PAZ COCHABAMflA BuloBula~ ( "\ ... cOfanJ !CllIIXl ~NTA CRUZ Piroimirl a m8 Na,aniiJl;;m;m CLo Pff10 P. a CJRio Grandf SUCRE o " ~Comln o POTOSI Mootl<l9<ldOA / GUOlrUY Camalinlli ~ A B""MVisto oil/110 I , N I ~ I TARIJA o ml~~ / o 50 100 150 200 km t , l \ OIL AND GAS FIELDS Ci)f,16uft [lmillt and J'i1J(jocialea FIGURE 7-3  Table 7-7 PRODUCTION OF HYDROCARBONS, 1970, 1977 AND 1978 PRODUCTION IN THOUSAND TONS VALUE IN MILLION PESOS (1) PRODUCT 1970 1977 1978 1970 1977 1978 Crude oil (2) 667 642 410 158 1,349 846 Gasoline and similar 492 860 907 304 1,292 1,385 Aviation fuel 12 56 66 11 181 211 Liquid gas 5 51 67 4 54 78 Natural gas 2,272 2,198 1,343 1,578 I I-' I-' Lubricants and grease 6 13 14 40 181 190 U1 I 1,182 3,894 3,662 517 4,400 4,288 (1) At 1977 prices. (2) As final product. SOURCE: Diez Afios de Estad!stica Petrol era de Bolivia (Ten Years of Petroleum Statistics in Bolivia), 1969/1978. For reasons of product definition, this table does not exactly coincide with others in this report.  Table 7-8 !111 INDUSTRIAL PRODUCTION, 1970, 1977 AND 1978 PRODUCTION IN VALUE IN THOUS1>.ND TONS HILLION PESOS (1) PRODUCT 1970 1977 1978 1970 1977 1978 Food, beverages, tobacco 502 934 1,023 1,840 3,302 3,474 r '--: Textiles and leather 89 101 103 1,921 2,387 2,485 Chemicals, petroleum products, rubber, plastics 1,798 1,127 1,263 1,071 1,313 1,394 Wood products 82 162 166 168 448 468 Non-metallic building materials (2) 204 494 510 272 371 385 Fabricated metals, rrachinery 12 45 48 68 285 303 Primary metals 21 26 7 182 193 Paper products, printing 7 20 21 68 180 183 Others 2 4 4 162 264 284 2,696 2,908 3,164 5,577 8,732 9,169 (1) At 1977 prices. (2 ) Excluding some bulky materials. SOURCE: La Econom!a Boliviana (The Bolivian Economy), Working Paper 46, Bolivia National Transport Study. -116-  In looking to the future, therefore, one needs to con- sider first the various possibilities before trying to specify which are more or less likely. It is advisable, particularly at the present time, to divide the future into the medium term, defined as the next five years, and the long term, till the end of the century. The Medium Term - In 1980, it has been shown that the Bolivian economy is in a situation of exceptional difficulty. The deficit on foreign trade has to be drastically reduced, and quickly, and the external debt has to be serviced without further net borrowing. Inflation is running high and the industrialized countries are in a state of serious recession. Bolivian agriculture is almost stagnant, the world tin ~arket has been affected by the recession and may be further h~t by American destocking, and Bolivian oil supplies are running down while oil prices soar. It is difficult to see how the country can pass through the next five years without reducing imports, by one means or another, and forcing more output into exports. This will mean almost certainly a reduction in living standards and a slump, or at least stagnation, in numerous industries. It is unlikely that GDP will grow rapidly during this period and the funds for public investment are likely to be restricted. The biggest hope of a solution lies in the export of nat- ural gas, but this cannot happen on a sufficient scale until a pipeline is built to Brazil and ultimately to Sao Paulo. This could be achieved by 1985. In any event, it is to be hoped that by 1985 the economy will be back in balance and a new surge forward may then begin. The Long Term - The possibilities of development in the long term are so numerous and wide-ranging that there is a real question of choice. It is not possible to do everything at once and there is no unanimity on what should be done first. There are, in fact, very different views on the best direction for development, whether it should be towards more mineral ex- ploitation, or towards agriculture, or towards industry. There are those who put political integration first, which implies regional dispersal of development funds. There are those who would move the population to the resources, i.e. away from the Altiplano, and those who would pile investment into the Altiplano and keep the people there. Some would seek develop- ment within the protective walls of either the Andean Group or a southern association, while others see more future in the much larger but more demanding markets of the northern hemi- sphere. -117-  · .,l Whichever way the economy develops, the crucial factor for transport planning will be the speed at which it happens. The growth of GDP and the general demand for transport of all sorts will depend partly on the continuing growth of popula- tion, which will be greatly influenced by the rate of emigra- tion, a highly unpredictable factor. Growth will also depend on the rapidity with which projects are implemented, markets established, people settled in new work and new places, and I· ·11 output targets achieved. It will depend, in other words, not only on correct economic decisions but on drive, hard work, sacrifice and good organization. These things are surely re- lated to political wisdom, leadership and stability. As one is entitled to take an optimistic or pessimistic view of these matters, the best approach is to consider first the various major developments that could take place by the end of the century and then to put them together in scenarios of varying degrees of realization. Mining - The prospects for tin mining are not good. For 30 years before the 1952 revolution, no major discoveries were made by the three big mining companies. After their nationalization, almost no exploration has been done by the National Mining Company - COMIBOL. Exploration and expansion have been frustrated by the high sales tax on minerals, and Bolivian mines also suffer from high costs of production and transportation. On the other hand, the Bolivian Geological Service (GEOBOL) has been set up by the State to look for new mineral deposits, and steps have been taken to encourage and support small mining companies. There is always the possibility of major discoveries in the Brazilian Shield. Also, there are still large -- though low-grade -- reserves of tin around existing mines. The most likely mining development, however, is the iron ore deposit at Mutun. Immense reserves of high-quality ore, easily extracted by open-cast mining, await increased exploitation. The mine is conveniently located for cheap water transport down the Paraguay River. If markets can be established, and so far they have not been, production could reach millions of tons a year. Petroleum and Natural Gas - There appears to be no reason to be optimistic about the discovery of more oil fields and it seems likely that Bolivia has little future as an oil producer. " The prospects for natural gas, though, are far better. It is expected that by 1985 a pipeline will be in place from Santa -118- PIPELINE ON RIO GRANDE RAIL ALTIPLANO SOUTH OF LA PAZ BRIDGE NEAR SANTACRUZ RIVER CROSSING NORTH OF SUCRE PIPELINES Cruz to Sao Paulo and that 400-500 million cubic feet a day will be exported to Brazil and Argentina, bringing enormous relief to the balance of payments. The future price of natural gas, however, is clearly a matter of speculation. Agriculture - Many large opportunities for agricultural development can be found in various parts of the country. The common need is for markets and transport. In general, the markets must be abroad, because Bolivia cannot generate a rap- id growth in demand for agricultural produce; and the markets are more likely to be in the northern hemisphere than in South America. Perhaps the leading agricultural opportunity is in wheat, because there exists a substantial and growing domestic demand yet to be satisfied. There are preliminary plans for a major irrigation scheme to develop 200,000 hectares of good agricul- tural land in the Villamontes-Sachapera region of Tarija, where wheat and dairy cattle could be produced for the home market and soya, corn, cotton, sorghum and groundnuts for export by rail through Argentina. A similar concept has been suggested for the Pilcomayo river basin where a hydroelectric installation could provide irrigation for 50,000 hectares of farm land and power for agro-industrial plants. It has been proposed that wheat be grown as a winter crop with cotton, groundnuts, soya, sunflowers and other crops in the summer. Apart from wheat and dairy products, most large agricul- tural schemes must be directed at export markets. The great- est potential lies in the Beni which could become one of the world's greatest sources of beef. At present, methods of production are primitive. This vast area is al- most totally devoid of transport except for small aircraft using rough grass air strips, and river boats plying up and down the Mamore and Beni Rivers. Much needs to be done to raise the quality of produce to international standards, to develop the means of transport and to establish effective marketing organizations, but the opportunity is there for an enormous and highly profitable export trade that could help transform. the economy of the country. The Alto Beni has attracted migrants because of its capacity to produce a wide range of crops for La Paz and for export. Transportation is a major obstacle. Yucca, corn for poultry, ramie, cacao, coffee, citrus fruits, avocados, sweet potatoes and mangoes can be produced. -119-  Another large area of great potential, somewhat remote from existing agricultural zones but not requiring irrigation or any other large investment, is the Tucavaca valley stretching westwards from Puerto Suarez. Wheat, corn, soya, rice, citrus fruits, timber and cattle could be produced in great quantities for export by rail through Brazil or by river to Rosario. All these projects, and others, need labor for their fulfilment. They need migrants from the Altiplano. But it has been suggested that an alternative solution to the problem of the people on the Altiplano would be to improve and develop the output of the Altiplano itself. Many proposals have been made to invest capital in improving the fertility and product- ivity of the area, encouraging dairy farming and wool produc- tion, reforestation, land reclamation around Lake Titacaca, fish ponds, horticulture and various agroindustries. Industry - It is common for non-industrialized countries to believe that industry is the key to prosperity. This may be true in countries with few resources except manpower, but it may not be so in countries rich in resources yet lacking in manpower. Bolivia's ventures into industry have produced too many expensive failures, such as the Cochabamba asbestos cement plant, the La Paz plate glass factory, the vegetable oil plant in Villamontes, the factories for making matches, batteries and drilling bits, the bismuth smelter and some of the oil refineries. There are doubts about the cement plants at Yacuses, Tarija and Sevaruyo. Nevertheless, there is no lack of grandiose and optimis- tic plans for further industrialization which, if they ma- terialize, will generate important demands for transport. Among them is the motor vehicle plant at Cochabamba designed to progress in the classic pattern from assembly work to part production. In association with this venture there would be a steel works using iron ores or pellets from Mutun. The sup- plies of natural gas would provide the basis for a petroche- mical industry, starting with a urea plant near Santa Cruz and proceeding to the production of olefin, phenol, polypro- pylene and polyethylene at Cochabamba. Less ambitious, but perhaps more feasible, plans for in- ,,, dustrialization are aimed at agricultural processing in order to raise the value-added element of many agricultural products, thus reducing the incidence of transport costs and making it easier to compete in world markets. The canning of fruit and vegetables, the production of juices, jams and jellies, the development of viniculture, instant coffee, cocoa products, tinned and preserved meat, leatherwork and wood products, are all ways of utilizing local produce, without need of heavy -120- industry or high technology, in order to penetrate foreign markets more profitably than by the simple export of the raw materials themselves. But even in these relatively simple industries success is not easy, because the export market demands high standards and is extremely competitive. One way of escaping the tough examination of world com- petition is to seek protection within the Andean Pact. The motor vehicle industry and several smaller industries are being launched in cooperation with the other members of the Andean Group. A semi-guaranteed Andean market would certainly help Bolivia to create a number of new industries; but the market is small and Bolivia will have to reciprocate by buy- ing other manufactures, possibly of low quality and high price, from the other members. Alternative Development Patterns As explained in Chapter 3, for planning purposes in this Study, the country was divided into 77 zones and transport de- mand was predicted as a function of, among other things, several parameters related to population and economic activity in each zone. Assumptions about future development must thus be trans- latable into their impact on the population and economic para- meters of the 77 zones. It is therefore necessary to produce development patterns to reflect, at the required level of detail, the expected socio- economic situations in the target years, 1989 and 1999. This is particularly difficult in Bolivia because of the state of political uncertainty and the wide range of development oppor- tunities. It is easier, however, to predict 1989 than 1999. Much of the development of the 1980s is already starte~ com- mitted or planned. Also, because of the current international recession, the early 1980s are expected to bring little econo- mic development. By 1999, however, the range of possibilities is so wide and varied that the scale and nature of development are much more difficult to predict. For this reason, it was decided to formulate one develop- ment pattern for 1989 and three alternatives for 1999. Clear- ly there is scope for error in the 1989 pattern but it is con- sidered small enough to ignore, whereas in 1999 the likelihood of error is large. Hence three contrasting patterns, called A, C and D, were selected (B was dropped because it proved to be very similar to C and thus provided no useful basis for fur- ther analysis) in order to test the sensitivity of transport projects to these alternative possibilities. They were chosen to reflect some broad differences of scale and direction and development. The size of the GDP is as unpredictable as its distribution. Hence a range of GDP assumptions was incorporated in the alternative patterns: Alternative C was given the official (Central Bank) assumption of GDP, Alternative A 10 percent lower and Alternative D 20 percent higher. -121-  • .• ~ A second element of variation was the emphasis on pri- mary or secondary production. This was related to the dif- , ''II ferences in GDP because secondary industrial development, i.e. manufacturing, is generally the more capital-intensive. There- I fore Alternative A has a greater emphasis on the development of primary production in the form of agriculture and mining, Alternative C introduces greater industrial investment and Al- ternative D assumes an even faster industrial growth. The third principal contrast is in regional distribution. Alternative A represents a continuation of recent trends, with growth concentrated heavily in a few areas, and other areas largely neglected. Alternative C incorporates more regional development intended to spread economic prosperity more evenly between different parts of the country. Alternative D contains ." the same regional development projects but, having a much greater GDP, concentrates its additional product in an industrial axis from La Paz through Cochabamba and Santa Cruz to Mutun. The 1989 Development Pattern - GDP is predicted to reach $b 117,500 million in 1989 (in 1977 pesos). Much of the de- velopment before then will consist of projects already in pro- gress or committed. They enable one to predict fairly well the state of economic development in 1989. Figure 7-4 indicates the 44 major projects included, of which 18 were agricultural. Schemes to be completed by 1989 are located at San Ignacio de Velasco, San Javier, Valles Mesotermicos, U11a U11a, Ingavi, Chimore, Cha11apata and Alto Beni. Other agricultural projects due to be under way but not completed by 1989 are at Abap6- Izozog, Gran Chaco,Misicuni, Rurrenabaque-San Buenaventura, Vi11amontes, San Jacinto and Tucavaca. Many of these develop- j .~ ments will continue right through the 1990s, especially those involving dams. Apart from major definable projects, there are other areas where agriculture w1ll stead1ly expand, 1n particular cattle-rearing in the Beni, cultivation of various foodstuffs in the Yungas and Valleys to feed the growing popula- tions of La Paz and Cochabamba, corn, sugar and horticultural production around Santa Cruz, and fruit production in Tarija. . In the mining sector there are 12 major projects of wh~ch 8 should be completed by 1989, at Catavi, Centenario, Re1aves E1 Kenko, Integral Bo1{var, Co1quiri, Sud Lipez, La Pa1ca and Machacamarca. At Mutun, the production of iron pellets is assumed to have commenced. There are also 14 major industrial projects, including the automobile project at Cochabamba, which will continue into the 1990s, cement plants in Oruro, east Santa Cruz and Tarija, paper factories in Santa Cruz and Tarija, steel in Santa Cruz, textile and glass factories, and plants for the production i. -122- ~ I '  LEG END • PRINCIPAL GROWTH CENTER • SECONDARY GROWTH CENTER .& INDUSTRIAL CENTER • MINERAL PROCESSING 50 0 100 200 Km. ~ NATURAL GAS PRODUCTION '!!!!!!I'~ I!oo' __~==!!!!I FOR EST PRODUCTS AGRICULTURAL DEVELOPMENT (,,> - IMPROVEMENT OF AGRICULTURE ~ ./1/ PRODUCTION AND PROCESSING OF CATTLE '1 ~ - MINERAL PRODUCTION ; 21 PROJECT NUMBER (SEE APPENDIX 7B) ... • if$> • '1> L___ ~ \­ 14 " ) 2 o 11 ) (\ (\ o A 1989 DEVElOPMENT PATTERN FIGURE 7-4 of vegetable oil and processed fruit. Most important is the development of natural gas for export, on which the growth of the whole economy depends. In addition to these major projects, which have been in- dividually incorporated in the development pattern, there will be innumerable smaller developments, which have been predicted in a more aggregated way. A prime objective in this work was to produce a detailed socio-economic forecast, zone by zone, consistent both with the macro-economic predictions, especially the GDP prediction, and with the major development projects that will contribute towards those predictions. 1999 Development Pattern A - The first of the three 1999 patterns is pessimistic about economic growth. GDP at 181,300 million pesos (1977) represents a growth of 54.3 percent since 1989, which is only 15.4 percent per head. This relatively low growth is partly the result of low investment and it also 1im-· its the budget available for future investment, which in 1999 is estimated at 35,400 million pesos. Major projects, presented in Figure 7-5, are largely in the agricultural sector with the continuation of the Abap6- Izozog, Gran Chaco, Misicuni, San Buenaventura, Vi11amontes, San Jacinto and Tucavaca developments. No major new agricul- tural projects are included. ~iringdevelopments continue at Uyuni, Fare11ones-Ke11wani and E1 Rodeo but not much expansion at Mutun is foreseen in this alternative. Little industrial investment is included. The major pro- jects are continuations from 1989, namely the automobile pro- ject, the Karachipampa sme1tinq complex and the Aquila chemi- cal works. In the all-important field of hydrocarbons a con- siderable increase in gas production is foreseen, while pe- troleum output declines. The lack of major projects in this alternative means that much of the economic growth, such as it is, consists of small unidentifiable deve10prents in all three sectors, agricul- ture, mining and industry. A general growth of production is incorporated, consistent with the macro-economic predictions. Because of the low growth in income per head, there is neces- sarily an emphasis on agricultural development. This development pattern does not reflect either the hopes or the expectations of the Study. It is a pessimistic view of the future containing a slow and unenterprising exploitation of the country's resources. It represents the minimum conceivable demand for transport facilities in 1999. -123-  1999 Development Pattern C - The most important of the three alternatives, Pattern C conforms with the Central Bank , ;, macro-economic predictions and can thus be regarded as the most likely or most feasible pattern of the future. It as- sumes that by 1989 the Government will be organized to make rational economic decisions to promote production in certain industries, particularly those utilizing national resources such as agriculture, livestock, minerals and hydrocarbons, '--1\ and that productive resources will be better distributed to permit a high level of employment. Various regions, e.g. Abap6-Izozog, Chan~-Piraf, Villamontes, San JuliSn and north La Paz, have been selected for agricultural development. , -t In 1975, Bolivia was a signatory to the International Strategy for Industrial Development which took as its starting pOint the pattern and structure of international trade. This resulted in the Action Plan of Lima which included a long- term plan for the promotion of economic harmonization and trade between developing countries, especially in the industrial ., field. Pattern C rests on the Lima objective of accelerating industrial growth in order to improve the pattern of interna- tional trade. Its broad lines may be expressed as follows: (i) State control of natural resources, and their exploitation in order to achieve industrialization; (ii) Strengthening of trade programs, with complementary agreements within the Andean Group and LAFTA(1); (iii) Large-scale colonization and promotion of farming in order to provide employment; (iv) Creation of steel, smelting, petrochemical and en- gineering industries to reduce the need for imports; (v) Exploitation of the potential value of natural re- sources, especially agriculture, by development of associated processing industries, e. g. in neat, timber, ci true fruits and milk; (vi) Establishment within the Andean Group of engineering, electrical and chemical industries and expansion of the range of manufactures for mass consumption such as foodstuffs, construction materials, textiles, leather and other light manufactures; \. , (1) Latin Arrerican Free Trade Association. On August 11, 1980 in M:>nte- video, this was replaced by a new organization, AIJ\DI (Asociaci6n Latinoarrericana de Integraci6n). -124-  1 ~ .. LEGEND 50 0 '---' 100 ZOOKm. I • • .to PRINCIPAL GROWTH CENTER SECONDARY GROWTH CE NTER INDUSTRIAL CENTER ~ • MINERAL PROCE§61NG NATURAL GAS PRODUCTION ZI PROJECT NUMBER (SEE APPENIX 7B) r. 71 FOREST PRODUCTS (. AGRICULJURAL DEVELOPM[NT I) LLL, PRODUCTION AND PROCESSING OF CATTLE MINERAL PRODUCTION /. / r· -/ ... o o 1'1 •L o c "0 .. • o V ~ \> <l ~ '" ~ ~. " , 0 :­ ~ ~ .. :-. ~ o " .\ R G E 1\ T ] :'J .\ 1999 DEVElOPMENT PATTERN A FIGURE 7-5  (vii) Intensive exploration for minerals, particularly in the Brazilian Shield, and for hydrocarbons; (viii) Incentives for small businesses and craft trades; (ix) Establishment of negotiating machinery for questions of foreign trade. The development projects in Pattern C were selected in accordance with the above objectives. As shown in Figure 7-6, it comprises all the major projects in Pattern A plus some important additions. The Rositas, Pilcomayo and Bala agricultural projects are included. Mineral development at Playa Verde and in the Precambrian area are included and output from Mut~n is expanded. In the industrial sector, the major additions are the petrochemichal, fertilizer, steel and palrr oil plants. In addition there is generally higher output from most sectors of production, especially industrial sectors and natural gas. 1999 Development Pattern D - The main purpose of Pattern D was to explore the implications of a much faster rate of growth. Income per head rises between 1989 and 1999 by 53.7 percent, compared with 28 percent in C and only 15.4 percent in A. The higher national product implies higher output in most industries, but especially in gas, processed food, building materials and non-ferrous metals, for domestic consumption, and in export industries such as meat, other animal products, non-ferrous minerals, iron and timber. The broad strategy in Pattern D is the sa~e as in C but the pace is faster and hence the emphasis on the industrial axis between La Paz and Mut~n is greater. Rising incomes lead to a much faster growth of demand for manufactures than for food, and especially for better housing and for transport. This high rate of economic growth is hardly feasible unless a much bigger share of industrial demand is met domestically: hence the emphasis on industrial development. No new major projects were added to this pattern, not because there would not be any but because they have not yet been conceived. Major expar.sion of numerous industries is involved but individual development projects, additional to those already incorporated in Pattern C, were not identified. A big growth in meat output and export from the Beni was assumed, together with other animal products. Wheat output from Tucavaca eliminates imports of this cereal. In the mineral sector there is a major development of Mut~n and in the output of building materials, also of non-metallic minerals. In the -125- industrial sector there is general growth but especially in flour and grain products, wood and wood products, chemicals, manufactured building materials and cement, steel and fabricated metals, and -- though still on a small scale -- in vehicles and I machinery. To make this pattern feasible, major expansion of exports was assumed in meat, animal products and timber through the Amazon, in minerals, iron pellets and chemicals down the River Paraguay, and in processed foodstuffs. Quantification of Development Projects The Bolivian economy is so little developed and the volumes of traffic so small that any major new development project could have a marked effect on transport demand in its '''t area. It was necessary, therefore, to take careful account of all such major projects included in the development patterns. For this purpose a questionnaire was completed for forty large projects likely to have important transport implications. The questionnaire elicited the following information: • location • investment cost • period of construction • scheduled year of completion • planned annual production levels from date of commencement • principal products • employment to be generated • state of project in 1989 • production in 1989 • inputs required, by type and tonnage • produotion by tonnage. This information was painstakingly obtained from government departments and public corporations, and is set out in Working Paper 47. Further information could be obtained from the Five-Year Plan 1975-80, which contains a list of 340 projects, with location and investment cost. Some of the biggest projects extend into the 1990s and even beyond, but naturally there was far less information about the 1990s than the 1980s. Socio-eoonomic Characteristics in 1989 and 1999 The last step in the projection of socio-economic development was to apply the macro-economic predictions and- development patterns to the 77 zones. This work is described in Working Paper 49. For each zone the following variables were obtained for the base year, 1977, and then projected to 1989 and 1999: -126- - ,  } ~ LEGEND 50 ~ 0 100 200Km . I • • PRINCIPAL GROWTH CENTER SECONDARY GROWTH CENTER • • INDUSTRIAL CENTER • MINERAL PROCE5S1NG NA,URAL GAS PRODUCIION 21 PROJECT NUMBER (SEE APPENIl( 7B) 7/,£/ FOREST PRODUCTS AGRICULJURAL DEVEl.OPME NT LLL/; PRODUCTION AND PROCESSING OF CATTLE MINERAL PRODUCTION ... j .~ ~ ". -( 14 o o III ) -----_ --­ ... ... N 2 o 18 I , C­ I V ~ ~~ V \) o ~. '\ :­ ~ ..... ., "l "­ o A RG E 1\ T 1 .\ 1999 DEVELOPMENT PAITERNS C AND D FIGURE 7-6  • Population • Dispersed Population • Economically Active Population • GOP per Head • Production of Goods in 39 Groups, in tons • Consumption of Goods in 39 Groups, in tons. For 1999 the variables were projected for the three alternative development patterns. Pattern C was used in the computer model. The other two alternatives were prepared for manual use and were not fully specified at the zonal level; some of the data were distributed between only 20 districts, each district being a group of zones. Table 7-9 shows the basic macro-economic projections which . had to be distributed between zones. The variables were first distributed by a method following past trends; as required, they were adjusted to coincide with the macro-economic projections. Then they were modified to reflect the detailed features of the development patterns. The distribution of population was made consistent with the location of production, and the size of the non-dispersed population in each zone or district was related to the amount of industrial and commercial activity. The si~e of the non-dispersed population has a big effect on the cash income of the zone and on its level of consumption and demand for trans- port of all kinds. For each zone, the production and cons~ption of each of the 39 commodity groups was calculated for 1977, 1989 and 1999 (Pattern C - the other alternatives were projected by district), and export and imports were balan=ed to give a consistent pattern. An important part of this work was the estimation of intermediate consumption, i.e. the consumption by one industry of the output of another. No input-output analysis was available for the Bolivian economy, so these interindustrial relationships had to be estimated as best possible from other sources including American experience. Table 7-10 shows the totals of supply and demand for each commodity group for 1977, 1989 and the three 1999 development patterns. Supply is defined as production plus imports; this equals demand, which is defined as consumption plus exports. The table is expressed in tons and shows an increase in total tonnage of 135 percent between 1977 and 1989 and further increases of 36 to 71 percent between 1989 and 1999. The very large increase during the former period is accounted for largely by natural gas and iron, which have high ratios of weight to value.Witixut these two almost new commodities, tonnage rises between 1977 and 1989 by 84 percent, compared with a rise of 71 percent in GOP. Other commodities showing exceptional growth during the 1980s are primary non-ferrous metals, fertilizers, rice and cement, with associated lime and gypsum. -127-  Table 7-9 SU~W~RY OF SOCIa-ECONOMIC VARIABLES 1977 and Projections for 1989 and 1999 1999 1977 1989 (1) A C - D Population 5,156 7,165 9,577 9,682 9,682 Dispersed population (1) 2,556 3,038 3,486 3,522 3,450 I (1) I-' Economically active population 1,660 2,401 3,354 3,391 3,391 N 00 I GOP ($b million) ( 2) 68,900 117,800 181,300 203,400 241,700 GOP per head, overall(2) 13,350 16,440 18,930 21,000 24,960 (2 ) GOP per head, dispersed 3,320 4,370 5,870 6,510 7,810 GOP per head, non-dispersed (2) 23,200 25,330 26,400 29,300 34,460 (3) 17,500 55,738 60,614 70,117 Goods production and consumption 41,048 (1) Thousands. (2) At 1977 prices. (3) Thousands of tons,includinq imports and exports. SOURCE: Forecast and Distribution of Socio-economic Variables, Working Paper 49, BolivJa National Transport Study • .. .i: ,4 .. .. - ';;.Q. ~ J,. .. - • • d • • •• 4 ~ ... .. ,  Table 7-10 SUPPLY AND DEMAND OF COMMODITY GROUPS 1977 anc1 Projections for 1989 and 1999 SUPPLY/DEMAND IN THOUSA~D TONS COMMODITY GROUP 1977 1989 1999A 1999C 1999D 1. Live animals 163.7 263.8 387.7 429.7 516 2. Meat and fish 145.5 265.4 381.6 407.0 591 3. Other animal products 68.0 104.1 158.0 160.0 234 4. Corn 403.1 510.1 561.1 641.0 680 5. Rice 100.5 272.9 469.5 475.5 497 6. Wheat 180.0 257.0 370.1 424.0 466 7. Other cereals 72.7 250.0 345.2 351. 7 372 8. Potatoes and other tubers 1049.1 1542.9 2122.6 2144.6 2150 9. Fruit 689.1 1027.2 1292.2 1445.9 1521 10.Vegetables and plants 193.2 303.9 358.5 400.5 422 11.Sugar cane 3556.0 6318.0 8455.2 8805.2 9245 12.0ther agricultural products 156.8 346.9 602.8 751.2 784 13.Crude oil 1739.8 2241.0 3667.2 4150.0 5150 14.Natural gas 2271.7 10091.0 12982.0 12975.0 13141 15.Bottled gas 51.7 150.2 265.5 315.0 380 16.0ils and lubricants 24.2 51.7 74.0 86.0 108 17.0ther refined petroleum products 871.5 1895.3 3300.0 3900.0 4900 18.Iron ore 7.1 1707.0 2000.0 2900.0 3900 19.Non-ferrous metallic minerals 325.7 510.2 858.6 911.5 1155 20.Non-metallic minerals 80.3 236.7 216.9 266.0 356 21.Sugar and confectionery 331.9 584.8 856.9 936.9 974 22.Flour and grain products 237.3 500.5 861. 6 861. 8 948 23.Beverages 289.1 658.3 784.7 831. 2 904 24.0ther food and tobacco products 247.0 455.5 480.1 529.8 569 24a.Food distribution 330.6 539.2 539.2 563 25.Textiles and leather products 110.3 179.5 254.5 269.5 341 26.Wood and wood products 505.5 973.8 1821. 8 1829.3 2370 27.Pulp and paper products 40.9 125.3 138.3 210.6 274 28.Fert111zers 17.5 86.7 144.5 193.4 232 29.0ther chemicals 116.4 263.4 299.9 670.5 920 30.Sand and aggregates 1870.8 2876.3 3499.4 3579.4 4650 31.Lime and gypsum 613.6 1849.9 2645.7 3174.0 4126 32.Bricks, ceramics, etc. 238.2 595.1 740.7 844.6 1098 33.Cement 369.1 1015.1 1194.6 1231.0 1569 34,35.Primary iron, steel and other 137.8 1562.3 1762.8 1964.0 2678 fabricated metals 36.Primary non-ferrous metals 29.5 135.9 189.5 201. 6 289 37.Vehicles 26.1 65.2 91.9 119.6 167 38.Machinery 61.9 211. 9 242.4 262.4 388 39.Miscellaneous manufactures 107.2 232.3 320.7 425.3 489 17,499.8 41,047.7 55737.9 60613.9 70117 SOURCE: Forecast and Distribution of Socio-economic Variables, Working Paper 49, Bolivia National Transport Study. -129-  In Table 7-11 the tonnage totals for production, con- sumption, imports and exports are shown, for 1977, 1989 and the three 1999 alternative development patterns. The fig- ures are shown with and without crude oil and natural gas, because these commodities have a dominating influence on the tonnages. The great growth in the production and export of natural gas is the most important feature in these pro- jections, as indeed it is expected to be in the future of the Bolivian economy. I I -130-  Table 7-11 PRODUCTION, CONSU~PTION, I~PORTS AND EXPORTS: 1977 and Projections for 1989 and 1999 (000 tons) PRODUCTION H1PORTS TOTAL EXPORTS CONSUMPTION GDP All Commodities ($bmil1ion) 1977 16,811 689 17,500 3,173 14, 327 68,800 1989 39,386 1,661 41,047 11,685 29,362 117,800 1999 A 52,251 3,487 55,738 15,071 40,667 181,300 1999 C 56,475 4,139 60,614 15,528 45,086 201,400 1999 D 64,748 5,369 70,117 16,404 53,713 241,700 All Commodities Except Crude Oil and Natural Gas 1977 12,806 682 13,488 519 12,969 1989 27,895 820 28,715 1,935 26,780 1999 A 38,172 917 39,089 2,571 36,518 1999 C 42,403 1,086 43,489 3,028 40,461 1999 D 50,510 1,316 51,826 3,904 47,922 Crude Oil and Natural Gas 1977 4,005 7 4,012 2,654 1,358 1989 11,491 841 12,332 9,750 2,582 1999 A 14,079 2,570 16,649 12,500 4,149 1999 C 14,072 3,053 17,125 12,500 4,625 1999 D 14,238 4,053 18,291 12,500 5,791 SOURCE: Forecast and Distribution of Socio-economic Variables, Working Paper 49, Bolivia National Transport Study. -131-  CHAPTER 8 RESOURCES AVAILABLE FOR TRANSPORT INVESTMENT CHAPTER 8 RESOURCES AVAILABLE FOR TRANSPORT INVESTMENT A major requirement of the Study is to recommend a year- by-year program of transport investments for the period 1981- 90 ,with an indicative program for the period 1991-2000 • Clearly, it is impossible to fulfil this task without first coming to a realistic assessment of how large will be the available budget for transport investment. It is not possible, either in Bolivia or in most other countries, to obtain official guidance on this vital question~ governments are not able or willing to commit themselves so far ahead. The question has therefore been answered by the Study team to the best of its ability. This chapter explains how. Discussion of the Problem Transport investment is made partly by government depart- ments, partly by publicly owned corporations and partly by private companies and individuals. The concern here is with the two public sectors, which in Bolivia account for a high per- centage of the total. No matter how the funds are raised -- by higher prices, taxes, domestic borrowing or foreign aid -- the more that is spent on transport facilities, the less that can be spent on other things, and in particular, since there is a practical limit to the proportion of national resources that can be appropriated for investment rather than consumption, the less that can be invested in other sectors. The question, how much could and should be allocated to transport investment, is therefore primarily a question of what proportion transport should get of a given budget for public investment. One might argue that this budget should be increased to permit a larger allocation to transport, but it would be difficult for a transport study to support such an argument, for the following reason. In theory, the trans- port planner can estimate the rates of return expected from transport investments and, if he can produce so many projects with rates of return above the test rate (1) that they exceed the available budget, he can claim that public money would be better spent on transport than in other sectors where such profitable opportunities do not exist. In practice, however, there are few other sectors where it is possible to estimate (1) The test rate is the rate set by government as the minimum acceptable rate of return. -133-  • rates of return on investment. In education, health, defense, sewerage, research and water supply -- all big spenders on investment -- it is practically impossible to estimate rates of return. The test rate is supposed to be the grand allocator, the simple mechanism that separates the investments that can be afforded from those that cannot, but because most public depart- ments cannot estimate their rates of return, there is no way of knowing what the test rate should be. Hence, in practice, the government has its investment budget and divides it between sectors by means of subjective judgements and economic calculations. It is then the task of each sector to invest its money wisely. '. Thus the transport sector cannot argue for a larger budget '. without claiming that marginal transport projects are more important than other projects for health, education, defense, etc., or than private sector investments, or than personal consumption; which are not arguments that the transport planner can or would wish to make. . " Consequently the size of the future transport budget becomes a question of passive prediction -- what will be, not what should be. Probably the strongest factor bearing on this question is the status quo. Because there is no objective way of determining the share of transport or any other sector, the fact that it was x percent last year is a powerful argument for keeping it at x percent this year. Of course there are other factors. In times of recession, for example, the transport . tl sector usually suffers because cuts in transport investment are less readily noticed by the public than cuts in other , .; services. For the purpose of this Study, the following way of fore- casting the future transport budget was adopted: examine past trends and extend them into the future, taking account of any , .1 good reasons for expecting a change in the trends. The first step was to obtain data on investment by mode in interurban transport for the last ten years. In fact, there has been rela- tively little other transport investment in Bolivia and it can be ignored for the present. Total transport investment was then expressed as a percentage of (a) public investment, (b) all investment and (c) gross domestic product (GDP). Investment in each mode was expressed similarly, since governments seldom decide the claims of each mode separately Having obtained time series relating both global and sec- toral transport investment to public investment, national invest- ment and GDP, it was necessary to do three things: -134-  (i) Predict the growth of GDP, national investment and public investment. Predictions of GDP were discussed in Chapter 7 and used in the formulation of develop- ment patterns. The predictions of national and public investment were made as percentages of GDP. (ii) Predict the percentage of transport investment to total public investment. If there are obvious growth sectors(or new sectors) in public investment, they must be taken into account. (iii) Look for new factors likely to change the past pattern. In Bolivia urban transport may soon begin to take a growing part of the transport budget, simply because the towns are growing much faster than the rural popu- lation and the needs of urban transport tend to grow as a power function of the urban population. It must also be expected that the transport budget will be hit by the expected national recession during the next few years. If GDP stops growing or grows slowly, while popu- lation continues to grow, government will find it difficult to maintain the level of social services. In such a time of financial stringency, one of the easiest budgets to cut would be the transport budget, especially since it appears to have enjoyed high priority in recent years. An important part of the investment budget comes, and will pr?bab1y continue to come, from foreign lending institutions. Th~s fact does not alter the budget problem. It is really not possible to forecast any change in the amount of foreign aid forthcoming except to say th~t a gradual increase (in real terms) is expected. A drastic change would arise only as a result of political upheaval. Otherwise, a steady increase in aid is to be expected. But lending agencies are not tied to transport projects. From a national point of view, transport investment cornes out of the total investment budget, regardless of the source of funds. The acquisition of foreign loans for trans- port projects should not be used as an argument for increasing the total investment in transport. Past Investments Details have been gathered of the annual investment expen- diture of the five dominant public agencies in the transport sector, i.e. SNC, ENFE, LAB, AASANA, and COMIXTA (2). These (2) The National Road Service, the National Railroad Company, the National Airline, the National Airport Administration and the Bolivian-Argentinian Joint Commission which is responsible for the new railroad construction north of Santa Cruz. -135-  '. figures are only available since 1973. They are shown in Table 8-1 and include highway maintenance and rail rehabil- itation costs; the transport investment program recommended in Chapter 18 also includes these items for which a trade- off relationship exists with the standard of initial con- struction. The rate of growth was high, averaging 38 percent per year, which compares with a general price inflation of about 17 percent. Hence one may say that transport invest- ment was growing at roughly l8 percent a year in real terms. The investment of ENFE shows a large rise in recent " years due to the rehabilitation of rail lines. Annual in- vestment by LAB. and AASANA was "lumpy", i.e. there were large fluctuations from one year to the next. This was because they are not very large enterprises and their investment pro- grams were greatly affected by one or two big projects or the lack of them. Over the whole period, 57 percent of the total went to SNC, 22 percent to ENFE, l2 percent to LAB, 6 percent to AASANA and 3 percent to COHIXTA. In Table 8-2, the figures of transport investment are compared with figures in public investment, national invest- ment and GDP. In contrast to the fluctuating record of trans- port investment, public investment grew quite steadily at an annual average of 32 percent, national investment at 29 per- cent and GDP at 25 percent. In real terms, if one accepts the index 9f general prices as an.index of inflation, the annual growth rates were: public investment, 13 percent; national investment, 10 percent; GDP, 7 percent; and transport invest- ment, 18 percent. The last three columns of Table 8-2 show the relation- I , ships between the other columns. The last column shows that national investment has obtained a growing percentage of GDP; a three-year moving average shows a steady growth from 16.3 percent to 19.2 percent of GDP. The second-to-last column shows that the public sector increased its share of national , , investment in 1974 and maintained a percentage averaging 58.0 percent since then. The preceding column shows the percentage of public investment that went to the five transport agencies. This percentage rose from 14.7 in 1973 to 35.0 percent in 1978 before falling back to 19.3 percent in 1979, with an overall average of 22.7 percent. Projection All future costs and prices are shown in 1977 pesos. On this basis, the GDP has been projected to reach 117,800 million pesos in 1989 and 201,400 million pesos in 1999 (3). (3) This figure corresponds to Development Pattern C. -136- OVERLAYING NEAR COCHABAMBA APPLYING SURFACE DRESSING HUARINA - TIQUINA HIGHWAY SURFACING  Table 8-1 TRANSPORT INVESTMENT, 1973 - 79 (Millions of current pesos) YEAR SNC(l) ENFE LAB MSANA COMIXTA TOTAL 1973 166 36 18 54 (2 ) 36 310 1974 345 113 104 72 (2 ) 15 649 1975 424 99 284 113 27 947 1976 662 303 32 85 85 1,167 1977 1,.164 442 136 87 46 1,875 1978 1,533 856 590 131 92 3,202 1979 1,559 478 66 76 2,179 (1) Includes road maintenance - see Table 6-1 for recent breakdown of costs. (2) Estimated. SOURCE: various. -137-  Table 8-2 TRANSPORT INVESTMENT AND NATIONAL RESOURCES, 1973 - 79 1 2 3 4 TRANSPORT PUBLIC NATIONAL RATIOS YEAR INVESTMENT INVESTMENT INVESTMENT GOP 1:2 2:3 ..1.:.i.. (millions of current pesos) (percentages) 1973 310 2,109 4,433 26,466 14.7 47.6 16.7 1974 649 3,635 6,468 44,339 17.9 56.2 14.6 1975 947 5,778 8,830 50,156 1G.4 65.4 17.6 1976 1,167 6,063 10,498 58,675 19.2 57.7 17.9 1977 1,875 7,502 12,776 68,842 25.0 58.7 18.6 1978 3,202 9,151 15,987 83,633 35.0 57.2 19.1 1979 2,179 11,305 20,366 101,614 19.3 55.5 20.0 " Growth .. 1973-79 603% 436% 359% 284% Average 22.7% 57.4% 18.2% ";1 "I -138- Because of the economic recession expected during the early 1980s, it is predicted that GDP will rise by only 3 percent per annum until 1985 and will then build up to the growth rate of 5.5 percent assumed for the 1990s. This projection applies to Development Pattern C. Patterns A and D have lower and higher levels, respectively, which will be considered separate- ly below. The projection of national investment is made at a rate rising to 23 percent of GDP, to allow for some continuation of the recent trend. The projection of public investment is made at a rate falling to 55.5 percent of national investment, i.e. a little less than observed in recent years, on the ex- pectation that the private sector will increase its role through industrial development. The historical trends for transport investment as a pro- portion of public investment show a definite increase with time. Using these would indicate that transport investment would form about 30 percent and this was used as a high value for projec- tions. However, in view of the country's financial position, it is uncertain that this rate can be achieved, and so a low rate of 20 percent was also used. Table 8-3 shows the projec- tions from 1981 to 2000 for both the low and high rates. From these, it can be calculated that the budget for transport in- vestment in the two decades 1981-90 and 1991-2000 will be in the following range, in 1977 pesos: ESTIMATED INVESTMENT LEVELS IN MILLION PESOS Transport Low Hi9h Total Public 1981-1990 23,600 35,400 118,000 1991-2000 41,700 62,600 209,000 The projection for the later period applies to Develop- ment Pattern C. It is assumed that GDP in 1999 will be 20 percent higher in Development Pattern D, i.e. 241,700 million pesos, and 10 percent lower in Development Pattern A, i.e. 181,300 million pesos. The proportion of GDP devoted to na- tional investment is 10 percent higher in D and 15 percent lower in A than in C. Table 8-4 shows the resulting projec- tions of transport investment. The budget ranges estimated for 1981-90 and 1991-2000 are as follows: -139-  • .. .. Table 8-3 TRANSPORT INVESTMENT AND NATIONAL RESOURCES, 1980 - 2000 Development Pattern C (1977 pesos in '000 millions) TRANSPORT •• NATIONAL PUBLIC INVESTMENT YEAR GDP INVESTMENT INVESTMENT Low High 1977 68.8 12.8 7.5 1.9 1.9 1978 75.8 14.5 8.3 2.9 2.9 1979 80.4 16.1 8.9 1.7 1.7 1980 82.8 16.6 9.2 1.8 2.8 1985 96.5 20.1 11.2 2.2 3.4 1989 117.8 25.3 14.0 2.8 4.2 1990 124.3 26.7 14.8 3.0 4.4 ,j 1999 201.4 46.3 25.7 5.1 7.7 2000 212.4 48.8 27.1 5.4 8.1 " , -140-  Table 8-4 TRANSPORT INVESTMFNT AND NATIONAL RESOURCES, 1980 - 2000 Development Patterns A and D (1977 pesos in '000 millions) DEVELOPMENT PATTERN A DEVELOPMENT PATTERN D National Transport InvestIrent National Tran.5J.X)rt Invest:rrent YEAR GDP Investment Low Hi~Fi GDP Investment Low Hi9: h 1977 68.8 12.8 1.9 1.9 68.8 12.8 1.9 1.9 I I-' ~ 1978 75.8 14.5 2.9 2.9 75.8 14.5 2.9 2.9 I-' I 1979 80.4 16.1 1.7 1.7 80.4 16.1 1.7 1.7 1980 82.8 16.6 1.8 2.8 82.8 16.6 1.8 2.8 1985 96.5 20.1 2.2 3.4 96.5 20.1 2.2 3.4 1989 117.8 25.3 2.8 4.2 117.8 25.3 2.8 4.2 1990 123.0 26.9 2.9 4.3 126.6 29.7 3.3 4.9 1999 181.3 35.4 3.9 5.9 241. 7 61.2 6.8 10.2 2000 189.3 37.0 4.1 6.2 259.7 65.7 7.3 10.9  " • TRANSPORT BUDGET IN MILLION PESOS Low High A C i5 A C 1:5 1981 - 1990 23,600 35,400 1991 - 2000 31,700 41,700 56,400 47,900 62,600 84,200 Conclusion It is estimated that the transport budget likely to be available in the decade 1981-90 is in the range of 23,600 to 35,400 million pesos. " A part of this should be allowed for urban transport. How much, it is difficult to say, perhaps five percent. This leaves 22,410 to 33,620 million pesos for non-urban transport. For the decade 1991-2000 the budget will be affected by the rate of economic growth, which is more unpredictable than in the earlier decade. Three low estimates of 31,700 million, 41,700 million and 56,400 million pesos, and three high estimates of 47,900 million, 62,600 million and 84,200 million pesos have been produced, each corresponding to Development Patterns A, C and D respectively. It is considered that 10 percent of these amounts should be allowed for urban transpor~ reflecting its growing relative importance. , " : ,d -142- i.4  CHAPTER 9 EXISTING AND FUTURE TRANSPORT DEMANDS CHAPTER 9 EXISTING AND FUTURE TRANSPORT DEMANDS As has been co~~ented upon elsewhere, the present transport infrastructure in Bolivia, while for the most part undeveloped, nonetheless offers considerable capacity in excess of the demands made upon it. This chapter describes transport demand in the base year (1977) and then analyzes the expected future demands as projected by the transport model. The first section is concerned with the transport of freight within the country and between Bolivia and her neighbors. The second section deals with the demand for person travel, and the third unites the two earlier parts, in discussing total transport demand in the major corridors. Freight Transport Demand The fol~owing description of the base year transport characteristics is based on data derived from a number· of sources. Rail movements were taken from tables supplied by ENFE, with some adjustments described in detail in Working Paper 17, and summarized in Appendix 3B. River and lake data were extracted from matrices of movements supplied by the Bolivian Navy. Data on movements by pipeline were taken from statistics published by YPFB (1). Information concern- ing movements by road was derived somewhat differently, the primary source being the roadside interview survey conducted around the major cities as described in Working Paper 1, and summarized in Appendix 3A. On the basis of this information, synthetic matrices of movements were produced, which inclu- ded estimates of movements not captured by the survey. These matrices form the basis for this discussion. Freight movements in Bolivia can be divided into three major categories: domestic traffic; international traffic Which, in view of the country's geographical position and limited domestic market, is of vital econonomic importance; and international transit traffic which does not involve Bolivian goods but does involve use of Bolivian infrastruc- ture and, in many cases, equipment. The base year situation and expected future developments will be discussed for these three categories. (1) Diez Anos de Estad!stica Petrolera de Bolivia 1969 - 1978 (Ten Years of Petroleum Statistics in Eolivia 196~ 19'8) , YPFB, La Paz,1979. -143- Freight Transport within Bolivia • .. Tables 9-1 to 9-3 present the tonnages and ton-kilometer of freight transported by each mode of transport, excluding .. gas and crude oil, for the tnree main years of analysis, 1977, 19a9 and 1999. The projections for the future year movements • were derived from the freight model system described in Chapter 3, which begins by analyzing future grmv-tll in production and consumption in the country and then allocates the result.ing tran~nnrt flows to modes and routes. In this table, and those that follow, volumes for four future situations are shown. The 1989 Committed Network assumes that only those pro- jects actually committed in 1980 are carried out -- in most respects the infrastructure of the transport system remains the same as at present. Also for 1989, the results of two further computer tests are shown. One is for the 1989 Test Network, which contains most of the candidate projects eval- uated by the Study, as detailed in Chapter 10. Only those candidate projects which duplicated others in the Test Net- work were omitted. The third set of computer results for 1989 are for the 1989 Recommended Network, which contains only those projects finally recommended by the Study for im- plementation in the 1980s; that is to say, this network rep- resents the recommended transport system for 1989. Finally, the results for the 1999 Test Network are presented. This network contains all the projects recommended for the 1980s plus additional projects which could be viable by the end of the century. The individual networks for each year are des- cribed in the last part of Chapter 10. Tables 9-1 to 9-3 show only the forecasts of 'normal' traffic; that is to say, movements linking the productions and consumptions of goods input to the transport model. For the committed network, this 'normal' traffic represents the entire volume of goods forecast to be using the transport system. The remaining networks contain various projects which improve the transport system, either by opening up new areas previously unexploited, or by reducing transport costs in existing corri- ,1/ dors of communication. These improvements can be expected to generate new traffic, essentially by stimulating the economy of the country leading to greater production and consumption, and therefore to more transport. The stimulus given will vary by region, depending upon the location of new transport pro- jects and the degree of improvement they bring about. For 14 this reason, generated traffic was estimated on a project-by- project basis, and it is not possible to present an overall 1'1\ table showing the total volumes of generated traffic. As a guide, though, generated traffice increased total road traffic volumes in the Test Networks by about 6 percent. In addition, the rail interconnection was estimated to have an important generating effect, increasing flows on the interconnection by just over 50 percent, adding 531,000 tons to the 1989 Test Network and 1,325,000 tons to the 1999 Test Network, and in- fluencing the traffic on all surrounding links. ill -144-  Table 9-1 EXISTING AND PREDICTED INTERNAL FREIGHT TONNAGES BASE YEAR (1977) 1989 COMMITTED 1989 TEST 1989 RECOMMENDED 1999 TEST MODE 000 tons Percent 000 tons Percent 000 tons Percent 000 tons Percent 000 tons Percent ---- Air 41 0.3 74 0.3 64 0.3 69 0.3 110 0.3 I I-' .c:- Road 12,821 93.3 22,262 87.4 22,057 86.8 22,136 88.8 33,105 82.8 UI I 2,185 1,574 4,871 12.2 Rail 241 L8 1,978 7.8 8.6 6.3 River 200 1.4 147 0.6 63 0.2 135 0.5 116 0.3 Pipeline (I) 438 3.2 992 3.9 1,038 4.1 1,015 4.1 1,774 4.4 Total 13,731 100.0 25,453 100.0 25,407 100.0 ;:4,929 100.0 39,976 100.0 (I) Refined petroleum products only. NOTE: This table does not include generated tonnages, which were calculated on a project-by-project basis.  Table 9-2 INTERNAL TON-KILOMETERS BY l-10DE BASE YEAR {1977} 1939 COMMITTED 1989 TEST 1989 RECOMMENDED 1999 TEST MODE JlQ~-km I>erCJmt 000 t-km Percent 000 t-k.ro Percent OJl~-k:m Percent 000 t-km I>ercent RAIL Agriculture 11,483 0.7 90,362 2.4 148,100 3.3 85,765 2.5 396,604 5.2 Other 74 f 017 805,283 1,481,996 630,769 18.3 2,943,177 38.7 Subtotal 85,500 5.5 895,645 23.6 1,630,106 36.4 716,534 20.6 3,339,781 43.9 ROAD Agriculture 547,362 35.6 907,746 23.9 970,399 21.7 870,396 25.3 1,553,294 20.4 Other 623,341 40.5 1,542,558 40.6 1,466,865 32.8 1,465,016 42.6 2,027,823 26.6 I I-' Subtotal 1,170,703 76.1 2,450,304 64.5 2,437,264 54.5 2,335,412 67.9 3,581,117 47.0 "" '" I RIVER Agriculture 76,644 5.0 59,508 1.6 19,705 0.4 33,689 1.0 19,661 0.3 Other 62,403 5.4 98,479 2.6 59,115 1.3 75,110 2.2 75 f 844 1.0 Subtotal 159,047 10.4 157,987 4.2 78,820 1.7 108,799 3.2 95,505 1.3 PIPELINE (1) 122,439 B.O 293,461 7.7 329,029 7.4 ,931 -!:..L 596,678 7.8 GRAND TOTAL 1,537,689 100.0 3,797,397 100.0 4,475,219 100.0 3,440,676 100.0 7,613,081 100.0 (1) Refined petroleum products only. ...ii :.Ii ,. . ,.; .... ... Jij .. ,..;. ... ' .. .. ... . .. .. . ' " • • ill .. ., .. • iii 15 •  GENERAL VIEW CARGO FOR BOLIVIA GRAIN FOR BOLIVIA (HELD UP FOR ONE YEAR FOR LACK OF TRANSPORT) ARICA PORT (CHILE)  Table 9-3 INTERNAL COMMODITY MOVEMENTS BY ROAD AND RAIL (Thousands of tons) 1977 BASE 1989 COMMITTED 1989 :rEST 1989 RECOMMENDED 1999 TEST COMMODITY Road RaH Road Ra11 Road Rail Road Rail Road Rail -- Agriculture 6,564 24 10,734 192 10,655 271 10,728 166 14,783 632 Mining 81 44 438 131 426 144 460 100 477 290 Hydrocarbons (1) 897 47 1,865 195 2,029 31 1,688 170 3,228 472 I ....... Food 1,301 20 2,069 106 2,043 132 1,773 179 3,689 280 "'I" ...,J Textiles 101 1 133 2 131 5 134 2 194 9 Wood and Paper 437 43 711 214 560 364 758 157 1,121 704 Chemicals 59 6 144 104 132 116 190 59 318 437 Building Materials 3,241 26 5,828 388 5,817 398 5,910 300 8,980 863 Metals 23 11 122 578 61 640 239 400 101 1,041 Vehicles & Machinery 117 19 218 68 203 84 256 41 214 143 -- -- -- TOTAL 12,821 241 22,262 1,978 22,057 2,185 22,136 1,574 33,105 4,871 (1) The majority of hydrocarbons are carried by pipeline.  'l'able 9-1 shows tonnages by mode. In 1977, road transport dominates in terms of total internal tonnage, with 93 percent of tile movements, follo\,led by pipelines anc1 rail which carries '. a very small percentage. These proportion do not change very significantly in future years apart from that of rail which is discussed below. Hore meaningful than simple tonnage movements, however, is the proportion of total ton-kilometers carried by each .. ~ mode. This is shown in TavIe 9-2. Although road still dom- inates, rail is now relatively more important because of the longer average haul length of 350 km (compared with only , '\ 91 km.by road). Hore surprisingly, river transport reveals itself to be as important as rail in supplying ton-kilometers, due to the very long average haul length of 750 km. Pipeline transport, not all of whicH is shown in the table, accounts for about 20 percent of all internal transport in terms of ton-kilometers. Table 9-3 shows the internal freight movements by the two main modes of road and rail and by commodity. The two most important commodities are ag!icultural p~oducts (37 percent of the total in 1977) and hydrocrabons (28 percent of the total). The bulk of hydrocarbons goes by pipeline although about 60 percent of internal petroleum products are distributed by road. virtually all agricultural pro- ducts move by road, as dc, surprisingly, 60 percent of inter- nal mineral movements. After agriculture, the most important commo~ity for road transport is building materials which form 25 percent of total road tonnage. More details of base year and projected internal traffic volumes are shown in Appendix 9A. . Internal Freight Flows in 1977 - The outstanding aspect of the base year situation is the extremely low proportion of over- I' all internal freight movement carried by rail, which is surpris- ing in view of the low standard of the road network and the availability of reasonably lengthy hauls. This situation may be partially, though not entirely, explained by the lack of a connection between the Andean and Eastern rail systems, the low level of availability of equipment on the railroads in the base '1' year (due to poor maintenance and lack of traction) and perhaps most importantly, the relatively high rail tariffs and the lonqer haul times of rail versus road, rail averaging 40 percent longer overall. -148-  Internal Frei2~t Flows.in ~989 - The mos~ strik~ng feature of the internal fre~ght proJect~ons for 1989 ~s the ~ncrease of rail transport, although its proportion still remains relatively low. The increase in rail's share of total movements may be attributed to two factors. Firstly, the projections in all but the 1989 Recommended Network were based on variable costs only: in the case of rail these form a lower proportion of total costs than they do for road transport, thus there has been a shift to rail. For the 1~89 Recommended Network, recommended tariffs were used which are substantia~ly lower than current tariffs for rail and higher for road transport. Hence the shift to rail is still observed although it is rather less marked. In addition there is a considerpble amount of industrial expansion projected for areas served by rail lines. This is most noticeable in the case of metals, where a large part of the production planned for the Mutun complex is expected to be ship- ped by rail to Santa Cruz. Comparing the projected movement pattern of the Test Net- work with that of the Committed Network, there is a slight increase in the proportion of total tonnage carried by rail. The largest increase in usage is projected for wood and paper. This may be largely explained by the inclusion of the candidate rail extension to Trinidad. However, there is a significant decline in rail carriage of hydrocarbons due to the inclusion of several candidate product pipeline projects parallel to rail links. The inclusion of the candidate rail interconnection between the Andean and Eastern systems in the Test Network has surprisingly little effect on rail's share of total normal ton- nage (although a substantial volume of traffic is generated -- see earlier in this Chapter), but it does have a major effect on the distribution of flows, as will be shown in the section on total demand in major corridors. In general, the Recommended Network volumes are similar to the Committed Network but with less emphasis :on rail transport. River transport for the most part retains its share (I percent) of total tonnage in the 1989 committed projection, but in the test case its share declines sharply as a result of the assumed construction of the Cochabamba-Trinidad road, and a rail line and pipeline connection between Trinidad and Santa Cruz. Also included is a road connection between Trin- idad and Guayaramer1n, which diverts traffic from the river route. In the Recommended Network, only the roads remain in this corridor so that volumes are between those of the Committed and Test Networks. -149-  • , Turning to the analysis of ton-kilometers, the increase in rail transport is even more marked than for the tonnages carried. Average haul length grows from 350 km to 450 kID over the period 1977-1989 and the construction of the rail interconnection has a dramatic impact, the average haul length rising to 750 km. In the 1989 Recommended Network, the in- crease in ton-kilometers is not so great because ~) tariffs for both road and rail are at a higher level than in either the 1989 Committed or Test Networks and b) less projects are included than in the 1989 Test Network. In terms of overall movement, product pipelines retain their importance with 7 to 8 percent of the total. The volume of internal air freight is expected to nearly double by 1989. The majority of this will continue to be meat transported from the Beni to the population centers. In the case of the Test Network, a slight decline in volumes is expected as roads reach some of the more remote parts, and transport of meat to the La Paz market by road in refrigerated trucks becomes possible. This trend will likely continue, with an increasing diversion from air to road transport during the 1990s. Internal Freight Flows in 1999 - The importance of the individual modes remains much the same as in the 1989 case, with rail increasing its share from6 to 8 percent to nearly 12 percent of total tonnage. River flows decrease further in percentage terms with the increase in road access to the Beni region and the continued operation of the pipelines included in 1989. In terms of ton-kilometers rail increases its share of total internal traffice from 21 percent in the Recommended Net- work to 44 percent; this is due to the projected increase in tonnage carried, and to the inclusion of the rail interconnec- tion. Road remains the most important mode but only by a narrow margin, with 47 percent of total ton-kilometers as opposed to rail's 44 percent. This represents a complete change from the base-year situation where road transport sup- plied 76 percent of total ton-km, and rail less than 6 percent. River declines further in importance with 1 percent of movements, while pipelines retain a steady 8 percent of inter- nal movement. In conclusion, the most important change projected to take place in the next twenty years regarding internal freight move- ments is the significant switch to rail which in base year car- ried an almost insignificant proportion of tonnage and which by I' 1999 is projected to be supplying nearly half of all freight movements internally. These projections, it must be stressed, are dependent upon certain critical assumptions being made regarding tariff policy for rail, which receives detailed discussion in Chapter 17. Road remains extremely important in moving large \, tonnages over relatively short distances (an average of 100 km). River, by 1999, is expected to decline significantly as a means of internal transport. I, -150-  SANTA CRUZ MONTERO LA PAZ - ORURO PAVED ROADS Freight Transport between Bolivia and Neighboring Countries As a consequence of her size and geographical position, movements to and from the exterior play a very important role in Bolivia's overall transport system. Tables 9-4 and 9-5 show Bolivia's foreign trade tonnages by commodity and by mode for the years 1977, 1989 and 1999. For the base year, this information was derived from data stored on computer tapes, received from ENFE and the National Statistics Institute (INE). For future years, an independent model was used which took the generalized trade flows (except by pipeline) between internal districts and external zones and allocated them by mode to place of entry or exit on the basis of observed 1977 data. Foreign Trade Flows in 1977 - Rail plays a much larger role in the movement ox xreight into and out of Bolivia than it does for internal transport, as one might expect with the longer hauls involved and the virtual absence of good roads in several export corridors. Rail carried some 81 percent of total non-pipeline import tonnage in 1977, and 58 percent of exports (71 percent of overall external non-pipeline movement). If pipeline is included, the proportion carried by rail falls, but still remains far larger than that carried by truck (25 percent as opposed to 3 percent). However, as might be expected, pipeline is the dominant mode, with nearly 70 percent of overall external tonnage and a similar proportion of overall ton-kilometers (between the seaports and the Bolivian origins and destinations). Turning to the commodities in external movements the predominance of hydrocarbons is extremely marked, accounting for 70 percent of total tonnages. No other individual commodity forms more than 7 percent of total external tonnage, minerals slightly outweighing agricultural products. Lake transport plays a significant role, in export traffic especially, with about 8 percent of total non-pipeline export tonnage. Virtually all of this is carried by rail between the lake and the Peruvian seaport. Air supplies about 1 percent of import tonnage, and an insignificant proportion of export tonnage. Foreign Trade Flows in 1989 - No radical change in the supply of transport by individual mode is projected for imports in 1989; rail's share of total non-pipeline tonnage falls marginally to 75 percent with road transport carrying some 18 percent. Regarding exports, however, there is a significant change as waterborne transport becomes an important element. About 43 percent of non-pipeline exports are expected -151-  Table 9-4 OBSERVED AND PREDICTED IMPORT TONNAGES BY MAJOR MODE (1) 1977 1989 1999 COMMODITY KlAD RAIL WA'IER FOI\D RAIL WA'IER 0ClAD RAIL WA'IER Jlgricul. ture 19,630 143,346 5 50,340 133,700 18,270 95,618 153,000 6,391 Minerals 591 2,799 310 5,150 967 883 5,485 1,598 Hydrocarbons 3,022 25,720 5,547 28,500 19,992 27,764 Food 14,846 93,879 13,800 19,139 86,600 5,281 31,517 61,780 5,816 I I-' U1 Textiles 4,342 4,292 5,425 10,750 1,428 15,530 h.l I Wood and Paper 7,074 14,091 4,792 25,000 2,356 5,020 Chemicals 10,226 30,242 9,963 18,017 41,550 9,995 12,243 22,380 3,834 Building Materials 6,396 104,033 1,284 15,000 324 3,201 12,155 266 loetals 16,293 83,046 9,981 21,563 95,400 7,405 27,863 115,050 9,373 Vehicles and Mach.inery 6,114 50,070 8,338 19,593 165,600 12,037 75,296 184,350 39,729 'lUl'AL 88,534 551,518 42,087 146,010 607,250 54,279 270,,397 602(514 67,007 Percent 13.0 81.0 6.0 18.1 75.2 6.7 28.8 64.1 7.1 U) Exclw.es pipeline and air tonnages. NOTE: This table indicates mode of arrival in the country and so totals for rail do not agree closely with those given by ENFE for imports. For example, traffic moving via Guagui is recorded in this table under water and not rail, even though it then mainly transfers to rail. Similarly, some traffic arrives by road at the frontier and then transfers to rail. Also, rail statistics for 1977 do not distinguish international transit traffic. Finally, ENFE and INE statistics do not agree exactly. -;> ~ ,:., -:; " .i . ., lOt .. •• ...4 . . It •  Table 9-5 OBSERVED AND PROJECTED EXPORT TONNAGES BY MAJOR MODE (1) 1977 19.89 1999 CCMmI'IY mAD RAIL WA'lER mAD RAIL WA'lER mAD RAIL .WA'lER .Agriculture 18,560 15,074 156,604 165,560 10,115 239,330 509,440 20,000 Minerals 24,006 197,518 79,865 18,434 137,960 30,748 36,235 270,550 74,727 Hydrocarllans 2,100 I Food 64,929 46,989 143,68B 100,750 194,4B6 242,640 10,000 I-' U1 IN Textiles 4 10 74 5,882 18,800 B09 9,000 33,000 7,050 I Wood and Paper 10,485 35,008 23,810 126,850 101 14,B19 167,340 3,000. Chemicals 18,072 4,659 8,982 10,780 9,103 40,466 23,800 7,000 Building Materials 55 8,657 94,250 6,069 29,047 49,550 4,500 lIEtals 281 811 8,613 30,560 764,754 9,336 108,370 757,986 Vehicles and Machinery 738 4(085 16,210 5,000 55,550 52,300 45,000 'I'OTAL 136,337 302,962 79,939 378,755 701,720 826,699 62B,269 1,456,990 929,263 Percent 26.3 58.3 15.4 19.8 37.0 43.2 20.9 4B.3 30.8 (1) Excltrles pipeline arrl air tonnages and potential agricultural exports on Paraguay River. to travel by water (as opposed to 15 percent in 1977)1 this radical change is brought about through the development of the Mutun complex, the products of which are expected to be exported via the Paraguay River. The proportion of tonnage carried by water may increase still further than shown, since these projections exclude any major agricultural developments in the eastern region of the country, the products of which would also use the Paraguay River. These are likely to be much more significant in 1999 than in 1989, but may still have some impact by the earlier year. , .~ As a result of the increase in river transport, rail's share of non-pipeline export traffic falls to 37 percent and that of road transport to only 20 percent. In absolute . ·ci terms, however, both modes will substantially increase their export tonnages, which are expected to be between 100 and 200 percent above the levels of 1977. Air played a very small part in the total of external freight movement in 1977, shipping small quantities of high value goods. It is possible, though unlikely by 1989, that its role will become more important with the growth in exports of agricultural products, in particular meat. This possibility receives more detailed discussion in Chapter 15. Foreign Trade Flows in 1999 - Obviously the projections for 1999 should be treated with care: the forecasts of foreign trade rely on a number of assumptions which mayor may not be shown to be correct as time passes. Much less reliance should be placed on these conclusions than on those for 1989. The position of the modes relative to one another is expec- ted to remain much the same as 1989, with rail handling by far the greatest import tonnage (64 percent) and carrying more export tonnage than other modes (48 percent). The figures shown in the table refer to non-pipeline transport and continue to exclude possible agricultural developments in eastern Bolivia, that might be exported via the paraguay River. Even when the construction of the rail interconnection is taken into account, the overall modal split does not alter significantly, since the major ex- pected shift is from one part of the rail system to another, not from one mode to another. Air is likely to playa larger part in export traffic by 1999. It should be noted that all foreign trade flow projections " assume a continuation of current policy on the part of Bolivia regarding the use of the available ports in the neighboring countries. Should a major switch take place from the use of one country's facilities to another, this could have a corres- ponding impact on the mode of transport used to carry I, -154- H' the goods: for example there is no rail access to the Peruvian port of Ilo, nor reasonable road access to the Chilean port of Antofagasta. Changes in the use of these facilities have clear implications in the division of transport demand between the modes. Alternative policies regarding usage of foreign ports have been examined when evaluating proposed projects in international corridors (see Chapter 15). More details of external movements are presented in Chapter 15 and Appendix 9B. International Transit throush Bolivia Bolivia shares a border with five countries, namely Peru, Chile, Argentina, Paraguay and Brazil. Thus, there exists considerable potential for transit traffic through Bolivia between all these countries. However, many share common bor- ders with direct road and rail connections -- the flows of which might naturally pass through Bolivia are the movements between Peru and Chile on one side, and Paraguay, Brazil and Argentina on the other. Several routes are available using the Bolivian rail system while the condition of the internal road system and the distances involved are such as to dis- courage transit traffic by truck. The main available routes are as follows: (i) Puno (Peru) - Guaqui - Cochabarnba - Santa Cruz - Corumba (Brazil) (ii) Puno (Peru) - Guaqui - Villaz6n - La Quiaca (Argen- tina) (iii) Arica (Chile) - Charana - Cochabamba - Santa Cruz - Corurnba (Brazil) \iv) Antofagasta (Chile) - Avaroa - Santa Cruz - Corurnb~ (Brazil) (v) Arica (Chile) - Char ana - Villaz6n - La Quiaca (Argentina) (vi) Antofagasta (Chile) - Avaroa - Villaz6n - La Quiaca (Argentina) The present transport between Peru/Chile and Brazil re- quires a transfer to road transport between Cochabarnba and Santa Cruz. Table 9-6, showing the volumes carried in 1979 on these routes, demonstrates that the transport consists almost entirely of metals, with some wheat and cement carried as well. The most important flow in terms of tonnage is that from Argentina to the Chilean port of Antofagasta, amounting to 65 percent of what is admittedly a small total. However, it is thought this flow was an unusual occurrence, brought about by temporary capacity constraints on the Argentinian rail system; the most stable flows are undoubtedly those of metal from Peru and northern Chile to Brazil. Accordingly, in projecting future transit traffic, particular attention has been given to these movements. -155-  Table 9-6 INTERNATIONAL TRANSIT TRAFFIC ON BOLIVIAN RAIL SYSTEM, 1979 ROUTE COMMODITY TONS Peru - Brazil Guaqui-Cochabamba-Santa Cruz-Corumba Copper '.and zinc 6,780 I Guaqui-Villaz6n (rail line to Co~ interrupted) Garlic 121 I-' V1 Chile - Brazil CTI I Charana-Cochabamba~Santa Cruz-Corurnba Copper, zinc, antimony 919 Avaroa-Cochahar:-ba-Santa Cruz-Corumba Copper 5,422 Argentina - Chile Villaz6n-Charana Cereals 301 Villaz6n-Avaroa Flour 27,600 Chile - Argentina Avaroa-Villaz6n Cement 1,209 Total 42,352 / AI iii ... .,0 .. q) ... '" ...ill ...,.. ...  Projecting transit traffic ten and twenty years into the future is extremely difficult, since it is a function of many uncertain developments, not least the economic futures of the various trading partners. A simple projection has been made, assuming that trade between the countries concerned will continue to grow at the same rate as the economy of the leading partner (Brazil in this case), as has been the case historically, and that the great part of this trade will use Bolivia as a transit corridor: this is more likely to be so in the case of Peru than Chile, where alternative routes to Brazil exist. Thus, only about one fifth of the total Chilean- Brazilian trade is assumed to pass through Bolivia. The likelihood of Bolivia becoming a transit corridor would be greatly strengthened by the construction of the rail inter- connection between Santa Cruz and Cochabamba; however, this project is not recommended until the 1990s. 1989 Transit Traffic - Based on the assumption that the Brazilian economy will grow at the approximate rate of 7 percent per annum, as it has for the past 5 years, trade with the relevant countries is projected at a similar rate. A total potential transit tonnage of 39 2 ,000 tons is thus predicted for 1989. Of this, 57 percent is Brazilian-Chilean trade, the remainder being trade between Peru and Brazil. 1999 T~ansit Traffic - Assuming a higher growth rate of 9 percent for the Brazilian economy after 1987, which is not unlikely if it frees itself during the 1990s from its present energy constraints, the potential transit volume through Bolivia will rise to over 900,000 tons by 1999. However, this figure should be treated with extreme caution as major changes in trading patterns may have occurred by that time. Also, population shifts in Chile and possible changes in relative tariff levels of competitive routes (either overland via Argentina or by ship) make this a very speculative figure. It is not considered likely that the construction of a rail interconnection in Bolivia would generate nlajor flows between Brazil and the Pacific ports. The relatively high costs of overland transport would certainly outweigh those of waterborne transport between Brazil's Atlantic ports and, say, Japan. Even for transport to and from Brazil's western states, a cheap route exists via the Paraguay River, thus making it unlikely for this traffic to pass through Bolivia. More details of transit traffic movements are presented in Appendix ge. Passenger Travel Table 9-7 summarizes passenger travel by each mode for the base year and demonstrates that passenger movements are overwhelmingly dominated by road which serves over 95 percent -157-  Table 9-7 PASSENGER TRAVEL, 1977 (Thousands of passengers) MODE INTERNAL % EXTERNAL % TOTAL % ROAD Cars 7,076 14.4 8 1.6 7,084 14.3 Buses 23,488 47.8 16 3.2 23,504 47.2 Trucks 15,929 32.4 49 9.9 15,978 32.2 Motorcycles 874 1.8 874 1.8 I I-' \.11 00 Subtotal 47,367 96.4 73 14.7 47,440 95.5 I RAIL 1,175 2.4 162 32.5 1,337 2.7 LAKE 2 0.4 2 AIR LAB, TAM 594 1.2 131 26.3 725 1.5 Other 19 130 26.1 149 0.3 Subtotal 613 1.2 261 52.4 874 1.8 TOTAL 49,155 100.0 498 100.0 49,653 100.00 ." :;: ,~ ~ ~. ~ 4 ~. ~ ~ • ~ of overall demand. Most road passenger movements in 1977 were made by bus; this is particularly true for movements between department capitals where bus accounts for nearly two thirds of all movements. The other major carrier was truck, and movements by car were quite low. Motorcycles account for an insignificant proportion of movements. It is interesting to examine more closely passenger travel by truck. About one third of all truck passengers were in groups of six or less. About 75 percent of the remainder, or 7.5 million passengers, travelled in trucks otherwise empty, devoted in fact to passenger transport. Of the remaining modes, rail is the most important for passenger transport. This is closely followed by air, the volumes reflecting the fact that many towns, particularly in the Beni and Pando, have no alternative access by road. In terms of geographical distribution of passenger trips, the departments of La Paz, Cochabamba and Santa Cruz dominate with over three quarters of all recorded passenger trips. Turning to external movements, the picture changes dramatically, with air carrying by far the greatest proportion of trips (52 percent). This is considerably more than the next most important mode, rail, with about a third of all trips. Road accounts for only 15 percent of external passengers, most of these travelling -- surprisingly -- by truck. Future Passenger Transport - Passenger travel forecast for 1989 reflects an annual increase of slightly over 2 percent. A certain shift from road to rail transport is projected by the transport model, especially in the case of the Test Network, which envisages a more extensive rail system. The rail passenger increases are similar to those projected for freight: higher than average increases are predicted in the Committed Network for the Eastern system and in the Test Network for the sections affected by the rail interconnection Santa Cruz - Cochabamba. The number of internal air passengers is expected to increase by over 20 percent per annum during the coming years. Major leaps in demand, such as that experienced in 1977 - 78, are unlikely in the future, since jet service now exists between nearly all major cities. Air transport offers the most rapid and convenient service between these cities, and will continue to provide the only access to many towns in the northern region of the country. -159-  External passenger flows are expected to increase in line with the general level of passenger movements internally. External air movements increase slightly more than internal with the likely growth in the tourist sector within Bolivia attracting more travellers from abroad. More details on passenger movements are presented in Appendix 9A. Transport Demand in Major Corridors In this section, the major flows of passengers and freight within the country are analyzed. They are discussed for 1977, 1989 (Recommended Network) and 1999 for road and rail. Internal, international and transit traffic are included in the flow • maps of this section. Road traffic volumes are shown as total daily vehicle flows, while rail volumes are in terms of annual tonnages. Road Transport - The largest generators of road traffic are the major cities of the country -- La Paz, Santa Cruz and Cochabamba. In 1977, La Paz generated the most passenger movements (7,030,000) followed by Santa Cruz (5,606,000) and then Cochabamba and Oruro. On a departmental basis, however, Santa Cruz generated more trips (4,529,000) with its larger dispersed population, followed by La Paz and Cochabamba. With regard to the generation of freight movements, Santa Cruz department dominates with over 40 percent of total tonnage (5,547,000), followed again by the departments of La Paz (2,151,000) and Cochabamba (1,917,000). When these figures are converted into vehicle flows and assigned to the 1977 network, it is clear that vehicle movements are largely concentrated in a central corridor running due south from La Paz to Oruro and then eastwards to Cochabamba and Santa Cruz. As shown in Figure 9-1, movements in other parts of the country are mostly very small, averaging 60 vehicles per day on unpaved roads. The most heavily travelled corridor of any length is that between La Paz and Oruro which has a good paved road connection. The highest volumes overall are found in the corridor to the north of Santa Cruz, which to some degree is due to the urbanized nature of that zone. With regard to 1989, the overall increase in volumes has been commented upon earlier. As evident from Figure 9-2, the rise in vehicle movements is well distributed throughout the system, the most notable increase being for the Potos! and Cochabamba areas. A substantial increase in truck traffic is projected for Potos!, partly due to the assumption in the Recommended Network that the roads from Potos! to Oruro, to Totora Palca and partly to Sucre would all be paved. -160- .'  Son de Moxos Villorroel 10 Puerto Paiios Salinas de Gorei Mendoza Forlin Villolon 19n DAILY ROAD VEHICLE MOVEMENTS FIGURE 9-1  Son Cerro Toronl L Villa Bermejo 1989 DAilY ROAD VEHICLE MOVEMENTS FIGURE 9-2  The overall level of generated truck traffic, brought about by the projects assumed in the Test Network, is not great, the variations between regions reflecting the amount of new road building assumed: the cities with the greatest increase in truck traffic are those now connected by the paved roads just described. The situation in 1999, presented in Figure 9-3, shows an overall increase in truck traffic of about 45 percent over 1989, considerably lower than the increase projected for the period 1977-1989 (80 percent). The most noticeable increases occur in the Sucre and Tarija regions; these may be explained by the assumption that Sucre would be the focal point of three new paved roads to the east, west and north, while Tarija would benefit from new paved roads from Potosi, Villamontes and Bermejo. Rail Transport - As has been remarked upon earlier, inter- national flows form a very important part of volumes carried on the rail system. For this reason, they have been distinguished in the three figures depicting base year flows and the projec- tions for the 1989 Recommended and the 1999 Test Networks. Flows have been projected for each of the ten major commodity groups but only total tonnages are shown in the figures for clarity. In 1977, freight flows were concentrated in the two main corridors: Viacha-Oruro-Uyuni and Santa Cruz-Coru~b§. External flows heavily outweigh internal traffic in both volume and distance travelled. As shown in Figure 9-4, the most heavily used part of the whole systerr is the section between Uyuni and R!o Mulato, carrying nearly 500,000 tons of which about 87 percent is external traffic. The lowest flows in the system are found on the Cochabamba-Aiquile line (8,000 tons), on the section between Santa Cruz and Santa Rosa (28,000 tons), on the stretch linking Sucre with Potos! (32, aaa tons) and on the link saving the lake port of Guaqui (50,000 tons, most of which is import traffic). These figures ~erely serve to emphasize the extremely low utilization of the Bolivian rail systew: many stretches carried less than one train a day in base year. For 1989, depicted in Figure 9-5, a large increase is pro- jected in the amount of rail traffic. The major proportion of the increase comes in internal traf~ic. This is particularly noticeable on the Santa Cruz-Corurr~a line which will be used for the transport of metal products from Mutun to Santa Cruz as well as movements from the new cement plant at Yacuses. The external flows experience a much lower percentage in- crease than internal movements, but still approximately double throughout the system. Major increases are projected for the -161- Viacha-Oruro section and larger increases on the links to the ports of Guaqui and Arica. Flows on the southern part of the network to Avaroa and Villaz6n are not projected to increase, reflecting the growing volume of Bolivia's trade with the coun- tries of the Andean Pact, particularly Peru. Major growth is expected on the Santa Cruz-Corumba line as exports to Brazil and other destinations from Santa Cruz increase. Figure 9-6 shows that once again a large increase in inter- nal flows virtually throughout the system is forecast for 1999. The effects of the inclusion of the rail interconnections from Cochabamba and Sucre via Mataral to Santa Cruz are particularly significant. These particularly increase flows on the La Paz- Oruro-Cochabamba sections, these cities now being directly linked to the east of the country. The peak internal flow is projected for the new section between Mataral and Santa Cruz. Owing to improved road connections, the rail volumes between Rio Mulato and Uyuni are projected to fall, but elsewhere on the system they rise by as much as three times. The highest volume of internal traffic is on the section linking Mataral with Santa Cruz (2.4 million tons), the lowest on the Sucre-Potosi line (136,000), where no increase is projected from 1989, again partly due to improved road connections. The inclusion of the rail connection Santa Cruz-Cochabamba in the Test Network brings about major changes in the distribu- tion of foreign trade flows as exports and imports to and from the Santa Cruz region use the Pacific rather than the more dis- tant Atlantic ports. Figure 9-6 demonstrates that large in- creases could occur on the Oruro-Cochabamba line and the links . I to the ports of Guaqui, Arica and Antofagasta. water Transport - For water movements some comment needs to be made upon the changes projected for this mode. In the base year a substantial amount of freight (approximately 200,000 tons) was moved on the Ichilo-Mamor~ river system, the major flow being between Puerto Villarroel and Trinidad. The flows projected on the Committed Network for 1989 show no change for the sections to Guayaramerin and Magdalena, and increase corresponding to the general growth in the Beni. Growth is not as high in the Puerto Villarroel-Trinidad section as there is competition from new roads in the south Beni. In the Recommended Network, a further road is assumed (Cochabamba-Trinidad) which reduces demand for river transport. In addition, the Test Network contains a road to Guayaramerin, parallel to the river, and a road to Magdalena which would be shorter and much quicker than the river route. In 1999 internal volumes decline still further on the Trinidad- Puerto Villarroel section, though doubling elsewhere on the system. Volumes between Trinidad and Guayaramer1n are thought likely to reach 100,000 tons by 1999 with the inclusion of ex- ternal traffic and the movement of cattle to the border town for export. -162- 1999 DAILY ROAD VEHICLE MOVEMENTS FIGURE 9-3  , o TRINIDAD \~ [T9!T] ~"'" LEGEND ( \ \ , , INTERNATIONAL AND TRANSIT TONNAGE , ,-----,.~ ~ \ 1'" r"'-T [:?:::::~:q DOMESTIC TONN AGE " ::;:;:_:_:.:_:4.;_;'.. , GUAOUI \ """ \ '-" '. " .... " -_/ / " \ ~ ;' , - - --- -~-----/) , ( ~ , /J/ I ~ '~'\ \ , I f.J - ,..-" "~''') .. CHARANA , ~ / MATARAL o ~ \ .~ AIQUILE \ '\ '~ ~ '\....,.,/ \, ---(\ ;' "v-; <. -.. . . . --~. \, I I - - l '.- / . - - - - --- -~ ""-." I :, . CORUMBA (BRASIL) ~ r \ I , \ N --- " I /' / '~-, ~ ) :- \ ; t,. ! II /"~.. _, J"'- ---J o l _.. 50 100 150 200 km \ ! I mm = 100,000 IOns 1917 ANNUAL TONNAGE flOWS BY RAil iI)Itt/(l? !J'mil/' and .;rI~joc.'ale(j FIGURE 9-4  I 't. , ) 11989 ( RECOMMENDED) t I -,. LEGEND \. • INTERNATIONAL AND TRANSIT TONNAGE W~ ; \ ~ <"- ", \ ,,-- --\- --------\ mmmmmi DOMESTIC TONNAGE \ ---, J - .-- -- - ----) \, --.-.' \ 1 ) . . r I ~-, SANTA CRUZ "- ,/" "l, .CHARANA 'y'" ".........., . " \ , "'" r \ \. ..\,~"'\ \,, . , ,,--'-, \ -/ 1_ -II' ' '--i, \ '\ SUCRE " VI ~, I --"'\ ,II ..... ~ , ) ,.-/ I I I ," ( ~ < ----_ ... -( ( ~ I I . I ! : , j "-.. ~-- -------.J , I . I '~ -- I o '0 100 150 200 kill \ ,, ./ I 1111111 100,000 tOni '\ - .- --...... V1Ll\AZON ~ r' / 1989 ANNUAL TONNAGE flOWS BY RAil i.i)ft/6ut .'Fmillt and A:.:.ocialpJ FIGURE 9-5  LEGEND .] \ , , International and Transit Tonnoge ----------...:::,. 1"'_ /~-\ Inter n al Tonnage \ '""'- -- ,J : ~CHA \\ ,, ! 7 , ~ CRUZ "l, \CHARA_~ \ ( I ') t'-u _____~~LR,i~st- , ',.-/ ~;nr I ,- i I ": " ',COR UM~A I {/~) (BRAZIL) " --- o 50 100 50 20e km "!ioiiiiiiooii mm = 100,000 !on 1999 ANNUAL TONNAGE FLOWS BY RAil <i)fil6ut ffmiti! and .;itJocciate(J FIGURE 9-6  CHAPTER 10 TRANSPORT PROJECTS SELECTED FOR EVALUATION CHAPTER 10 TRANSPORT PROJECTS SELECTED FOR EVALUATION There is no shortage of plans for the development of transport in Bolivia. They range from quite minor schemes such as the regravelling of some existing roads, to ambitious major investments, such as the interconnection of the two ~ail s~stems. A few projects are currently being built, 1nclud1ng two new paved roads and a new rail line, but most projects are only plans or concepts, often with only an approximate idea of their cost, potential benefits and priority. For road and rail alone, the projects which have been suggested would cost nearly ~50,OOO million pesos. In contrast, the entire budget for inter-urban transport investments is unlikely to exceed 34,000 million pesos for the next ten years (1), (of which nearly 13,000 million pesos are already committed), and 55,000 million pesos for the 1990s. Clearly, many projects have to be postponed, reduced in scope or rejected in.select- ing an investment program for the decade 1981-1990. This chapter first examines the deficiencies of the exist- ing transport system, then identifies transport projects which are completely committed and goes on to present the candidate projects selected for detailed evaluation by the Study. Fin- ally, the transport networks defined for the computer model are presented. Deficiencies of the Transport System The main population centers of the country are separated either by severe, and sometimes unstable, mountainous terrain or by plains subject to seasonal flooding and containing large rivers which are liable to change course from one year to another. It is not surprising, therefore, that the transport system has many limitations. However, it would be wrong to describe all these limitations as deficiencies, with the implication that they have to be corrected. Given the terrain, the often tortuous alignment of the mountain railroads is an acceptable solution to the problem of linking population centers at an economic cost. Given the conditions of landing strips and the volume of trade, the use of 1940s vintage aircraft may be an acceptable low cost solution to the problem of lifting meat supplies from the low lying Beni to the 4,000 meters high Altiplano. There are deficiencies, though, and in the two examples just quoted, the generally low level of maintenance of the rail track in one, and of the aircraft in the other, is a cause for concern. In the rest of this section, the principal deficiencies of each mode, as well as some of the more insoluble limitations, will be identified. (1) See Chapter 8. -163-  .. • Railroads - The railroad system is narrow gauge with infrequent passing tracks but, even with a large increase in current volumes, there are unlikely to be capacity problems in the near future. '. The principal deficiency of the railroad is the poor quality of the track which is mainly the result of insuffi- cient maintenance. ENFE acknowledge that at current levels of maintenance, the track is gradually deteriorating. The problem is made worse by the fact that the initial construction of the railroads was not to a high standard. In particular, sleepers were laid directly on to the soil base without ballast. A program of rehabilitation and ballasting is now under way with World Bank support, and on such rehabilitated lines maintenance requirements are reduced considerably. By the end of 1979, 433 km of the total ENFE system of 3,643 km had been rehabili- tated. Inadequate prov~s~on of drainage structures in the or~g~n­ al design has made some lines vulnerable to flood damage 1 an example is the section of the Santa Cruz-Corumb~ line near Robor~ which was washed away late in 1978. Other lines pass through unstable mountainous zones subject to earth slides and disruptions during the wet season. Especially affected are the Cochabamba-Oruro line and the Atocha-Villaz6n line. Realign- ments of the track in these zones could alleviate these problems. , '1 Six of the nine departmental capitals have rail connections, ,I the exceptions being Cobija and Trinidad in the north and Tarija, in the south. However, the rail system is divided into two parts, the eastern and the western systems which are not con- nected except by a detour through Argentina. Therefore, Santa Cruz is isolated from the other departmental capitals with '1 rail connections. There are projects to remedy this situation, with proposals for four separate interconnections as well as links to the departmental capitals of Trinidad and Tarija. There are problems in the operation of ENFE which stem from lack of equipment and a low serviceability rate, the latter due to inadequately equipped and staffed maintenance facilities. Action is currently being taken to improve these aspects with funds from the World Bank Third Railway Project Loan. Highways - The principal deficiency of the highway system is the poor quality of road surfacing. Only about 1,300 m, of paved roads are open to traffic, with another 120 km under construction. Unfortunately, many of the existing paved roads have deteriorated badly. For example, the Cochabamba-Santa Cruz highway, built in the 1950s and which accou~ts for.over a t~i7d of the paved roads in the country, now requ~res maJor rehab~l~­ tation. The deterioration is attributable to insufficient maintenance but has been accelerated by the high(illegal)axle , "\ loads prevailing on large trucks. -164-  The rema~n~ng roads are surfaced either with gravel or earth. In the basic road network modelled in the computer, there are over 6,000 km of each type. Although such roads can be satisfactory for low volumes of traffic, the surfaces in Bolivia are remarkably poor. Oversized stone is used on the gravel roads, so that even frequent grading is not suf- ficient to maintain a smooth surface. In fact, grading and other maintenance operations are not performed frequently enough so that the unpaved surfaces are often very poor. There is also a lack of bridges and minor drainage structures. In the central region, most rivers can be cross- ed at all times, except for periods of a few hours after heavy rain. Similarly, the lack of drainage structures does not necessarily make routes impassable, but it does confine them to reasonably rugged vehicles. However, there are some routes in the central region which are impassable throughout the wet season of December to March. In the flat northern and eastern regions, many rivers can be crossed only be ferries, but here the size of the rivers, and their likelihood to alter course, make the provi- sion of bridges prohibitively expensive in many cases, for the current levels of traffic. Also in these flat regions, there is a lack of drainage structures. These need to be much lar- ger than in the mountainous regions in order to pass the large volumes of flood water in the wet season. The San Borja to Trinidad road, for example, cuts across many water courses and is on a high embankment to raise it above flood levels, but it lacks a number of small bridges and many drainage structures, and consequently it can be interrupted in the wet season for long periods. The unpaved mountain roads, which form a large part of the main road system, are generally narrow and poorly aligned. However, volumes are low with no unpaved road reaching more than 400 vehicles per day in 1977, and with only about 500 kms with more than 200 vehicles per day, so that the poor align- ment and narrow surfaces cannot be regarded as such serious deficiencies as the poor surface. Lack of capacity is defi- nitely not a problem tOday on Bolivian roads (see Appendix lID). Mainly in the north and east, there are large areas with no road connection at all, and these are therefore entirely dependent upon river and air transport. Given the very small size of these communities, it is not clear that lack of roads can properly be classed as a deficiency. Indeed, the linking of all these remote areas by means of a road system is proba- bly a luxury which Bolivia cannot afford for many years, although the selective and gradual development of such roads will be necessary to exploit fully the resources of these regions. -165-  There are fewer missing links in the road system of the central region. In all cases, the gaps occur in areas of severe terrain and separate comparatively small communities. Air Transport - The most glaring deficiency in air trans- port is the lack of a proper air navigation and communications system. The equipment in use is old and unreliable and does not form a complete network. With some major exceptions, such as the operations of Lloyd A~reo Boliviano (LAB), the same neglect of air safety evident in air communications extends to other parts of the air sector. Maintenance of the non- regular aircraft fleet is poor and the system of inspection • and enforcement of basic regulations regarding the airworthiness of aircraft and the fitness of crew leaves much room for improvement. Most departmental capitals have, or are building, airports with paved runways. The two exceptions are potosi, where the gravel runway is being improved, and Oruro, where the existing gravel runway is sufficient for present requirements. The run- way facilities are, or will be, adequate in all cases with the exception of Cochabamba where the runway is too short for the optimum operation of modern passenger jets (a plan exists to correct this). With the rapid growth in air passenger movements, pressure is now being felt on the terminal facilities at a num- ber of departmental capital airports. Mainly on urban planning grounds, Santa Cruz airport is now being rel?cated to Viru Viru • and there is also pressure tQ relocate the a1rports at La Paz and Cochabamba. An alternative project for Cochabamba, pro- posed for safety reasons, is a runway on a new alignment at the existi~g airport. Airfields in other parts of the country are, for the most ?art, primitive. Th~ basic construction of the airstrips 1S poor and maintenance is negligible. This effectively restricts their use to certain types of 1940s vintage aircraft, although F27 passenger aircraft can be operated from some air- fields. These airfields are liable to be closed for periods in the wet season and poor systems of air communication and navigation, mentioned earlier, result in severe restriction or cancellation of operations in poor weather. There are several daily flights between the three major cities of La Paz, Cochabamba and Santa Cruz but many other centers, including departmental capitals, have flights only once or twice a week. While more flights would certainly generate more traffic, it is not clear whether the demand for travel is currently suppressed by inadequate services, nor whether the provision of more services could be economically justified. This question is examined in more detail in Chapter 14. -166- COCHABAMBA - VILLA TUNARI COCHABAMBA-SANTACRUZ (NEAR SANTA CRUZ) POTOSI - SUCRE (NEAR SUCRE) EARTH SLIDES ONTO HIGHWAYS  Rivers - River transport is the least developed of all the transport modes in Bolivia. The basic facility is the river itself, and little has been done to enhance its natural attributes. River ports are generally crumbling, muddy banks with all transfer of cargoes by manual labor. The abundance of labor and the instability of the river banks have dis- couraged the introduction of mechanical handling but the lack of such equipment need not necessarily be classified as a deficiency. Probably the main deficiency is the presence of hazards in the river caused by sand bars and embedded tree tr- unks. These restrict downstream navigation to daylight hours, generally limit speeds and are sometimes the cause of major accidents and loss of boats. There are plans to clean the Ichilo-Mamor~ river between Puerto Villarroel and Guayaramerfn over a period of years. The wide meanders of the river system also result in long journey times, and there are ambitious plans for canals to shortcut some of the bigger river loops. These canals would also have the advantage of bypassing many of the river obstruc- tions which would otherwise have to be cleared. There are major rock obstructions in the northern part of the river system, making it impossible to navigate between the Mamor~ river and the western river system (the Rivers Beni, Madre de Dios and Orthon). More important, the rock obstructions also block navigation from all the Bolivian rivers to the main Amazon system in Brazil. In the southeast, Bolivia has access to the Paraguay River but at present has no proper port facility. Puerto Busch is a metal barge tied up to the river bank and has no land access in the wet season. At Quijarro a port has yet to be developed so that exports now have to use the Brazilian port of Corumba. Lack of an adequate port and suitable access facilities is a hindrance to the exploitation of the vast Mut~n iron ore reserves and this must be regarded as a deficiency. Silting on the Paraguay River is also a problem, although the solution to this lies mainly in the hands of Brazil and Paraguay. Lake Titicaca - Of the two main lake services, the rail roll-on/roll-off ferry between Puno (Peru) and Chaguaya near the Matilde mine (Bblivia)is adequate, especially considering that production at the Matilde mine has declined over recent years. -167-  The main deficiency of the Guaqui-Puno service is in the cargo handling facilities. Cargo is transferred between rail wagon and boat by ancient steam cranes in both Guaqui and Puno, and this is a relatively costly and time consuming • process. Possibly, unit handling techniques (pallets and containers) might be encouraged to reduce time and costs. Unfortunately, the difference in gauge between Bolivia and Peru rules out a simple roll-on/roll-off ferry, as between Puno and Chaguaya, although this could be accomplished if combined with a rail wagon change-bogie facility. Solutions to these problems must be considered in relation to plans to improve the road and complete the rail link around the lake, and this is discussed in Chapter 15. Pipeline Transport - This is a very specialized type of transport, but once a pipeline is constructed, its operation can be straightforward. In fact, the operations of the Nation- al Hydrocarbons Company (YPFB) are relat~vely sophisticated with multiple-product pumping and frequent modifications of the pumping program. These follow reviews of supply, demand and transport costs (including costs by other modes, where appro- • priate), to ensure optimum use of the pipeline system. All major centers are connected to refineries by petroleum product pipelines and further development of this system must wait on growth in demand. A trunk gas pipeline system is currently being considered by YPFB. • Possibly the major deficiency of the pipeline system is the lack of import and export lines. The line to Arica proba- bly must be reversed in order to import heavy crude oil, which Bolivia lacks. This leaves one export facility to Argentina for the disposal of Bolivia's excess of light gasoline products and, unfortunately for Bolivia, Argentina can supply its own needs for these products. Hence, Argentina is prepared to offer only a low price for the Bolivian product. Committed Projects A certain number of transport projects were taken to be completely committed and were therefore not examined in detail by the Study. The definition of a committed project required some care. In general, projects were only considered committed when financing was completely assured. This includes many projects in construction, but not all, since many projects proceed as and when funds are available with no definite timetable. It was considered that the Study could usefully comment on priori- ties in these cases, and so such projects were included in the list of candidate projects for evaluation. -168-  Projects considered as committed by the Study are illus- trated in Figure 10-1 are are discussed below. Committed Railroad protects - The line from Yapacanl to Rio Grande, which is part 0 the projected railroad from Santa Cruz to Trinidad, was taken to be committed by the Study, al- though not without a certain amount of controversy. It was a condition of World Bank funding for the Third Railroad Project (which includes this Study) that no further work should be done on the Santa Cruz-Trinidad line without proper feasibility study employing consultants. However, by the time the Study evaluations had started, an agreement had already been signed between the governments of Bolivia and Argentina, committing Argentinian funds to the project, and construction work has now commenced. There seemed little point, therefore, in eval- uating this section of the line. Instead, the Study concen- trated on the as yet uncommitted final section of line from R!o Grande to Trinidad. Completion of the third phase of the railroad rehabilita- tion program is committed with World Bank assistance and a Study of the proposed Motacucito-Puerto Busch line is also committed. Committed Highway Projects - The following projects are in construction and are fully committed and financed, with construction expec.ted to be completed well before 1989: l. La Paz - Cotapata (paved 47 km) 2. Bella Vista - R~o Quiquibey (gravel - 67 km) 3. Rio Quiquibey - Rio Yucumo (gravel - 42 km) 4. R10 Yucumo - San Borja (gravel - 50 km) 5. 6. R10 Yucumo - Puerto Salinas Quillacollo - Confital (gravel (paved - 113 - 105 km) km) 7. San Ramon - Trinidad (gravel - 367 km) 8. .. San Ramon/San Julian - San Javier (gravel - 46 - km) 9. 10. .. San Javier - Concepcion Potos~ - Tarapaya (gravel (paved 62 19 km) km) 11. potosi - Uyuni (gravel - 240 km) 12. Guanay - Mapiri - Apolo (earth - 210 km) 13. Sorata - Mapiri (earth - 205 km) 14. Puerto Rico - Conquista (earth 60 km) 15. Conquista - Blanca Flor (earth 65 km) 16. Cobija - Porvenir - Chive (earth - 152 km) 17. Pto Siles - San Joaquim - San Ramon (earth 89 km) ~e following three projects were also considered to be fully committed with financ~ng assured (by the Interamerican Develop- ment Bank - IDB), although construction has not yet started: 18. Cotapata - Santa Barbara (paved 38 km) 19. Confital - Caihuasi (paved 58 km) 20. Chimore - Yapacan! (paved - 153 km) -169-  Some projects suggested by the National Road Service (Servicio Nacional de Caminos - SNC) were not included as committed since funding appeared uncertain. These are: • Eteramazama - Rio Ichoa ( 80 km) • • Santa Rosa - Riberalta (420 km) • Charazani - Apolo (180 km) • • San Ignacio de Velasco - Concepcion (178 km) •• For the same reason, the completion of the San Borja - Trinidad road was not included as a committed project as • there was no definite funding to complete bridges and drain- age structures. In addition to highway construction projects, a major road maintenance project for the three districts of Cocha- bamba, La Paz and Santa Cruz is committed with World Bank funding. A feeder road project is also committed with USAID funding, although the project is currently suspended. Committed Air Projects - The following projects were considered to be committed: • • Construction of Santa Cruz (Viru Viru) airport • Construction of Cobija airport • Construction of Puerto Su~rez Airport • Construction of Riberalta Airport • Improvements at Potosi Airport • Completion of Tarija airport improvements • Runway repairs at Trinidad , • Air Navigation and Communication Equipment Plan. • Projects for airports in the Beni at San Borja and Santa , Ana de Yacuma, even though approved and partly financed by • the World Bank in 1978, have made so little progress since then that they were not still considered to be committed. • These projects are all discussed further in Chapter 14. Candidate Projects A list of projects for evaluation was compiled for each mode based mainly upon plans provided by each of the modal organizations. The initial list of projects was circulated to many organizations interested in transport development. These included each of the modal organizations, the development corporations for each departmen~ and various government min- • istries. Based upon comments received, and upon a better understanding of the problems as evaluation proceeded, a re- , vised list was made, which is included in Appendix lOA. Each project was defined by the location of the end points and was , numbered for ease of reference during the evaluations. -170-  lU£NO _ _ '''''[0110''0) _ _ _ _ UNf'iI'V[O ROAD IN COif~"UCTION +0++++++ ItAI. LINE ASSLlftD "MAIICIN!; _ PAYEOffOAO\' ~ "All UN( > COWPlETED SINCE r~77 - - - UIIIP.w£DRIJ,lO/ r· Q o Q COMMITTED TRANSPORT PROJECTS U)fillu.. [fmil" and Jlooocialeo FIGURE 10-1 QUILLACOLLO CONFITAL LA PAZ -COTAPATA MAJOR ROAD CONSTRUCTION PROJECTS  The projects chosen for each mode are discussed below: Candidate Rail Projects - Candidate rail projects are illustrated in Figure 10-2. The basic characteristics and estimated costs are shown in Table 10-1. The principal rail project, and one of the most impor- tant projects to be evaluated by the Study, is the rail system interconnection linking the western network at Aiquile (east of Cochabamba) to the eastern network at Santa Cruz, and pass- ing via Mataral (Projects 100 and 101). Almost 400 km in length, and expected to cost about 10,000 million pesos, this would be one of the most ambitious projects to be included in a future transport plan. It would link the two rail sys- tems and, at the same time, provide the main missing link in a trans-Continental railroad linking Brazil with Chile and Peru. Other possible interconnections between the two systems include lines from Tarabuco (east of Sucre) to Mataral (Pro- jects 103, 104 and 104A), from Tarabuco to Cuevo (near Boy- uibe) (Projects 103 and 105), from Cochabamba to Yapacani (Project 116) and an interconnection in the south which would pass via Tarija (Projects 108 and 109). The last section of the Santa Cruz - Trinidad railroad (Project 102), from the projected terminus on the Rio Grande of the section currently under construction, was included at an estimated cost of just over 1,000 million pesos. Two pro- jects were considered linking the iron ore deposits at Mutun to Motacucito on the Santa Cruz - Corumb~ line (Project 106), and to Puerto Busch on the Paraguay River (Project 107). These projects depend upon the strategy selected for develop- ing Mutun and for providing a port on the Paraguay River. A small project in the west was included, linking Guaqui to Desaguadero on the Peruvian border (Project Ill). This project is entirely dependent upon Peru constructing the line from Puno to Desaguadero, alongside lake Titicaca (Project 114). A branch line was considered to Camiri (Project 117), which in the major town near the Santa Cruz - Yacuiba line without rail connection. A rail project was considered linking Bermejo in Bolivia with Oran (Project 115), a rail terminus in Argentina. Although the benefits to this project would accrue to Bolivia, construction would be almost entirely in Argentinian territory. A new rail project was included from La Paz to Rurrena- baque in northern La Paz (Project 110). This would be a repeat of the now abandoned and dismantled railroad into the Yungas from La Paz, and would be a very expensive undertaking at an estimated 17,000 million pesos. -171-  '. Table 10-1 CANDIDATE RAIL PROJECTS MAXIMUM MINIMUM PROJECT TERRAIN LENGTH GRADIENT RADIUS ESTIMATED COST (km) (%) (m) (millions of 1977$b) 100 Aiquile-Mataral Mountainous 184 2 130 4,462 101 Mataral-Santa Cruz Mountainous 207 2 130 5,016 102 Rio Grande-Trinidad Flat 149 0.8 300 983 103 Tarabuco-Zudanez Mountainous 61 3 77 300 104 Zudanez-Vallegrande Mountainous 150 2 130 2,730 104a Mataral-Vallegrande Mountainous 55 2 130 994 105 Zudanez-Cuevo Mountainous 302 2 130 7,308 106 Motacucito-Mutun Flat 26 0.8 500 184 107 Mutun-Puerto Busch Flat 107 0.8 1,000 1,059 I 108 Balcarce-Tarija Mountainous 250 2 130 10,500 I-' -..J 109 Tarija-El Palmar Mountainous 180 2 130 4,356 N 110 La Paz-Rurrenabaque Mountainous 500 3 130 16,600 I 111 Guaqui-Desaguadero Flat 25 0.8 500 100 112 Oruro-Cochabaroba improvements Mountainous 12 465 113 Robor~-Taperas improvements Flat 102 360 114 Puno-Desaguadero Flat 150 585 115 Or~n-Bermejo Flat 50 0.8 500 200 116 Cochabamba-Yapacani (1) Flat to Mountainous 380 9,720 117 Camiri branch line (1) Flat to Undulating 45 300 (1) Very rough estimate. SOURCE: Costos de Construcci6n de Ferrocarriles (Costs of Railroad Construction), Working Paper 38, Bolivia National Transport Study • • 11 It" If. 1B • ., •••• II ••••••••• !Ii if. 5- * ........ 4 .... ,. 50 0 '---' NOTE : NUMBERS REFER TO 100NTIFlCATION IN TABLE 10 ­ I ... o () t r ------~"'/ \> -, I"'OTOI (. 'i. .,~ o ,\ II I I; X T I ~ ,\ CANDIDATE RAIL PROJECTS FIGURE 10-2  Two major realignment schemes were included to avoid unstable regions on the existing lines from Oruro to Cocha- bamba (Project 112 - in fact three separate schemes; see Chapter 13) and from Santa Cruz to Corumb! (near Robor~) (Project 113). The Oruro - Cochabamba line has always been affected b~ earth slides in the wet season. The Corumb~ line near Robore was badly damaged by floods in the 1978/79 wet season and it is believed by some that conditions in that area were fundamentally changed so that the line in now per- manently unstable and liable to disruption each wet season. Not included in the map, but evaluated by the Study, was the rehabilitation of the existing lines of the rail system which involves ballasting of the track, replacement of sleepers and some small bridges, and minor realignments. The equipment needs of ENFE over the next ten years were also evaluated. Finally, the Study considered a pro- posal for the electrification of the railroad system. Candidate Road Projects - Candidate road projects are shown in Figure 10-3 and are listed in Table 10-2. Over 13,000 km of road are included, which may suggest a bias towards road projects at the expense of other modes. How- ever, this is not really so. Most of the projects in the central, southern and western regions of the country are improvements of existing highways which, with the exception of parts of the limited paved system, are generally in poor condition. Proposed improvements range from simple regravel- ling to major realignment and construction. Indeed, each candidate project consists of a number of alternatives from which one was selected as a result of the forecast traffic levels and economic evaluation. For this reason, it is not possible to present a compact table of design characteris- tics and cost for each project. These are discussed in detail in Chapter 11. A rather different type of higl.way project was considered in the Pando and Beni departments and in northern La Paz and eastern santa Cruz. These regions have few roads and vigorous attempts are being made to correct this. New roads are under construction to Trinidad from La Paz and San.ta Cruz, gravel- ling is in progress on the Santa Cruz - Concepci6n road and new penetration roads are being built in the regions of Cobija and Puerto Rico in Pando, and to Apolo and Ixiamas in La Paz. However, vast regions still have no regular road access and the cost of developing a network in these areas on the scale shown the Figure 10-3 would be very high --' in the region of 15,000 million pesos. In general, the distances are very great and the communities to be linked very small. -173-  Table 10-2 CANDIDATE ROAD PROJECTS PJ\NOO tEPARlMENT Pa:OCSr IEPARl'MENT ClltQUISACA IEPARl'MENT 600 Nacebe-Riberalta 650 Uyuni-Tupiza 710 Sucre-Padilla 601 HunaitlS:-cachuela Esperanza 651 Cuchu Ingenio-M:>jo 711 Padilla-MUyupampa 602 cachuela Esperanza-Fortaleza 652 M:>jo-Villaz& 712 Villa Serrano-Pucara 653 Potos!-Tc:rratas 713 Boyuibe-VillCllDltes LA PAZ IEPARl'MENT 654 Potos!-Sucre 714 Penas-Fortln Villaz& Ixianas-au.~ 655 Potos!-Uyuni. 715 Azurduy-Palos Blanoos 610 657 Uyuni-Avaroa (formerly OllagueJ 611 San BiJenaventura-Ixiamas 658 Uyuni-Quetena TARIJA IEPARl'MENT 612 Apolo-'lUnapasa 613 Adlacachi -charazani 720 Iscayachi-M:>jo 614 Achacachi-Sorata 721 Padcaya-Be:mejo BENI lEPARl'MENT 615 Huarina--1lchacachi 722 Tarij a-VillCllDltes 616 Tiquina-Cbpaoabana 670 Cruce Blanca Flor-Ri.beralta 723 VillCllDltes-Yacuiba 617 R!o Seoo-Desaguadero 671 Cruce Blanca Flor-Santa Rosa 724 Zapatera-canbar! 618 Viac:ha-Corocoro 672 Blanca Flor-Cruce Blanca Flor 725 Trarentinal-Carapari (near Yacu1ba) I ..... 619 La Paz-I..aIri:Jate 673 Cruce Blanca Flor-Puerto Siles ..,. -...I 620 Ulduavi- Inquisivi 674 Puerto SHes-Guayaranerfn SJ\NTA CRJZ IEPARIHNT I 621 Inquisivi-Panduro 675 Trinidad-San Miguel de Gasser 622 Inquisivi -Independencia 730 Mataral-Vallegrande 676 San Miguel de Gasser-Santa J\na de YaClllla Patacamaya-Puerto Japcn~ 731 Santa Cruz-Boyuibe 623 677 San Miguel de Gasser-San RaOCn 624 Santa BAt:bara-Bella Vista Santa Ana de YaClIlla-Santa Ihsa 732 Santa Cruz-Cbtoca-Puerto Pailas 678 625 Viacha-Glarafia 733 Adapt rail bridge over R!o Grande to take road 679 San R:m'5n-Magdalena 626 Olarazani-Ap:>1o Magdalena-San Ignacio de Velasoo as well 680 627 Lairbate-Ia Plazuela Magdalena-ca.sarat:e 734 Pail&-Ias TrcIloos 681 Trinidad-San Ignacio de M:>xos 735 Pail.6n-San Jose de Chiquitos 682 ORtJR) IEPARlHNT San Borja-San Ignacio de M:>xos 736 San J~ de Chiquitos-<:OJ:'U'llb! 683 737 San Rafael-San Mat!as 684 EteranazCllla-San Ignacio de Moxos 630 Puerto J~-Tamto Quenado 738 MutCn-Puerto Busch 685 Bridge over the R!o McmJ~ 631 caraoollo-Tamto Quemado 739 M.lyupanpa-Ipat! 686 San Joequ1il-La Ibrquilla 632 Oruro-Caihuasi 740 Val1egrande-Ipit! 633 Madlaca:narca-Unc!a 741 Bridge over R!o Grande at Puerto Ba1egas CDC.H1IBAMBA IEPARl'MENT 634 Machaasnarc:a-Olallapata 742 Orioepci.&-San IgnaciO de \elasoo 635 O1allapata-lJyuni 690 Quillaoollo-Independeneia 743 MutGn-Quijarro/'Puerto S\l.1rez 636 O1allapata-Tarapaya 691 Ind:!penc:'k!ncia-I.equepalca 744 San Jos(S de Chiquitos-Fortln Ravelo 637 Ule!a-Sucre 692 Vint0-i30l1var 638 Oruro-Sabaya-Pisiga 693 R!o caine-Anzaldo EXTERNAL 639 Santiago de Huarf-Llica 694 Epizana-Sucre 695 Aiquile-Palizada 750 Desaguadero-Puno (Peru) 699 Bolivar-Anzaldo ., JIll lit • • • It :j ., til • j • 4 • 4 •• • 4 • ~ • • II! .. w • • • • ~ •• • ,  1 NOTE : NUMBERS REFER TO IDENTIFICATION ~ IN TABLE 10- 2 50 0 100 200Km . I I I t, ,.. -/ . ("l ~ d o (') III l> 2 o 1J ( \ l> \. o " ", ,'. \' \' \ . \', \', \\ \) \. '\ ~ (. " () 1 \ -- \ ' ." ,~ :­ " C AUOE l\: T r ;>,:,\ r CAN DIDAlE ROAD PROJ EClS , FIGURE 10-3  Since the communities to be served are by no means isolated, being served by river and air transport, the strate- gic and regional integration reasons for providing these road links did not appear overwhelming. Perhaps a rich country could afford to build such a network, although the experience of Brazil is cautionary, but it was concluded that Bolivia is definitely not in such a position. However, it was agreed that the provision of penetration roads is essential to the development of the resources of these regions, provided that they are built with a complementary feeder road system and that the pace of road development is matched by the set- tlement and development of the newly opened regions. Thus the evaluation of these roads d~ds highly on the analysis of the potential agricultural and mineral resources of these regions, and on the location of markets, and less on the need to complete links of a national highway system. This analysis, termed the producer surplus analysis, is described in Chapter 12 together with a detailed description of the penetration roads considered. Bridge projects were generally considered separately from the road candidate projects and only two major river crossings, the River Mamore west of Trinidad, and the Rio Grande east of Santa Cruz, were given separate project numbers. The evalua- tion of bridges is described in Chapter II, with the exception of the Rio ~amor~ crossing (Project 685) which is discussed in Chapter 12. Candidate Air Projects - Thirty-six separate airport pro- jects were identified for evaluation. They are listed in Appendix lOA and illustrated in Figure 10-4. A range of costs were estimated for each airfield and this is discussed in detail in Chapter 14. Other special projects selected for evaluation are the use of wide-bodied jets to carry foreign trade, and the use of dirigibles for cargo transport. Plans for the re-equipment of LAB (Lloyd A~reo Boliviano, the National Airline) fleet and for the modernization of the non-regular aircraft fleet were also examined. Candidate River Projects - River transport candidate pro- jects are shown in Figure 10-5 and are listed in Table 10-3. On the Paraguay River, the main issue is the choice of a location for a new river port, partly to serve the needs of the Mutun project but also to provide an outlet for other Bolivian exports. Two locations were identified, Puerto Busch (Project 201) and a port at Quijarro (project 203), which were compared with the alternative possibility of continuing to use, and developing, the facilities in the Brazilian port of Corumba. -175-  Table 10-3 CANDIDATE RIVER PROJECTS PROJECT ESTIMATED COST 6;nillions of 1977$b) (1) 200 Improve condition for navigation on the Paraguay River 24 201 New port on Paraguay River at Puerto Busch 900 203 New port near Paraguay River at Quijarro 680 (2) 205 Improve Bermejo River from south ot Bermejo to the Paraguay River 5,580 205(a) Build canal from south of Bermejo to the Paraguay River (3) I 206 Clear Mamore River of rock obstructions between Guayaramerfn and Villa Bella (3) ....... 207 Bypass Cachuela Esperanza (between Riberalta and Villa Bella) (3) ....... 0'\ 208 Clear obstructions on Madeira River between Villa Bella and Porto Velho(Brazil) (3) I 209 300 ( 4) Construct all-season canal from Mamore River to Trinidad 210 Provide new port facilities at Puerto Villarroel 46 211 Provide new port facilities at Trinidad 46 212 Provide new port facilities at Guayaramerfn 46 214 Equipment for river cleaning between Puerto Villarroel and Guayaramer!n 30 (5) 215 Port on Rfo Grande at terminal of rail line currently under construction from Yapacan! - 46 (1) Between Corumba and Puerto Busch. (2) Includes cost of dredging the Tamengo Channel. (3) Cost estimates not available. (4) Assumes average 7 meter depth of excavation. (5) Cost of boat plus equipment • • • • • • • • • • • • 4.4 Ittl , •••••••• 4 1It •••••• .i.4 .. "  ) 50 0 ~ 100 ZOOKm . I I I ... SAN JOSE OE CHIQUITOS 523 ROBORE o 522 ,r---------- III ) 2 o ... , G ~ \> \J ~ ~ v ~ ~ (> \J F \ ~ ~ o '\--- ~ o .\ R G E~TIN.\ CANDIDATE AIRPORT PROJECTS FIGURE 10-4  ) •• 50 0 ~ 100 ZOOKm . I I I ~ : NUMBERS REFER TO IDENTIFICATION IN TABLE 10 - , :0 " ;~ t'l / " J 0/ ~ .J ~. , J al ~ l_ . / r' r ... ~ .. f' , 7 ~ d trl ) ) 1 .-1-- .!L '( J o () z o ' .. " I / ----------- , ~, )~ ... . . \ C \.> ( ~ .~.' t/-' ~ ~~ ~ y. ~ '-' . " \\ \) ~. . \oJ ) i ,.. () ! \ -. ~ \ , » \\ " i '. ~ o " R EP l ' B AHGEX T I;:-JA CANDIDATE RIVER PROJECTS FIGURE 10-5 Issues of sovereignity are involved here, but the Study con- fined itself to a comparison of costs. An important factor in evaluating the alternative port locations is the cost of the necessary land connections. Other associated issues con- cern the cost of river improvements on the Paraguay River (Project 200) and the possible limitation imposed on barge train configurations by a Brazilian bridge over the Paraguay River at Puerto Esperanza, between Corumba and Puerto Busch. ; Many projects were identified on the Mamore River system between Puerto Villarroel and Guayaramerfn. One project deals with the river cleaning proposals, a scheme already started with help from the Belgian government (Project 214). Other pro- jects concern port improvements at Puerto Villarroel (Project 210), Trinidad (project 211) and Guayaramer!n (project 212), and also on the Rfo Grande at the terminus of the rail line now under construction (Project 215). At Trinidad, a scheme for a canal linking the Mamore to the center of Trinidad for all- year use was the subject of another evaluation (Project 209). North of Guayaramer1n, several ambitious projects were identified to clear obstructions in the river between Villa Bella and Guayaramerfn (project 206), between Villa Bella and Riberalta at Cachuela Esperanza (Project 207), and between Villa Bella and Porto Velho in Brazil (Project 208). Finally, two ambitious projects were identified in the south to permit navigation from just south of Bermejo to the Paraguay River. One involves the dredging of the Bermejo River (Project 205), the other the possible construction of a canal to bypass the Bermejo River (Project 205A). Both projects would be almost entirely within Argentina and would be linked to other projects to develop that reqion of Argen- tina. Thus, they would not be under the' control of Bolivia but could be exploited by Bolivia if Argentina decides to go ahead with them. Miscellaneous Candidate Projects - The rema~n~ng candi- date projects are shown in Figure 10-6 and are listed in Table 10-4. Roll-on/roll-off ferry facilities were considered at Guaqui on Lake Titicaca (Project 300). Associated with this is a wagon bogie change plant for conversion between narrow and standard gauge (Project 301). An alternative wagon bogie change plant project was considered for Desaquadero (Project 302) for the proposed junction of the Bolivian and Peruvian rail systems. Two petroleum product pipeline projects were identified, one from Villamontes to Tarija (Project 400) and the other from Santa Cruz to Trinidad (project 402). All other major cities already have product pipeline connections. - -177-  ''I Table 10-4 MISCELLANEOUS CANDIDATE PROJECTS PROJECT ESTH1ATED COST (millions of 1977 pesos) 300 Rail roll-on/roll-off facilities at Guaqui 14 301 Change of wagon bogie plant at Guaqui 84 302 Change of wagon bogie plant at Desaquadero 84 400 Product pipeline: Villamontes to Tarija 510 402 Product pipeline: Santa Cruz to Trinidad 516 800 Hovercraft Ferry at Puerto Banegas 80 (1) 801 Motorized pontoon Ferry at Puerto Banegas 3 (1) .. 802 Mineral Conveyor Belt: Mutun to Quijarro 500 . (1) For two craft. • . • • .. III • • .. III -178- .. • III •  NOTE : NUMBERS REFER TO IOENTIFICATION IN - TABLE 10- 4 I r' o :' ~'------" 1i n III "'-- RIIO _ _ _ --~ OUTU 8 2 ~ ,,..----- , ~ , 2. o 1) c \ ~ ~ ,/ -=;/ /:. til o .\ 1\ G E .\ T I X .\ MISCELLANEOUS CANDIDATE PROJECTS rJ)fi"ult [lmil/' and Jliw;ciale() FIGURE 10-6  A hovercraft ferry (project 800) was considered for the crossing of the Rio Grande at Puerto Banegas east of Santa Cruz. Two craft would be required at an estimated cost of $b40 million each. As an alternative, a motorized pontoon was evaluated at a cost of $b3 million for two craft. A mineral conveyor belt was considered linking Mutun with Quijarro (project 802) at an estimated cost of $bSOO million. The Computer Transport Networks Three transport networks were defined which were the basis of the computer transport model forecasts of person and freight movements used in the evaluation of candidate projects. They were: (i) 1989 Committed Network (ii) 1989 Test Network (iii) 1999 Test Network. In addition to these networks, a 1989 Recommended Network was defined containing all projects recommended by the Study (see Chapter 20). This was used to establish the final traffic estimates on each link and was defined, built and tested after the main evaluations were completed. It is not discussed fur- ther in this section. Although all the candidate projects discussed in this chapter were the subjects of evaluations, not all projects were included in the transport networks defined for the computer model tests. For those projects which were included, a decision had to be made about the level of design tested, and this was particularly important for road projects. Projects were selected for each computer network after a preliminary screening, which is reported in Working Paper 53. ~his was a necessary process to eliminate projects whose viability was considered very unlikely by the year of the computer network, and whose inclusion would be likely to complicate the evaluation of remaining projects. Similarly, design levels of projects were set to represent a feasible network by the year in question, and this was largely decided by considering likely budget availability and probable priori- ties. The screening process was not too rigorous, since it was not meant as a substitute for the full-scale modelling and evaluation process. Sufficient projects were included to allow considerable scope in the selection of alternative plans based on the results of the evaluation. Each network is briefly described below. -179-  11 .. 1989 Committed Network - This network contains the exis- ting links of the transport system, including new links opened since the base year 1977, together with those projects listed earlier as completely committed. •• 1989 Test Network - This network contains all the candi- date projects for testing using the computer model. The road candidate projects involving construction on a new alignment which were included in the 1989 Test Network are illustrated '. in Figure 10-7. In addition, it was assumed that all unpaved roads would be regravelled to provide a better surface. Most . of the development road candidate projects in northern La Paz, Pando, Beni and Santa Cruz departments were included, even though it was clear that only a small proportion would prove .. to be feasible. They were included because it was thought that the computer test would show up any significant network advan- tages of each route, which could be used to supplement the more • important producer surplus analysis. At the same time, volumes forecast on these routes were not expected to be large and so would not distort the forecasts on the routes in the central, western and southern parts of the country. This assumption proved to be correct. Apart from the development roads, the 1989,Test Network includes p~ving of the routes f~om Oruro to Potos~, Sucre (via Potos~), Tarija and Villazon, from Santa Cruz to Yacuiba (apart from one section), and some roads near La Paz including .. the routes to Desaguadero, Achacachi and Bella Vista. Some new gravel roads were included east of La Paz, south of Cocha- bamba and east of Tarija. Non-road projects included in the 1989 Test Network are also shown in Figure 10-7. Both the rail interconnections from Aiquile to Santa Cruz via Mataral and from Sucre to Mataral .. were included, but the Tarabuco-Cuevo line and the inter- connection via Tarija were omitted. Also omitted was the rail line from La Paz to Rurrenabaque which preliminary analy- sis showed to be completely unfeasible by 1989. Both the petroleum product pipelines were included. .. River projects, apart from the projects ~o cl~ar rock obstructions south of Villa Bella, were not 1ncluaed, although they \V'ere later evaluated on the basis of the model forecasts. Air projects were considered separately in the air .. • model (see Chapter 14). .. • The 1999 Test Network - It was considered that as far as non-road projects were concerned, sufficient projects had been included in the 1989 Test Network to fill an investment .. • • -180- • •  ... ... ' ~ .} LEGEND _ PAVING PROJECT 50 0 , _ _ _ NEW GRAVEL ROAD I ....... RAIL LINE _ •• _ PRODUCT PIPELINE ••••••• RIVER PROJECT ." ttl "1) (", !:xl 'C" .... ("'l ... d . ~ () ) r-----·' Z Q 11 ) ~ ~ ~ () \) ~ ~F ~ ~ ~ \) () '\ ~ () ~ ~ Q .\ R G E N T I : ' l l A THE 1989 TEST NETWORK: CANDIDATE PROJECTS O)fe,/Iu. 9'mill. and A66ociale6 FIGURE 10-7 program for at least the next twenty years, when taken together with other investment possibilities for these modes. Investment possibilities beyond this were evaluated but no more non-road projects were included in the 1999 Test Network. All road candidate projects involving new construction in places where there are no roads, or only primitive tracks, were included in the 1989 Test Network, although sometimes at fairly low levels of design. Hence, projects for inclusion in the 1999 Test Network all concerned upgrading of either the alignment or the surface of roads already in the 1989 Test Net- work. The selection of projects for the 1999 Test Network was made after further project screening. The improvements included are shown in Figure 10-8. Essentially, the paving projects represent the completion of the trunk road system in the central part of the country, including the links to Trinidad. -181-  1 ~ • • z.. ... ..;.. _ RECONSTRUCTION AND PAVING 50 0 , 00 1 2ooKm . _ _ _ PAVING ONLY I I ,0 ~ v ~ y (", c :- -/ ~ ~- ~' / r' r n . j , d ~ / o () 11\ ) Z o "U C ) I \ I \J ~ () \L-< \ \ .~ y, ~ ~ - ~ ~ \J \J ~f ? 0 () i / ""'-- -\ ~ ~ ~ ; '--.." 'l o " ..\RG EX·TI:\r..\ THE 1999 TEST NETWORK: ADDITIONS TO THE 1989 TEST NETWORK eo %tlut [lmil/' and VIIo4«ial FIGURE 10- 8  TABLE OF ABBREVIATIONS $b Bolivian Pesos- SUS 1.00 = $b 25.00 (1981) SUS 1.00 = $b 20.00 (1977) ADT Average Daily Traffic AADAA Customs Administration AASANA National Airport Administration ADEPTA Association of Private Air Transport Firms COMIBOL National Mining Company COMIXTA Bolivian-Argentinian Joint Commission CNT National Transport Center CEPAL Economic Commission for Latin America CODETAR Taiij~ Department Development Corporation COFADENA National Development Corporation of the Armed Forces CORDEBENI Beni Department Development Corporation CORDECO Cochabarnba Department Development Corporation CORDECH Chuquisaca Department Development Corporation CORDECRUZ Santa Cruz Department Development Corporation CORDEOR Oruro Department Development Corporation CORDEPANDO Pando Department Development Corporation CORDEPAZ La Paz Department Development Corporation CORDEPO Potosi Department Development Corporation DGCP General Directorate of Port Captaincies DGMM General Directorate of Merchant Marine DGTA General Directorate of Road Transport DGTTA General Directorate of Air Transport and Aerial Works DPC Directorate of Planning and Coordination ENFE National Railroad Company FNB Bolivian Navy GNP Gross National Product GEOBOL Bolivian Geological Service GDP Gross Domestic Product HDM Highway Design and Maintenance Standards Model IBRD International Bank for Reconstruction and Development (The World Bank) IBTA Bolivian Institute of Agricultural Technology IDB Interamerican Development Bank INE National Statistics Institute INALCO National Cooperatives Institute INTAL Institute of Latin American Integration IRR Internal Rate of Return LAB Lloyd Aereo Boliviano LAC Lineas Aereas Canedo LAI Linea Aerea Imperial LPG Liquid Petroleum Gas LINABOL Bolivian Shipping Line MTCCA Ministry of Transport, Communications and Civil Aeronautics NPV Net Present Value NTS National Transport Study SHN Hydrographic Service of the Navy SN Structural Number (Used in Road Pavement Design) SN Modified Structural Number SNC National Road Service SNT National Transit Service TAM Transportes Aereos Militares UNDP United Nations Development Programme USAID United States Agency for International Development YPFB National Hydrocarbons Company