Peru - Public Expenditure Review for the Infrastructure Sectors

Confidential 17808 - PE PERU Public Expenditure Review for the Infrastructure Sectors RETURn TO LATIN AMERICA ANO CARIBBEAN IN FO1MAii ER C 41 March 4, 1994 LAl Department Infrastructure Operations Division Latin America and the Carribean Regional Office FILE COPY PERU Public Expenditure Review for the Infrastructure Sectors Table of Contents Page No. I. INTR ODUCTION ........................................1 A. Overview of Investments in Infrastmcture ....................... 1 B. Structure of Expenditures and Sectoral Priorities ................... 2 U. ANALYSIS OF SUBSECTORS ...............................3 A. The Road Investment Plan.................................3 B. Other Transport: Ports, Airports, Railroads, Tren Electrico ............ 8 C. Water Supply and Sanitation .............................. 13 D. Regional Housing InvestmentPlan........................... 17 ANNEXES Annex A: Planned Investment in Infrastructure Annex B: Review of Tren Electrico Project Annex C: Detailed Analysis of Road Budget Annex D: Assessment of the Capacity of the Peruvian Construction Industry Annex E: Detailed Analysis on Water and Sanitation PERU - Public Expenditure Review I. Introduction A. Overview of Investments in Infrastructure The role of the public sector and the structure of infrastructure investments have changed dramatically in Peru since 1990. During the three preceding decades, Peru followed a state-led development strategy. This was characterized by heavy investments in civil works (particularly highways); the entry of public enterprises into such sectors as railroads, ports, and airport operations, and tight regulation of the private sector (i.e, trucking, urban bus transit). Many of the investments were misdirected; focusing on large scale new infrastructure while neglecting maintenance and rehabilitation of existing networks. State enterprises were inefficient. Underpricing in utilities prompted a vicious cycle of revenue shortfalls and declining service quality. The result was a staggering maintenance backlog and a chronic lag in the provision of services to the poor in rural areas, small cities, and marginal urban areas. The capacity of the state to carry out this dominant role in infrastructure provision eventually collapsed in the late 1980's--the result of overall economic decline and civil strife; as well as the inappropriate sectoral policies in specific infrastructure subsectors. Since 1990, there has been a fundamental shift in the Government's approach to infrastructure services. State enterprises in many subsectors (airlines, telecommunications, urban bus transit) have been privatized, and private providers (in intercity trucking, for example) have been deregulated. Within the infrastructure activities that remain in the public sector, there has been a shift in the composition of expenditure, away from new civil works toward rehabilitation, and efforts to introduce commercial incentives into the management of public enterprises. This report reviews the proposed Government strategy for the infrastructure sector. (Infrastructure is defined as roads, other transport (aviation, ports, rail, urban mass transit), and water supply and sanitation.) In particular, the report focuses on proposed expenditures on infrastructure in the 1993-97 public investment plan. The report assesses the Government's spending priorities in the major infrastructure subsectors and identifies institutional and financial reforms required to improve the efficiency of infrastructure spending. 1 B. Structure of Expenditures and Sectoral Priorities The Government of Peru is planning to spend about $4.2 billion on infrastructure between 1993 and 1997, about 35% of the total public sector investment plan. This is a significant increase over the previous plan period. The overall level of spending on infrastructure is consistent with the recommendations of an IMF mission which visited Peru in November 1993. (The mission suggested reducing the capital investment budget of the Ministry of Transport, Telecommunications, Housing, and Construction (MTCC) by about 15%.) The sectoral composition of infrastructure expenditure during these five years is expected to change, reflecting the changing role of Government in the infrastructure sectors. Expenditure allocations in road infrastructure will increase (from 48% of the infrastructure budget in 1993 to 71 % of the same in 1997) while expenditures on aviation, ports, rail, and communication sectors will decline as the state's role in these sectors diminishes. Overall, the transport sector alone will account for $2.6 billion, about 22% of public expenditures over the five year period. Water and sanitation expenditures over the same year are expected to be around $ 1 billion (close to 9% of the total public sector investment plan). We support the sectoral shift in the composition of Pa VIlc Investment In Infrastructure: infrastructure spending, . as it is consistent with the privatization agenda . supported by the - BankWe have concerns 4 about the composition of - expenditures within individual subsectors, however. While we endorse the emphasis in road spending on . . . - - 6 'a rehabilitation, we have doubts about the choice of links to be rehabilitated and the standards to be employed in the proposed works. In the water sector, we are concerned that too much is to be invested in new production facilities and extensions to the distribution network and too little in rehabilitation. And in both sectors, we have concerns about underlying institutional and financial issues. In the roads sector, we are concerned about the public sector's ability to manage the proposed rehabilitation program, the Government's commitment to maintenance, and the adequacy of revenues, particularly for maintenance of the secondary and feeder road network. In the water sector, 2 we note past management and financial problems in the sector, and are only cautiously optimistic regarding the Government's program of decentralization and tariff reform. II. ANALYSIS OF SUBSECTORS A. THE ROAD INVESTMENT PLAN Road Network. The Peruvian road network comprises 70,000 km of roads, of which 11% are paved, 19% stabilized, 23% earth, and the remaining 47% are trails. Peru's road density -- 54.4 km per 1000 square km, 3.20 km per 1000 population, and 1.91 km per 1 million GDP -- is half that of its neighbors, Ecuador and Chile, partly because of the large unexplored areas in the Selva, the high concentration of economic activity in the Costa, and the extreme difficulty and high costs inherent in building (and later maintaining) long roads through the extremely rugged Sierra and tropically forested Selva. The network is not connected, with missing links in important corridors. I Conditions of the road network, which had deteriorated to very low levels in the 1980's (by July 1990, about 44% of the national network was rated "poor", 44% received a "fair" rating, and only 12% was in the "good" category) have improved recently following increased investments in rehabilitation and maintenance. The national network by December 1992 had improved with about 29% rated as "poor", nearly 54% rated "fair", and the remaining 17% in "good" condition. Conditions in the secondary and feeder road systems, which fall under the responsibility of the departmental and local governments are in very critical condition, having not benefitted from recent rehabilitation investments. Additionally, the lack of all weather roads coupled with poor maintenance practices render large segments of the secondary and feeder road system (even the national network in some corridors) impassable or very costly to use during the rainy season. The central highway which carries a large percentage of goods and passenger traffic is totally closed due to mudslides for up to three days a year and accumulated closure reaches up to 7 days per year. With a country where other forms of access are impossible, these closures are quite significant. The condition of bridges is even more severe, as they have not been properly maintained, and the weight limits on many of them are being violated due to the growth and change in traffic composition. Government Policy for Roads The Government's stated objective in the roads sector is to strengthen the national, departmental, and local road networks to maximize accessibility, 1.The network is classified as: (i) the national road network of about 15,000 km; (ii) a secondary engineered road system of 15,000 km; and (iii) a feeder road system of about 40,000 km, much of which consists of tracks. 3 foster regional and national integration, and facilitate transport of goods between productive and consumption zones. With these objectives in mind, the Government of Peru has articulated the following priorities for investment in the roads subsector according to the submitted five year plan (1993-1997): (a) Rehabilitation, improvement, and maintenance of the national road network. The principal criterion for prioritizing investments is the Average Daily Traffic (ADT). Additional considerations include factors such as concentrated production zones, network integration, and homogenization of road quality. (b) Rehabilitation and construction of bridges on the national network. (c) Rehabilitation, improvement, and maintenance of departmental and local road networks, including bridges, following the prioritization criteria outlined above. (d) Construction of new roads, giving priority to feeder roads that would improve the connectivity of the road network. There are two plans for the investments in roads: (i) a five year plan from the Ministry of Economy and Finance (MEF Plan) covering the years 1993-1997, and (ii) a budget for 1994 prepared by the Ministry of Transport, Telecommunications, Construction, and Housing (MTCC Plan) and ratified by congress (see Table 1). The two plans differ in a number of respects (see Table 1). MEF's budget is considerably larger than MTCC's ($313 million for 1994 under the MEF plan, compared to $ 106 million for the same year under the MTCC plan). It is more modest in physical terms (866 km under the MEF plan as compared to 1762 km under the MTCC plan) and it assumes higher unit costs for construction (about $ 200,000 per km compared to $ 50,000 under the MTCC plan). These discrepancies indicate different assumptions on the type of works envisaged, the design standards, and priorities for investment. These also indicate a serious lack of coordination between planning and budgeting units in the ministries, and emphasize the necessity for more rational planning and cost estimation within the MTCC. 4 Table 1: Underlying Assumptions for the MEF and MTCC Budgets Item MEF MTCC Content of Program (Length in Km, excluding IDB, IBRD) 866 Km 1762 Km Budget allocation (for 1994, US$) $313 mn $106 mn Average unit costs of construction and/or rehabilitation 200,000 50,000 (US$/km) The MEF Plan for 1993-1997 We have analyzed the MEF plan. The budgeted level of expenditures for roads under the MEF plan is $ 2.2 billion over five years (including $1.7 billion under the sectoral budget and $ 0.5 billion under the regional budget for the same period ). It is largely devoted to rehabilitation: expenditures on rehabilitation constitute 70% of the proposed budget. This level is appropriate given the extensive deterioration of the highway system in the 1980's, and the consequent need for rehabilitation. Annual savings in vehicle operating costs resulting from the proposed rehabilitations are estimated at $300 million per year (see Annex A). Priorities and Design Standards While endorsing the emphasis on rehabilitation in the MEF plan, we have some concerns about the choice of road links to be rehabilitated and the standards to be used in rehabilitation. The MEF Plan includes a number of rehabilitation works on roads with very low traffic volumes. Of the 12,000 km of planned works on the National Road Network, some 4600 km (38%) are on roads with average daily traffic volumes of less than 75 vehicles per day. The standards to be used in rehabilitation also appear to be excessive. The cost of works on these road links under the MEF plan is about $ 200,000 per km. (compared to a range of $60,000 to $130,000 for proposed Bank financed projects.) Over the next five years, rehabilitation works on these low traffic volume roads consume about one third of the entire road budget, a level which appears excessive. We have calculated the economic rates of return on questionable links- -i.e, those that have very low traffic volumes (average daily traffic of less than 75) or unusually high unit costs of works planned (more than US$ 100,000 per km). We have also calculated ERRs on all new construction works. (Annex C presents a detailed analysis of these estimates.) Based on this analysis, we propose an reduced program with lower design standards. According to our calculations, the Government could save approximately US$ 275 million if the recommendations proposed are followed. This would reduce the road budget by about 16%. Maintenance We also have concerns about the Government's commitment to maintenance. Over the five year span, the MEF plan allocates 15% of the 5 highway budget to routine maintenance (an annual average of $70 million). This allocation seems to be acceptable, given the large allocation to rehabilitation in the budget. (It is also roughly consistent with the terms of the Bank financed Peru Transport Rehabilitation Project, which requires the Government to spend at least $50 million annually on maintenance for the national highway network alone.) We nevertheless have concerns about the willingness of MTCC and Congress to implement the MEF's proposed spending plan for road maintenance over the longer term. Under the MTCC 1994 budget, recently approved by Congress, only $ 20 million is allocated for maintenance. The Government expects that another $ 20 million per year will be raised from tolls, but this appears extremely optimistic. If the budgeting practices that prevail in 1994 persist, maintenance will be severely underfinanced in the outer years of the investment plan. Table 2: Breakdown of MTCC Road Budget for 1994 Activities Budget Amount Share of Budget US$ (million) (%) 1. Rehabilitation of paved network a) IDB 122 50.4 b) IBRD 14 5.7 c) MTCC 10 4.1 2. Rehabilitation of unpaved roads by MTCC 15 6.2 3. Maintenance (SINMAC) 20 8.3 4. Paving 25 10.3 5. Upgrading to gravel roads 6 2.5 6. Construction of gravel roads 11 4.5 7. Reconstruction of paved roads 1 0.6 8. Construction of new bridges 15 6.2 9. Studies 3 1.2 Total 242 100 Institutional and Financial Issues The scale of the proposed increase in road expenditures raises concerns about the public sector's capacity to administer the planned rehabilitation program and the capacity of the local construction industry to execute them. MTCC MTCC's capacity to undertake the contract management and supervision required by the large volume of works in the MEF plan is questionable. In the last five years, the Ministry has been buffeted by 6 changes in the allocation of responsibilities for road maintenance. Historically, MTCC's Directorate General of Highways (DGC) had overall responsibility for the planning, design, construction, and maintenance of the entire road system, working through the central office of Lima and 24 Departmental Directorates. The Regionalization Law (D.L. 572) transferred to the new regional governments, the jurisdiction over all roads, including the departmental directorates, their equipment and staff. As a consequence, MTCC's staffing level dropped from 10,600 in 1989 to 5,600 in 1992. The change in mandate was partly reversed in 1991, under a new legislative decree (D.L. 676) which reassigned to MTCC the responsibility for standardizing, managing, improving, and maintaining the national road network while the secondary and feeder road networks remain respectively under the control of departmental and local governments. Nevertheless, the disruption has prompted the departure of much of MTCC's more senior, technical staff; a condition which has yet to be rectified. Construction industry The proposed increase in expenditures on road rehabilitation will also strain the capacity of the domestic construction industry, and require an increased participation by foreign contractors. This has already prompted opposition from the domestic construction firms. It is nevertheless expected that a high degree of foreign involvement will continue to be necessary to implement the level of expenditure proposed by the MEF (see Annex D for an assessment of the capacity of the construction industry to carry out the planned works). Local finances The assignment of responsibilities for the secondary and feeder road network to local government (under D.L. 676) has also created a new institutional and financial problem, with possible adverse consequences for the maintenance of this component of the network. As a result of the decree, subnational governments are responsible for about 55,000 km (78%) of the total road network. The resources available to these governments are limited. While at the national level, revenue from road users (gasoline taxes, taxes on vehicle imports and ownership) have historically exceeded government expenditures on roads by a factor of 10, local governments have no similarly buoyant revenue source. (Local governments are largely dependent upon a centrally controlled property tax, and intergovernmental transfers). In the absence of an adequate source of local revenue (or a recentralization of responsibilities for the secondary network) maintenance of the secondary network may be severely underfunded. Summary of Recommendations for the Road Sector We support the major increase in spending on road rehabilitation that is proposed in the MEF plan. We would, however, advise a rescheduling of planned investments on very low traffic volume roads until the demand for them reaches the recommended levels in attachment C2, and a reduction in design standards. These recommendations, if followed, could result in up to a 16% reduction in the proposed road budget, which is well in line with the IMF recommendations. 7 In order to avoid a repetition of the cycle of investment- deterioration-and rehabilitation that has characterized the road sector in the past, it is critical that the Government devote sufficient resources to maintenance. At the national level, this requires that the Government allocate the level of resources to maintenance called for in the MEF plan. At the local level, it requires that the Government resolve the road financing hiatus that now exists. To address the management constraints on the execution of the proposed rehabilitation program, the Government will need to strengthen MTCC's capacity to administer construction programs, and to plan and prioritize capital investments, and must continue to permit the participation of foreign rmis in road construction. B. Other Transport: Ports, Airports, Railroads, Tren Electrico A number of public sector enterprises have the major responsibility for investments in the ports, airports, and railway sub-sectors. These sectors are classified by MEF under the rubric of "other transport". The System of Ports Peruvian ports have recently become competitive, in relation to ports in neighboring countries, following many years of inefficient operation. This has a significant effect on the international competitiveness of Peruvian products as ocean freight accounts for 95% of all Peruvian international trade. Movement of trade is accomplished through a system of 27 ports, of which 22 are maritime, four are river ports, and one is a lake port. About 16 of these ports are publicly owned (by the Empresa Nacional de Puertos Peru S.A. -- ENAPU), five are privately operated, the remaining being in the hands of public industrial companies. Sizeable reductions in the costs of shipping have been observed as a result of recent reforms, such as the deregulation of tariffs and the dissolution of the Comisi6n Coordinadora de Transporte Marftimo (CCTM) which held a monopoly on stevedoring activities. Stevedoring costs have been reduced by up to 80%, cargo handling costs have declined by about 50% for containerized cargo and about 30% for handling break bulk cargo. An agency to oversee private sector development in ports has been established following the overall privatization of state-owned companies (Legislative Decree 674 and Law 26120). This agency is called the Comit6 Especial de Privatizaclon -- CEPRI-ENAPU. Government Policy for Ports The privatization committee CEPRI-ENAPU has defined appropriate mechanisms for privatizing Peruvian ports, considering the profitability and operating characteristics of maritime and fluvial ports. Priorities for 8 privatization are given to those ports that are the most profitable such as Callao, Chimbote, and Paita. Public investments in ports are to be undertaken only to supplement private investments. Investment Priorities for Ports Following the policy guidelines above, the Government plans to place priority on the execution of feasibility studies for the development of the port sector whereby the state plays a supportive role to the private sector in port- related investments. The 1993-1997 investment plan for ports includes an allocation of about US$ 105 million. Nearly 74% of these (US$ 77,700) are for the port of Callao. The investments in the other ports total about US$ 27,600 (43% of the five year allocations to ports) all programmed for the year 1993. No new investments are planned for the ports subsector after 1993 except for US$ 60 million for construction of a Post-PANAMAX dock at Callao planed for 1996 and 1997. The investments in ports as summarized in Table 3 below, seem to be in line with the Government's overall policy for this subsector as well as policies relating to trade and international competitiveness. In particular, the concentration of investments in the port of Callao, which accounted for 55% of the total freight tonnage carried by ENAPU ports. Table 3: Planned Investments in Ports 1993-1997 Investment Component Amount US $ (000) Share of Port Budget Callao 77,700 73.8% Chimbote 8,400 8.0% Iquitos 6,200 5.9% lb 5,000 4.7% Salaverry 4,700 4.5% Paita 1,700 1.6% San Martin 1,600 1.5% Total 105,300 100% Note: Amounts exclude supervision which accounts for about US$ 2.5 million from 1993-1997 The System of Airports Peru has about 305 airports and landing strips, of which 59 are managed by the Corporacfon Peruana de Aeropuertos y Aviaci6n Comercial 9 (CORPAC), 113 by municipalities and the Ministry of Defense, and 133 by private operators. Peru's airport facilities are not sufficiently equipped to meet the challenges of an export economy and satisfy the role of national integration. Past maintenance neglect has led to a deterioration in the condition of runways. Notably, Lima airport, which was constructed 30 years ago for a smaller volume of traffic, is now handling about double the design volumes (about 70% of freight tonnage and 55% of passenger traffic is concentrated in Lima). The runway at Lima is in very bad condition and in need of urgent repair. The condition of navigation equipment is no better shape. Privatization of the auxiliary services of CORPAC is envisaged by the agency overseeing these activities CEPRI-CORPAC. Government Policy for Aviation The Government has three main policies with respect to airport infrastructure as summarized below: (i) strengthen the national system of airports to maximize accessibility and foster national integration; (ii) invest in rehabilitation of runways and navigation equipment that will remain in the hands of the state after the privatization of CORPAC; and (iii) give concessions to the private sector for the management and operation of airports that are commercially profitable. These policies derive from the classification system developed by the General Directorate for Air Transport (DGTA) aimed at adapting airport facilities to current civil aviation needs. This system, known as the Sistema Nacional de Aeropuertos (SNA) classifies airports using the following criteria: (i) strategic and integration purposes; (ii) airport equipment; (iii) air transport services; (iv) air transport demand; (v) air operations; and (vi) socio-economic development. The role of the state is expected to remain in the case of categories (i), (ii), (v), and (vi). Investment Priorities for Airports The number one priority for investments in airports is the International Airport "Jorge ChAvez" at Lima. This includes a total of about US$ 30 million over the five years, making up 26% of the expenditures in the aviation sector. About US$ 22 million is from a World Bank Loan covering rehabilitation works on the runway and taxiway and the installation of lighting and electrical equipment. The breakdown of investments into rehabilitation, improvement, and new construction are shown in Table 4. The allocation of investments seems to be adequate, especially since all the new construction works are planned for the years 1995-1997. We 10 have the following recommendations: (i) investments in new construction should be jointly evaluated with the investments planned under the regional budget (US$ 38 million for aviation projects in the 1993-1997 budget) to ensure an integrated system of airports; (ii) priorities should be given to those airports where there already is a high degree of private interest in the improvement and rehabilitation of the airports; and (iii) the investments should be related to the performance of CORPAC. Table 4: Composition of Aviation Expenditures - Investment Component Amount in US$ (000) Share of Aviation Budget Construction 19,600 17% Rehabilitation 41,700 36% Improvement 40,780 35% Navigation and Communication 12,800 12% Total 114,880 100% The System of Railways Railways play an important economic role in Peru, transporting minerals, petroleum derivatives, fertilizers and other bulk commodities. These minerals represent a substantial part of total Peruvian exports. Peru's railway system (2,074 km) is made up several companies, not all of which are physically integrated. The national railway company ENAFER (Empresa Nacional de Ferrocarriles) is the largest operator, dividing its operations into the Central (475 km), Southern (915 km), and South-Eastern (172 km) railways. CENTROMIN, a state-owned mining company currently undergoing privatization, owns and operates two lines (272 km) which connect with ENAFER's Central Railway. A private mining company, Southern Peru Copper, also operates a line (240 kam). A line previously operated by Cap Casagrande has been closed. The condition of ENAFER's locomotive and wagon fleet has deteriorated significantly during the last few years due to lack of necessary maintenance and overhaul as a result of financial constraints. Locomotive availability is about 32% for the Central Railways (56% in 1983) and 42% for the Southern Railway (47% in 1983). The condition of track is no better, about 85% of railway track in the central corridor (one of the largest export corridors) is rated in poor condition. Government Policy for Railways The Government intends to privatize the only remaining Government- owned railway, ENAFER. To this end, the Government has formed a special privatization committee - CEPRI-ENAFER. The privatization strategy for ENAFER would be coordinated with that of its principal client CENTROMIN. The Government has given priority to investments in the 11 rehabilitation of track on the Central and Southern railways, as well as the rehabilitation of about 47 locomotives. These investments are deemed necessary to make ENAFER a commercially viable company that would attract the private sector. Over the five year period a total of US$ 107 million is planned in the railway subsector. The remaining include works on the three systems currently operated by ENAFER (see Table 5). Table 5: Composition of the Investments in Rail Investment Component Amount US$ (000) Share of Railway Budget Tren Electrico 26,469 25% Rehabilitation of track 60,000 56% Rehabilitation of rolling stock 4,500 4.3% Improvement of zig-zags 900 0.84% Retirement of personnel 10,000 9.3% Emergency repairs 2,000 2% Workshop upgrading and 800 0.74% equipment Safety program 100 0.1% Telecom and signalling 300 0.3% Other 1,400 1.3% Total 106,969 100% With the exception of the category of works classified under "other", which include planned investments on the Matarani-Ilo section of the Southern Railway and the some works in Saracocha on the South-Eastern railway, the overall investment plan for the ENAFER system, from the point of view of types of works, is within the general recommendations of the World Bank for this sector. The Bank is financing a US$ 42 million railway privatization and rehabilitation project (the Bank share is US $ 32 million while the Government is US$ 11 million). These works make up about 52% of the planned investments in ENAFER's system and have been economically and financially justified. The remaining US$ 39 million (48% of the investments in the ENAFER system) need to be economically and financially justified before inclusion in the program. Tren Electrico The tren electrico is an uncompleted urban rail system, initiated by the previous administration and intended to connect southern suburbs with central Lima. The present government now agrees that this is a poorly conceived project, and has halted civil works. About US$ 26 million (24%) 12 is included in the public investment plan for the first phase of the Tren Electrico. The Public Investment Program for 1993-1997 does not foresee subsequent government expenditures for the tren eldctrico, as it envisions the sale to private concessionaire. We feel the prospects for private operation are minimal, and that with train operations soon to be initiated, there is now a danger of committing this and future governments to recurrent subsidy payments. A detailed analysis of the investments planned under the tren electrico project can be found in Annex B. Based on this analysis, it is recommended that only those works that have already been contracted out be paid for (about US$ 3.8 million). The remaining investments should be halted until the alternatives--which include selling the equipment and dismantling the structures--have been evaluated. Summary of Recommendations for "Other Transport" We generally support the proposed public expenditure program for ports and airports, and the Government policy of reducing the role of the state in these sectors. We recommend that the Government leverage its investments in airport infrastructure through joint ventures with the private sector, where private interest has been shown. (We also note that a number of questionable investments in new airports are included under the regional budget heading of the public investment plan, as discussed below.) In the railways sector, we support the Government's intention to privatize ENAFER, as well as Its decision to coordinate the privatization strategy with that of CENTROMIN. While we agree with the need to invest sufficient resources in ENAFER to render the railway commercially attractive, it is not clear that all the works proposed in the public investment plan are justifiable on this basis. These should reevaluated. We are extremely concerned about the Government's proposal for the tren electrico. While we would support the privatization of the system, we believe that the prospects for fmding a buyer are minimal, and that the Government, in undertaking to start the service on a pilot basis, may be obligating itself to heavy long term operating subsidies for a system which is not economically justifiable. We recommend that any further investments be halted until the alternatives have been evaluated. C. Water Supply and Sanitation Sector Background The cholera epidemic in Peru has brought to light the deplorable condition of its water and sanitation sector. As water and sewerage conditions have deteriorated in the past decade, the incidence of waterborne diseases has increased significantly, and became endemic among the poor. Service coverage by 1991 was 58% in water supply and 42% in sewage collection, as reported by the Interamerican Society of Sanitary Engineers (AIDIS). In urban centers (population greater than 2000) coverage reaches 74% and 62%, in water and sewerage respectively. However service 13 coverage figures, as understood in Peru and in other countries in Latin America, are related to territorial network coverage and hide dramatic disparities on the continuity of the service, and in overall service quality indicators. In fact, the quality of service has deteriorated steadily as water service is not being provided on a continuous basis and water rationing prevails in major urban centers like Lima-Callao. A survey carried out in 1989 revealed that 30 percent of the population in urban centers other than the three major cities, had a water service less than 10 hours a day and only 20 percent had service more than 20 hours a day. The steady degradation in service quality in the last years is partly due to the low level of infrastructure investment in the recent past. During 1985- 1989, investments in water and sanitation infrastructure represented only 0.15 percent of GDP, one of the lowest in Latin America, where countries with better service have been investing around one percent of GDP. (Ironically, the first ten weeks of the cholera outbreak represented US$1 billion in emergency health expenditures and lost earnings from exports and tourism. This is approximately four times the investment on the water and sewerage sector between 1981 and 1988.) But the poor operational performance in the sector is the result not only of low level of investment, but also of weak management, and misguided tariff policy. In the past, service tariffs have been set following short sighted political and social criteria, and in general, their levels were insufficient to cover operating costs. Sector companies survived by sacrificing service quality, where the reduction of maintenance expenses, was one of the first targets. Cash deficits created by current operations have been covered through government transfers, delayed payments to suppliers, and failure to service debt. Capital investments have been financed, mainly, from user contributions (in new neighborhoods and Pueblos Jovenes), donors, and to a lesser extent government transfers and concessionary loans from FONAVI. Investment Priorities in the Water and Sanitation Sector The Government is planning to spend $ 1 billion in the water and sanitation sector (see Table 6). More than 60% of these expenditures will be carried out by water companies owned by the central and municipal governments; the remainder through the direct expenditures of central and regional governments.' 2. The investment program includes two IDB projects. The first for approximately US$75.0 million is targeting institutional development (80% of the project) for approximately 15 regional water companies outside Lima. Consultants to assist the GOP ( Proyecto Especial Programa nacional de Agua y Alcantarillado) were hired in December 1993, after a delay of more than two years. The project is scheduled with priority A for Board approval at the end of 1994. The second IDB project in the pipe line will prioritized investments in rehabilitation and expansion of selected water and sewerage systems.The investment program includes the proposed US$260.0 million project "Lima water Privatization and Rehabilitation Project', scheduled for Bank Board approval in July 1994. A second WB proposed financed project (continued...) 14 Table 6: Investment Program in the Sanitation Sector 1993-1997 Institutional Level 1993-1997 Investment Program 1993 1994 1995 1996 1997 Total Central Government (IDB projects) 1.1 34.8 24.7 89.0 115.0 264.6 Regional Government 1/ 1.6 6.0 36.9 53.8 43.8 142.0 Companies (central and regional) 77.8 142.3 142.1 152.7 97.0 612.0 Total 80.5 183.1 203.7 295.5 255.8 1018.6 Priorities While water supply is an important sector for Peru, we have several concerns about the allocation of this $1 billion. First, the proposed budget concentrates too much on new production facilities and expansion of the networks, and too little to rehabilitation of the existing networks. The need for new production facilities is questionable. Most water supply systems in Peru would have production capacity in excess of demand, if rates of unaccounted for water were reduced to reasonable levels. For example in the case of Sedapal, it is estimated that 50% of the water is either lost or wasted. It is clear that reductions in unaccounted-for-water, if successful, can substantially eliminate the water shortage at the production level. Second, a number of proposed extensions to the distribution network do not appear well justified. Data is insufficient to evaluate all the water projects proposed under the plan. We have reviewed the investment programs for regional companies and municipal services, which are presented separately in the regional government investment program. In general, both central and regional investment programs are not described or justified consistently and they do not correspond to a country-wide strategy for the sector. (Future sector work with the Ministry of the Presidency (as the new top sectoral institution) and IDB should help the government formulate a comprehensive strategy for the sector, and eliminate such inconsistencies and lack of clarity in its investment programs.) Moreover, as a group, the allocations for sewage treatment works also need to be carefully evaluated. A number of expensive sewage treatment plants are proposed under the budget. It is not clear that, under Peru's present budget constraints, sewage treatment merits priority over water supply. Rather than investing so heavily in new production and distribution works, public expenditures should be concentrated on the rehabilitation of existing physical systems and on water demand management (particularly in the coastal cities). The severity of maintenance neglect in the last decade has led to serious water losses due to leakage and is largely responsible for the cholera epidemic (which resulted from the infiltration of contaminants into a 2.(...continued) (US$300.0 million, FY 97) to reduce current pollution levels in the Lima metropolitan area and improve water management practices is only partially included. 15 deteriorated piped system.) As described in Annex E, we estimate the costs of a fully funded rehabilitation program at $400 billion, or 40% of the proposed Plan investment in the water sector. Institutional and Financial issues in the Water Sector The reallocation of expenditure from new capital works to rehabilitation will slow the rate at which the Government is able to achieve the objective of universal coverage. (Rough estimates put the costs of achieving universal coverage at the full $1 billion--see Annex E.) While we support the objective of universal coverage, we would argue that system expansion should be tied to institutional and financial improvement in the sector. Our calculations suggest that a large share of new investment could be financed (or in the short term, repaid) from increased tariff revenue, subject to institutional and financial reform. The poor performance of Peru's water supply companies reflects an historical absence of commercial incentives and the lack of autonomy over tariffs. Until 1989, central government's Servicio Nacional de Agua Potable y Alcantarillado (SENAPA), was responsible for water supply and sewerage for most urban Peru, acting as a holding company for 14 subsidiary companies in charge of operations in the largest cities or departments and providing service directly in 10 departments through 10 Operational Units which provided services to small and medium size towns.' Tariffs were controlled by the Comision Reguladora de Tarifas de Agua Potable y Alcantarillado (CORTAPA), subject to ratification by the Ministry of Economy and Finances (MEF). As a consequence, the management autonomy of field operations was limited, and tariffs were manipulated for political and social objectives. The institutional and financial structure of the sector is now changing. SENAPA was formally abolished in December 1992, and responsibility for its various operating agencies (with the exception of Lima-Callao and Trujillo) are to be transferred to the regional and local governments. CORTAPA is to be abolished, and tariff setting authority is to be decentralized to the municipalities (subject to a national tariff policy.) The decontrol of tariffs and the commercialization of the water supply utilities is expected to substantially improve their financial viability. Estimates of potential revenue from tariffs (as described in Annex E) suggest that tariff revenue would be sufficient to cover operating and maintenance costs, plus debt service, and to make a significant contribution to capital investment on a pay as you go basis in most of the major urban water systems. As a result, it is expected that where government funding will be required, it can be provided on a loan, 3.Some 185 municipalities, representing about 10 percent of the total urban population, administered and still administer their services directly or through cooperatives or other forms of management. Rural communities (below 2000 inhabitants) and disperse population comprising 32% (6.8 million) of total population were once assisted by the National Rural Sanitation Program under the Ministry of Health, but the Program has long been abolished.) 16 rather than grant basis, and that significant contributions to capital investment can be made from internally generated revenues, reducing the sector's claims on the resources of the public investment plan. Summary of Recommendations for Water Supply and Sanitation We support the proposed increase in expenditure on water supply, but believe the Government should reorient its investment priorities in this sector, focusing more on the rehabilitation of existing works, and less on the construction of new production and distribution systems. Noting past management and financial problems in the sector, we endorse the Government's efforts to restructure the sector, increasing management autonomy and deregulating tariffs. D. Regional, Housing Investment Plan Regional Development Some infrastructure expenditure proposals also appear in the regional development and housing categories of the public expenditure plan. The regional budget totals $1.9 billion, of which $636 million is programmed for infrastructure investments (34% of the total regional budget). Table 7: Composition of the 1993-1997 Regional Investment Program (thousands of US $) Sector 1993 1994 1995 1996 1997 Total Transport 10728 36694 121797 153796 179050 502065 Sanitation 1594 6631 37050 44505 44326 134106 Housing 0 97 0 0 0 97 Sub-Total 12322 43422 158847 198301 223376 636268 %of 19% 27% 35% 36% 36% 34% Regional Regional 64166 163500 448002 558370 627337 1861375 Total 1/ Regional investments in the sanitation and roads subsector have been evaluated as part of the sectoral discussions presented earlier in this chapter. More detailed analysis is necessary for the package of regional investments as a whole, as well as of the investments in airports, ports, rail, and infrastructure in industrial parks. Works included under the regional budget fall under sectors for which there are also sectoral budget allocations: about 79% of the planned investments are in the transport sector, 21% in the sanitation sector, and 17 0.1% in the housing sector (see Table 7). While we have no basis for analyzing the merits of these investments, it appears that these may be works which are justified on security or political grounds, rather than on their economic merits. Among the more prominent items are $13 million allocated to the construction of a new airport (Huanchao); and more than $ 12 million explicitly programmed for integration purposes (carreteras de integration de Oxapampa). Maintenance is also under-budgeted: only $2.3 million is budgeted for the maintenance of the 50,000 km of roads included under the regional budget, and only $583,000 for the maintenance of 113 municipal airports. Housing The official budget allocation for housing is modest, averaging $70 million over the five year period ($ 50 million annually under the sectoral budget and $20 million annually under the regional budget). Most of these expenditures will be made through public enterprises (ENACE and Creditos Supervisados) and will finance the construction of core housing and service sites. The majority of Government expenditures on housing and residential infrastructure do not appear in the budget. Funding by FONAVI ($300 million annually) and FONCODES ($500 million) account for the majority of Government expenditure on community infrastructure. Concerns have been expressed about the terms and criteria by which these funds are allocated. We understand that these will be addressed in another section of the PER. Summary of Recommendations on Regional Development and Housing While we have little basis for analyzing the merits of the regional investments, it appears that these may be works which are justified on security or political grounds, rather than on their economic merits. The allocation of expenditures under the "housing" heading may also be vulnerable to the same criticism. Further information is required to permit a more systematic analysis of expenditures under these two headings. 18 PERU - Public Expenditure Review ANNEX A Planned Investments in Infrastructure Table 1: Infrastructure Policies 1963-1993 Table 2: Proposed Infrastructure Investment Program 1993-1997 Table 3: Sectoral Composition of Infrastructure Investments 1993-1997 Table 4: Sub-Sectoral Composition of Public Expenditures 1993-1997 Table 5: Regional Expenditures in Transport Public Investment in Infrastructure: Trends by Sub-Sector Public Investment in Transport: Trends by Sub-Sector Pattern of Maintenance Deferral Under Different Administrations 1981-1989 Pattern of Maintenance Deferral Under Different Administrations 1989-1997 Table 1: Infrastructure Policies 1963-1993 Period Policy 1963-1968 high priority on development of transport infrastructure for integration purposes 1968-1980 creation of state-owned enterprises and large investments by these enterprises 1980-1985 large expenditures in transport sector especially new construction 1985-1990 state-owned enterprises in financial crisis, neglect of maintenance 1990-1993 reducing state role in service provision, investments in rehabilitation and maintenance of infrastructure Table 2: Proposed Infrastructure Investment Program 1993-1997 (millions of US $) Sector 1993 1994 1995 1996 1997 Total Transport 312 416 510 606 724 2568 Roads 181 313 332 386 482 1694 Aviation 25 21 38 20 17 121 Ports 49 1 1 31 31 113 Rail 33 30 17 15 15 110 Tren Electrico 13 14 0 0 0 27 Regional transport 11 37 122 154 179 503 Communication 218 91 31 9 9 358 Water & Sanitation 81 184 205 287 256 1013 Housing 43 58 57 45 33 236 Total Infrastructure 654 749 803 947 1022 4175 Total Public Capital 1803 1978 2506 2672 2750 11709 Source: Ministry of Finance Table 3: Sectoral Composition of Infrastructure Investments 1993-1997 Sector 1993 1994 1995 1996 1997 Total Transport 47.7% 55.5% 63.5% 64.0% 70.8% 61.5% Water & Sanitation 12.4% 24.6% 25.5% 30.3% 25.0% 24.3% Other 39.9% 19.9% 11.0% 5.7% 4.1% 14.2% Total Infrastructure 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Source: Ministry of Finance Table 4: Sub-Sectoral Composition of Public Expenditures 1993-1997 Sector 1993 1994 1995 1996 1997 Total Transport 17.3% 21.0% 20.4% 22.7% 26.3% 21.9% Roads 10.0% 15.8% 13.2% 14.4% 17.5% 14.5% Aviation 1.4% 1.1% 1.5% 0.7% 0.6% 1.0% Ports 2.7% 0.1% 0.0% 1.2% 1.1% 1.0% Rail 1.8% 1.5% 0.7% 0.6% 0.5% 0.9% Tren Electrico 0.7% 0.7% 0.0% 0.0% 0.0% 0.2% Regional transport 0.6% 1.9% 4.9% 5.8% 6.5% 4.3% Communication 12.1% 4.6% 1.2% 0.3% 0.3% 3.1% 0.0% Water & Sanitation 4.5% 9.3% 8.2% 10.7% 9.3% 8.7% Housing 2.4% 2.9% 2.3% 1.7% 1.2% 2.0% Total Infrastructure 36.3% 37.9% 32.0% 35.4% 37.2% 35.7% Source: Ministry of Finance Table 5: Regional Expenditures in Transport US$ ('000) Region Roads Airports Ports Bridges Rail TOTAL Arequipa 38,232 24,646 62,878 A.A. Caceres 29,201 29,201 Chavin 33,363 33,363 Grau 11,267 4,000 15,267 Inka 72,133 72,133 La Libertad 70,566 9,200 7,600 3,015 90,381 Wari 12,355 183 12,538 Loreto 2,412 583 136 3,131 Mariategui 82,304 82,304 Maranon 48,725 40,683 89,408 San Martin 527 527 Ucayali 12,600 790 13,390 TOTAL 413,685 38,429 49,209 3,015 183 504,521 Source: Ministry of Finance Public Investment in Infrastructure: Trends by Sub-Sector (US$ Equivalent) 800 700 - 600 - 500 - 2 400 - 300- 200 - 100 - 0 1993 1994 1995 1996 1997 Time (years) a Transport + Water & Sanitation + Telecom a Housing Public Investment in Transport: Trends by Sub-Sector (US$ Equivalent) 600 500 - 400 - 0 300 - E 200 - 100 - 0 1993 1994 1995 1996 1997 Time (years) a Roads + Aviation * Ports a Rail x Tren Electrico v Regional Transport Pattern of Maintenance Deferral Under Different Administrations 1981-1989 Pattern of Expenditures 1981-1995 2 2 21 I: - 1981 1982 1983 1984 1985 TIme (Yeare) 0 khinternnce + Total Pattern of Expenditures 1985-1989 2. 1 2- 1.9 1.6 - 1.5 - 1.4 - 09 - 0.6 - 0.5- 0.4- 0.3 - 0.2 - 0.1- 19s 1988 1987 1988 1989 Tim (Yare) 0 MmIntenance + Total Pattern of Maintenance Deferral Under Different Administrations 1989-1997 Pattern of Expenditures 1999-1992 2.2 2- 1.6- 1.2 - 0.8 - 0 4 0.2 - 0 1989 1990 1991 1992 Tim cYar@3 O M&Intenance Total Pattern of Expenditures 1992-1997 800 500 - 400 300 200- 100 N 93 2§k 1995- 1996 1997 Tima CYamr93 1 Kbint'nmne 4 Tate! PERU - Public Expenditure Review ANNEX B Review of Tren Electrico Project Background Bl. The national Government decided in 1986 to build a 21-km rail rapid transit line between Villa El Salvador at the southern edge of Lima's metropolitan area and Hospital Dos de Mayo near the historical center. Civil works were halted in late 1990 following the change of government administration. The 9-km section from Villa El Salvador to Puente Atocongo was about 90% complete at that time; in addition there were about 2.5 km of elevated structure and 2 km of bridge supports between Puente Atocongo and Avenida Javier Prado. B2. In 1992 and 1993, the Autoridad Autdnoma Tren Eldctrico (AATE) signed several contracts totalling about US$9 million (16 million Soles from the national treasury) to complete the workshop at Villa El Salvador and the 9-km section to Puente Atocongo. This reflected a decision to get at least some use from the past investments with a minimum of additional expendituresY The Government's contract with Tralima, the consortium responsible for the electromechanical and some construction aspects of the project, continued to operate at the same time under bilateral funding. B3. Most electrical and mechanical equipment has been built in Italy, with funding from an attractive US$100 million creditY By the end of January 1994, 10 train cars (costing US$1.2 million apiece) had arrived in Lima; the remaining 22 train cars are due to be delivered by May 1994. A recently built warehouse at Villa El Salvador is reported to contain US$20-30 million worth of electrical equipment for substations, transmission, signalling etc. B4. Estimated Investments. As indicated in Table BI, about $200 million have been invested in Section I (Villa El Salvador - Puente Atocongo) of the rail system This initial section will be of limited use for passenger services without its continuation into inner Lima. The construction of Section II (Puente Atocongo - Hospital Dos de Mayo) would result in a sufficiently long rail line to be attractive to some public transport passengers, although it still would serve less than 5 percent of Lima's travel demand. The extension of the line is estimated to cost another US$364.5 million, resulting in an overall cost of US$593 million. Y 1/ o0y five & . pl...d .. ewim wm .v bes. bul. md dw tot suswnfay rduced, so-fri degs. Zi 1.5% .inm lesverepsymo im 2o years with 10 years of gwc. 2/ Tim towl Psm I cost f USS28.7 millon iniaues contracus sQ to be completed mad some pmvious wwb in de Puente Atoomp - Hospitl Dos do Mpyo S The argle ivoestmet f US28 m1lim per km of mogty elevoted repid transi lime, including mMling stock ad other oquipmen is t de low and of hentionl eqermoo (US$25-70 miman; Source: Alipot md Thoeamn. Soy MAsw AWd Twls in Drwkpbg Cowls, Trmsport and Road Research Laboratory Ctractor Report 188, 1990). 2 Table Bi: INVESTMENTS IN TREN ELECTRICO US$ million National External External Total Treasury Credit Grant Section I (1987-1994) - 9 km - Civil Works 67.4 - - 67.4 - Electromechanical Works 6.9 100.0 10.5 117.4 - Studies and Technical Assistance 2.0 - 11.5 13.5 - Administration 9.3 - - 9.3 - Taxes and Duties .1L- - .1. Subtotal 106.7 100.0 22.0 228.7 Section H* - 12 km - Civil Works 108.6 5.2 - 113.8 - Electromechanical Works 43.2 160.0 - 203.2 - Studies and Technical Assistance 1.0 13.1 - 14.1 - Administration 8.2 4.5 - 12.7 - Taxes and Duties 20.1 - - 20.7 Subtotal 181.7 182.8+ - 364.5 Total 288.4 282.8 22.0 593.2 Section I: Villa El Salvador - Puente Atocongo Section II: Puente Atocongo - Hospital Dos do Mayo * estimated for future year + includes US$10 million in domestic loans SOURCE: Autoridad Autdnoma Tren El6ctrico, Lima, January 1994 B5. Planned Operation of Section I. Current plans are to start train operations south of Puente Atocongo in 1995. As the rail section is in a remote area with relatively low population density, AATE proposes an integrated train-bus operation, whereby train passengers would transfer at Puente Atocongo to one of three bus lines providing services to central and northern Lima.- Transfer passengers would buy a combined train-bus ticket for 53 USCents, equivalent to the current bus fare between Villa El Salvador and inner Lima. Bus services in corridors parallel to the rail line would continue to operate. B6. In line with government policy, the operation of the train-bus system would be offered to the private sector, as a 7-year franchise. Apart from supplying 74 articulated buses (US$13.3 million), the private operator would have to pay a US$5 million franchise fee for the use of the rail facilities. AATE estimates that the combined train-bus operation would yield a cash profit, part of which would be placed into a fund for the future extension of the rail line. (These financial estimates appear quite unrealistic, as discussed below). S he* M . *m wmta wA tu.: (a) th Vin Exprem bney. to ich m sm rwmly rehicd by th Micipahy to keep s amber of bees below unaeptable conetion ivol; (b) Avui&h Avacidn were Sectim I of the m lie wold uimautly be buiht ad whem AATE props for the ann aUSs11.5 mu11- bwvmy improveme; .d (c) Avad& Panovion, the =mAormy bypem4 Unas m... &oe.y p.Wuhmd am. 3 1994 Budget B7. The 1994 Budget is shown in Table B2. It includes US$ 10.2 million from government resources, complemented by about US$5 million from the bilateral credit.9 About one third of the local funds are committed to ongoing construction projects scheduled for completion before mid-1994. The remainder is for administration, pre-operation, and civil works contracts (US$ 2.4 million) yet to be signed. Table B2: 1994 BUDGET FOR TREN ELECTRICO million Soles National External External Total Treasury Credit Grant Administration 1.0 - - 1.0 Goods and Works - Civil Works 4.6 - - 4.6 - Construction Supervision 1.2 - - 1.2 - Capital Goods - 4.2 4.- Subtotal 5.8 4.2 0.4 10.5 Pre-Operation - Personnel 1.3 - - 1.3 - Machinery and Equipment .24... - - 21 Subtotal 3.4 - - 3.4 Total 10.2 4.2 0.4 14.8 SOURCE: Autoridad Aut6nomna Tren El6ctrico, Lima, January 1994 Exchange rate for 1994 is 2.52 B8. The items under "pre-operation" refer to the preparation of commercial train operations. This would include the contracting of about 100 persons whom Tralima would provide the necessary technical training to operate the rail system. The pre-operation phase is scheduled to commence in the second half of 1994. I/ is undereood td. of Jamary 1994, ab"t 89% of d USSIO in95an credit had bem used. 4 Public Expenditures after 1994 B9. The Public Investment Program for 1995-1997 does not foresee further government expenditures for the tren el6ctrico. However, if AATE's current plans were to be implemented, additional public funding could not be avoided, including: a. Construction of a busway on Avenida Aviacidn, estimated to cost US$11.5 million. b. Additional pre-operation costs. "Dry runs" of empty trains are now envisaged for about January-April 1995. More personnel would also need to be contracted to eventually reach the planned staffing level of 576 persons for the train operations. c. Purchase of 74 articulated buses for US$13.3 million, in case no private operator is interested to run the combined train-bus system. d. Operating subsidies, as discussed further below. Assessment of Planned Train-Bus Operation BIO. Demand Forecast. AATE estimates that the combined train-bus operation would serve 169 million passengers in 1998, or 463,000 passengers per day. About 32% of these would only use the bus, and the remainder both train and bus. The directional peak-hour volume by train would be 11,492 passengers. The peak volume is the basis for the proposed train frequency of 10 trains per hour. B11. The estimated all-day volume of 315,000 passengers implies that the directional peak-hour flow would represent only 3.7% of the total daily two-directional volume. This is not credible, considering that this percentage usually ranges between 10% and 20% The projected peak-hour volume (although still overestimated considering the area served by the rail line) is in fact quite a meager demand for any rail rapid transit line. On the other hand the projected 24-hour and all-year volumes are unrealistically high. B12. This is evident from the international comparison shown in Table B3: Lima's train usage would approach that of Sao Paulo (one of the busiest metros in the world) and be greater than that of all other Latin-American systems - quite unbelievable for the truncated and isolated location of the Villa El Salvador - Puente Atocongo stretch. Until the results of an updated and realistic demand analysis are available (see below), it is concluded that current demand forecasts overestimate train usage by a factor of 5 to 20(!). B13. Cost Estimates. The projected peak and all-year traffic volumes would severely distort any cost analysis by underestimating operating costs per passenger and, at the same time, overestimating passenger flows and fare revenues. AATE's estimates imply that the trains would run full not only during the peak, but all day long. Not surprisingly, the cost projections shown in Table B4 would make Lima the financially most attractive urban rail system in Latin America whereas, in all probability, exactly the opposite would likely be the case. 5 Table B3: PASSENGERS CARRIED BY RAIL RAPID SYSTEMS City Year Route-km Million passengers Million passengers per year per year and per km Mexico 1988 132 1450 11.0 Porto Alegre 1986 27 36 1.3 Rio de Janeiro 1987 34 108 3.2 Santiago 1986 26 139 5.3 Sao Paulo 1988 29 465 16.0 Lima 1998 9 115 12.8 SOURCES: AATE (for Lima) and The Performance and Inpact of Rail Mass Transit In Developing Countries, Transport and Road Research Laboratory Research Report 278, 1990. Table B4: OPERATING COSTS AND DEFICITS OF RAIL RAPID SYSTEMS US Cents City Year Route-km Operating Costs Operating Deficit _I _ I per passenger per Passenger Mexico 1988 132 10 6 Porto Alegre 1986 27 35 28 Rio de Janeiro 1987 34 23 11 Santiago 1986 26 11 0 Sao Paulo 1988 29 14 6 Lima 1998 9 8 0 SOURCES: AATE (for Lima) and The Performance and Impact ofRail Mass Transit In Developing Countries, Transport and Road Research Laboratory Research Report 278, 1990. Short-Term Issues B14. While the present Government agrees that the tren el6ctrico is a poorly conceived project, it has (understandably) opted for completing at least a short section to salvage as much as possible from the past investment. With train operations soon to be initiated, there is now the danger of committing 6 this and future governments to recurrent subsidy payments - for virtually no social or environmental benefit - which could eventually exceed the public funds already spent by the previous Government. The immediate issues that should be addressed are: a. whether and/or when to start train operations; b. whether any reductions should be made in the expenditures programmed for 1994; and c. whether the contract with Tralima should be renegotiated or terminated. B15. Starting Train Operations. The current policy is to transfer the tren el6ctrico to a private operator, thereby avoiding the need for government subsidies. The Government should pursue this policy without committing itself to any subsidies or financial guarantees. As it is unlikely that a private firm would be interested in such a proposition, it would be unwise for AATE to initiate operations now (even the "dry runs"), which the Government would likely have to subsidize thereafter. Therefore, the hiring of operating staff and the start of train operations should not begin until a satisfactory franchise agreement has been signed with the private operator. B16. Reductions in the 1994 Budget. For the above reasons, it would not be necessary in 1994 to authorize the US$ 3.4 million programmed for "pre-operation". Similarly, the approximately US$ 2.8 million budgeted for installing the available electrical equipment could be postponed until a clearer picture emerges on whether any private operator is interested in the system. About US$ 3.2 million are committed to ongoing construction works, and it is probably unrealistic to stop them now. However, stopping the work on the San Juan station might be considered, as it will be of little use even if a private train operator is found (that station, which is located 1.6 km south of Puente Atocongo, will be even more under-utilized than other stations, as few people would climb 30 feet and wait for a train - just to take a 2-minute train ride and then transfer into a city-bound bus). B17. Tralma Contract. The postponement of the electrical installation and pre-operation phase would require a similar rescheduling of the expatriate personnel that are meant to assist in this operation. If a decision were to be reached to postpone indefinitely any further work on the system, Tralima may well be interested in renegotiating and possibly cancelling the remainder of its contract without penalty payments by the Government, thereby saving the last US$ 5-10 million of the bilateral credit. Longer-Term Options B18. If no private operator is interested in the train-bus operation, the basic question will have to be answered on what to do next with the tren el6ctrico. Several alternatives are possible - none of them very attractive - ranging from continuing with heavy capital investments to dismantling the current system. B 19. Extend Rail Line. To many, the most plausible long-term option is the extension of the rail line to Hospital Dos de Mayo and beyond. This would at least make the remote section south of Puente Atocongo more useful. The associated capital and operating costs would be enormous, however, and hardly be an economic priority for the next 10-20 years when busways could cope with the travel demands in Lima's generally low-density environment. 7 B20. Train-Bus Operation. AATE's proposal is probably the right approach if the objective is to start train operations. To make it work, heavy operating subsidies will probably be necessary for the reasons indicated above. Even with those subsidies, the trains risk to run empty much of the time, which could become an embarrassment to the decision-makers responsible for authorizing the train operations. If the objective is to transport passengers at a reasonable level of comfort, and with a minimum of government subsidy, the train-bus operation does not appear to be the best solution. B21. Sell Equipment and Dismantle Structures. An extreme solution which, in strictly financial terms, might be attractive. The revenues from selling the 32 train cars, the electrical and signalling equipment, as well as the rails and ties would probably exceed the cost of knocking down the elevated structures. Also, no operating subsidies would ever be required for train operations. It would be difficult, however, to explain the demolition of the structures that had been erected in the last five years at great expense. B22. Convert Line into a Busway. A decision might be taken to consider a rail rapid transit system for the long-term future, say in 20 or 30 years when urban densities might justify such a system and when Peru's economy might have grown sufficiently to afford its high costs. The tren eldctrico could then be incorporated into the future rapid transit system. In the meantime, its elevated structures and right-of-way would be converted into a busway integrated into Lima's existing busway system via Avenida Tomls Marsano. This option, which would also imply the sale of the electrical and mechanical equipment acquired for the tren el6ctrico, would have to be studied in greater depth to determine its operational and economic feasibility. B23. Who Should be Responsible for Transit Services? At present, the central Government owns the tren el6ctrico and is responsible for its potential operating subsidies. According to the Constitution, the local authorities are responsible for urban transport. The question thus arises whether it would be appropriate to transfer the ownership (and the burden of operation) of the tren el6ctrico to the Municipality of Metropolitan Lima. Conclusions and Recommendations B24. While a 9-km section of the tren eldctrico line will soon be physically completed, recurrent subsidies would be required to operate passenger servicesY It therefore appears unlikely that a private operator would come forward to run the system without a government subsidy. Even then, the trains would run empty most of the time, which would make any subsidy pointless on either social or environmental grounds. B25. It is recommended that no further funds be committed to this system, at least until a satisfactory franchise agreement has been signed with a private operator to run the system without subsidies. In order to offer the tren eldctrico operation to the private sector, the passenger flow analysis needs to be improved substantially. This analysis must quantify the attractiveness and expected patronage of parallel bus services that would compete with the train. 2/ MT's pisenger aid financal forests conclude tha the proposed aim-bs operatimon would produce a ftncia proit The wek link of AAT 's amalsis is th pseinger foarecant, which is based anon omnal allocation of trip productions from the tim mcro-zonos in southern Lim md which does no tko .com.s of comping paralel be. ervion. AATs reoope thas and ha recommanded to update md mefine the passenger demand analysis. his World link no6e estimat tat passeger demand has been over-estiamted by a factor of 5 to 20. 8 B26. A rigorous study of existing and future passenger flows is the first step needed to give decision-makers the information which they require to reach a deliberate and substantiated decision on what to next with the tren elctrico. If the analysis results in attracting a private operator, the current strategy could be pursued. If not, the Government will have to choose from among several options, some of which were outlined above (paras 18-23). It is therefore recommended that the proposed study not only examine the passenger demand and finances for the proposed train-bus scenario, but that it also investigate all options for dealing with public transport in Southern Lima. TRENUW.NT OWFebruay 18. 19N4 PERU - Public Expenditure Review ANNEX C Detailed Analysis of Road Budget C1. Introduction The IMF is suggesting, for the 1994 budget, to reduce the funds allocated to the Ministry of Transport by about 15%. Expenditure allocations to the transport sector will increase over time from 17% in 1993 to 26% of the total expenditure program in 1997. Investments in roads which make up about 62% of the transport investments in 1993 will make up 86% of the same in 1997. This annex presents an assessment of the 1994 budget and the 1993-1997 investment program. The assessment is done in terms of the priorities for the road component of the transport budget. Recommendations based on the identified priorities are made, and a suggested program of investment is put forward. C2. Assessment of the 1994 Road Budget There are two versions of the 1994 Road Budget: (i) MEF version' which we were asked to assess; and (ii) MTCC version which has been approved by congress. A summary of the two budgets is shown below (expressed in US$ million): Table Cl: Comparison Between Suggested Budgets Component MEF MTCC IMF IDB Rehabilitation Program 122 (122) (122) IBRD Rehabilitation Program 14 (14) (14) MTCC Road Program 238 106* 138 Total 374 242 274 Note: * Approved by Congress. As can be seen from Table Cl, large discrepancies exist between the MEF and MTCC programs. The MTCC plan is 35% lower than the MEF plan, while the IMF recommendation is 27% lower than the MEF plan. Therefore, the MTCC plan, in terms of the overall size of the budget, is within the IMF recommendations. However, there are other differences between the budgets, which reflect variations in the: (i) content of the programs (and therefore priorities); (ii) budget allocations; and (iii) assumed unit costs of construction and rehabilitation (and therefore, standards assumed), as shown in Table C2 below. 1. Based on a document prepared by MEF and transmitted to us by Cheikh Kane on December 8, 1994. 2. Based on a letter from the Vice Minister Juan Castillo to Asif Faiz, January 5, 1994. In particular, the MTCC program is recommending works on a more extensive part of the network, comprising about 1762 km compared to almost half that (866 Ki) under the MEF Plan. The MEF Plan, however, is assuming a much higher unit cost of construction and rehabilitation at about US$ 200,000 per Km. This unit cost is four times higher than that assumed under the MTCC Plan, indicating that the assumptions on the types of works to be carried out are different. The higher unit cost assumed under the MEF Plan may be due to the higher volume of upgrading and improvement works, as well as new construction, than assumed under the MTCC Plan. We consider some of these works of an unnecessarily high standard and recommend an overall lowering of standards and therefore costs. A more detailed analysis of these budgets is performed in the following sections. Table C2: Underlying Assumptions for the MEF and MTCC Budgets Item MEF MTCC Content of Program (Length in Km, excluding IDB, IBRD) 866 Km 1762 Km Budget allocation (US$ million, excluding IDB, IBRD, and 177 87 maintenance) Average unit costs of construction and/or rehabilitation 200,000 50,000 (US$/km) The discrepancies in Tables Cl and C2 above suggest a serious lack of coordination in the Programming and Budgeting of road activities in the country. There is a need for more rational planning and cost estimation within both the MEF and MTCC, as well as a clear delineation of priorities (construction versus improvement and rehabilitation) both of which affect the assumptions on unit costs. Finally, the separate treatment of regional roads within the budgeting process, in addition to the other deficiencies above, indicate a lack of a coordinated transport plan for the country. C3. The 1994-1997 Road Budget Evaluation Criteria Our analysis of the five year investment program is based on the following criteria for evaluation: (i) identification of the types of projects envisaged by the five year plan distinguishing between new construction and rehabilitation or improvement works; (ii) verification of the extent in length and location of each link in the investment plan; (iii) assessment of the level of traffic on each road link to be improved, rehabilitated, or maintained; (iv) measurement of the proposed unit costs of works planned on each link; and (v) estimation of vehicle operating costs on each road link. Economic rates of return are calculated for all links in the road investment program. Special emphasis us given to those links that: (i) have very low levels of traffic (average daily traffic ADT of less than 75)3; (ii) unusually high unit costs of works planned (more than US$ 100,000 per Km); and (iii) new construction works. Attachment Cl presents a sample estimate using such inputs. Road links with rates of return less than 12% are identified and two alternatives suggested for these links: (i) the level of traffic at which the proposed level of investment would be justified; and (ii) the level of investment which would be appropriate for the traffic on these links. Excluding IDB and IBRD rehabilitation projects which comprise some 4,800 Km of roads to be rehabilitated (or adequately maintained through a pilot maintenance project on 620 Km), the total program covers mostly the National Road Network; that is a length of approximately 12,000 Km of roads. About 4,600 Km (38% of the 12,000 Km of roads in the MEF Plan) classify as very low traffic links. The unit cost of works for such links in the MEF Plan is about US$ 200,000. Such a high cost would imply extremely high standards of construction and paving. Some 1,900 Km of roads (41% of 4,600 Km) are "new" projects involving construction works in the Sierra and Selva. On about one third of these links the traffic volumes are of the order of 20 vehicles per day. Based on our analysis (see Figure C1), about 24% of the road links in the proposed budget have economic rates of return of less than 12%. These links need to be reevaluated before inclusion in the program. We have done such an analysis for 28 links with either low traffic volumes, high unit costs, and involving new construction is given in Attachment C2. The economic rates of return and recommended alternatives for those links with less than 12% rates of return are also included. As can be seen from this analysis, the Government could save approximately US$ 275 million, if the recommendations proposed are followed. This would reduce the road budget by about 16%. Concerning the new construction works, should any of these links be considered for inclusion in the program (on the grounds of social or national integration objectives) they should be preceded by careful environmental studies. The units costs should be limited to less than US$ 130,000 per Km, which implies lowering the design standards. The cost savings per Km with such a recommendation are in the order of 30%. The remaining program, about 7,500 Km, concerns roads which generally carry more than 100 vehicles per day, while the average cost of rehabilitation is of the order of US$ 90,000 per Km. A typical rehabilitation cost from studies carried out for the IBRD project is US$ 130,000 per Km (in the Sierra and Ceja de Selva). Unit costs on the Pan American Highway (Costa) average some US$ 60,000 to 70,000 per Km. Therefore, the unit cost assumed for these links are acceptable. About 15% of these links had rates of return between 12 and 20%, and 61% had returns above 20%. This part of the road program is very well justified. 3. Existing traffic volume is only one of the variables affecting the economic rates of return. Generated and diverted traffic as a result of improvement in road condition is another important variable. As a result, the rates of return on these road links are underestimated. The analysis, however, serves as an objective prioritization tool in the face of limited budgets and high uncertainty. Figure Cl Integrity of the 1993-1997 Road Investment Plan: Economic Rates of Return on Roads 70 60 - 50 - C 40 - 20 - loo 01 < 12 13-20 >20 Range Integrity of the Overall MIEF Road Budget To measure the overall integrity of the road budget, calculations of the savings in user costs (Vehicle Operating Costs VOC) was done. The user costs on the current network, before the planned investment was calculated and compared to the user cost after the investment. The results are summarized in attachment C3. The vehicle operating costs without the planned investments under the five year plan are of the order of US$ 1.1 billion per year. User costs are more than three times higher than the average annual investment over the five year period. Planned investments, if properly executed will reduce the operating costs to about US$ 0.8 billion per year. The savings to the users of the planned investments are therefore close to US$ 0.3 billion per year. The savings in user costs after the investment are just about equal to the average annual investment on roads under the five year plan. With any reduction in the proposed budget, following the recommendations made above, the investments in road rehabilitation and improvement will be well justified. Remaining questions have to do with the capacity of the construction industry to undertake these works within the above mentioned implementation period, which is covered in Annex E. Evaluation of the MTCC 1994 Budget The MTCC Plan which has been approved by congress was also evaluated, as there is a high likelihood that it would be implemented. The percentage distribution of contemplated activities under the MTCC budget for 1994, which was approved by congress is shown in Table C3. Table C3: Breakdown of MTCC Road Budget for 1994 Activities Budget Amount Share of Budget I US$ (million) (%) 1. Rehabilitation of paved network a) IDB 122 50.4 b) IBRD 14 5.7 c) MTCC 10 4.1 2. Rehabilitation of unpaved roads by MTCC 15 6.2 3. Maintenance (SINMAC) 20 8.3 4. Paving 25 10.3 5. Upgrading to gravel roads 6 2.5 6. Construction of gravel roads 11 4.5 7. Reconstruction of paved roads 1 0.6 8. Construction of new bridges 15 6.2 9. Studies 3 1.2 Total 242 100 A global assessment of the distribution of allocations is 68% to rehabilitation, 24% to "new" construction, and 8% to maintenance. This allocation seems to be acceptable except for the fact that one of the conditions of the IBRD loan is that the Government should (as from January 1995) allocate at least US$ 50 million per year, or a figure to be agreed with the Bank, to maintenance. This issue is discussed in more detail in the main body of the report. Most of the road investments under the 1994 budget are essentially on the National Network. Needed investments in the system of secondary and feeder roads fall under the regional budget which is discussed in the main body of this report. Recommendations Based on our evaluations, we recommend the following: (i) a clear alignment of priorities for the transport sector, and consequently, for the roads sub-sector to the budgetary allocations in the 1994-1997 plan. In particular, the content of the road program should reflect the priorities listed in the supporting text from MEF; (ii) a clarification of the type of works, standards, and hence unit costs of programmed works under the MEF and MTCC plans (i.e., between the budgeting and implementation of the plans). We recommend analysis of the type done on the 28 links identified in Attachment C2 with the suggested minimum levels of investment. (iii) rescheduling of planned investments on very low traffic volume roads until the demand for them reaches the recommended levels in attachment C2 or when budgetary constraints are relaxed. These recommendations, if followed, could result in up to a 16% reduction in the proposed road budget, which is well in line with the IMF recommendations; (iv) a more critical assessment of the feeder road system to support and enhance the returns on investment on the national road network. A first step in this direction is to explicitly integrate the planning, programming, and prioritization for the subnational road network within the process for the national road network; and (v) strengthening of the programming, budgeting, and prioritization processes within and between MTCC and MEF. This would include institutional capacity building in both ministries to handle issues such as cost estimation, network analysis, environmental assessment, traffic projection, condition evaluation, economic analysis, and investment prioritization. 1. Cajabamba-Cajamarcea Length: 123 Present Condition: Gravel and Fair Terrain: Mountain ADT: 54 Investment: $18,219,000 Before Proj00t After Project Maintenance Total Total TOTAL Maintenance Total Total TOTAL NET Year ADT VOC/km Par Km Maintenance VOC (VOC + maint) VOC/km Per Km Inveetment Maintenance VOC (VOC+rmaint+lnv) BENEFIT 1 54 04605 3,500 430,500 1,116,404 1,546,904 0.219 0 $16,219,000 0 530,928 16,749,928 -15,203.024 2 56 0.4605 3,500 430,500 1,101,060 1,591,560 0.219 0 0 552,165 552,165 1,039,395 3 58 0.4605 3,500 430,500 1,207,503 1,638,003 0.219 0 0 574,252 574,252 1,063,761 4 61 04605 3,500 430,600 1,255,803 1,686,303 0219 0 0 597,222 597,222 1,089,081 5 63 0.4005 3,500 430,500 1,306,035 1,736,535 0 219 1,500 184.500 621,111 805,611 930,924 6 66 04605 3,500 430,500 1,358,276 1,788,776 0219 1,500 184,500 645,955 830,455 968,321 7 68 0.4605 3,500 430,500 1,412,607 1,843.107 0.219 1,500 184,500 071,794 856,294 986,814 a 71 04605 3,500 430,500 1,469,111 1,899,011 0.219 1,500 184,500 698,65 883,165 1,016.446 9 74 0.4005 3,500 430,500 1,527,876 1,958,376 0.219 1,500 184,500 726,612 911,112 1,047,264 10 77 0.4605 3,500 430,500 1,588.991 2,019,491 0.219 1,500 184,500 755,676 940,176 1,079,314 11 80 04605 3,500 430,500 1,652,551 2,083,051 0.219 1,500 184,500 785,904 970,404 1,112,647 12 83 0.405 3,500 430,500 1,718,053 2,149,153 0.219 1,500 184,500 817,340 1,001,840 1,147,313 13 86 0.4605 3,500 430,500 1,787,399 2,217,899 0.219 1,500 184,500 850,033 1,034,533 1.183,365 14 90 0.4605 3,500 430,500 1,85,895 2,289,395 0.219 1,500 184,500 884,035 1,068,535 1,220,860 15 94 0.4605 3,500 430,500 1,933,250 2,363,750 0.219 1,500 184500 919,390 -2,139,904 4,503,654 Base ERR 2.22% Alternatives: Minimum ADT required to have ERR of at least 12%: 130 Investment required to have ERR of 12% at the same ADT: $8,300,000 CAJCAJ.XLS Of the 28 links analyzed the summary results are as follows: Current For a min. 12% return Length investment ADT ERR Investment ADT Remarks Name of the link in km (in US$m) 1. Cajambambe-Cajamarca 123 $16.22 54 2.22% $8.30 130 2. Cutervo-Socoto 30 $5.75 69 5.92% $3.90 110 3. Socota-Cavico 43 $8.95 69 4.84% $5.60 120 4. Huancabamba-Ayabaca 22 $2.45 21 3.34% $1.25 54 New Constr 5. Rio Pampas-Andahuaylas 118 $24.51 118 5.85% $16.50 185 6. Andahuaylae-Abancay 139 $28.68 108 -3.08% $8.10 535 7. Abancay-Umatambo LV 119 $24.51 140 3.51% $13.80 275 8. Pte. Paucartambo-DV.P. Bermudez 132 $42.13 190 1.57% $19.80 440 9. Tocache-Pts. Pizana 39 $7.01 61 -1.51% $2.44 235 10. Juanjul-Terapoto 134 $39.76 260 5.69% $26.50 405 11. Satipo-Mazamari 21 $2.00 114 12.72% n.a. n.a. New Constr 12. SV. Olmos-Huancabmba 143 $30.72 304 21.61% n.a. n.a. 13. Pte. Balzas-Achamaqul 157 $31.86 20 -4.47% $7.40 140 14. Achamaqul-Chachapoyas 14 $0.55 36 15.36% n.a. n.a. 15. Tarapoto-Yurimaguas 131 $13.06 76 11.76% $12.80 78 16. Shorey-Laguna-Saceacocha 73 $25.38 213 1.73% $11.90 480 17. Lag. Saceacocha-EMP 5N Huayabamba 261 $12.45 30 3.99% $7.50 112 New Constr 18. Yanac-EMSN Pt. Huicte 261 $73.28 41 -1.05% $23.00 160 New Constr 19. Yanahuance-Pte. Tui 40 $8.18 112 1.56% $3.87 270 20. Huayllay-Cerro de Pasco 41 $5.73 100 9.71% $5.00 118 21. Pampano-Rumichaca 120 $16.33 77 2.83% $8.80 170 22. LV Apacheta Gde. Ayacycho 98 $22.21 100 2.91% $11.90 205 23. Puqui-LV Sol do Oro 114 $23.62 77 -1.28% $8.40 275 24. LV Sol do Oro-EM3S El Olivio 130 $36.13 142 9.95% $32.00 163 25. Abacay- Cuzco 81 $26.83 142 1.75% $12.80 320 26. Urcos- Pto. Maldonado 355 $58.88 107 14.83% n.a. n.a. 27. Yauli-Combapate 114 $23.85 70 -1.92% $7.80 278 28. LV-Julia 143 $29.70 215 1.34% $13.80 530 Total $640.71 473, I Assumptions: Total excl. $548.56 1. Across the board Salvage value of 20% was used except for New Construction - 30% 2. ADT has a fixed growth rate of 4% per annum 3. VOC before and after differs according to the road condition and terrain 4. Maintenance is assumed to be fixed at $3500/km for before and $1500/km for after scenario 5. Analysis period of 15 years SUMMARY.XLS Program andlor Projects Length Surface Terrain COST COST Stages andlor Principle Componente km Type Condition Type Before VOC After VOC 001 Longitudinal Highway Coastal or Panamerican 3,05 001 North Panamericana 1,170 Lima- Dv. Ancon 44 P F/G F 56,531,200 50,878,080 ,Av. Ancon-Chancay 39 P G F/M 11,098,517 10,011,419 Chancay-Huacho 67 P F/G F 29,160,142 26,244,128 ·uacho-Pativilca 54 P B F 17,026,917 12,394,594 ,Pativica-Huarmey 91 P B R 20,801,857 15,141,988 Huarmey-Santa 152 P FIG R 20,868,830 18,.731,934 Santa-Pte. Moche 117 P B F 24,613,795 17,917,395 Dv. Otuzco-El Milagro 98 P B F 17,902,170 13,031,726 El Miagro-Chepen 39 P F F 5,719,965 4,881,694 Chepen-Dv. Morrope 106 P F/B F 20,650,246 16,225,193 Dv. Morrope-LV El Cruce 83 P F F 4,589,571 3,916,962 LV El Cruce-Piura 95 P G F 6,238,726 5,938,787 Piura -Suilana 34 P F F 9,198,788 7,850,690 Suilana-Las Lomas 85 P F F 338,297 .288,719 Las Lomas-Pte. Macara 27 P F M 106,996 101,438 Ramal Sullana-La Tina 39 E B , M 213,739 80,712 002S South Panamericana 1.341 Lima -Pucusana 58 P G F 26,296,866 25,032,594 Pucusana-Quilmana 64 P G F 10,193,930 9,703,837 Quilmana-Pte. Jahay 53 P F/B F 12,187,350 9,575,775 Pte. Jahay-Pte. Huamani 54 P B F 22,334,426 16,258,148 Pte. Huamani-Ica 74 P B F 11,416,803 8,310,761 Ica-TV C4 33 P F F 5,798,463 4,948,688 TV C4-Palpa 60 P F/B F 6,727,417 5,285,828 Palpa-Dv. Lomes 124 P B F 6,967,868 5,072,198 Dv. Lomes-Dv. Acari 9 P B F 652,270 474,814 Dv. Acari-Pto. Viejo 81 P B F 5,870,426 4,273,325 Pto. Viejo-km. 715 5 P F . F 309,082 263,786 Km 715-Pte. Haway 166 P F M 10,221,775 8,744,409 Pte. Haway- Acc. Microondas 118 P B M 10,627,264 7,809,969 Acc. Microondas-Dv. Mollendo 85 P B F 4,843,871 3,526,053 Dv. Mollendo-Sta. Rosa 69 P F/B F/M 3,477,167 2,735,809 Pte. S. Rosa-Pte. Montalvo 99 P G F/M 2,714,660 2,614,349 Pte. Montalvo-Oda. Honda 63 P G F 3,520,259 3,351,015 Qda. Honda-Pte. Camiara 10 P F F 536,024 457,469 Pte. Camiara-Tacna 80 P F F 4,288,195 3,659,753 Tacna-La Concordia 36 P G F 3,110,291 2,960,757 LEGEND Surface Type: Paved, Gravel, Earth Surface Condition: Good, Fair, Bad Terrain: Flat, Rolling, Mountainous Program and/or Projects Length Surface Terrain COST COST Stages and/or Principle Components km Type Condition Type Before VOC After VOC 001A Var Sullana-Aguas Vedes 275 Empaf. 1 N Sultana-Dv. Talara 75 P F/B F 8,954,253 7,035,485 Dv. Talara-Cancas 100 P/G G F 10,225,840 7,988,938 Cancas-Tumbes 73 P G F 5,708,425 5,433,981 Tumbes-Aguas Verdes 27 P G F 1,783,361 1,697,622 001B Variante Lambaywque-Dsv Sech 246 Lambayeque-Dv. Olmos 90 P B R 8,311,116 6,049,787 Dv. Olmos-Limite Vial 65 P B F 4,504,334 3,278,890 LV-Km 172 35 E B F 5,974,357 2,050,899 Km 172-Piura 56 P/G B F 4,900,081 2,668,238 001C Var. Pasamayo 23 Dv. Ancon-Pte. Chancay 23 P B M 6,829,668 5,019,119 003 Garretera Langitudinul de la Sierr 346 003N Longitudinal de la Sierra North 1,713 La Oroya-Dv. Tarma 22 P F R 1,923,619 1,963,528 Dv. Tarma-Huayre 33 P F R/F 2,746,091 2,345,499 Huayre-Carhuamayo 30 P F F 2,370,193 2,022,837 Carhuamayo-Chirchin 29 P F R 2,130,072 1,820,726 Chircin-Dv. Cerro de P. 11 P F R/M 957,818 819,060 Dv. Cerro de P-Huanuco 110 P B R/M 7,581,324 5,545,719 Huanuco-Huallanca 155 G G/F R/M 5,568,394 4,437,177 Huallanca-LV Abra Yanashaila 28 E F M 761,104 480,136 LV A. Yanashalla-Aquia 38 E B M 5,999,815 2,265,665 Aquia-Conococha 48 E F/B M 6,057,671 2,861,892 Conocoha-Huaraz 79 P G/B M 5,651,501 4,679,315 Huaraz-Tres Cruces 136 P/G G/F M 6,537,836 6,195,817 Tres Cruces-Cabana 63 E F/G M 261,798 186,260 Cabana-Mollepata 125 E B M 782,925 295,650 Mollepata-AFngasmarca Angasmarca-Cachicadan 63 E B M 706,981 266,972 Cachicadan-Huamachuco 121 G B M 1,223,371 596,228 Huamachuco-Cajabamba 54 G B M 644,241 313,980 Cajabamba-Cajamarca 123 G F M 889,729 654,569 Cajamarca-Chota 153 G F/B R/M 2,644,540 1,518,705 Chota-Cutervo 67 G F/B M 772,044 443,369 Cutervo-Socota 30 E B M 540,218 203,999 Socota-Cavico 43 G F/B M 488,413 280,485 Cavico-Tambo 86 P/G G/F M 2,440,259 1,905,687 LEGEND Surface Type: Paved, Gravel, Earth Surface Condition: Good, Fair, Bad Terrain: Flat, Rolling, Mountainous Program arid/or Projects Length Surface Terrain COST COST Stages and/or Principle Components km Type Condition Type Before VOC After VOC Tambo-Huancabamba 15 E B M 82,207 31,043 Huan.-Ayabaca 22 E B M 120,570 45,530 Ayabaca-Vado Grande 29 E B M 855,215 322,948 003S Longitudinal do la Sierra South 1,484 J:a Oroya-Huancayo 125 P B R/F 18,991,406 13,824,375 Huancayo-LV Nahulmupuquio 39 P/G B F/R 7,206,213 4,012,448 LV Nahulmupuquio-La Mejorada 57 G F R/M 862,783 627,479 jLaMejorada-Mayoc LV 52 G B M 1,135,611 553,457 LV Mayoc-Ayacucho 76 P/G F/B R 3,277,412 1,824,876 Ayacucho-Rio Pampas 143 G B R/M 1,679,374 1,221,363 Rio Pampas-Andahuaylas 118 G/E F/B R/M 2,472,304 1,437,332 Andahuaylas-Abancay 139 G/E F M 2,178,054 1,479,433 Abancay-Limatambo LV 119 G F/B M 2,800,254 1,641,843 Limatambo LV-Cuzco 78 P/G G F/M 4,916,541 4,496,837 Cuzco-Urcos 46 P B F/R 3,657,492 2,662,390 Urcos-La Raya LV 133 P/G/E F/B R 4,747,216 2,799,008 La Raya LV-Puno 211 G/P F/B F/R 23,311,747 15,365,417 Puno-Dasaguadero 148 P F/G F 130,728 117,656 003A Var. Huancavelica 264 Izcuchaca-Sta. Ines 264 G/E F/B M 6,412,083 4,373,491 003 B Pte. Stuart-Huancayo Pte. Stuart-Huancayo 39 P F F 2,381,117 2,032,160 005 Carretera Lontgitudinal Marginal de la Sel 1,736 005 N Marginal de la Solve North 1,494 S. Luis Shuaro-Pte. Paucartambo 8 G B M 745,750 363,452 Pte. Paucartambo-DV. P. Bermudez 132 G F/B M 4,215,509 2,471,634 Bermudez-Pto. Inca 107 G G/B R 1,649,722 1,001,839 Pto. Inca-S. Alejandro 89 G B F 3,857,074 1,725,928 S. Alenjandro-LV Abra la Divisional 104 P/G F/B R/F 8,311,342 4,799,662 LV Abra la Div-DV Tingo Maria 27 P B M 1,794,379 1,318,688 Tingo Maria-Tocache 153 G/P F/B F/R 16,938,347 11,930,726 Tocache-Pte. Pizana 39 G F R 302,181 215,347 Pte. Pizana-Juanjui 138 G/E F/B M 5,831,536 2,910,379 Juanjui-Tarapoto 134 G F R/M 4,546,185 3,293,599 Tarapoto-Rioja BIRF 134 P/G F/G R/M 6,843,120 5,969,808 Rioja-Rio Nieva 100 G B R/M 6,146,837 2,940,039 *Rio Nieva-Las Pomacochas 43 G B M 2,556,339 1,245,869 Pomacochas-Bagua BIF 96 G G/B R/M 5,578,403 2,668,156 Bagua-El Reposo 24 G B R 1,362,408 638,709 .El Riposo-Pte. 24 del Julio 11 G F R 410,783 292,742 LEGEND Surface Type: Paved, Gravel, Earth Surface Condition: Good, Fair, Bad Terrain: Flat, Rolling, Mountainous Program and/or Projects Length Surface Terrain COST COST Stages and/or Principle Components km Type Condition Type Before VOC After VOC Pte 24 de Julio-Chamaya 15 G F M 1,040,830 765,734 Chamav-Jaen 18 P C M 1,437,233 1,362,572 Jaen-S Igacio 110 P B M 11,725,406 8,616,993 S. Ignacio-R. Canchis 12 E B M 3,069,066 1,158,948 lamal de la Marginal Norte 50 Ramal Pte. Paucart-Oxapampa 50 G F R 4,013,832 2,860,432 005S Marginal de la Selva South 192 S. LuisShuaro-R. Chanchamayo 11 G B M 1,067,669 520,344 R. Chanchamayo-Satipo 111 G F M 4,490,440 3,303,593 Satipo-Mazaman 21 G F R 304,086 216,705 Mazamari-P o. Prado Pto. Prado-I io Tambo Rio Tambo- itzcarraid Fitzcarriad-MIaiu 49 G B M 1,129,545 550,500 Manu-Inam ari Inambari-Ba timore Baltimore-R o Heat ITOTAL LONGITUDINAL ROUTES 8,252 Total Transnor 1928 002 Trans Paita-Huancabamba 269 Paira-Empaime 49 P G/B F 3,888,914 3,208,354 Empaine-Piura 10 P F F 767,201 654,766 Piura-Desvio Olmos 67 G F M 2,728,395 2,007,266 Dv. Olmos-Huancabamba 143 G B M 8,790,473 4,284,166 004 Bayovar-Ayar Manco 678 Pto. Bappo-Bayovar 67 P F F 4,907,629 4,188,408 Bayvor-Cruce Emp. in Bid Cruce Emp-Olmos Olmos-Desvla Chiclayo - 66 P G M 1,563,706 1,482,475 Dv Chiclayo-Emp.3N El Tambo El Tambo-Pucara Pucara-Cavico Cavico-Chamaya 215 P G/F R 7,945,555 7,149,229 Chamaya-Pte. 24 de Julio Pte. 24 de Juloi-El Reposo 16 G B R 1,186,314 556,155 *El Reposo-Nazaret-Punta de Carretera 110 G F/B R 3,133,828 1,772,382 Punta Carr.-Ayar Manco 204 G F/B R 5,811,826 3,286,962 .006 Pimentel-Cochabamba 181 LEGEND Surface Type: Paved, Gravel, Earth Surface Condition: Good, Fair, Bad Terrain: Flat, Rolling, Mountainous Program and/or Projects Length Surface Terrain COST COST Stages and/or Principle Components km Type Condition Type Before VOC After VOC Plntentel-Lnp 001N 12 P F F 2,564,788 2,188,914 Imp 001N-Chiclayo 64 P F R 9,384,810 8,021,871 Chiclayo-LV Cumbil E B R 0 0 LV Cumbil-003N Cochabamba 105 G F/B M 370,622 217,303 508 PACASMAYO-YURIMAGIAS 800 EmplN-Tembladera 45 P B F 3,323,894 2,419,600 Tembladera-Pte. Yonan 5 P F F 315,010 268,844 'Pte. Yoonan-Cajamarca 127 P B R/M 4,295,695 3,142,290 Cajamarca-Pte. Balzas 164 G/E B R/M 4,207,320 1,751,923 Pte. Balzas-Leymebamba 89 E F/B M 371,304 175,419 Leymebamba-Achamaqui 68 G F M 182,179 134,028 Achamaqul-Chachapoyas 14 G F M 67,513 49,669 Chachapoyas-Rodriguez de M 85 G F M 979,211 720,401 Rodriguez de M.-Omia 15 E B M 336,658 127,130 Sortor-Emp 5N 37 E B M 511,772 193,257 Caizada-Tarapoto 20 G B M 307,359 149,796 Tarapoto-Desvio Juanjut 38 E F M 451,163 284,612 Dv. Juanjui Em5n-Alianza 28 G B M 430,303 209,714 Allanza-Shanusi 65 G F R 627,479 447,169 Shanusi-Yurimaguas RAMAL DE LA RUTA 8 Achamaqui-Pedro Ruiz 39 G F M 397,043 292,102 TOTAL TRANSCEN 3.281 010 SALAVERRY-JUANJUI 475 Salaverry-Emp. 1N Alto Moche 5 P G F 416,801 396,762 Alto Moche-Desvio Trujillo 37 P G R 2,181,571 2,061,484 Dv. Trujillo-Shorey 84 G F M 2,396,723 1,763,257 Shorey-Laguna Sacsacocha 73 G F M 2,082,866 1,532,354 Lag. Sacsacocha-Pta. Carretera E F M Pta. Carretera-Huicungo 276 G G M 572,203 507,730 Huicungo-Emp. 5N Huayabamba G G M 012 CHIMBOTE-HUICTE 475 Emp. 1N Santa-Chuquicara* 71 E F M 23,048,386 14,539,870 Chuquicara-3N Yungaypampa 60 E F M 403,048 254,259 Yunggaypampa-DV Sihuas (Yanac) 83 E F M 272,291 171,773 Yanac-Sihuas Punta Carret. Sihuas-Huacrachuco 261 E B R 2,687,235 968,655 Huacrachuco-Pampayacu Pampayacu-Em5N Pt. Huicte LEGEND Surface Type: Paved, Gravel, Earth Surface Condition: Good, Fair, Bad Terrain: Flat, Rolling, Mountainous Program and/or Projects Length Surface Terrain COST COST Stages and/or Principle Components km Type Condition Type Before VOC After VOC 014 PATIVILCA-CONOCOCHA 122 Emp IN Pativ -Conoc (Emp 3N)* 122 P F R 3,337,568 2,852,859 014A CASMA-HUARAZ 149 Emp. IN Casma-Yautan 35 P G R 1,609,701 1,521,094 Yautan-3N Huaraz* 114 E F R 5,080,581 3,095,784 016 HUACHO-PUCALLPA 653 Emp.1N Huaura-Sayan 45 P G/B R 4,959,200 4,686,217 Sayan-Oyon 93 G/E F M 1,594,600 1,173,139 Oyon-Abra Uchucchacua 31 E B M 962,737 363,551 A. Uchucchacua-Yanahuanca 52 E B M 1,614,913 609,827 Yanahuanca-Pte. Tusi 40 G F M 600,118 441,504 Pte. Tusi-Ambo 32 G B M 750,604 365,818 Ambo-Huanuco 24 P G R 2,904,623 2,744,736 Huanuco-Tingo Maria* 96 P FIG R 4,585,440 3,919,504 Tingo Maria-Dv. Tocache 15 P G M 2,540,833 2,408,841 Dv. Tocache-Abra La Divisoria 64 G G F 3,573,940 2,812,053 Abra La Divisoria-Von Humboldt 75 P B F 4,415,478 3,214,208 Von humboldt- Pucallpa 86 P F F 4,318,511 3,685,625 018 LIMA-CANTA-CERRO DE PASCO 299 Lima-Canta 103 P/E F/B M 32,489,223 16,361,720 Canta- LV Abra La Viuda* 45 E F/B M 1,103,694 696,256 Abra La Viuda-Abra La Cruzada 64 G F/B M 1,474,577 1,084,838 Abra La Cruzada-Huayllay 46 G B M 1,599,886 779,728 Huayllay-Emp.3N Unish Unish-Emp. 18A Yanamate 41 G B M 829,061 404,055 Yanamate-Cerro de Pasco 020 LIMA TARMA CHANCHAMAYO 294 Lima-Ricardo Palma 39 P G/B F 24,832,901 23,639,011 Ricardo Palio-Cupiche 15 P F M 2,936,352 2,511,957 Cupiche-CoCachacra Cocachacra-Matucana 23 P F/B M 4,502,406 3,851,668 Matucana-San Mateo 19 P B M 5,305,039 3,898,670 -San Mateo-La Oroya 79 P G/B R/M 13,866,982 13,103,662 La Oroya-Dt. Tarma* (Las Vegas) 34 P B R 2,177,459 1,585,005 -Las Vegas-Tarma* Tarma-Emp. 5 Chanchamayo 85 P/G F/B R/M 13,417,692 8,033,738 * 022 CANETE-HUANCAYO 285 Canete-Yauyos* 135 G/E B F/M 5,934,780 2,607,633 Yauyos-Lv. Tinco 57 E B M 2,915,613 1,101,001 * Lv. Tinco-N. Bueno 24 G F R/M 1,005,998 716,918 LEGEND Surface Type: Paved, Gravel, Earth Surface Condition: Good, Fair, Bad Terrain: Flat, Rolling, Mountainous Program and/or Projects Length Surface Terrain COST COST Stages and/or Principle Components km Type Condition Type Before VOC After VOC N. Bueno Em 3S Huancayo 69 G B R 4,396,545 2,061,140 024 PTO. S. MARTIN-AYAC-S. FRANCI 529 Pta. Pejerrey-Paracas 14 P G F 4,132,477 3,933,801 Paracas-Pisco 20 P G F 5,903,539 5,619,715 'isco-EmT. IS 5 P G F 1,538,139 1,464,190 Dv. Pisco-Dv.Huancavelica 15 G F R 1,329,899 947,744 Em. 1S-Pampano 64 G F R 2,170,518 1,546,806 Pampano-Huaytara* 32 G F M 330,065 242,827 Huaytara-Rumichaca 88 G F M 907,679 667,775 Rumichaca-LV Apacheta Gde. 17 G F M 173,070 127,327 LV Apacheta Gde.-Ayacycho* 98 G B R 1,892,233 887,096 Avacucho-Chacco 16 P B R 416,485 303,166 Chacco-Dv.Tambo* 46 G B M 2,344,018 1,142,392 Tambo-S. Francisco 114 G B M 5,809,089 2,831,144 TOTAL TRANSSUR 3,283 026 SAN JUAN-INAPARI 1,293 San Juan - Emp 1S 40 P B F 1,513,027 1,101,395 Emp. 1S - Nazca 20 P F R 214,635 183,464 Nazca - Puquio 135 G B M 3,275,802 1,596,510 Puquio - LV Sol de Oro 114 G F M 1,175,857 865,072 Sol de Oro - Abra Huashuccasa 0 LV Huashuccasa - Em.3S El Olivo 130 E B M 4,817,599 1,819,233 El Olivo - Abancay 45 G F M 855,972 629,735 Abancay - Lmte. Vial Limatambo 29 E B M 1,074,695 405,829 Limatambo - Cuzco 52 G F M 989,124 727,693 Cuzo - Huarcapay 84 E B M 2,345,643 885,767 Huarcapay - Urcos 45 G F/B M 644,993 474,518 Urcos - Quincemil 115 E F M 1,922,287 1,212,658 Quincemil - Pto. Maldonado 240 E/G F M/F 3,580,562 2,343,300 Pto. Maldonado - Iberia 177 E B R 1,422,344 512,705 Iberia - Inapar 67 E B R 538,401 194,075 028 AREQUIPA-TINTAYA-COMBAPATA 406 Arequipa - Yura 24 P F R 1,929,583 1,836,814 Yura - OV (Negromayo) 193 G B F 1,158,679 518,475 Negromayo - Yauli 75 G G M 536,441 475,997 Yauli - Sta. Lucia 19 G F M 175,615 129,199 Sta. Lucia - Combapata 95 G F M 890,801 655,358 RAMAL SICUANI-EL DESCANSO 44 Sicuani - El Desc (Emp. Ruta 28) 44 G F M 147,351 108,405 LEGEND Surface Type: Paved, Gravel, Earth Surface Condition: Good, Fair, Bad Terrain: Flat, Rolling, Mountainous Program and/or Projects Length Surface Terrain COST COST Stages and/or Principle Components km Type Condition Type Before VOC After VOC 030 MATARANI-JULIACA-INAMBARI 648 Matarani - Em 1S 58 P B M/F 5,703,029 4,191,152 Emp 1S - Emp. 30A Reparticion 25 P G M 4,097,709 3,884,841 Reparticion - Arequipa 17 P G M 2,786,442 2,641,692 Arequipa - LV 138 G/E F/B R/M 10,544,254 6,900,690 iV - Juliaca 143 G F M/R 4,118,446 3,029,920 Pucara - PC (San Gaban) 267 G/E F F 4,125,855 2,824,246 .032 HUMALZO-PUNO 153 Humalzo - Pte. Vizcachas 26 E F R 208,571 127,090 Pte. Vizcachas - Emp. 3S Puno 127 G/E B M 1,588,261 648,321 034 ILO - DESAGUADERO 317 Ilo - Emp. 1S 47 P G F 2,233,718 2,126,328 Emp. 1S-Montalvo 48 P F F 2,264,004 1,932,211 Montalvo - Moquegua Moquegua - Torata BIRF 23 G/E F/B R 513,102 335,800 Torata - Em034 Humalzo BIRF Humalzo - L. Suches 31 E B F 326,958 117,857 L. Suches - LV 29 G F R 1,631,826 1,162,910 LV - Mazocruz 56 G F R 3,151,112 2,245,620 Mazocruz - Em 35 Liave 83 G F R/M 7,099,542 3,328,330 RAMAL DE LA RUTA 34 0 Mazocruz - Picupinchuni Pichupinchuni - Desaguadero 036 TACNA - MAZOCRUZ 236 Emp. 1S Tacna - Tarata 87 E B M 7,061,201 2,666,467 Tarata - LV Pte. Chilliculco 92 G B M 5,785,633 2,819,713 Pte. Chilliculco - Mazocruz 57 G B R 594,316 278,621 040 TACNA-PACHIA-COLLPA 186 Tacna - Pachia 16 P G F 583,066 555,034 Pachia - Collpa 170 G/E G/B M/R 5,570,229 3,645,438 TOTAL ROUTAS TRANSVERSALES 8,492 SUB TOTAL CARRETERAS RED VIAL NACIONAL 16,74 Plura Ruta s/n: Sullana-San Ignacio Junin 857: Junin San Ramon Junin 741: Concepcion-Ingenio y -(104) Jauja Amazonas Ruta s/n: Orakusa-Chiangos-Pte. Nieva Arequipa 101: Camana-Ilo-Los Banos LEGEND Surface Type: Paved, Gravel, Earth Surface Condition: Good, Fair, Bad Terrain: Flat, Rolling, Mountainous Program and/or Projects Length Surface Terrain COST COST Stages and/or Principle Components km Type Condition Type Before VOC After VOC Puno 109: Sina-Yanahuaya Pasco 109: Pozuzo-Codo Pozuzo Arequipa 590: Orcopampa-Antabamba Lima 722: Mala-Calango Ayacucho 101: Negro Mayo-Pampachirl junin 104: Satipo-Concepcion Lima 112: Cleneguilla-Huarochiri-S. Lorenzo Quinti Huancavelica 620 y 595: Acobamba-Lircay Puno 100: Juliaca-Huancane Junin 574: Huasahuasi-Huayonioc Ica 527- Ica-Conatrama-Carhua SUB-TOTAL REDES DEPARTAM. Y VECINALES TRANSITABILIDAD Y OTROS, sin especificar Ruta PROGRAMA DE PUENTES BIRF PROGRAMA DE PUENTES CAF PROGRAMA DE PUENTES: OTROS SUB-TOTAL PUENTES ESTUDIOS, ADMIN. Y SUPERV. incl. BID y BIRF FONDO DE CONSERVACION VIAL (FCV) TOTAL SUB-SECTOR CARRETERAS 1,092,301,946 803,797,079 LEGEND Surface Type: Paved, Gravel, Earth Surface Condition: Good, Fair, Bad Terrain: Flat, Rolling, Mountainous PERU - Public Expenditure Review ANNEX D Assessment of the Capacity of the Peruvian Construction Industry D1. Introduction In the context of its assistance program to Peru, the World Bank is conducting a Public Expenditure Review. The objectives of the review are to: (i) assist the Government in prioritizing investments in accordance with economic and social objectives; (ii) assess the absorptive capacity to administrate and implement works; (iii) address resource mobilization issues; and (iv) increase the effectiveness of public expenditures. D2. Objectives of the Studv This study addresses mainly the second objective of the public expenditure review; assessing the absorptive capacity to implement works. It discusses some of the constraints and obstacles preventing the domestic construction industry of Peru from assuming a more effective role in the country's construction and maintenance programs. Recommendations for assisting the industry are put forward and suggestions for further study are made. While the following sections attempt to present a comprehensive analysis of the aspects affecting the capacity of the construction industry in Peru, the analysis is brief and focused on the main issues deriving from the proposed 1993-1997 budget. D3. The Economy and the Construction Industry in Pern The study of the construction industry of Peru and its environment is based on a careful consideration of the following subjects: (i) the structure of the industry; (ii) demand for construction activity; (iii) the capacity of the industry; and (iv) a brief description of the obstacles facing the industry in terms of regulation, financing, and security. D3.1 Structure of the Construction Industry The structure of the construction industry in Peru is shaped by a number of factors: (i) public investment patterns in terms of level of expenditures and sectoral priorities; (ii) accelerated migrations to urban areas; and (iii) changes in government policies such as regulations on imports and exports, labor regulations and price controls. Public Investment Patterns. The nature of works undertaken historically have fluctuated with the pattern of public expenditures. The public sector played a dominant role in generating demand for construction during the twelve years of military rule (1968-1980). Large civil engineering construction works were undertaken in highways, power generation and irrigation structures, airports, ports, and telecommunication structures. Building construction was also substantial as the Government increased its direct participation in the economy. A large number of public housing schemes, factories, office and farm buildings were constructed. Government intervention became more pervasive during the second Belaunde (1980-1985) and the subsequent Garcia (1985-1990) administrations. During this period, the construction industry enjoyed large contracts which were quite evenly split between the civil works and building sectors. Under the expansionary fiscal policies of the 1980's, large expenditures were undertaken in the transport sector. These investments financed mainly new construction works to create employment, and very little maintenance and repair work was done. Peru's hyper inflation, accompanied by an unprecedented decline in production and real wages, effectively destroyed both public sector revenue-raising capacity and formal financial intermediation. The resulting fiscal crisis necessitated a drastic cut in public investment levels, producing a chronic lag in the provision of basic services such as water supply and sanitation and deferred maintenance of infrastructure networks. The structure of demand facing the construction industry during this period reflected this situation. Works were small and fragmented (water storage tanks instead of water distribution networks for example), deriving from communities trying to cope with infrastructure deficiencies. The structure of demand for construction activity is projected to change again following the planned investments in the 1993-1997 budget, as is discussed in later sections of the paper. Urbanization. The rapid urbanization in Peru in the last two decades created a large demand for housing, water supply and sewerage works, as well as urban roads and streets, as the cities of Peru tried to handle the overload of immigrants from rural areas. Urban centers (more than 2,000 people) account for 70% of the total population of Peru. Furthermore, the urban population is unevenly distributed. In 1990 some 67% of the urban population was concentrated in 16 cities with more the 100,000 inhabitants, while 26% was living in some 410 towns of less than 50,000 people, and 7% in intermediate size cities. The nature of demand for buildings and basic services emanating from this population distribution forces a concentration of the construction industry into large size firms in large cities such as Lima, and small-scale informal firms in smaller cities and marginal areas of large cities. Other factors affecting the size and degree of informality of the industry are described under section 3.3. Participation of the Industry in the Economy. The construction industry in Peru is the fourth largest contributor to the economy (see table 1 below). In 1991 this sector contributed to 8.6% of the GDP, compared to service at 43.5%, manufacturing at 25%, commerce at 9.8% and agriculture at 6.1 %. The contribution of this sector has historically been between 5 and 9%, with peaks in 1982 at 8.9% and 1988 at 9.0%. This puts Peru in the higher range of contribution of the construction industry to GDP, which ranges between 3 and 8% for most developing countries (Henriod, 1984). In 1992, the construction industry in Peru employed 300,000 people, constituting 3.7% of the national labor force; compared to 33% in agriculture, 27.3% in services, 16.1% in commerce, 10.4% in manufacturing and 4.4% in transportation. Although this percentage makes the industry the sixth largest employer, it has been a source of steady employment in the range of 3.4 to 3.9% since 1960. The labor productivity in the construction industry has declined dramatically: from 778,000 intis in 1970 to 657,000 intis in 1991. However this level of productivity is higher than that of the whole economy (419,800 intis in 1991), and is the fourth highest after electricity, gas and water at 2,331,800 intis, mining at 1,646,200 intis and manufacturing at 919, 300 intis during the same period. Table D1: Contribution to the Economy by Sector, 1991 Sector Contribution to Contnbution to Nommal GDP (%) Employment (%) productivity per Capita (1980 = 100) Agriculture and 6.5 33.4 105.5 Fishing Mining 2.4 2.4 41.1 Manufacturing 25.0 10.5 68.7 Electricity, Gas 0.6 0.3 101.6 and Water Construction 8.6 3.7 73.1 Commerce 9.8 15.8 53.5 Services 43.5 26.8 64.6 Other 3.6 7.1 67.7a Total 100 100 £ whole economy. Sources: Series Estadisticas, 1970-1991, World Bank Calculations, Peru en Numeros, 1991 Anuario Estadistico Participation of the Main Sectors of the Construction Industry. The informal sector represents about 50% of the labor force in the construction industry (Peru en Numeros, 1992) and contributes to 30% of the output. Other estimates indicate that about 60% of the urban labor force in Peru is informal, and that the construction industry employs about 5.4% of the informals (see Lubell, H. The Informal Sector: In the 1980's and 199Q's., OECD, 1991). Investments in road works (which were less than 20% of the total output from the construction industry in the 1980's) were carried out by the formal sector of the construction industry through a variety of options: (i) 23% by force account units; (ii) 35% by the army; and (iii) the remaining 42% by private contractors. This makes the share of the formal private construction industry relatively low. Due to a number of obstacles, the construction industry in Peru faces capacity, efficiency and modernization constraints. The lack of steady, large size contracts which are the main source of activity for the construction industry (in developing countries up to 80% of the construction works originate from the public sector, according to Henriod, 1984) coupled with the political uncertainty deriving from fluctuating government policies, hinders attempts at modernization and specialization. The informal sector, which in most countries complements the formal sector through subcontracting of small works, acts in Peru as a competitor for these works as they are the only source of activity for the industry due to the fragmented and unsteady nature of demand. The concentration of the construction industry into large size firms in urban areas, resulting from urbanization patterns, as well the high mobilization costs imposed by the geography of Peru, further impede efficiency and constrain the capacity of the industry at the national level. D3.2 Demand for Construction Activity Construction activity is characterized by two important features: (i) demand is subject to considerable fluctuations with serious repercussions on the utilization of resources; and (ii) construction depends on a wide range of factor combinations (particularly of capital and labor) than can be tailored to suit each finished product'. The output of construction fluctuates considerably more than that of the economy as a whole. This tendency is inherent in the demand structure of capital goods industries where relatively small changes in demand by consumers will cause the production capacity to be expanded or contracted at a considerably higher rate. In addition, important public sector investments tend to concentrate during periods of rapid economic growth, thereby accentuating the cyclical variations. Peru's construction industry is no exception to this rule as can be seen in Figure 1. Construction works are usually divided into three categories (see International Recommendations for Construction Statistics, United Nations 1968): (i) new construction and capital repairs on buildings (including factories, office buildings, hospitals, schools, power stations, etc.); (ii) civil engineering works (including roads, highways, airports, dams and irrigation, water and sewage, etc.); and (iii) current highway repairs and maintenance. The historical contribution of these three types of works to construction value added is shown in Table D2. Table D2: Relative Contribution of Different Works on Construction Value Added 1976 1977 1978 1979 1980 19 198 Value added 308 09 40919 73308 83581 1596W537_246 53~17 vn ~ ~ -64r =5T 85944 1221 298 314~na put in place: (i) Buildings 43560 61049 85811 160765 305265 550906 114736 9 (ii) Civil works 20118 22470 31825 59883 111978 217446 na (iii) M&R 1175 2415 3615 .863 16861 145511 n Note: All measures in million of current soles. Value added estimates refer to the contribution of construction to the GDP in producers' prices. Value of construction put in place refers to Gross Fixed Capital Formation in buildings and other construction in all sectors of the economy. The value of land improvement is included. Source: Construction Statistics Yearbook, United Nations, 1983 From this table, we observe that the value of the buildings makes up a large share of the value of construction put in place. Between 1976 and 1982, buildings have steadily contributed between 67 and 71% to the value of construction put in place. The table however indicates non-linear relations between the value of construction put in place and the I A lot of this discussion is derived from The Construction Industry: Issues and Strateries in Developin Countries, Ernesto Henriod, The World Bank, 1984 value added in construction, suggesting a structural change in the profitability of the industry (see figure 2). The construction industry has obtained a higher value added for a low value of construction put in place, indicating that it has succeeded in passing on the costs of inflation to its clients and employees (as is demonstrated later). The structure of demand facing the construction industry in the 1993-1997 period will be dramatically different from that in the last decade due to the size, sectoral composition, and nature of works planned. The Government of Peru is planning large investments in infrastructure, especially in rehabilitation and maintenance of the highway network. The planned investments in the road sector are $ 180 million for 1993, around $ 313 million in 1994; compared to $ 94 million in 1992, $ 8 million in 1991 and $1.2 million in 1990. The successive doubling in planned investments raises concerns about the capacity of the Peruvian construction industry to undertake these works. The following section develops a methodology to measure the capacity of the construction industry, and evaluates the impact of the planned investments on the structure of the industry. D4. Methodology for Measuring the Capacity of the Construction Industry Three methods are used to assess the capacity of the construction industry: (i) a gross output method, where historical measures of net aggregate output from the construction industry are evaluated to obtain the relative contribution of different types of works on construction value added; (ii) a net output method where the changes in value added in the construction industry are compared to the GDP growth rate; and (iii) an input method, where we estimate the aggregate production function of the industry. The three methods and their respective results are described below. A combination of analysis approaches was used to arrive at measures of the capacity of the construction industry. D4.1 Gross Output Method In this method the relative contribution of: (i) civil works; (ii) buildings; and (iii) highway maintenance and repairs on the construction value added is estimated. The value of the buildings put in place (as was shown in table 2 earlier) contribute significantly to the value added in construction. While the value of construction put in place in terms of road maintenance and repairs is low (see Table 2), its contribution to the value added is high (five times the contribution of buildings). The estimated coefficient of civil works is negative, suggesting that there is a crowding out effect, whereby construction firms perform less profitable civil works instead of more profitable building and repair works. However the paucity of the data precludes any definite conclusions, and we rely on theoretical and empirical works to support this hypothesis. Using these results we aggregate civil works and maintenance & repairs to obtain the relative contribution of road works and buildings to construction value added. The gross fixed capital formation in buildings, using this approach, contributes to 82% of the value added in construction, the remaining 18% deriving from road works (construction, reconstruction, maintenance & repairs). A summary of the regression results is shown in table D3 below: Table D3: Estimation Results from Alternative Models Variable Model 1 Model 2 Modl 3 Constant 14161.1 4336.7 12022.4 (3.95) (1.11) (1.91) X1 = buildngs 0.481 0.489 0.492 (111.64) (73.38) (39.03) X2 = civil works -0.424 0.108 (-2.66) (3.74) X3 = M&R 2.638 (3.50) Note: (t-statistic) As can be seen in this table, the coefficient for new constructions and capital repairs on buildings (Xl) does not change significantly from one model to the other. The analysis and discussion in the subsequent section use the results of the second and third model. D4.2. Net Output Method In this method, the growth rate of the construction value added is regressed against the GDP growth rate (which was shown earlier in Figure 1). From this analysis, we see that a 1% GDP growth results in a 1.54% growth in the construction value added. Estimation of the Capacity of the Industry We assume that the maximum output attainable in the short term (the coming 2 to 3 years) equals the maximum output produced during the last decade. The following justify this assumption: (i) the stock of capital is the sizing factor to determine the capacity to undertake construction works. This means that the substitution of labor to capital is only efficient for small scale works, and given the structure of the industry described under section 2.1, this assumption is justified; (ii) in the short term, the stock of capital is assumed to be fixed meaning that no new equipment will be purchased. This assumption is justified by the length of the delay in importing foreign equipment and the severe financial constraints facing the industry; (iii) the capacity of the industry to undertake works is therefore capped in the short run by the available stock of equipment; Includes highway new construction, reconstruction, repairs & maintenance. (iv) the maximum volume of works was undertaken with a certain stock of equipment which was probably used at its maximum capacity and which has certainly been depleted for the lack of maintenance?; and (v) it is therefore a reasonable assumption to consider the maximum output attained as a cap of the feasible output. By inspection of the data we find the maximum net output was reached in 1988 (corresponding to Alan Garcia's investment program) and was worth $ 451 million in 1991 constant terms (see Figure 3). In 1982, the industry came close to this capacity with a net output value of $ 439 million. Estimation of the Net Output of the Construction Industry The supply of construction works at any given point in time is the sum of buildings, civil works and maintenance and road repair works deriving from public and private demand. From the Gross Output Method described above, we found that changes in GDP capture the variation in demands on the construction sector deriving from new construction and capital repairs on buildings. Public investments in road construction, reconstruction, maintenance and repairs, due to their historical fluctuations in both level and structural allocation, appear as a perturbation on the net output of the construction industry. We therefore estimate the total net construction output by adding the following two components: (i) estimated demand from the building sector as derived from GDP growth; and (ii) public demand obtained from undertaken and planned investment in the road sector. The results are shown in Figure 3, which demonstrates that the capacity of the construction industry will be reached in 1994 if the proposed investment in the 1993-1997 budget is undertaken. This point is reached two years earlier than when a moderate' growth in planned investment is assumed. D4.3. The Input Method The capacity of the industry is also estimated from its aggregate production function, as described below (see Andersson, T.D., Profit in Small Firms, Avebury, 1987): Y = f(K,L,M) where: (i) Y is the net construction output measured in terms of value added; (ii) K is the stock of capital including equipment; This is exemplified by the state of the fleet of mechanical equipment for carrying out force-account road maintenance and rehabilitation. Out of 346 units of heavy machinery (tractors, ,mixers, pavers etc), about 75 percent of them (258) are inoperable. The situation is worse in the case of mobile equipment (platform trucks, tankers, trailers etc) where only 58 percent out of a fleet of 162 vehicles is operable. The condition of the private fleet is also poor due to the lack of construction activity in the last decade. By moderate we mean a steady growth in investment rather than the planned doubling after every year since 1992. (iii) L is the labor force measured in thousands of employees; (iv) M is the consumption of materials such as cement and bitumen; and (v) fo is the set of feasible production possibilities, including substitution between capital and labor. The production function approach assumes that capital is fixed in the short term. This assumption was justified earlier under section 4.2. Evidence supporting this assumption in the case of the construction industry derives from the limited capacity of a firm to purchase new equipment in the short run, as was mentioned under 3.2. Furthermore, we assume that in the short run, the materials used are non substitutable. Figure 4 provides a justification for this assumption. As can be observed from this figure, in the last decade, the fluctuations in the production rates of bitumen and cement and construction value added have been closely correlated. Such a pattern suggests that there has been no or relatively weak substitution in the materials used for construction. This leaves labor as the only input which affects the construction value added in the short run. Labor statistics in Peru by economic sector from 1950 to 1992 were used to evaluate the effect of labor inputs on construction value added. Table D4 compares the construction value added to the size of the labor force employed in the construction industry over the last 20 years. Table D4: Construction Value Added and Labor Force 1970 1975 198_ 1985 1990 Construction Value 15T 244 225 254 247 Added (million 1979 Nuevo Soles) Labor force 171 198 225 244 280 (thousand) I I I I I Source: Series Estadisticas, 1970-1991 and Cuentas Nacionales, 1950-87, Peru en Numeros, 1992, Anuario Estadistico. The production function using these data is presented in Figure 5. As can be seen from this figure, the actual value added fluctuates between the maximum capacity (defined as the highest net output from the industry in the last 20 years) and the estimated value added. This result indicates that over the last 20 years, the construction industry in Peru, has operated above the production feasibility set. However, the industry has only reached the maximum capacity at two points (1982 and 1988). To reach capacity requires a lot of labor, and consequently higher salary costs and costs of formal employment. Since the actual value added is always above the production function, it appears that the industry has been able to employ more people at lower costs or higher productivity. The estimated production function in Figure 5 acts as a long-run lower bound, and is always above the total cost of labor. This result is in concordance with the effect shown in Figure 2 suggesting that the construction industry has operated at significant profits, and has managed to pass on the costs of the inflation and other costs on to its clients. It also appears from Figure 5 and Table D4 that the industry has been able to increase its labor force while decreasing the total cost of labor. This is explained by the dramatic decline in real wages (except in the informal sector), illustrated in Table 5. Table D5: Real Wages in Lima from 1980-1992 in 1991 Nuevo Soles 1980 1984 1988 1992 Minimum Wage 2 6. 125.7 51.8 Wages(a) 9 865. 722.8 588.2 Daily Wages 20.7 14.5 11.4 7.6 Income Index from 97.9 108.6 176. 117.2(b) Informal Services (a) Weighted average of employees and employers (b) 1991 value Sources: Peru en Numeros, 1992 D5. Summary of FIndings and Recommendations D5.1 Capacity of the Construction Industry The results of the analysis in Section D4 show that, under the current structure the Peruvian construction industry does not have the capacity to meet the investment demands presented by the proposed 1993-1997 budget. The demands from this budget can only be met through structural changes in the industry. Such changes could derive from factors such as the share of construction works undertaken through force account and the military, the participation of foreign contractors in small size works, and the availability of financing to firms in the construction industry. Heavier reliance on the public sector through force account work will be impossible due to the erosion of the capacity of this sector as a result of the collapse of the state in the late 1980's and the staff reductions undertaken to reform the public sector. As a result reliance will have to be placed on the private sector. Under this arrangement, the public sector would have to manage numerous contracts with local firms, requiring efficient administrative and institutional arrangements. The administrative capacity to manage the planned investment under whichever role the public sector would have is a serious concern. In any case, strengthening of the public sector as well as assistance to the local construction industry will be required to allow them to meet the demands of the planned investments efficiently and in a cost effective manner. A possible assistance structure could follow the AGETIP model, whereby a private agency would: (i) manage contracts privately executed for the public sector (both central and local government agencies); (ii) coordinate and manage the allocation of works to the formal and informal sectors; and (iii) strengthen the contractors through training. As other Peruvian industries, construction firms face severe constraints that impede modernization. A recent World Bank report titled Peru: A Private Sector Assessment, December 1993 cites political uncertainty as a major worry of the business community. Other factors cited include financing: real interest rates are high, close to 3% a month and maturity is short, from 90 to 180 days. These constraints would also need to be addressed. D5.2 Further Work This study limited its scope to the assessment of the capacity of the Peruvian construction industry to undertake works planned in the proposed 1993-1997 budget. Additional issues that need to be considered in order to formulate a strategy and action plan to be undertaken by the Government and the industry include but are not limited to: (i) an assessment of the procedures for procurement in the public and private sector, particularly regulations for bidding, contractor registrations and negotiation of construction and maintenance works; (ii) an evaluation of contracting practices and their suitability for a fragmented and non modernized construction sector; (iii) a review of the policy and legal framework affecting activities in the construction sector; and (iv) a survey of the ownership and organization of construction enterprises: state (force account and parastatal) and private (individual, family, corporate) particularly outlining the problems and constraints affecting the efficiency of each type of organization. D6. Reference 1. Andersson, T. D., Profit in Small Firms, Avebury, 1987 2. Henriod, E., The Construction Industry: Issues and Strategies in Developing Countries, The World Bank, 1984 3. Instituto Nacional de Estadistica e Informatica, Peru, Series Estadisticas, 1970-1991 4. Instituto Nacional de Estadistica e Informatica, Peru, Cuentas Nacionales del Peru, Sectores Institucionales, 1950-1991 5. Lubell, H., The Informal Sector: In the 1980's and 1990's, OECD, 1991 6. United Nations, Construction Statistics Yearbook, 1983 7. United Nations, Construction Statistics Yearbook, 1983 8. United Nations, Industrial Statistics Yearbook, 1990 9. Webb, Richard and Fernandez Baca, Graciela de Valdez, Peru en Numeros, 1991, Anuario Estadistico, Lima: Cuento S. A., May 10. Webb, Richard. and Fernandez Baca, Graciela de Valdez, Peru en Numeros, 1992, Anuario Estadistico, Lima: Cuento S. A., May 11. World Bank, Peru: Strategy for the Infrastructure Sectors, LAlIN, Yellow Cover Report # 12412, June 1993 12. World Bank, Peru: A Private Sector Assessment, LA1CO, Green Cover Report #12096, December 1993 Figure 1: GDP Growth Rate and Growth in Construction Output 1971-1992 10 ------~-- - - -- 0 02 -30 - -- - - -40 - -- year GDP ---o- Construction output Figure 2: Construction Value Put in Place and Value Added in Construction 1,200,000 -- L1,000,000-- 800,000 400,000 C= 05 200,000 0 0 100,000 200,000 300,000 400,000 500,000 600,000 VaLue Added in Construction (mittion current SoLes) Figure 3: Impact of Road Investments on Capacity of the Construction Industry 6.00E+08 - 5.OOE+08 -1 ---- Cd ac ty 4.OOE+08- 3.00E+08- E 2.O0E+08- 1.00E+08 - 1 .OOE+008 1971 1972 1973 1974 1975 1976 1977 1978 1979-1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 year O Construction Value Added U Value Added from Road Investment Figure 4: Growth in Construction Value Added, Cement and Bitumen Production 60.00 - - 40.00 - - 20.00 0.00 1 1983 1984 19 1986 197 98 1989 1 *20.00 -40.00 *60.00 year Bitumen 0 Cement - Construction Value Added Sources: Industrial Statistics Yearbook 1990, Volume II: Commodity Production Statistics, 1981-1990; Series Estadisticas, 1970-91 and Cuentas Nacion.' s, 1950-87 reported in Peru: A Private Sector Assessment, LA1CO, The World Bank, June 1993 Figure 5: Measuring the Capacity of the Peruvian Construction Industry 1980-1991 350 - - 300 200 150 100- Cl 50 0- I1 I I 1980 1982 1984 1986 1988 1990 1992 Estimated value added Total Labor Cost - Maximum capacity -o- Actual Value Added PERU - Public Expenditure Review ANNEX E Detailed Analysis on Water and Sanitation Introduction El. The objective of this annex is to review the 5-year investment program for the water and sanitation sector presented to the Bank for the Government of Peru. The program provided by the GOP is poorly supported, and at the present time they can do very little to improve it. According to our knowledge, specific investments proposals are mostly a partial "wish list" presented to the Ministry of the Presidency, by Sedapal and some other regional companies from the already extinct Senapa system. We have also reviewed the investment programs for other regional companies and municipal services, which are presented separately in the regional government investment program. However, we could not conclude whether investment programs for rural areas (32% of the population) were included in this regional program or not. In general, both central and regional investment programs are not evaluated consistently and they do not correspond to a country-wide strategy for the sector. We also found some possible overlaps between the central and regional government programs. Future sector work with the Ministry of the Presidency (as the new top sectoral institution) and IDB should help the government formulate a comprehensive strategy for the sector, and eliminate such inconsistencies and lack of clarity in its investment programs. Sector Background E2. The cholera epidemic in Peru has brought to light the miserable condition of its water and sanitation sector. As water and sewerage conditions have deteriorated in the past decade, the incidence of waterbome diseases has increased significantly, and became endemic among the poor. Service coverage by 1991 was 58% in water supply and 42% in sewage collection, as reported by the Interamerican Society of Sanitary Engineers (AIDIS). Similarly, in urban centers (population greater than 2000) coverage reaches 74% and 62%, in water and sewerage respectively. However these figures are not supported by reliable statistics and could be grossly misleading. Sewage treatment is almost nonexistent. Domestic, industrial, and mining wastes are dumped untreated into rivers and along the Pacific Ocean coast. E3. Service coverage figures, as understood in Peru and in other countries in Latin America, are related to territorial network coverage and hide dramatic disparities on the continuity of the service, and in overall service quality indicators. The quality of service has deteriorated steadily as water service is not being provided on a continuous basis and water rationing prevails in major urban centers like Lima-Callao. A survey carried out in 1989 revealed that 30 percent of the population in urban centers other than the three major cities, had a water service less than 10 hours a day and only 20 percent had service more than 20 hours a day. The steady degradation in service quality in the last years is closely related to the level of infrastructure investment. During 1985-1989, investments in water and sanitation infrastructure represented only 0.15 percent of GDP, one of the lowest in Latin America, where countries with better service have been investing around one percent of GDP. Ironically, the first ten weeks of the cholera outbreak represented US$1 billion in emergency health expenditures and lost earnings from exports and tourism. This is approximately four times the investment on the water and sewerage sector between 1981 and 1988. E3. The sector long tradition of poor operational performance is the result of wrong, economic and financial, policy choices. The economics of the water supply and sewerage industry were not recognized, and the financial health of the sector was not pursued. Sector revenues have been insufficient to cover operations and finance investments. Service tariffs have been set following short sighted political and social criteria, in general, their levels were insufficient to cover operating costs. Sector companies survived by sacrificing service quality, where the reduction of maintenance expenses, was one of the first targets. Paradoxically enough, labor cost was not reduced, it increased in absolute terms, consuming a larger share of operating cost, reaching 70% in some companies. Cash deficits created by current operations have been covered through government transfers, delayed payments to suppliers, and failure to service debt. Capital investments have been financed, mainly, from user contribution (new neighborhoods and Pueblos Jovenes), donors, and to a lesser extent government transfers and concessionary loans from FONAVI. E4. The institutional set up for the sector is changing rapidly. The regionalization process initiated in 1989 has transformed the water sector institutional organization, shifting responsibilities from the central government to the regional and local level. The Servicio Nacional de Agua Potable y Alcantarillado (SENAPA), created in 1981 under the Ministry of Housing and Construction (MHC) to operate or supervise sector operations in urban centers, had transferred, with the exemption of Lima-Callao and Trujillo, its subsidiary companies and operational units to local governments. SENAPA, acting as a holding company, supervised the operations of 14 subsidiary companies in charge of operations in the largest cities or departments, and operated directly the services in 10 departments through 10 Operational Units which provided services to small and medium size towns. The SENAPA system, formally abolished in December 1992, comprised 35 percent of the urban centers and 90 percent of the urban population. Some 185 municipalities, representing about 10 percent of the total urban population, administered and still administer their services directly or through cooperatives or other forms of management. Rural communities (below 2000 inhabitants) and disperse population comprising 32% (6.8 million) of total population were once assisted by the National Rural Sanitation Program under the Ministry of Health, but the Program has long been abolished. Because of the low social expending in rural areas in the last few years, service to this segment of population is estimated to be the worst. E5. Since late 1992, the responsibility for the water and sanitation sector in urban areas, including the function of setting national policies, have been transferred to the Ministry of the Presidency (MOP). The Ministry of the Presidency is also responsible for Lima's Water Utility Company (SEDAPAL). Responsibility for defining and enforcing water quality standards continues with the Ministry of Health (MH). The Ministry of Agriculture (MOA) is responsible for water resources management. In December 1992, the Government enacted legislation creating the Superintendencia Nacional de Servicios de Saneamiento (SNSS) reporting, administratively, to the Viceminister of Infrastructure of the Ministry of the Presidency. The SNSS is conceived as the new apex sectoral and regulatory institution, however, its financial autonomy and management independence to perform its regulatory functions, it is not yet clearly established. The Superintendency is initially staffed with about 20 employees transferred from the abolished SENAPA, and its been financed through a contribution of up to 3% of water companies revenue, mainly from SEDAPAL. ES. The SNSS has assumed the responsibility for preparation of sector laws and regulations. A new basic law of sanitation services has been prepared and its approval is expected by January 1994. Several essential bylaws and decrees, to complete sector regulatory framework, are in advanced degree of preparation, and their presentation to the Congress has been scheduled. Detailed regulatory instruments are under preparation in the following areas: (i) tariffs; (ii) private sector contracting; and, (iii) service and environmental standards. However, while the new legislation has been implemented, the highly ineffective, Comision Reguladora de Tarifas de Agua Potable y Alcantarillado (CORTAPA) still regulates the tariff setting process. CORTAPA submits the tariff adjustment to the Ministry of Economy and Finances (MEF) for final approval. Application of the new sector law implies the elimination of CORTAPA, and a new tariff setting system with participation of SNSS, Municipalities, service operators and consultants. Tariff levels will be approved, ultimately, by municipalities. Sector Finances E6. A gross estimate of the order of magnitude of sector financial requirements could be calculated as follows. Peru's population requires for consumption approximately 1100 Hm of water per year, including 40% of physical losses. Operation and maintenance could be estimated at about US$120 million (US$O.11/M). Investments could be estimated at approximately 0.75% of GDP or US$200 million. Investment levels and maintenance of operational assets in the last 10 years have been minimal, therefore the backlog of investments, neglect and growth induced, are sizeable. These three components add up to approximately US$2 billion, a conservative figure, considering the long neglected sector investment. We believe the government's investment program is much too small, even compared with this conservative estimate. The investment program presented by the government is attached to this note for comparison (see Table El). Estimated Sector Revenue Requirements (1994-1998)' (US$ million) Operation & Maintenance 600.0 Investment 1000.0 Estimated Backlog of Rehabilitation A., Total 2000.0 E7. The investment program includes two IDB projects. The first for approximately US$75.0 million is targeting institutional development (80% of the project) for approximately 15 regional water companies outside Lima. Consultants to assist the GOP ( Proyecto Especial Programa nacional de Agua y Alcantarillado) were hired in December 1993, after a delay of more than two years. The project is scheduled with priority A for Board approval at the end of 1994. The second IDB project in the pipe line will prioritize investments in rehabilitation and expansion of selected water and sewerage systems. 1996. E8. The investment program includes the proposed US$260.0 million project "Lima water Privatization and Rehabilitation Project", scheduled for Bank Board approval in July 1994. A ' At December 1993 US$ second WB proposed financed project (US$300.0 million, FY 97) to reduce current pollution levels in the Lima metropolitan area and improve water management practices is only partially included. Projected Disbursement (1994-1998) (IDB + IBRD) Projects (US$ million) 11DB 1 75.0 IDB 2 200.0 IBRD 1 200.0 IBRD 2 Total 500.0 E9. Additionally, grants from internatonal contributors (Germany, France, Canada, and Japan) and concessionary financing and Grants from FONAVI and FONCODES will contribute, particularly in the poorest areas. We estimate for period 1994-1998 US$125 million from donors and US$250.0 million from concessionary financing (FONAVI and FONCODES) E10. Based on the previous (an estimated) figures, the balance of revenues and expenditures for the sector shows the extraordinary importance of revenue generation from tariffs, which is something that is frequently overlooked. The new sector law and regulation under preparation should ensure sufficient sector revenues to match the investment needs. Financial independence of the water companies, in turn, will reduce the financing burden of the government. Summing up: Revenue (US$ million) Tariff Revenue 1100.0 2 Multilateral Institutions 500.0 Fonavi 250.0 Grants 100 Sub-Total 1950.0 Expenditures O&M 600.0 Investment 1000.0 Rehabilitation A0.Q, Sub-Total 2000.0 2based on US$180.0 million for 1994, and yearly increases of approximately 10% thereafter Setting Investment Priorities El1. With limited financing resources, investment priorities must be set clearly to assure effective investments.Recent studies in Lima had demonstrated that investments priorities should be concentrated in rehabilitation of existing physical systems and in water demand management (particularly in the coastal cities). Apparently, most water supply systems in Peru have capacity in excess of an efficiently managed water demand. However, the poor maintenance an inadequate commercial practices had conduced to large amount of losses and wastage, resulting in a chronicle shortage of water. For example in the case of Sedapal, it is estimated that 50% of the water is either loss or wasted. It is clear that reduction of unaccounted-for-water, if successful, can substantially eliminate the water shortage, while bringing more revenue to sector utilities without major investments in capacity expansion. In the past, most investment had concentrated in water and sewerage network extension and very little attention is given to their maintenance. Sewerage disposal systems are expensive and should be carefully evaluated and their priority should be adequately weighted. E12. In light of the previous discussion, criteria to set investment priorities should be emphasized: I programs to develop the institutional and commercial capacity to generate revenue o programs for assisting the poor population in cities and rural areas o concentrate on rehabilitation rather than expansion o projects for final disposal of sewage should be carefully evaluated Table El Summary of 5 Year Investment Program (Central and Regional) In the Sanitation Sector Presented by the Government of Peru nstitutional Level 1993-1997 Investment PtOA 1993 1994 1995 1996 1997 Total Central Government (IDB projects) 1.1 34.8 24.7 89.0 115.0 264.6 Regional Government 1/ 1.6 6.0 36.9 53.8 43.8 142.0 Companies (central and regional) 77.8 142.3 142.1 152.7 97.0 612.0 Irotal 80.5 183.1 203.7 295.5 255.8 1018.6 1/ There seems to be some overlap between central and government investment programs. Some the regional government programs are also included in the companies' investment program administered by the central government. CATALOGUERSIFOLE s FP MROGAgPHMS CONFIDENTIAL CONFIDENTIAL Report No.: 17808 PE Repoit No.: 17808 PE Type: ER Type: ER