Report No. 30379-CO Colombia Recent Economic Developments in Infrastructure (REDI) Balancing Social and Productive Needs for Infrastructure (In Two Volumes) Volume II: Main Report November 1, 2004 Finance, Private Sector and Infrastructure Unit Latin America and the Caribbean Document of the World Bank DISCLAIMER This volume i s a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed inthis paper do not necessarily reflect the views of the ExecutiveDirectors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. ABBREVIATIONS AND ACRONYMS APTDEA Andean Trade Promotion and Drug Acta de Preferencias Arancelarias Andinas y de Eradication Act Erradicacidn de Drogas BOMT Build-Operate-Maintain-Transfer Construir-Operar-Mantener-Transferir BOT Build-Operate-Transfer Construir-Operar-Transferir BRT Bus RapidTransit Sistema Integrado de Trarisporte Masivo CAF Andean Development Corporation CorporacidnAndina de Foment0 CAR Autonomous Regional Corporations Corporaciones Autdnomus Regionales CNG Compressed Natural Gas Gas Natural Comprimido CONPES National Council of Economic and Social Consejo Nacional de Politica Economicu y Social Policy CORMAG Magdalena Valley Autonomous Regional CorporacidnAutdnoma Regional del Rio Grande Corporation de la Magdalena CRA Regulatory Commission for Water and Comisidn de Regulacibn de Agua Potable y Sanitation Saneamiento CFWG Regulatory Commission for Electricity and Comisidn de Regulacidn de Electricidad y Gas Gas CRT Regulatory Commission for Comisidn de Regulacidn de Telecomunicaciones Telecommunications DANE National Administrative Department of Departamento Administrativo Nacional de Statistics Estadistica DGM General MaritimeDirectorate Direccidn General Maritima DNP National Planning Department Departamento Nacional de Planeacidn EAAB Bogota Water and Sewerage Utility Empresa de Acueducto y Alcantarillado de Bogota' EAP East Asia and Pacific Asia del Este y Pacqico EEB Bogota Electric Utility Empresa Elkctrica de Bogota' EMCALI Cali Municipal Enterprises Empresas Municipales de Cali EPM Medellin Public Enterprises Empresas Phblicas de Medellin ETB Bogota Telecommunications Utility Empresa de Telecomunicacionesde Bogota' FAER Rural Electrification Fund Fondo deAsistencia a Electrificacidn Rural FAZNI Non-Interconnected Zones Fund Fondo de Asistencia a Zonas No Interconectadas FC Communications Fund Fondo de Comunicaciones FDI Foreign Direct Investment Inversidn Extranjera Directa FINDETER Municipal Development Bank Financiera de Desarrollo Territorial FIS Social Investment Fund Fondo de Inversidn Social FNR National Royalty Fund Fondo Nacional de Regalias FONADE National Fundfor Development Projects Fondo Financier0 de Proyectos de Desarrollo FSSRI National Solidarity Fund Fondo de Solidaridad pura Subsidiosy Redistribucidn de Ingreso FTA Free Trade Agreement Acuerdo de Libre Comercio GAP1 Investment Project Support Group Grupo de Apoyo a Proyectos de Inversidn GDP Gross Domestic Product Product0 Interno Bruto IDB Inter-AmericanDevelopment Bank Banco Interamericano de Desarrollo INCO National Institute for Concessions Instituto Nacional de Concesiones INVIAS National Roads Agency Instituto Nacional de Vias IIRSA Initiative for the Regional Integration of Iniciativapara la Integracidn de la Infrastructure in South America Infraestructura Regional en Suramerica ISA National Electric Transmission Utility Interconexidn Ele'ctrica ISAGEN National Generation Utility Generacidny Comercializacidn de Energia ITU InternationalTelecommunications Union Unidn Internacional de Telecommunicationsunicaciones LPG LiquidPetroleum Gas Gas Licuado de Petroleo MC Ministry of Communications Ministerio de Comunicaciones MD Ministry of Defense Ministerio de Defensa MDG Millenium Development Goals Metas de Desarrollo del Milenio MEM Wholesale Electricity Market Mercado Ele'ctrico Mayorista MME Ministry of Mines and Energy Ministerio de Minas y Energia MMAVDT Ministry of Environment, Housing and Ministerio de Medioambiente, Vivienda y Regional Development Desarrollo Territorial MT Ministry of Transport Ministerio de Transporte OECD Organization for Economic Co-operation Organizacion para la Cooperacidn y el andDevelopment Desarrollo Economic0 OPS PanAmerican Health Organization Organizacidn Panamericana de la Salud PPA Power Purchase Agreement Acuerdo de Compra de Energia PRAP Public Administration Reform Program Programa de Renovacidn de la Administracidn Publica PSP Private Sector Participation Participaciondel Sector Privado SIC Superintendence of Industry and Commerce Superintendencia de Industria y Comercio S I N National Interconnected System SistemaInterconectado Nacional SISBEN Beneficiaries Selection System Sistema de Seleccidn de Beneficiarios SPR Regional Port Society Sociedad Portuaria Regional SPT Superintendence of Ports and Transport Superintendencia de Puertos y Transporte SSPD Superintendence of Public Utilities Superintendencia de Servicios Phblicos Domiciliarios UAEAC Civil Aviation Administration Unit Unidad Administrativa Especial de la Aeronautica Civil UPME MiningandEnergy Planning Unit Unidad de Planeacidn Minero-Energe'tica VAT Value Added Tax Impuesto a1 Valor Agregado WHO World Health Organization Organizacion Mundial de la Salud ZNI Non-Interconnected Zones Zonas No Interconectadas Acknowledgements This report was edited by Vivien Foster based on contributions from a team of Bank staff and consultants during January-June 2004, and with close cooperation from a team of government counterparts from DNP. It also draws on discussions with a wide range of individuals in government agencies, private firms, financial institutions, think tanks, academia, and international organizations. A list of the team members who contributed to the various chapters and annexes follows below. In addition, Marc Heitner and Bjorn Wellenius acted as advisers to the team on issues relating to the telecommunications and natural gas sectors respectively. Chapter 1: Vivien Foster and Tito Yepes Chapter 2: Jose Barber0 Chapter 3: Vivien Foster with inputs from Tito Yepes Chapter 4: Rafael Herz with inputs from Tito Yepes Chapter 5: Clemencia Torres The team acknowledges the extremely close collaboration of government counterparts from the Infrastructure and Energy Directorate of DNP, under the leadership initially Alejandro Alvarado and later of Mm'a Constanza Garcia. The capable coordination of Francisco Bema1 and Daniel Torres are gratefully recognized. Moreover, the team acknowledges the participation and collaboration of Lyda Esquivel, Hernan Molina, Jean-Phillipe Penning, Mauricio Millan, Guillermo Cruz, Erika Bibiana Pedraza, Carlos David Beltran, Alejandro Castaiieda, Margarita Caycedo, Juan Camilo Granados, Cesar Peiialoza, Leonard0 Diez, Adriana Perez, Juan Esteban Gallego and PaulaNieto. The report is closely based on ten background papers. These include five sectoral reports on electricity (Carmenza Chain and Juan Manuel Rojas), gas (Luis August0 Yepes), telecommunications (Ciro Mendoza), water (Diego Fernhndez), and transport (Germhn Ospina) as well as six reports on cross-sectoral thematic issues, namely competitiveness (Jaime Maldonado and Giovanna Sardi), cross-subsidies (Marcela Melendez), rural infrastructure policy (Mauricio Cardenas and Marcela Melendez), public finance for infrastructure (Israel Fainboim and Carlos Rodriguez), private finance for infrastructure (P. S. Srinivas) and institutional and regulatory issues (Carlos Caballero, Manuel Ramirez, and Alejandro Jadresic). All of these background papers are available to the public and can be downloaded from the following website: www.worldbank.org/redicolombia The team would like to thank the peer reviewers JosC Luis Guasch, Steven Webb, and Cecilia Briceiio-Garmendia for providing useful comments. The report also benefited from helpful comments and discussions with Eva Maria Uribe, Cesar Pifieros, Claudia Mora and many other persons from the SSPD; Cesar Caballero, Sergio Acosta and Adriana Ruiz from DANE; Ismael Concha and Jose Vicente Dulce from UPME; Maria Inks Agudelo, Jorge Cardona, Andrds Arias, Alejandro Gamboa, and Luis Fernando Villota from the Ministry of Finance; and Manuel Jose Maiguashca from the Ministry of Miningand Energy. Finally, it is important to note that the analysis and recommendations presented in this document remain the opinion of the World Bank, and are not in any way intended to represent the views or policies of the Colombian authorities. Table of Contents 1 . COLOMBIA'S INFRASTRUCTUREININTERNATIONALPERSPECTIVE ...........1 INTRODUCTION....................................................................................................................... 1 INFRASTRUCTUREFINANCING........................................................................................... 2 INFRASTRUCTUREAND SOCIAL OBJECTIVES.............................................................. 10 INFRASTRUCTUREAND GROWTH ................................................................................... 16 CONCLUSIONS....................................................................................................................... 24 2 . THE CONTRIBUTIONOF INFRASTRUCTURETO ECONOMIC GROWTH ......26 INTRODUCTION..................................................................................................................... 26 THE CURRENTSTATUS OFINFRASTRUCTURE............................................................. 27 Electricity.............................................................................................................................. 27 Generation........................................................................................................................ 29 Transmission..................................................................................................................... 30 Distribution....................................................................................................................... 31 Retail................................................................................................................................. 34 Natural gas ............................................................................................................................ 35 Telecommunications............................................................................................................. 38 Water..................................................................................................................................... 41 Transport............................................................................................................................... 44 A VIEW FROMTHE PRODUCTIVESECTORS................................................................... 50 The importance of infrastructurefor the productive sectors ................................................. 51 The incidence of infrastructurein selected value chains....................................................... 55 TOWARDS AN INFRASTRUCTUREFOR GROWTHAGENDA ....................................... 59 A growthbasedvision of infrastructureneeds...................................................................... 59 Strategic prioritiesby infrastructuresector........................................................................... 62 Infrastructureinvestment plan .............................................................................................. 68 Electricity ......................................................................................................................... 68 Transport.......................................................................................................................... 70 Other sectors .................................................................................................................... 73 CONCLUSIONSAND RECOMMENDATIONS.................................................................... 75 3. THE ROLEOF INFRASTRUCTUREINTHE ACHIEVEMENT OF SOCIAL OBJECTIVES .............................................................................................................................. 79 INTRODUCTION..................................................................................................................... 79 ACCESS TO INFRASTRUCTURESERVICES ..................................................................... 80 Policiesto promote access .................................................................................................... 89 Electricity ......................................................................................................................... 89 Natural gas....................................................................................................................... 92 Telecommunications ......................................................................................................... 95 . Water ................................................................................................................................ 99 Transport........................................................................................................................ 101 Cost of reachinguniversalaccess ....................................................................................... 105 AFFORDABILITY OF INFRASTRUCTURESERVICES ................................................... 109 Utility services .................................................................................................................... 109 Evaluation of cross-subsidy scheme............................................................................... 109 Simulation of reforms ..................................................................................................... 124 Urbantransportation ........................................................................................................... 127 CONCLUSIONSAND RECOMMENDATIONS.................................................................. 128 4. LEGAL. REGULATORYAND INSTITUTIONAL FRAMEWORK FOR INFRASTRUCTURE ................................................................................................................ 132 OVERVIEW OF LEGAL. REGULATORYAND INSTITUTIONALFRAMEWORK.......132 INTRODUCTION................................................................................................................... 133 Legalframework ................................................................................................................. 134 Line ministries ................................................................................................................ 135 Regulatory Commissions................................................................................................ 136 Superintendenceof Public Utilities (SSPD)................................................................... 139 CROSS-CUTTINGREGULATORYISSUESFOR UTILITIES .......................................... 141 Conflicts of interest within the State................................................................................... 141 Fuzzy boundaries betweenpolicy andregulation............................................................... 142 Intemal structure of the Regulatory Commissions.............................................................. 145 Lack of a regulatory appeals channel.................................................................................. 147 Problems faced by the SSPD .............................................................................................. 149 Intemal structure of the SSPD........................................................................................ 149 Supervision of utilities .................................................................................................... 150 Intervention of failing enterprises .................................................................................. 152 Anti-trustregulation............................................................................................................ 154 SECTOR-SPECIFIC REGULATORY ISSUES FOR UTILITIES ........................................ 157 Electricity............................................................................................................................ 157 Lack of competition ........................................................................................................ 1.58 Weak incentives to invest in new plant ........................................................................... 159 Financial distress of distribution utilities....................................................................... 160 Naturalgas .......................................................................................................................... 160 Telecommunications........................................................................................................... 161 Barriers to competition................................................................................................... 162 Artificial legal dichotomy ............................................................................................... 162 Complex institutionalframework ................................................................................... 163 Water...........:....................................................................................................................... 164 Decentralization............................................................................................................. 164 Tariff regulation ............................................................................................................. 16.5 Eficiency, quality and competition................................................................................ 166 Role of SSPD .................................................................................................................. 167 TRANSPORT SECTOR FRAMEWORK .............................................................................. 168 Legalframework................................................................................................................. 168 Institutionalframework....................................................................................................... 169 Regulatory framework ........................................................................................................ 172 CONCLUSIONS AND RECOMMENDATIONS.................................................................. 176 Cross-cuttingrecommendations.......................................................................................... 177 Sector-specificrecommendations ....................................................................................... 179 5. PUBLICAND PRIVATEFINANCING FORINFRASTRUCTURE ......................... 182 INTRODUCTION................................................................................................................... 182 PUBLICFINANCING............................................................................................................ 183 Review of the adequacy, efficiency andequity of public spending.................................... 183 Electricity ....................................................................................................................... 187 Natural gas..................................................................................................................... 189 Telecommunications....................................................................................................... 189 Water.............................................................................................................................. 190 Transport........................................................................................................................ 192 Review of budgetary practicesfor financinginfrastructure................................................ 194 Public investmentprocedures......................................................................................... 194 Contingent liabilities ...................................................................................................... 197 Expandingfiscal space................................................................................................... 201 PRIVATE SECTOR FINANCEOF INFRASTRUCTURE................................................... 205 Past trends in privatefinancingof infrastructure.............................. ............................. 205 Electricity ....................................................................................................................... 206 Natural gas ..................................................................................................................... 208 Telecommunications....................................................................................................... 208 Water ........................................................................................ ................................ 209 Transport.................................................................................. ................................ 210 Prospects for continued privatefundingof infrastructureprojects..................................... 213 Credit enhancements...................................................................................................... 215 Domesticfinance ............................................................................................................ 216 INVESTMENT PRIORITIESAND FINANCINGNEEDS ....... ............................. 222 Electricity ....................................................................................................................... 226 Natural gas ..................................................................................................................... 227 Telecommunications.............................................................. ..................................... 228 Water .............................................................................................................................. 229 Transport........................................................................................................................ 230 CONCLUSIONSAND RECOMMENDATIONS.................................................................. 231 6 . A WAY FORWARDFORTHEINFRASTRUCTURESECTORS ............................ 235 INTRODUCTION................................................................................................................... 235 INTEGRATION OF RECOMMENDATIONS...................................................................... 237 PRIORITIZATION OF RECOMMENDATIONS.................................................................. 241 FINANCIAL IMPLICATIONS OFTHE REFORMAGENDA ............................................ 244 INSTITUTIONAL,RESPONSIBILITIESFOR THE REFORMAGENDA.......................... 246 List of Figures Figure 1.1: Overall financing for infrastructure investments. 1980-02............................... 3 Figure 1.2: Composition of infrastructure investment, 1999-00......................................... 4 Figure 1.3: Public investment ininfrastructure, 1980-02 .................................................... 4 Figure 1.4: Private investment in infrastructure, 1980-02................................................... 5 Figure 1.5: Evolution of country risk premium, 1987-03.................................................... 6 Figure 1.6: Composition of private investment, 1999-00.................................................... 6 Figure 1.7: Accumulated value of asset sales to private sector, 1988-99............................ Figure 1.8: Population agreeing privatizationbeneficial for country, 1998/02...................78 Figure 1.9: Percentage of market served by private operators ............................................ 9 Figure 1.10: Household coverage of social infrastructure, circa 2000.............................. 11 Figure 1.11: Household coverage gap between urban and rural areas, circa 2000 ...........14 Figure 1.12: Equity inhousehold coverage of basic infrastructure services, circa 2000 ..16 Figure 1.13 Endowments of productive infrastructure ...................................................... 17 Figure 1.14: UnderstandingColombia's energy consumption patterns ............................. 19 Figure 1.15: Comparisons of internet density, 1997/04 ................................................... 20 Figure 1.16: Prices of electricity and telephony services .................................................. 21 Figure 1.17: Labor productivity of telecommunications sector, 1996/01 ......................... 22 Figure 1.18: Quality of ports, 2002 ................................................................................... 23 Figure 2.1: Ownership concentration for different segments of the electricity industry...30 Figure 2.2: Evolution of retail electricity prices................................................................ 34 Figure2.3: Breakdown of telecommunications sector turnover........................................ 40 Figure2.4: Evolutionof fixed and mobile telephony........................................................ 40 Figure 2.5: Evolution of telecommunications tariffs ......................................................... 41 Figure2.6: Institutionalbreakdown of water sector.......................................................... 42 Figure2.7: Transportation by mode .................................................................................. 45 Figure2.8: Concentration of vehicle flows ....................................................................... 47 Figure 2.9: Perception of energy services.......................................................................... Figure 2.10: Perception of transport services inColombia relative to other countries .....53 54 Figure 2.11: Firmsatisfaction with services...................................................................... 55 Figure 3.1: Service coverage b y region ............................................................................. 82 Figure 3.2: Access to infrastructure services across income quintiles .............................. 83 Figure 3.3: Distance of households from all-season road.................................................. 87 Figure 3.4: Access to transportation infrastructure............................................................ 88 Figure 3.5: Accessibility of rural towns ............................................................................ 88 Figure3.6: Commuting behavior across quintiles........................................................... 102 Figure 3.7: Geographic distribution of benefits from Transmilenio as o f Dec. '03 ........103 Figure 3.8: Average expenditure on household utilities by income quintile...................110 Figure 3.9: Average utility subsidy receivedby income quintile.................................... 110 Figure 3.10: Weight of utility subsidy b y income quintile.............................................. 111 Figure3.11:Evolution of expenditure of subsistence basket of utilities......................... 112 Figure 3.12: Relationshipbetween strata quintiles .......................................................... 114 Figure 3.13: Physical consumption of services by quintile ............................................. Figure 3.14: Convergence towards target tariff ........................................................ : 117 ......116 Figure 3.15: Percentage tariff increases required to reach target levels.......................... 118 Figure 3.16: Financial balance of cross-subsidy schemes............................................... 119 Figure 3.17: Relative balance between beneficiaries and contributors ........................... 120 Figure 3.18: Distribution of population across strata ...................................................... Figure 3.19: Household expenditure and consumption of public transport services.......120 127 Figure 5.1: Total investment ............................................................................................ 183 Figure 5.2: Breakdown o f infrastructure investment ....................................................... 184 Figure 5.3: Composition of investment financing by sector............................................ 185 Figure 5.4: Operational results of loss-makingdistribution utilities ............................... 187 Figure 5.5: Joint venture investment by Telecom ........................................................... 190 Figure 5.6: Public investment intransport....................................................................... 192 Figure 5.7: Dispersion intransport investment across departments ................................ 193 Figure 5.8: Approved future budgetary appropriations ................................................... 197 Figure 5.9: Payment obligations related to PPA's in the electricity sector ..................... 199 Figure 5.10: Private participation in infrastructure.......................................................... 205 Figure 5.11: Asset allocation of pension funds, 2003 ..................................................... 219 Figure 6.1: Integratingrecommendations for growth and competitiveness .................... 239 Figure 6.2: Integratingrecommendations on social welfare and equity.......................... 240 Figure 6.3: Conceptual framework for prioritization of measures.................................. 242 ListofBoxes Box 1.1. Methodological issues with internationaldata on investment .............................. 3 Box 1.2: Internationalbenchmarkingof infrastructure ..................................................... 13 Box 2.1: Efficiency analysis of electricity distribution utilities ........................................ 33 Box 2.2: Evaluation of development of logistics............................................................... 48 Box 3.1: Impact of civil conflict on infrastructure coverage............................................. 85 Box 3.2: Will Colombia achieve the MDGs for water and sanitation?............................. 86 Box 4.1: SSPD experience with intervention of failing companies ................................ 153 Box 4.2: Colombian port reforms.................................................................................... 174 Box 5.1: The cost of government "rescue operations" .................................................... 188 Box 5.2: Debate on the treatment of public investment infiscal accounts...................... 201 Box 5.3: Macroeconomic Consequences of Using Public Capital Spending Costs to Reduce Budget Deficits ........................................................................................... 203 Box 5.4: Credit enhancement options.............................................................................. 216 Box 5.5: Methodology for the sectoral financial balances .............................................. 225 Listof Tables Table 1.1: Colombia's social infrastructure endowment relative to other countries. 200112 Table 1.2: Effort at expansion of social infrastructure over time ...................................... Table 1.3: Household coverage concentration coefficients. circa 2000 ............................ 14 Table 1.4: Endowment of productive infrastructure relative to other countries ................15 18 Table 1.5: Comparative performance of electricity distribution. 2001 ............................. 22 Table 1.6: Comparative survey on the quality of transport infrastructure. 2002...............23 Table 2.1: Final energy consumptionby source................................................................ 28 Table 2.2: Overview of vertical integration in the electricity industry.............................. 28 Table 2.3: Electricity generation by source ....................................................................... 29 32 Table 2.5: Market concentration in telecommunication services ...................................... Table 2.4: Performance of electricity distribution utilities................................................ 38 Table 2.6: Government ownership of telecommunications operators ............................... 39 Table 2.7: Performance of major water utilities ................................................................ 43 Table 2.8: Road network by category ................................................................................ 45 Table 2.9: Infrastructure intensity by productive sector .................................................... 51 Table 2.10: Top 10infrastructure intensive manufacturing activities............................... 52 Table 2.11: Infrastructure intensity for representative productive activities ..................... Table 2.12: Comparison of the three value chains ............................................................ 53 55 Table 2.14: Trade flows expected to increase or decrease after FTA ............................... Table 2.13: Infrastructureintensity in the primary growth-leading sectors ...................... 60 61 Table 2.16: Regional integration priorities........................................................................ 68 Table 2.15: Key areas for action on infrastructure services .............................................. 63 Table 2.17: Generation needs until 2011........................................................................... 69 Table 2.18: Summarypower sector investment needs ...................................................... 69 Table 2.21: Overall summary of investment needs. priorities and financing sources.......74 Table 2.20: Summary o f transport investment needs ........................................................ Table 2.19: Investment needs for primary road network................................................... 71 73 Table 3.1: Evolution of service coverage in urban and rural areas over time ...................81 Table 3.2: Percentage of customers with service expressing satisfaction with quality .....84 Table 3.4 : Likely delay inreachinguniversal access ....................................................... Table 3.3: Distribution of unserved population across categories..................................... 84 86 Table 3.5: Overview of electricity coverage patterns........................................................ 90 Table 3.6: Financialbalance of rural electricity funds ...................................................... 92 Table 3.7: Summary of universal access policy for electricity.......................................... 92 Table 3.8: Comparison of alternative gas subsidies on incentive to connect .................... 94 Table 3.9: Patterns of fuel usage across geographic areas................................................. 95 Table 3.10: Summary o f universal access policy for gas .................................................. 95 Table 3.11:Resultsof universal access programs............................................................. 97 Table 3.12: Summary of universal access policy for telecommunications ....................... 99 Table 3.13: Summary of universal access policy for water and sanitation ..................... 101 Table 3.14: Summary of universal access policy for transport ....................................... 105 Table 3.15: Cost of reaching universal access across sectors.......................................... 106 Table 3.16: Overall summaryof investment needs. priorities. and financing sources....108 Table 3.17: Distortions in socio-economic stratification over time ................................ 113 Table 3.18: Targeting performance of stratification system............................................ 115 Table 3.20: Summary of impact of reforms within current legal framework..................118 Table 3.19: Evolution of subsidy and surcharge rates..................................................... 126 Table 3.21: Targeting performance of urban transport subsidies.................................... 128 Table 4.1: Comparison of the three Regulatory Commissions ........................................ 138 Table 4.2: Overview of institutional framework for transport sector .............................. 171 Table 4.3: Overview of regulatory framework for transport sector................................. 174 Table 4.4: Summary of diagnostic and recommendations by sub-sector ........................ 176 Table 5.1: Overview of multilateral infrastructure finance 2001/15 ............................... 186 Table 5.2: Investment in local telephony by geographic area ......................................... 190 Table 5.3: Public investment in water 1998/01 ............................................................... Table 5.4: Historical behavior of budgetary appropriations for the transport sector.......191 196 Table 5.5: Expenditures from specific investment funds (budget execution) ................. 197 Table 5.6: Payment of guarantees under toll road concessions (US$m) ......................... Table 5.7: Evaluation of extent of commercial operation of public enterprises..............200 204 Table 5.8: Private participation inthe electricity sector .................................................. 206 Table 5.9: Private sector participation inthe natural gas sector ...................................... 208 Table 5.10: Private sector participation inthe telecom sector......................................... 209 Table 5.11: Private sector participation inthe water sector............................................. 210 Table 5.12: Private sector participation inthe transport sector ....................................... 211 Table 5.13: Risk allocation in successive "generations" of toll roads ............................. 211 Table 5.14: Performance improvementsinports following PSP .................................... 212 Table 5.15: Private investment intransport ..................................................................... 213 Table 5.16: Projects proposed for private participation................................................... 214 Table 5.17: Domestic commercial bank lending to infrastructure projects (excluding urban transport) ........................................................................................................ 217 Table 5.18: Exponential growth of pension assets .......................................................... Table 5.19: Maturity profile of fixed incoe holdings of pension funds, 2003 .................218 220 Table 5.20: Overview of social and productive investment financing needs..................222 Table 5.21: Public investment marginover maintenance and rehabilitation obligations224 Table 5.22: Financial balance for electricity sector: average yearly resources and investment for the 2004/08 period........................................................................... 226 Table 5.23: Financial balance for natural gas sector: average yearly resources and investment for the 2004/08 period........................................................................... 227 Table 5.24: Financial balance for telecom sector: average yearly resources and investment for the 2004/08 period........................................................................... 228 Table 5.25: Financialbalance for water sector: average yearly resources and investment for the 2004/08 period ............................................................................................. 229 Table 5.26: Financial balance for transport sector: average yearly resources and investment for the 2004/08 period........................................................................... 230 Table 5.27: Summary o f overall financial balance for infrastructure sectors: average yearly resources and investment for the period 2004/08 ......................................... 231 Table 6.1: Overview of recommendations by sector and b y theme ................................ 237 Table 6.2: Classification o f recommendations according to prioritization matrix ..........243 Table 6.3: Classification of challenges related to longer term recommendations ...........243 Table 6.4: Financial implications of the reform agenda .................................................. 245 Table 6.5: Institutional activities for the reform agenda.................................................. 247 Colombia At A Glance: RED1EnergyIndicators iFinal energy consumptior 4 no, Final energy consumption (Tcal '000s) 246.4 238.7 Energy consumptionper capita (Gcal) 6.2 5.6 Energy intensityof GDP (US$/kgoe) 4.1 Electricity consumption(GWh) 4,524 Natural gas consumption (Mm3) Number of players Electricity production I Electricitygeneration 4% I Electricitytransmission 9 Electricitydistribution 30 Electricity retail 36 Extent of PSP (Yo) Electricitygeneration 0 45 Electricitytransmission 0 8 Electricitydistribution 0 52 IGeneration llju l 6 l i G G d Electricity retail 0 45 Herfindahlindex Gas transoortation 8 Gas distribution 20 Gas retail 21 Extent of PSP (%) Gas transportation 0 Gas distribution 0 100 93 100 100 72 77 83 87 93 95 al interconnectedsystem terconnected zones 84 90 94 36 34 39 72 76 80 2 Access to natural gas (Yo) In proximityto distribution network 53 79 Among total urban population 40 58 Among total population 36 Average electricity connection charge (US$) 300 Average natural gas connection charge (US$) 400 ColombiaAt A Glance:RediEnergyIndicators (Cont) Evolution of accessto electricity I 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1993 1997 2903 I Price trends g 200 180 2.7 160 11.2 k p 140 120 0 100 2- c 80 60 ; 5 40 20 0 1994 1996 1998 2000 2002 -+e- Commercial +Residential lndustnal I I Affordability of elecbicity bills in urban areas 12% 7 0 With subsidy Distribution of operational performan- ~~~~~~ ~~~~ 40% Distnbutionlosses Colombia At A Glance: RED1Telecom Indicators Concentration of market power Other Celluiai LD LDinrl Local Total Value natlonal added MConcentration largest two el Heifindahlindex Public sector ownership 3 0 ~-.. Pricetrends r 251 Telecom expenditure 004 ?e E 0035 003 __-_______ <:; OM5 8 002 P 0015 001 ow5 0 I 2 3 4 5 h c m q u n t l e ColombiaAt A Glance: RED1TransportIndicators 1 Modalsplit of domesticfreight (ton- , k m ) ~ I i 79% owatemay 1 Modal splitof internationaltrade (tons) 1 4%- 1 Modal split of inter-urbanpassenger transport i Modal split of urban commuter transport (mns) % Public 0 Pnmte transpolt On fwt I i 2 3 4 5 "Journeys Incomewintile -CExpendllure ColombiaAt A Glance: RED1Water Indicators 1 ~ ~~~ Evolutionof access to water connections 100 T - ..I I 90 -' 80 - 70 - 60 - 50 - 40 ~ 30 - 20 - 10 - 0. Urban Rural National I ~ Evolution of access to improved sanitation 90 v) 80 2$ 70 60 B 50 5$D4 0 30 20 1a 10 0 Urban Rural National 1 I Colombia at a glance: RED1Water indicators(cont) 1 AWoidsbllllyotwater bills in urbanarea9 1 subsidy): All water consumers (Yoincome) 3.7 First quintilewater consumers(% Distributionof operational (15m3/m), performance All water consumers (Yoincome) 0.5 ....-..--. First quintilewater consumers ("1. Unaccountedfor water ("1) 40 Working ratio (%) 80 Meter coverage ("1.) 88 j Distributionof comnercial performance Labor costs (US$/m3) 0.07 iHouseholdswith continuouswater service ("10) 67 Customersexpressingsatisfactionwith service Revenue collectionefficiency J Urban areas ("A) 84 Rural areas (Yo) 75 Urbanwater effectively disinfected(YO) 57 Urbanwastewater with at least primary Publicexpenditureon water (% sector 35 Net cost of cross-subsidyprogram ("1. sector turnover) 20 1. COLOMBIA'SINFRASTRUCTUREIN INTERNATIONALPERSPECTIVE INTRODUCTION 1.1 Before embarking on a detailed analysis of Colombia's achievements and challenges in the infrastructure sector, it is important to put the country's situation in a wider international perspective. International comparisons can be very helpful in identifying the country's particular strengths and weaknesses. However, inevitably, they are always subject to limitations in the quality and consistency of internationally comparable indicators for the infrastructure sectors. Inparticular, it i s important to note that there are often significant lags in the compilation of international databases. As a result, the most recent indicators available often date from two or more years ago, which can be a problem in sectors (such as telecommunications) that are experiencing rapid evolution. Furthermore, in order to permit cross-country comparisons, indicators are sometimes adjusted to account for exchange rates or taxation in ways that subsequently make them difficult to compare with domestic data sources. For both of these reasons, the data presented in this chapter should not therefore be used to draw inferences about the current absolute level of infrastructure indicators for Colombia, but rather to gauge their level relative to other countries at the most recent point in time for which internationally comparable data exists. Detailed information about current infrastructure indicators for Colombia i s provided inthe other chapters of this report. 1.2 Colombia's performance on a range of infrastructure issues is compared to that of a peer group of nine other middle income Latin American countries and where possible to four East Asian `tigers'. The Latin American countries can be divided into two groups. The first i s a group of upper-middle-income countries (namely, Argentina, Brazil, Chile, Mexico and Venezuela) with per capita incomes in the US$3,000-6,000 range. The second i s a group of lower-middle-income countries, with per capita income closest to Colombia inthe US$l,500-2,500 range (namely, Dominican Republic, El Salvador, Guatemala and Peru). Where data availability permits, Colombian infrastructure indicators are also compared to those of three East Asian `tigers'; namely Indonesia, Philippines, and Thailand. The per capita income for these countries ranges from less than US$l,OOO for Indonesia and Philippines to around US$2,000 for Thailand. 1.3 Benchmarking is performed on levels of infrastructure finance, and on the country's resulting endowments of social and productive infrastructure. The 1 benchmarking exercise has three facets. The first looks at the volume of financing flows for infrastructure, and their composition across sectors, and public versus private sources. Thereafter, attention turns to examining the country's endowments of infrastructure, distinguishing between social infrastructure (defined as household coverage of water, sanitation, electricity and telephony, availability of public telephones, density of the paved and unpaved road network, and public transport) and productive infrastructure (defined as energy production and transportation, quality of paved roads and overall teledensity). 1.4 It is worth observing that the distinction between social and productive infrastructure is not a standard one, but has been developed for the purposes of this study. The distinction i s used as a device to distinguish those infrastructure services provided primarily to the household sector with the main impact being on quality of life, from those provided to the productive sector with the main impact being on productivity and competitiveness. The distinction i s inevitably a somewhat artificial one, given that the same physical infrastructures can serve both productive and social needs, and there are thus a number of gray areas whose precise allocation between the two categories i s clarified in the text throughout. Nevertheless, the distinction proves to be helpful in terms of facilitating the presentation of the material. 1.5 The results suggest that Colombia is an outlier along a number of dimensions of infrastructure provision. Colombia's performance in the infrastructure sectors departs significantly from the Latin American average. Some of the key findings can be summarized as follows, and will be developed in much greater detail below. First, Colombia has had an outstanding record in sustaining financing for infrastructure, based on relatively high levels of public support, and a substantial flow of private capital. Second, given its income level, and general geographic characteristics, Colombia performs above average in the equitable provision of social infrastructure, but lags Significantly behind its peers in the creation o f productive infrastructure. Third, Colombia's telecommunications sector stands out as being unusually dominated by public operators and finance, and exhibits a number of performance deficiencies, particularly with respect to internet penetration, cost of long distance calls, and efficiency of enterprise performance. INFRASTRUCTUREFINANCING 1.6 Inspite of the recent economic crisis, Colombia hassustained one of the most stable flows of infrastructure investment among Latin American countries (Figure l.l(a)). From 1980-95, Colombia was investing 2-3 percent of GDP in the infrastructure sectors. This percentage surged to around 4 percent o f GDP during the mid-l990s, when the boom for private participation in infrastructure was at its height, falling back to around 3 percent with onset of the economic crisis in the late 1990s, subsequently remaining at around that level. This makes Colombia one of the only countries in Latin America (alongside Chile) that has consistently maintained infrastructure financing above 2 percent of GDP. Indeed, Chile and Brazil are the only two countries to have consistently maintained infrastructure investment above Colombian levels, although 2 Brazil has not been able to do so in recent years. The figures presented here and in the remainder o f this chapter are subject to the caveats described in Box 3.1. 1.7 Colombia's level of investment in infrastructure is lower, but much more stable, than that observed inEast Asian countries in the same income range (Figure l.l(b)). Comparisons with countries outside the region are hampered by the absence of a global database on overall infrastructure investment. Nevertheless, some investment data is available for East Asian countries in the same income bracket as Colombia. The comparison suggests that Colombian investment in infrastructure i s towards the lower end of the range found in countries such as Indonesia, Philippines and Thailand, in the mid-l990s, but has proved to be very much less volatile over time. Figure 1.1:Overall financing for infrastructure investments, 1980-02 7% c I ~~ ~ ~~ (a) Against Latin American peer group (b) Against East Asian peer group Source: Adapted from Easterly and Serven, 2003 Source: EAP InfrastructureFlagship, 2004 Box 1.1:Methodological issueswith international data on investment 1.8 The composition of Colombia's infrastructure investment has been skewed towards energy and away from telecommunications and transportation (Figure 1.2). When compared against its peers in Latin America, Colombia stands out as having an 3 unusual degree of concentration of its infrastructure investments in the energy sector, which accounts for over half of total infrastructure investment during the 1990s. Furthermore, it has invested relatively little of its infrastructure budget in the telecommunications sector. A likely explanation for this i s the energy supply crisis experienced b y Colombia's predominantly hydroelectric system in the early 1990s following a protracted drought. This triggered a period of intense investment in thermal generation capacity in order to diversify the country's generating portfolio. Figure 1.2: Composition of infrastructure investment, 1999-00 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Source: Adapted from Easterly and Serven, 2003 1.9 Colombia's public financingfor infrastructure has been relatively stable but shows a long term decline (Figure 1.3(a),(b)). The vast majority of Latin American countries experienced a sharp decline in public financing for infrastructure investments towards the end of the 1980s' falling from 2-4 percent of GDP to around 0-2 percent of GDP. Colombia is alone in bucking this trend, managing to sustain public finance for infrastructure at around 2-2.5 percent of GDP. Comparing against the East Asian peer group, Colombia's level of public investment in infrastructure i s towards the middle of the range, but i s considerably more stable than that observed for other countries in the comparator group. Figure 1.3: Public investment ininfrastructure, 1980-02 I +Phiippines Thailand 1/ +Colombia (a) Against Latin American peer group (b) Against East Asian peer group Source: Adapted from Easterly and Serven, 2003 Source:EAP Infrastructure Flagship, 2004 4 1.10 Private financing of infrastructure in Colombia started comparatively late, but reached relatively high levelsby the mid-1990s (Figure 1.4). From 1992 onwards, most Latin American countries were able to attract private infrastructure finance worth about 1percent of GDP, or 3 percent in the case of Chile. However, this surge in-private investment was for most cases not large enough to fully compensate the simultaneous decline in public investment. Moreover, the availability of private finance declined substantially after 1998. In Colombia, private financing of infrastructure did not begin to take-off until 1995. Flows peaked at close to 2 percent in 1997, and then began to tail-off duringthe economic crisis of the late 1990s, although they never fell back down below 1 percent of GDP. Since 2000, there has been a modest resurgence with flows reaching 1.3 percent of GDP in 2001/02. Having been one of the strongest regional performers on private finance in the late 1990s, Colombia i s currently towards the lower end of the Latin American peer group. Private capital flows are also comparatively low by East Asian standards. Figure 1.4: Private investment ininfrastructure, 1980-02 1D% 3 9% 8% +hdcnrsia - 8% Philippims 5= 5% Thaland 6 4% E 3% 2% (a) Against Latin American peer group (b) Against East Asian peer group Source: Adapted from Easterly and Serven, 2003 Source: E A P Infrastructure Flagship, 2004 1.11 Notwithstanding, in proportional terms, Colombia has remained more dependent on public finance than other countries in the region. Although Colombia achieved relatively high levels of private finance, sustained levels of public investment meant that private sources never accounted for more than 40 percent of investment financing in any given year. This i s in marked contrast to most other middle income countries in the region, which financed over 50 percent of their investments from private sources, and as much as 90 percent inthe case of Venezuela. 1.12 Since losing its investment grade in 2000, Colombia has one of the highest country risk premiums in the region, although this has not prevented a recent resurgencein private finance (Figure 1.5). The country risk premium i s the difference between the rate o f return on dollar denominated sovereign debt and the risk free rate. Colombia's achievement of investment grade rating between 1995 to 1998 undoubtedly contributed to the high levels of private investment achieved during that period. Following the onset of economic recession in 1998, as well as the deteriorating security situation, Colombia lost its investment grade rating seeing its country risk premium 5 bounce back to 12 percent, close to the level of the early 1990s, and relatively high compared to other countries in the region. Notwithstanding, the current level of the country risk premium, Colombia has recently begun to experience a modest resurgence in private finance. Figure 1.5: Evolution of country risk premium, 1987-03 I I II +Chile -43- Malaysia IThailand I Mexico Philippiner ~ -ai- 4-Colombia - Peru - Braril -4- Argentina -+-Venezuela Source: Sirtaine et al., 2004 1.13 The sectoral composition of private investment in Colombia differed significantly from that of other Latin American countries, with a much lower participation of the telecommunications sector (Figure 1.6(a)). Whereas in most countries, around half or more of private sector investment during the 1990s flowed to the telecommunications sector, in Colombia only 22 percent of private sector investment went to telecommunications. Instead, Colombia received a relatively high proportion of its private investment in the energy (40 percent) and transport (33 percent) sectors, in common with Chile and the Dominican Republic. Figure 1.6: Composition of private investment, 1999-00 Q ET&" Energy BTransporl U k randsew e g e o m r e Greenfieldp q e d Comeaim (a) By infrastructure sector (b) By PSPmethod Source: PPIDatabase, World Bank Source: PPI Database, World Bank 6 1.14 Moreover, the mechanisms for private participation in Colombia, in common with many East Asian countries, tended to emphasize greenfield projects and concessions over asset divestitures. Whereas many Latin American countries received 40-60 percent of their private capital flows in the form of revenues for asset sales, in Colombia divestiture accounted for less than 2 percent of private capital flows. This i s an important difference because i t means that almost the entirety of private capital flows into Colombia resulted in the physical expansion of infrastructure, and creating only minimal fiscal revenues. In this respect, Colombia has much more in common with East Asian countries such as Indonesia, Malaysia, Thailand and the Philippines, which channeled private capital flows primarily towards greenfield projects, and concession contracts (Figure 1.6(b)). This tendency can also be observed in the accumulated value of asset sales during the decade 1988/98, which for Colombia amounted to no more than 5 percent of GDP, compared to between 6-12 percent in most of the other middle income countries in the region (Figure 1.7). Figure 1.7: Accumulated value of asset sales to private sector, 1988-99 Peru 3 Bra2 il Argentina H Salvador Guatemta lukxlco Vener ueb coIorrbE Jomnlcan Republic Chik 0 2 4 6 8 10 12 14 16 Accurrulated value o f privatization (% GDP) Source: Inter-American Development Bank 1.15 In common with a broader trend around Latin America, the Colombian publichas become relatively skeptical about private sector participation. Throughout the Latin American region, public support for privatization has dwindled during the last few years. The results o f successive Latinobarometro polls show a declining share of the population around Latin America agreeing that privatization has been beneficial for their countries, falling from the 46 percent in 1998 to a 22 percent in 2002 (Figure 1.8). Until 2001, opinion in Colombia mirrored this regional trend, with support for privatization declining from 39 percent in 1998 to 14 percent in 2001. Nevertheless, in the last two years, the opinion trend in Colombia has begun to reverse itself, with the level of support for privatization climbing back to 25 percent in 2003. Indeed, the percentage of the population in Colombia favorably disposed to privatization is now a little higher than in the region as a whole, although it undoubtedly remains a minority. 7 Figure 1.8: Population agreeing privatization beneficial for country, 1998/02 co c. c 40 Ua, c 35 0 (OLAC a co 30 L W 25 4- 0 20 Colombia II) cn m c 15 W 10 W 2 a 5 I 0 1998 2000 2001 2002 2003 Source: Latinobarometro, 2003 1.16 Overall, private sector participation in Colombia stands close to the regional averages for electricity and water, but well below the regional average for telecommunications. The following series o f maps illustrate the extent of private sector participation in countries across the region according to the percentage of the market that i s supplied by private operators (Figure 1.9). The maps show that, as far as the electricity and water sectors are concerned, private sector participation in Colombia i s close to the regional average. Private operators account for 13 percent of urban water consumers, and 50 percent of urban electricity consumers, and generate around 60 percent of the country's electricity. Moreover in common with the majority of countries around the region, Colombia has chosen not to privatize electricity transmission. In the case of telecommunications, Colombia stands out as having low levels of private sector participation compared to its peers, except in mobile services. Less than a quarter of fixed line and long distance customers receive services from private operators, compared with around 70 percent across the region as a whole. 8 Figure 1.9: Percentage of market served by private operators Gener2hon Transmission Distribution (a) Electricity Source: World Bank, 2004 L O d L.ongDistance Cellular c (b) Telephony Source: World Bank,2004 9 (c) Water Source: World Bank. 2004 INFRASTRUCTURE AND SOCIAL OBJECTIVES 1.17 This section reviews the extent to which Colombia's achievements in infrastructurefinance have translated into good social infrastructureoutcomes. For 'these purposes, social infrastructure i s defined to include household coverage of water, sanitation, electricity and telephony services, as well as public payphones. Furthermore, the density of the overall road network (paved and unpaved) i s used as a crude measure of accessibility throughout the national territory. 1.18 By Latin American standards, Colombia has achieved high household coverage of social infrastructure; particularly for water and sanitation, though less so for mobile telephony (Figure 1.10). Colombia's coverage of water and sanitation services (defined as household connections to water and sewerage) i s very good, outperforming all of the upper-middle-income countries except for Chile, and standing some 10-15 percentage points ahead of most o f the Latin American countries in the same income bracket. With regard to electricity, Colombia lags behind the upper-middle- income group but leads countries in the same income bracket by around 10 percentage points. Colombia's weakest performance on coverage i s in the telephony sector, and was largely attributable to its deficit in mobile telephony as o f the year 2002. Whereas Colombia's fixed line coverage i s ahead o f that of Mexico, in mobile coverage at that time lagged behind countries such as the Dominican Republic, El Salvador, and Guatemala. This i s largely due to Colombia's relatively late start in mobile telephony, and penetration has in fact increased by about 50 percent between 2002 and 2004, following the entry of a third mobile provider. In payphone density it has one of the lowest scores of its peer group, although this comparison is weakened by the absence o f data on unofficial telephone shops, which are widespread throughout the region. 10 Moreover, such figures do not capture the country's noteworthy success in rural telephony, which do not greatly affect the national statistics but have a major reach among the rural population (see Chapter 3 for further details). Colombia's road density, is relatively low in absolute terms. However, bearing in mind that a large proportion of the Colombian territory is sparsely populated and covered in tropical rainforest, road density i s comparable to that of other countries with similar characteristics (such as Peru and Venezuela), although still significantly behind that of Brazil. Figure 1.10: Householdcoverageof socialinfrastructure, circa 2000 _- - - -- I_ Chle , aaterrda I I I I I I I I LI 2 C o b h a Braz 1 Mexco Mnezuda :an Rpubic kgemna Peru E Sabdor Oormican Rpubic I 0 10 20 30 40 50 60 70 80 BO100 (a) Water, circa 2000 (b) Improved sanitation, circa 2000 Chlk Chile tbkxco Brazil Braz11 Mexico Argentna Argemina lknezueh Venezuela Colorrbe Comnioan Republic Peru Colombia E Salvador E Sabdor katerreh Guatemla Oomncan Republc 0 10 20 30 40 50 60 70 80 g0 100 IL Peru 0 10 20 30 40 50 60 70 SO 90 10 0 (c) Electricity, circa 2000 (d) Teledensity, 2002 E Salvador M2XlCO [bmnican Republic Peru Bazi E Sabador hhxico Chle Guatemla Braz1 Mnezuela kgentlna Colon'bia Venezuela Cobrrbia Chle Oormican Fepubic kgerrtina Peru 0 10 20 30 40 50 60 70 80 00 100 Per 10,000inhabtants 0 10 20 30 40 50 BO 70 8090 100 b p e r kri2x100 (e) Pay phones, 2002 (f') Road density, circa 2000 Source: OPS for water and sanitation, OLADEfor electricity, ITUfor telephony, and WDI for roaddensity. Note: Coverage statistics are not always for the same year across countries, but represent the most recently available data for each country and are always from the period 1998/02 11 Teledensity statistics are based on lines in use. However, as of 2000, Colombia had 3.7 lines per 100 inhabitants of unused fixed line capacity. Similar information is not available for the other countries. 1.19 Moreover, by global standards, Colombia's coverage rates are significantly higher than those achieved by other countries facing similar economic conditions (Table 1.1). It is also interesting to benchmark Colombia's performance on service coverage against wider international standards. This was done using cross-country regression analysis for a sample of 80 developing countries, considering two model specifications. The first i s a simple model that controls solely for variations in per capita GDP. The second i s a much more complete model that attempts to capture all of the factors that contribute to the difficulty of service expansion in any given country, including per capita GDP, urbanization, income distribution, population density, growth of urban areas, and average household size. Full details of the model specification are provided in Box 1.2. The results differ significantly for the two models. On the basis of the simple model, Colombia has coverage of electricity, water, and sanitation services that are about 20 percent higher than would be expected given its level of GDP per capita, but perfoms substantially worse than might be expected in terms of road density. As is the case of many other large Latin American countries, Colombia's population and economic activity are concentrated ina relatively small percentage of its land mass. Thus, the Andean region that occupies 45 percent of the national territory has 78 percent of the road network. As a result, road density in the Andean region i s four times higher than road density in the rest of the country. However, when the full range of economic conditions are taken into account, Colombia's coverage o f electricity, water, and sanitation are in fact very close to or slightly better than what the model would predict. Moreover, Colombia's road density turns out to be better than might be expected once low population density i s taken into account. The only deficit is with regard to public pay phones, where Colombia falls 25 percent below what the model predicts. However, this result may be misleading given the proliferation of phone shops, both in Colombia and other countries, which provide an equivalent public service but do not appear in the official statistics. Table 1.1: Colombia's social infrastructure endowment relative to other countries, 2001 ______ Simplemodel Full model +5.6 +2.1 -25.4 ,._~. ~- ~ ~ Source: Own calculations based on World Development Indicators. Note: Table reports percentage deviation between observed and predicted coverage statistics. Thus, a positive value indicates that Colombia has achieved coverage levels superior to those predicted by the model, while a negative value indicates the opposite. 1.20 These achievements reflect relatively high rates of coverage expansion sustained by Colombia during the late 1990s (Table 1.2). Over the period 199900, Colombia's effort at expanding service coverage was equivalent to reaching an additional 2 percent of its population each year with water, sanitation and electricity services, and an additional 3 percent of its population each year with telephony services. Compared with other Latin American countries, this i s a relatively fast growth rate. The only exception i s 12 the case of telephony, where countries such as Brazil and El Salvador have achieved much higher rates of expansion over the same time period. Box 1.2: International benchmarkingof infrastructure Colombia's endowments of social and productiveinfrastructurewere compared against wider intemational standards by means of a cross country linear regression analysis for a sample of 80 developing countries. Two model specifications were considered.The first is a simple model that controls solely for variations in per capita GDP. The second is a much more complete model that attemptsto capture the factors that contribute to the difficulty of service expansion in any given country, including per capita GDP. urbanization, income distribution, population densily, growh of urban areas, and average household size. Data was taken fro the World Bank's database of World Development Indicators. The full details of the specifications used for the models relating to social and productive infrastructureare presentedin the table below. Socialinfrastructure ProductiveInfrastructure Improved Improved Electncty Tota Elenr c ly San:tanon Water Generaton Telephone Telephone Mooile Paved Access Payphones Tota. Roads Access Access Capacity Lines Manllnes Subscroers Roaas SIMPLE MODEL GOP 0.01 0.00 000 0.01 0.01 0.00 0 07 0.04 003 000 (695)"' (11.52)"- (5.96)"' (A 94)'-' (5 96)"' (5 60)"' (12.08)''- (11.16)'*- (9.50)-' (6 62)'" Constant 4040 0.21 010 58.33 66.10 0.19 4533 32.90 1146 -003 (9 69)"- (1.58) (1 56) (18.40)"' (29.77)'" (373)"' (2.99~~" (341)"- (1 46) (045) Observations 109 104 115 101 105 1I 8 116 118 116 108 R-squarea 0.31 0.57 0.24 020 026 0.21 0.56 0.52 0.44 0.29 FULLMODEL GDP 000 000 0.M 000 0.00 000 0.06 002 004 0 00 (0.48) (8.58)"' (2 321'. (012) (0.67) (005) (686)"' (483)"' ( 6 . 8 r ' (1 23) Jrban zat.onRate 0.77 000 0.00 0.25 0.35 0.01 1.22 1.00 0 18 0.00 (5 15)"' (0.57) (0.83) (1 39) (3.03)"' (2.05)" (1.69)' (2.48)" (041) (0.88) Gtni .0.40 0.02 .0.01 0.05 -0.01 -0.01 -2.46 -2.14 -025 -0.01 (1.74)' (1.76)' (2.66)'*- (0.19) (0.08) (1.38) (221)" (3.45)"' (038) (3.10)"' PopiLaton0ens.ty 0.02 0.00 0.00 0.00 0.05 000 0.00 4.03 0.W 000 (1.08) (0.33) (6.62)'-' (0.10) (320)'-' (1.85)' (0.04) (053) (052) (2.45)'- G m hof Uroan Population -9.82 -001 -0.04 -5.21 -3.52 4 11 -13.99 -12.79 -1.80 -0.05 (6.53)'*' (010) (1.82)' (2.99)". (3.01)"' (4.62)'" (1.931' (3.19)"' (0.41) (2.51)** tiasehold Size 060 -0.09 .0.05 -1.97 1.55 -0.08 -30.58 -19.07 -10.58 .0.02 (0.28) (0.78) (1.47) (0.79) (0.97) (2.40)" (2.93)'-' (3.31)"' (1B8)' (0.91) Constant 69.59 -018 0.88 85.86 59.39 118 310.30 242.54 64.84 0.76 (4.73)"' (0.24) (4.01)"' (5.05)"' (5.26)"' (5.06)'" (4.35)"' (6.06)"' (1.51) (4.45)"' ~ Observatans 86 82 87 73 77 88 87 88 87 83 R-squared 0 76 068 0.57 037 046 0.64 0.78 0.79 065 0.44 ADsolLte value oft stats!cs in parentwses * sign.ficamat 109.. ** significant at 5% *** s gnifcam at 196 13 Table 1.2: Effort at expansionof social infrastructure over time Water Sanitation Electricit Telephony Argentina 1.4 Brazil 1.1 Chile 1.6 Colombia 1.9 2.0 3.1 El Salvador 3.1 3.3 4.3 1 Guatemala Mexico 1.1 2.6 1.7 2.6 Perii 2.1 2.1 1.8 2.6 Venezuela 1.8 1.9 i 1.7 1.8 I1 Source: Own elaboration basedon household survey data Note: Reports annual percentage of population that received new service each year from 1995/00. 1.21 However, Colombia does not perform so well with respect to coverage of basic services in rural areas, except for the case of water (Figure 1.11).Coverage of basic service in urban areas in Colombia i s close to universal, with access rates between 97-99 percent for water, sanitation and electricity, when sanitation is broadly defined as access to sewerage or a septic tank. Comparative statistics on coverage of sewerage in urban areas are much harder to obtain. However, it is known (for example) that Colombia's urban sewerage coverage of over 80 percent compares favorably with 70 percent coverage levels in Argentina, a country with much higher income levels. While coverage of water services in rural areas i s highby regional standards, at 73 percent, rural coverage of sanitation (51 percent) and electricity (64 percent) i s about average with respect to other countries inthe Latin American peer group. Figure 1.11: Householdcoverage gap between urban and rural areas, circa 2000 Guatemla II Chik Gobrrbia E Salvador Oomnican Republc llratemb a l e M X l C O [bmncan Republc E Sabdor bbrrbE Venezuela Argentna B n z l Peru Peru B n z iI Bqenana &xco 0 10 20 30 40 50 BO70 80 90 100 I 0 10 20 30 40 50 80 70 80 90 100 (a) Water (b) Improved sanitation 14 Oarmican Rpublc Venezueb kgermna h z i l Guatemb E Salvador I 0 10 20 30 40 50 60 70 80 90 100 ~ ~~ ~~ (c) Electricity Source: OPS for water, and OLADE for electricity. Note: Coverage statistics are not always for the same year across countries, but represent the most recently available data for each country and are always from the period 1998/02 1.22 Coverage of electricity, sanitation and telephony services in Colombia are among the most equitable inthe region. This conclusion can be observed graphically in Figure 1.12, which shows that access rates among the lowest income quintiles are very high inColombia compared to other countries inthe peer group, with the exception of the water sector where Colombia performs close to the average in equity terms. These impressions are confirmed by the concentration coefficients for coverage of services reported in Table 1.3. These coefficients summarize the overall distribution of connections in a single statistic that takes a value of zero for universal access and increasing values o f up to one as coverage becomes increasingly inequitable. The equity of telephony coverage in Colombia i s particularly noteworthy, in that the distribution of household telephone connections inColombia i s almost as equitable as it is inChile, even though Chile presents higher overall telephone coverage than Colombia. Table 1.3: Household coverage concentration coefficients, circa 2000 ---i ~~- - -~ Water Sanitation Electricity-. .-~- -- Argentina +o.oo ~~ +0.01 +0.09 Brazil +O. 18 +0.03 +0.28 Chile +0.03 +0.08 +0.01 +0.21 Colombia +0.06 +0.07 +0.02 +0.22 El Salvador +0.24 +0.07 +0.26 Guatemala +o.10 +0.38 +0.12 +OS6 Mexico +0.05 +0.15 +0.01 +0.36 Perii +o. 11 +0.09 +o.11 +0.46 Venezuela +0.02 +0.05 +o.oo +0.23 L-._ - ~- Source: Own calculations based on LSMS data Note: Coverage statistics are not always for the same year across countries, but represent the most recently available data for each country and are always from the period 1998/02 15 Figure 1.12: Equity in householdcoverage of basic infrastructure services, circa 2000 I00 100 90 90 80 Chile 80 --&Brazil 70 !Z 60 B 70 -r--Colombia Chile 60 50; -+-Guatemala E 50 " t C o l o m bia -Mexico 9 40 -El Salmdnr 30 +Guatemala 20 10 1 10 -Mexico 0 1 -Peru I I 0 1 2 3 4 5 1 2 3 4 5 Quintile of inmme I Quintile of incom e I (a) Water (b) Sanitation -Brazil Chile t i o l o mbia f 50 -El Sabador -E I Saldor 2 4 0 -Guatemala - ~ G u d mda 30 - -Mexico -Mexico 20 10 Peru 0 1 2 3 4 5 Quinttle of income (c) Electricity (d) Telephony (fixed plus mobile) Source: Adaptedfrom data inWorld Bank, 2003 usinghouseholdssurveys 1998102 INFRASTRUCTUREAND GROWTH 1.23 This section reviews the extent to which Colombia's achievements in infrastructure finance have translated into its endowment of productive infrastructure. For these purposes, productive infrastructure i s defined as electricity generation capacity, paved roads and telephone mainlines, measures, which are normalized against the size of the workforce in order to facilitate international comparisons. It i s also relevant to benchmark the cost and efficiency with which productive infrastructure services are provided against other countries in the region and beyond. 1.24 Colombia's endowments of productive infrastructure are relatively small in comparison with its peers in Latin America and East Asia (Figure 1.13). Colombia lags behind many middle income countries as regards its endowment of electricity generation capacity per worker, and kilometers of paved road per worker. The result for electricity generation capacity i s puzzling given the observed concentration of 16 infrastructure investments in the energy sector over the last decade, but this may simply reflect a relatively low starting point. As regards telephone mainlines per worker, Colombia lags behind the upper-middle-income countries, but leads the group of lower middle-income countries. If physical infrastructure stocks are valued at efficient unit prices, following the methodology developed by Fay and Yepes (2001), Colombia's stock of productive infrastructure has a total value equivalent to 210 percent of GDP. This puts Colombia behind other Latin American countries, such as Argentina, Guatemala, Mexico and Venezuela with stocks of productive infrastructure in the range of 360-780 percent of GDP, as well as the East Asian tigers whose stocks represent 320- 1000 percent o f their respective GDP. Figure 1.13 Endowmentsof productive infrastructure (a) Generation capacity per worker (b) Telephone mainlines per worker 8 1200 a 8% n N 6% 2 1000 4% 4 800 2% 600 0% 2 -2% 5f2 -p 400 -4% .6% 26 200 -8% r0 0 .lo% c3Smck20m -Growth rate 1880100 (c) Kilometers of paved road (d) Infrastructure stock as percent GDP Sources: Easterly and Serven, 2003; Fay and Yepes, 2003 1.25 Moreover, Colombia's position with respect to its peers has been deteriorating consistently since 1960. Repeating the same exercise of relative endowments of productive infrastructure for the same group o f countries in 1960, shows that Colombia's ranking was systematically higher then than in the year 2000. Thus, Colombia has slipped from 7th to 10thplace in electricity generating capacity, from 3rd to 7th place for telephone mainlines, and from 8th to 12th place for paved roads. This indicates that the rate of growth of these stocks o f productive infrastructure in Colombia has been significantly smaller than in other countries from the peer group (Figure 1.13). 17 Indeed, the East Asian tigers, in particular, show very high rates of growth of productive infrastructure stocks duringthe last forty years. Electricity ' Fixed Mobile Overall Internet Pavedroad generation teledensity teledensity teledensity density density capacity I j I Simple model -23.9 +46.2 -22.1 +20.5 -43.1 -94.4 Full model I -40.4 +53.8 I -31.8 +19.0 -45.7 1 -346.1 Source: Own calculations basedon World Development Indicators, see Box 1.1 Note: Table reports percentagedeviation between observed and predicted infrastructure endowments. Thus, a positive value indicates that Colombia has achieved coverage levels superior to those predicted by the model, while a negative value indicates the opposite. 1.26 This conclusionis confirmed by a global cross-country analysis of productive infrastructure endowmentscontrolling for a wide range of economic and geographic factors (Table 1.4). It i s also interesting to benchmark Colombia's performance on service coverage against wider international standards. As described above for the case of social infrastructure, this was done by means of cross-country regression analysis for a sample of some 80 developing countries, using both a simple model that controls solely for variations in per capita GDP and a more sophisticated one that controls for a wide variety o f economic and geographic factors. Full details of the model specification were provided in Box 1.1. The results indicate that Colombia's endowment of electricity generating capacity, internet density, mobile telephony and paved road density are all well below what would be expected given the characteristics o f the country. B y far the largest deficit occurs with respect to the density of paved roads. Only with respect to fixed teledensity does Colombia substantially outperform the predictions of the model. In fact, the superior performance with regard to fixed teledensity i s large enough to offset the deficit in mobile telephony, so that overall, Colombia has teledensity 20 percent higher than predicted by the model. 1.27 The results obtained for electricity generation capacity warrant further explanation and need to be interpreted with caution. The fact that Colombia's electricity generation capacity per worker is lower than might be expected, should not be taken as an indication that Colombia faces a particular shortage of capacity, as this i s not currently the case. Rather it i s a reflection o f the fact that the Colombian economy i s not very intensive in the use of electricity. A potential explanation for this i s that Colombia i s more reliant on other sources of energy. In order to evaluate this hypothesis, the same cross-country analysis performed for electricity generating capacity i s repeated for overall primary energy consumption (Figure 1.14).The results turn out to be very similar, indicating that both for electricity and overall energy Colombia uses about 40% less than would be expected given its level of GDP. The most plausible explanation of this finding lies in the structure of Colombia's economy, which i s relatively focused on primary products. To the extent that Colombia succeeds in diversifying its economy towards manufactured goods, a higher demand for energy may result in the future. However, the results obtained from these indicators in no sense justify a supply-led expansion of energy production, since supply should rather follow demand. Indeed, a mere expansion 18 of energy production capacity would not of itself do anything to alter the structure of the economy. Figure 1.14: Understanding Colombia's energy consumption patterns I=I Im5 Imu Im5 =ria I I& Issl 1% * (a) Primary energy consumption (b) Electricity generating capacity Source: Own elaboration 1.28 Colombia's shortfall inpaved road density also merits further reflection. The cross-country model developed above evidently cannot take into account the spatial distribution of population and commerce within a country. In the respect, it i s relevant to note that Colombia's urban and economic centers are heavily concentrated within one half of its territory, while the other half i s largely unpopulated and covered by environmentally sensitive areas. Indeed, if the model i s re-run (as an illustrative exercise) entering Colombia with only 50% of its full geographic area, the residual for paved road density rises dramatically from -346% to +125%. Nevertheless, Colombia i s by no means the only country in the cross-country comparison group to present a relatively large area of low population density within its territory. This characteristic i s shared by most South American countries that border on the Amazon basin, and can also be observed in many other larger developing countries around the world. Hence, it i s misleading to make such an adjustment only for the case o f Colombia. Furthermore, this finding should not be interpreted as an open-ended imperative to pave additional road, since paving decisions should be guided by traffic volumes on specific corridors, but simply as a pointer to where the country's weaknesses in terms of infrastructure endowments lie. In practice, the quality as opposed to the quantity of paved roads i s likely to be even more important from an economic growth perspective. However, unfortunately, there i s no international database of consistent road quality indicators against which to compare the quality of Colombia's roads. 19 Figure 1.15: Comparisons of internet density, 1997/04 9 0% 8 0% z:12 a2001 7 0% 10 5z 60% $s 86 FJ2002 2003 +- 2 4 5 2 0 1997 iwa 1999 2000 2001 2002mobservec DPredictec (a) Against regional peers (b) Against model predictions Source: International Data Corporation Colombia, WDI indicators and own elaboration. 1.29 The observed shortfall in internet density represents a significant concern. Figure 1.15 (a) illustrates the extent to which Colombia is lagging behind its Latin American peers with regard to internet density. This i s true both in terms of the absolute level of penetration and in terms of the rate of expansion of internet density, which has been less than 1% per year in Colombia compared with 1.5% to 2.0% per year in the rest of the peer group. Furthermore, over time, the gap between Colombia's internet density and the density that it should have given its level of GDP has been widening rather than narrowing (Figure 1.15(b)). 1.30 The prices of electricity by Colombian utilities tend to lie close to the regional average, while for telephony service relative costs differ significantly by service (Figure 1.17). Electricity prices in Colombia, which range from US$0.075-0.080 per kilowatt-hour depending on customer group, lie very close to the regional average. For telephony services, there i s a mixed picture. Local telephony and mobile calls cost substantially below the regional average. Moreover, the price of mobile calls has fallen further in the last two years as a result of intensifying competition. Trunk calls cost somewhat higher than the regional average, while international calls cost 50 percent more than elsewhere in the region. While Colombia's trunk call costs were comparatively cheap as of the mid-l990s, market liberalization and intensifying competition in a number of countries across the region (notably Argentina, Brazil, Dominica1 Republic and Venezuela) has allowed them to bring their charges below the Colombian level, while charges in Colombia have remained fairly static. With regard to international calls, there i s a marked dip in the Colombian price series in the year 2000, following the entry of new competitors into this market. However, new market equilibrium has subsequently emerged with prices bouncingback to pre-competitive levels. 20 Figure 1.16: Prices of electricity and telephony services Chik Argentna [bmnican Fkp Peru Guaemh E Salvador Q l O K b E mhdustrial Brazil 0 5 10 15 20 MComrcia OD0 0.20 0.40 0.60 0.80 1.00 1.20 1.40 OCelulal Riceof elenriciy(US$cem per !Oh) nksidentkl Cosper 3 mnute localcal(US$) 0fixed (a) Electricity, 2002 (b) Local teleDhonv. 2002 +Gletemla - R -e-Chle -- Rlppnes Vmezueh [bmnican Rep Drmican Fe A G u - r r e h 4- Brazil _i_ hdoneja +Prgrmra -Cdorrbe - Wlm hdonesia Peru * Thalmd (c) Peak hour trunk call, 1995/03 (d) Internationalcall to US, 199903 Source': OLADEfor electricity, ITUand Tarifica for telephony data. 1.31 Performance of telecommunications and electricity distribution utilities does not compare very favorably with other countries in the region. Colombia's telecommunications sector performed well below the regional average on labor productivity as of 2001, although the trend has been improving (Figure 1.18). Regarding the electricity distribution sector, Colombian utilities perform 25 percent better on average than other regional utilities on labor productivity, but their distribution losses are twice as high (Table 1.5). A disaggregation of the Colombian performance statistics across different groups of utilities reveals considerable heterogeneity, with labor productivity being twice as high in the larger utilities compared to the smaller utilities. Distribution losses are above the regional average, across the board. However, again the larger utilities perform 50 percent better than the smaller ones. 21 Figure 1.17: Labor productivity of telecommunications sector, 1996101 I - +Argentina 700 A n Brazi 600 ,f mile 500 I O I b I o h i a '! 400 -s- bsta 300 Rca -8-hkxico - 200 R r u 100 I+- I--- P /- Paraguay Q I O !lQQ6 ' 1907 - 1998 1ggg 2000 2001 Source:ITU,2004 Total numberof connections Distribution losses per employee ( percent) Argentina 818 12 Bolivia 12 Brazil 803 19 CostaRica 3 Panamti 279 16 1,201 10 Average (6 countries) 715 12 Colombia - Total 1,100 25 - Largeprivate 1,800 18 - Largepublic 700 19 Source:OLADE and RED1Electricity SectorReport, 2004 1.32 In a recent World Economic Forum survey, major industrialists ranked the quality of Colombia's transport infrastructure below that of its peers. A recent international survey of the quality of transport infrastructure provides subjective ratings of the quality of transport infrastructure in many developing countries based on surveys of large industrial users (Table 1.6). Notwithstanding the limitations of such subjective ratings, the results suggest that the quality of Colombian transport infrastructure compares poorly with that of its peers, scoring below average on each of the sub-sectors. Colombia's overall transport infrastructure quality rating i s 3.1 on a scale of 1 to 7 on which the peer group scores an average of 3.6. 22 Table 1.6: Comparative survey on the quality of transport infrastructure, 2002 Overall Road Port Air transport Railroad infrastructure infrastructure infrastructure infrastructure infrastructure quality quality quality quality I quality 1 Argentina 3.9 5.0 3.7 3.9 2.5 Brazil 3.8 4.4 2.2 Chile 4.8 4.6 2.2 Colombia 3.1 3.7 1.4 DominicanRepublic 3.9 4.6 1.6 El Salvador 3.8 4.1 2.6 5.0 1.1 Guatemala 2.8 3.9 2.6 3.7 1.4 Mexico 3.6 5.0 3.3 4.6 2.4 Peru 2.5 4.3 1 2.3 3.0 1.7 Indonesia 3.7 3.9 4.1 3.2 Philippines 2.3 3.5 3.7 2.4 3.9 I 1.5 1i Thailand 4.9 5.3 4.5 5.6 1 3.7 Average I 3.6 1 4.4 I 3.3 4.4 2.1 I Source: Global Competitiveness Report 2003/04, World Economic Forum Notes: Survey based subjective evaluation on scale from 1-`underdeveloped and inefficient' to 7-`as developed and efficient as the world's best'. 1.33 However, more detailed information reveals a more optimistic picture of the quality of port infrastructure. The evidence cited above comes from a general worldwide survey. A more detailed survey of Latin American ports based on a similar methodology, placed the ports of Cartagena and Barranquilla among the best in the region on four performance dimensions of cost, speed, reliability, and security. These ports perform substantially better than those of Callao and Guayaquil in neighboring Andean countries. Figure 1.18: Quality of ports, 2002 Source: Universidad Politecnica de Valencia, IIRSA-CAF, 2003 Notes: Survey based subjective evaluation on scale from 1-`very poor' to 7 - `excellent' 23 CONCLUSIONS 1.34 Colombia has sustained relatively high and relatively stable levels of finance for infrastructure relative to Latin American peers, with both public and private sectors playing an important role. Colombia, along with Chile and (until recently) Brazil, i s one of the few Latin American countries to have consistently maintained total infrastructure finance above 2 percent of GDP. Colombia's finance for infrastructure stands out as being relatively stable, relative to Latin American and (particularly) East Asian peers. Public finance for infrastructure has been experiencing a gradual decline over the last twenty years, beginning to dip below 2 percent of GDP since the economic crisis of the late 1990s. Private finance took-off relatively late around 1995, reaching a peak of 2 percent of GDP in the late 1990s, and more recently falling back to the 1 percent level. Nevertheless, the share of private finance in the overall financing of infrastructure remains among the lowest in the region. Overall, infrastructure finance has been quite heavily skewed towards the energy sector. The nature of private participation in infrastructure in Colombia differs significantly from that of other countries in the region, in the predominance o f greenfield projects and concessions over asset divestitures. Moreover, the participation of the telecommunications sector in the total flow of private investment has been relatively small reflecting the fact that Colombia's telecommunications sector has remained largely under public ownership. 1.35 Colombia presents impressive achievements in terms of access to social infrastructure, with high and equitable coverage of electricity, water and sanitation. By regional standards, Colombia has very high coverage of basic household services, in particular water, sanitation, and electricity. Indeed, even at the global level, coverage rates are higher than would be expected given the country's economic, social and geographic conditions relative to those of its peers. Coverage of basic services expanded more rapidly during the last decade in Colombia, than in other Latin American peers, resulting in a relatively equitable distribution of access across the income distribution. However, Colombia has a relatively large access gap between urban and rural areas, lagging behind many of its peers particularly in the provision of rural electricity and sanitation services. 1.36 However, Colombia appears to lag behind its peers in the accumulation of productive infrastructure, in particular paved roads, internet access and electricity generation capacity. Comparisons of productive infrastructure are conceptually more difficult than those made for social infrastructure, given that they reflect other factors that are more difficult to control for in a cross-country regression. Thus, energy use i s strongly driven by the underlying structure of the economy, and paved road density i s driven by the internal spatial distribution o f economic activity. Nevertheless, and subject to these important caveats, Colombia's endowments of productive infrastructure appear to be relatively small in comparison with peers in Latin America and East Asia, and relative to what might be expected on the basis of the country's economic, social and geographic conditions relative to those of its peers. Moreover, Colombia's position in relation to its peers regarding productive infrastructure, has been deteriorating steadily since the 1960s. The largest shortfalls are in paved roads, internet access and electricity 24 generation capacity. Density of fixed telephone lines i s higher than expected, while mobile density (which was lagging behind the peer group) has been catching-up during the last couple of years. Prices of Colombian infrastructure services are generally competitive, with the exception of long distance and international telephony. However, performance of telecommunications and electricity distribution utilities lags behind those of regional comparators. 1.37 Therefore, Colombia needs to enhance its performance on productive infrastructure, while preserving its achievements inthe social sphere. On the basis of this comparative assessment, the challenge for Colombia would appear to lie primarily in reorienting its infrastructure investments towards the productive sectors, without jeopardizing its strong performance in social infrastructure. Given that the overall level of resources devoted to infrastructure will be difficult to increase, this shift will primarily have to be funded by improvements in the efficiency o f social investments, as well as new sources of finance for productive infrastructure. While the private sector can play an important role in addressing the deficits in energy and telecommunications, the key challenge of financing improvements in the road network will necessarily remain a predominantly public responsibility, given that the limited scope for toll roads has already been largely exploited. 25 2. THE CONTRIBUTIONOFINFRASTRUCTURETO ECONOMIC GROWTH INTRODUCTION 2.1 Infrastructure services make an important contribution to the productivity and competitiveness of the productive sector, and hence to economic growth. Infrastructure has multiple impacts across the economy, with its relevance to production process going well beyond what i s captured directly by its share of GDP. It i s often overlooked that infrastructure services are not only consumed by the residential sector, but constitute key inputs into the productive process. The infrastructure needs of the productive sector are an important, though often overlooked, component of overall demand. Indeed, according to a recent estimate, infrastructure in developing countries i s consumed in approximately equal shares by households as a final consumption item and by firms as an intermediate consumption item (Prud'homme 2004). The quantity and quality of productive infrastructure are thus important determinants of a country's competitiveness. Energy and telecommunications are strategic inputs into many productive sectors. Road transport networks play a key role in facilitating the movement of goods along domestic production chains, while the quality and efficiency of sea'ports, airports, and border crossings play a critical role in enabling flows of international trade. This chapter deals primarily with the contribution of infrastructure to the productivity and competitiveness of firms, and hence ultimately to the economic growth of the economy. The focus i s therefore on productive infrastructure, as defined in Chapter 1, namely energy production, road networks and telecommunications. 2.2 The first part of the chapter takes a supply side perspective, providing an introductory overview of the infrastructure sectors in Colombia. This will provide a basic characterization of infrastructure asset endowments, industry and market structures, ownership arrangements, and enterprise performance. Colombia's infrastructure endowment i s generally good, but shows weaknesses particularly in the road sector. The reforms of the 1990s promoted private sector participation and market liberalization. However, in practice, substantial vertical and horizontal integration remains across different segments of the electricity and telecommunications industries. Competition i s universal, but typically takes place between public enterprises, while market power is heavily concentrated in a handful of national and municipal enterprises. The natural gas industry has expanded enormously over the last decade, following a comprehensive national plan. The water sector i s completely decentralized, and has witnessed major 26 improvements in financial sustainability, although quality of service remains deficient outside of the maincities. 2.3 The second part of the chapter takes a demand side perspective, examining how infrastructure impinges on the productivity and competitiveness of key productive sectors. The discussion draws upon an ad hoc survey o f large industrial users of infrastructure, as well as in-depth qualitative interviews with firms at different stages along three strategically important production chains. Infrastructure services are estimated to represent 16 percent of production costs across the economy, and as much as 15-25 percent in certain infrastructure intensive sectors. About 25 percent of infrastructure services by value are provided within firms, most commonly in the form o f own electricity generation and truck fleets. Firms identify road quality and security in Colombia as substantially worse than in neighboring countries. The difficulties in accessing ports, particularly Buenaventura, present serious problems for exporters. 2.4 The final section integrates demand and supply side perspectives, to define an infrastructure investment agenda grounded in the economic growth scenario. The analysis concludes that Colombia will need to invest an average of US$l,OOO million per year on transport and energy sector needs alone. About 80 percent o f this sum will need to go on the road network. Notwithstanding significant anticipated private sector participation, the bulk of this investment i s likely to have to come from public sources. THE CURRENTSTATUS OF INFRASTRUCTURE 2.5 This section provides a supply side perspective on infrastructure in Colombia, with a brief sector-by-sector overview detailing the country's physical infrastructure endowments, together with their economic and institutional organization, and the main issues arising in each case. Thereafter, separate treatment i s given to a number o f important cross-sectoral issues, including regional integration o f infrastructure networks, as well as infrastructure inthe context of competitiveness and free trade. Electricity 2.6 National energy consumption has been stagnant over recent years, and continues to be dominated by oil and biomass; the share of natural gas has grown considerably. Colombia's energy consumption has been stagnant over recent years, reflecting the overall downturn in the economy (Table 2.1). As a result, energy consumption per capita has fallen from 6.2 gigacalories in 1997 to 5.6 gigacalories in 2001. As of 2001, Colombia obtained US$4 of GDP for every kilogram of oil equivalent energy consumption, which i s close to the average for middle income countries in Latin America. The Colombian economy remains primarily reliant on oil and oil derivatives. Moreover, biomass sources (primarily firewood and bagasse) represent almost 20 percent of energy consumption, and continue to be significantly more important than electricity, which represents only 13 percent o f the total. Natural gas i s a comparatively new source 27 of energy for Colombia, but has rapidly increased its share in the national portfolio from 6 percent in 1997 to almost 10 percent in 2001, recently overtaking coal as the fourth most important source of energy for the country. Table 2.1: Finalenergy consumption by source 2001 , ercent ercent Oil andby-products .%k+k-- %2ki-+k Firewood,bagasse 43,034 17.5 43,770 18.3 Electricity 30,474 1 12.4 30,284 12.7 Naturalgas 15,028 6.1 22,838 9.6 coal 19,281 7.8 14,504 6.1 ~ Others 21,494 i 8.7 25,273 10.6 Total 246,384 , 100.0 238,709 t100 .o Source: Energia en ( ilombia 1997-2001 2.7 Colombia's integrated national power grid serves one third of the country's territory, covering 96 percent of the population. Colombia's electricity supply relies on the National Interconnected System (SIN), and numerous isolated local systems in the Non-Interconnected Zones (ZNI). SIN encompasses one third of the territory, covering 96 percent of the population. The ZNI accounts for the remaining two thirds of the national territory, where 4 percent of the population lives, and faces a very low population density of three people per square kilometer (see Chapter 3 for further details). 2.8 Notwithstanding the 1990s reforms, the sector is still characterized by vertical integration and horizontal concentration, and dominated by public enterprises. Duringthe mid-1990s, Colombia underwent a major reform of its electricity sector, adopting the essential elements of the British model with the deregulation of the wholesale price and the creation of a wholesale market (MEM) and competitive pool, as well as the formal unbundling of generation, transmission, distribution, and commercialization activities. Notwithstanding, the Colombian power sector has evolved along very different lines during the last decade. Although there has been significant private sector participation, public enterprises continue to be dominant in all segments of the industry. A functioning spot market has been established, however competition i s limited by the concentration of market power in a handful of large generators. Moreover, in spite of the fact that the industry has been formally unbundled, the vertical organization of the sector i s highly heterogeneous, with vertically integrated, partially integrated, and completely unbundled companies operating side by side in the market (Table 2.2). Table 2.2: Overview of vertical integration inthe electricity industry G+T+D+R 24 13 Source: Adapted from Chahin and Rojas, 2004 Notes: G-generation, T- transmission, D-distribution, R-retail 28 Generation 2.9 The generation portfolio is dominated by hydropower, with increasing participation of thermal generation since the mid-1990s. The SIN draws upon 93 power generation plants, totaling 13.6 gigawatts of capacity as against system peak demand of 8.1 gigawatts. These plants are generally in good condition, with only 4 percent of the portfolio close to the end of its useful life. At present, hydroelectric schemes represent 66 percent of the generating capacity, but typically produce 77 percent o f the country's electricity (Table 2.3). The remaining 34 percent of capacity i s thermal, predominantly fueled by natural gas, but typically accounts for only 23 percent of electricity output. Just under 1 percent o f electricity generation comes from auto- generation or co-generation by industrial users. Finally, the ZNI i s served by more than 1,000 small diesel generators, many of them indeteriorated condition. Table 2.3: Electricity'generationby source -. Capacity I Production GW 1 % GWh 1 % Hydropower 1 1 1 ~ 9,022 66.4 1I134,878 77.1 10,365 22.9 7,958 17.6 - Coal 1,983 4.4 1 ~ 1 I I Total 13,592 100.0 145,242 100.0 Source:UPME, 2002 reportedinChahin and Rojas,2004 2.10 During the 1990s, Colombia succeeded in improving security of supply by reducing reliance on hydropower. Most of Colombia's hydroelectric installations lack multiyear storage capacity and are hence reliant on each year's rainfall. This makes the country particularly vulnerable to meteorological events, in particularly the El Niiio phenomenon and associated droughts, which tend to come around once every four to five years, and can last for between four and six months. A particularly severe drought experienced in 1992/93, led to rationing of eight hours per day causing major economic damages. Thereafter, the country adopted policies to promote the development of natural gas based generation, reducing the dominance of hydropower in the generation portfolio from 80 percent of capacity in the early 1990s to 66 percent today. These measures allowed the country to overcome the droughts of 1997/98 and 2002/03 without any major disruption to power supplies and associated economic losses. Official projections of short and longer term expansion paths for the electricity geperation portfolio, provided by the Mining and Energy Planning Unit (UPME), suggest that by the year 2011 the weight of hydropower in the overall portfolio i s likely to have fallen further into the 60-65 percent of capacity range. 2.11 The ownership of the generation sector shows relatively high concentration, and continues to be predominantly controlled by national and municipal enterprises. Generation assets were historically held by national generators (ISAGEN, Corelca) and by vertically integrated municipal utilities such as those in Bogota, Cali and 29 Medellin. Following the drought of 1992193, Power Purchase Agreements were signed with six private companies for a total of 1,535 megawatts of thermal generating capacity. Thereafter, the capitalization of the Bogota and Atlantic coast electric utilities, led to a further transfer of generating assets to the private sector. The combined effect of these changes was to put about 55 percent of generating capacity in private hands. However, the two largest public generation groups (held by the national government and the city of Medellin) control more than 50 percent of the market. Overall, the four largest groups account for more than 80 percent (Figure 2.1(a)), with a Herfindahl index of 0.2 indicating a moderate degree of concentration inmarket power. Transmission 2.12 The transmission network is generally in good condition, but suffers from frequent terrorist attacks. The S I N transmission grid comprises 12,000 kilometers, and operates at 220 kilovolts or above, with 87 substations organized in 13 areas. Technical standards are quite good, and the assets are well maintained. The main problem faced by the grid i s terrorist attacks, which destroyed 483 transmission towers in2002, during 127 separate attacks. However, the number of towers attacked declined to 326 in 2003. The radial topology of the transmission network amplifies the disruption caused by such attacks, which could be significantly attenuated by some strategic reinforcements in the design of the network. In addition, some regional links are weak, leading to problems with stability and frequency regulation. Although there are currently no transmission capacity constraints, new lines should enter in service to avoid shortages in the center- east region. 2.13 The network is primarily operated by two national public enterprises, one of which is also responsible for administering dispatch and clearing financial transactions. Two national public enterprises, ISA and Transelca, manage more than 80 percent of the transmission network, with the remaining assets owned by some half a dozen vertically integrated municipal utilities (Figure 2.1 (b)). Over 90 percent of the transmission grid i s controlled by public enterprises. The transmission company ISA i s also in charge of operating the SIN, including the administrative dispatch of generation and the financial exchange of transactions on the wholesale energy market. Figure 2.1: Ownership concentration for different segments of the electricity industry 11% National holdings national holdings BMedellin holding: med dell in holdings oBogota holdings (Endesa) oUnion Fenosa ~ B o g o t aholdings (Endesa) oAES nothers 0Others (a) Generation (b) Transmission 30 BNational holdings "ationat holdings mMedellinholdings Medellin holdings Wogota holdings BBogda holdings (Endesa) (E ndesa) oUnion Fenosa OUnion Fenosa OCali holdings OCali holdings nothers 25% mothers (c) Distribution (d) Retail Source: CREGdatareportedinChahin and Rojas, 2004 Distribution 2.14 The electricity distribution sector is currently dominated by three large players. Colombia has 164,000 kilometers of distribution lines (excluding low level connections) and 30,000 mega-watt amperes of transformation capacity, managed by 30 utilities. Electricity distribution accounts for just over a quarter of electricity sector turnover. This segment has traditionally been dominated by municipal utilities, often serving broader regions centered on large urban municipalities. During the 1990s, a number of major distribution utilities were privatized or capitalized (see Chapter 5 for details). The existing utilities are controlled by 14 economic groups. The three largest players are Union Fenosa (with Electrocosta and Electrocaribe), Endesa (in Bogota) and Medellin, which together account for 72 percent of the market, with the two major private operators supplyingjust over half of the market (Figure 2.l(c)). 2.15 Distribution utilities present deficient operational performance; although there is considerable variation around the sector average. The average performance of the distribution utilities i s poor with distribution losses of 25 percent, billing delays of 123 days, and average outages of 12 hours per month (Table 2.4). Moreover, there i s considerable variation between utilities. The utilities serving Bogota, Medellin, EPSA (power sector company serving the Pacific Coast) and (to a lesser extent) Cali perform relatively well. The two utilities serving the Atlantic coast show particularly poor performance with distribution losses in excess of 30 percent, and outages of around 60 hours per month. The remainder of the small public municipalities, perform somewhat better, although still very poorly in absolute terms. Although distribution losses have been improving since the early 1990s, 11 utilities exhibit a deteriorating trend with respect to this indicator, and less than a quarter of utilities have losses below the benchmark level of 15 percent established by the regulatory commission. 31 Table 2.4: Performance of electricity distribution utilities Large pub Private Small Overall Medellin Cali Bogota EPSA Atlantic Caribe public average Coast Operational efficiency - Distribution losses 11.2 18.2 10.4 13.4 30.3 33.8 27.8 24.5 (percent) - `000s connectionsper 1.2 1.2 4.5 1.3 0.9 0.8 0.8 0.8 worker Commercial efficiency - Average billing delay 105 160 110 160 170 200 160 123 (days) Quality of service - No. of interruptionsper 0.03 0.05 0.10 0.10 0.58 0.51 0.40 0.39 month - Hoursof interruptionper 1.73 4.39 5.84 4.95 67.47 55.26 14.36 12.43 month ource: SSPD, CREGdata reported in__ and Rojas, 2004_ ~ _ _ Chah 2.16 Comparative efficiency analysis indicates a high variance of performance between distribution utilities, as well as increasing returns to scale. The results of a simple comparative efficiency analysis for 26 electricity distribution utilities in Colombia i s reported in Box 2.1. They indicate a high degree o f dispersion in efficiency levels, with more than 50 percent of firms being less than 80 percent as efficient as those on the frontier. Furthermore, there i s evidence that a significant number of utilities are producingbelow the minimumefficient scale. 2.17 Performance trends indicate that 42 percent of distribution utilities present infrastructure in poor or very poor condition against 58 percent in good or very good condition. Based on an analysis of the levels and trends in the operational performance indicators, it i s possible to draw conclusions about the underlying condition of the distribution infrastructure. On this basis, eight utilities are classified as having good or very good quality infrastructure (Bogota, Cali, Cartago, Medellin, Pacifico, Pereira, Putumayo, Tulua), and a further 11 as having poor or very poor quality infrastructure (Arauca, Bajo Putumayo, Boyaca, Caqueta, Caribe, Cauca, Costa Atlantica, Cundinamarca, Nariiio, Sibundoy, Tolima). 2.18 The operational and financial weaknesses of the distribution utilities, undermines the sustainability of improvements in the upstream segments. Distribution remains the weakest link of the power supply industry. Poor performance of distribution companies often leads to payment arrears that ultimately jeopardize the financial health of upstream activities (irrespective o f their own performance), creating pressure for government to bail out the insolvent utilities (see Chapter 5 for details). 32 Box 2.1: Efficiency analysis of electricity distribution utilities 2.19 Retail electricity prices have increasedsubstantiallyin peso terms during the last decade, although in dollar terms they have barely remained stable. The last decade witnessed major efforts to raise electricity prices towards cost recovery levels. Thus, between 1994 and 2002, prices rose by 50 percent for industrial customers, 125 percent for commercial customers, and over 250 percent for residential customers (Figure 2.2(a)). However, the falling value o f the peso, largely wiped out the value o f these gains indollar terms (Figure 2.2(b)), with dollar prices falling substantially for commercial and industrialcustomers, and rising only slightly for residential customers. Electricity tariffs also fall within the national cross-subsidy framework, described in Chapter 3 below. Due to significant adjustments in the cross-subsidy framework over time, price trends for specific customer groups may vary significantly from the overall average. 33 Figure 2.2: Evolution of retail electricity prices = 140 120 f:100 80 60 s 40 20 I 1934 1935 19% 1337 1998 1939 2013 22'3 a332 +,,-&"mid d.Resdertid lnlrstnd ~~~ ~ ~ (a) Inpeso terms (b) Indollar terms Source:Adapted from Chahinand Rojas, 2004 2.20 There are some problems with continuity of service in the SIN, but the situation is far worse and deteriorating in the ZNI. The terrorist attacks on the transmission network lead to low tension and power oscillations and were responsible for 0.4 percent of power demand going unsatisfied in 2002. Regarding more routine supply interruptions, the most serious problems are concentrated in the Atlantic region, which experiences supply losses of the order of 60 hours per month. In the ZNI service quality i s very poor, with current continuity of only six hours per day, except in a few of the larger towns. Retail 2.21 Large unregulated customers account for 30 percent of power demand, while 2 percent of regulated demand has switched power retailer. There are 36 power retailers in the country, and 5 to 15 active within each department. All of them are vertically integrated to some degree (Figure 2.1 (d)). Moreover, the market i s highly concentrated with almost 90 percent controlled by the two largest groups (split almost equally between public and private operators), and a Herfindahl index of just under 0.2. Power retailers in Colombia are free to compete for customers at all levels of demand. However, a size threshold i s established, below which regulated tariffs must apply, and above which tariffs are unregulated and can be negotiated freely between customers and their suppliers. This threshold has been gradually reduced over time from a starting value of 2 megawatts in 1994 to its current level of 0.1 megawatts as o f 2004. The number of unregulated customers has been growing at just under 1 percent per year reaching 3,500 in 2003. These customers represent a total demand which is close to 1,250 gigawatt hours, equivalent to 30 percent of the market. Competition for regulated customers has been more intense than anticipated and tends to be concentrated at the boundaries between service areas. Since competition began, over 2 percent of regulated demand has moved over to new entrants. 2.22 Summary. Electricity infrastructure i s generally adequate; although deficient operation and maintenance of local distribution networks make power supply somewhat unreliable in certain areas of the country. Electricity generation and distribution activities 34 are more or less evenly divided between public and private sectors. Large vertically integrated national and municipal utilities control at least two thirds of the market in each segment of the production chain, limiting the extent of effective competition in generation and retail. Natural gas 2.23 Colombia has significant (but declining) natural gas reserves,concentrated at two fields in Guajira and Cusiana. Until 2003, exploration and production of hydrocarbons were the exclusive right of the public enterprise Ecopetrol, in association with private operators on a shared cost and revenue basis. This changed following Decree 1760/3, which created the National Hydrocarbons Agency (ANH) with responsibility for licensing exploration activities, leaving Ecopetrol to compete with private investors for exploration licenses. As of 1997, Colombia had proven natural gas reserves of 6.9 trillion cubic feet (or 33 years), which have fallen to 4.2 trillion cubic feet (or 19 years) up to 2002, as a result of declining exploration activity. The bulk of these reserves are concentrated in two fields. The first i s the Guajira field (dry gas) on the Atlantic coast, which has been operating since the 1970s and produces over 500 million cubic feet per day. The second i s the Cusiana field (associated gas) in the Andean foothills, discovered in the early 1990s. Currently producing 70 million cubic feet per day, Cusiana is scheduled to expand its production to 180 million cubic feet per day by 2005, thereby liberating a comparable volume of gas from the Guajira field for potential export to neighboring countries. 2.24 In 1991, the government published a strategic plan for scaling-up the use of natural gas in the interior of the country. Prompted by the discovery of the Cusiana field, the government published a strategic plan for expanding the use of gas on a massive national scale. There were two central motivations behind this policy. The first was the need to promote gas-fired electricity generation in order to reduce the vulnerability o f the power supply system to periodic droughts. The second was the large potential benefits that were identified for residential customers switching to natural gas for domestic cooking and heating; in particular for the estimated 340,000 low income households relying on highly flammable petroleum-based cooking fuel (known as cocinol). The key elements of the plan were the extension of the gas transportation network into the interior of the country, followed by rapid expansion of natural gas distribution networks in the main urban areas. The plan relied heavily on the private sector to finance and operate the new gas network. 2.25 A modern gas transportation network has been rapidly developed. The gas transport network i s 5,600 kilometers in length, and comprises two pipelines. The first east-west linking the Guajira field with the major cities on the Atlantic coast (Santa Marta, Barranquilla and Cartagena), and the other north-south bringing gas to the main cities of the central Andean region (Bogota, Cali and Medellin). Both networks are interconnected. There are eight gas transportation operators, but two of them control 97 percent of the gas network: Promigas, a private firm operating the Atlantic coastal pipeline; and Ecogas, a public enterprise established in 1998 to oversee the development 35 of the central Andean pipeline. About half of the transportation assets of Ecogas are under Build-Operate-Maintain-Transfer contracts (BOMT) with the private sector. There i s substantial excess capacity on the transportation network, with an average utilization rate of 50 percent in the western region of the country. This i s due to over-optimistic estimates about the rate of growth of electricity demand, and hence of thermal generation based on natural gas. However, the regulatory regime prevents the cost of this excess capacity being passed on to consumers via tariffs. The condition of the infrastructure i s good, given that the assets are relatively new and are being well maintained by private operators. The few potential local transportation bottlenecks that exist in the network are linked to power generation demands, and can be readily solved with additional compression. 2.26 Gas distribution has expanded rapidly reaching 77 percent of the potential residential market in 327 municipalities. With regard to distribution, the government decided to privatize the handful of distribution utilities that existed in the early 1990s, and bid out concessions for exclusive service areas to support development of networks from scratch in 126 municipalities. As of today, there are 20 natural gas distribution companies serving 327 municipalities. The development of a separate retail sector i s incipient, with one established player (Dinagas) and a recent second entrant (Surtigas). All of these utilities are private with the exception of MedellinPublic Enterprises (EPM) that accounts for only a small share of the market. The largest operator i s Gas Natural serving the metropolitan area of Bogota. Residential connections to the natural gas network rose fourfold over the decade 1994 to 2003. At present, it i s estimated that 77 percent of potential residential clients have connected to natural gas. About 30 percent of the market i s in Bogota, and a further 25 percent on the Atlantic coast. Overall, consumption of natural gas increase by 50 percent during the 1990s, reaching 392.2 thousand cubic feet per day by 2003. 2.27 Although vertical integration in the natural gas sector is limited, other constraints restrict the development of competition. The legal framework permits limited vertical integration, allowing existing transporters and distributors to participate in exploration and production up to a market share limit of 25 percent. However, there is no vertical integration between transportation and distribution activities. While these measures help to avoid the concentration o f market power that could arise from vertically integrated firms, there are a number of other considerations that restrict the development o f competition. Important factors are the small size of the sector as a whole, its natural geographic segmentation between the Atlantic coast and the interior, as well as the presence of significant scale economies. 2.28 However, use of natural gas remains marginal inthe industrial sector, where itfaces intensecompetition from coal and heavy fuels. Natural gas is at a disadvantage against coal and heavy fuels in the competition for industrial customers. This i s primarily due to underlying differences in the marginal cost of these fuels. Moreover, underlying price differentials are exacerbated b y relatively high and distance-related transportation costs, which make the fuel particularly expensive in the western regions of the country (Cali, Medellin) that are furthest removed from the natural gas fields. Thus, industrial 36 users in these cities are able to obtain coal supplies at 60-70 percent o f the price of natural gas. 2.29 Further use of natural gas as a fuel for new thermal generation plants, will necessitate the resolution of a current mismatch between upstream costs and downstream revenues. Given the lack of competitiveness o f natural gas for use in the industrial sector, electricity generation remains the major non-residential use of natural gas. However, there are fundamental incentive problems in developing the natural gas infrastructure needed to meet the demands of the next generation of thermal electricity generation plants. In particular, in order to finance the associated pipelines, gas transporters typically require take-or-pay contracts from their client generators that cover at least 60-70 percent of the maximum possible demand under the contract. However, since thermal plant i s primarily dispatched in low hydrological scenarios, generators do not have sufficient certainty in their revenue stream to allow them to enter into such arrangements. While there i s clearly a need to explore more creative contractual approaches, an ultimate resolution to this problem may require regulatory recognition that the fixed costs of thermal generation include upstream investments ingas transportation. 2.30 Two other potential uses of natural gas are Compressed Natural Gas (CNG) for vehicles, and potential export to neighboring countries. In addition to the domestic residential and industrial uses of natural gas noted above, CNG for transportation and export to neighboring countries, represent two additional applications whose full potential i s currently being explored by the authorities. There i s an incipient CNG market for vehicle use, and vehicle conversions are expected to reach 10,000 during 2004. Colombia has the possibility of exporting natural gas to Panama (for gas-fired generation), western Venezuela (for use in oil fields) and/or Ecuador. Any such arrangements would necessitate a significant extension o f the transportation network, entailing major investments. 2.31 Inparallel with the expansion of natural gas, the productionand distribution of Liquid Petroleum Gas (LPG) has been actively promoted as an alternative for more isolated localities. Untilthe early 1990s, supply of LPG was scarce and unreliable, being distributed primarily based on political and administrative criteria. The national plan of 1991, established the expansion of LPG as a strategic complement to natural gas in order to cater for smaller towns and urban peripheries where natural gas distribution networks could not be developed. This led to the establishment o f a national infrastructure for production, importation and distribution o f LPG, and the development of a wholesale and retail industry, mwng the fuel available in 614 municipalities across the country. 2.32 Summary. Colombia i s close to completing its ambitious national plan to take natural gas into the central Andean region, with near universal access in the main urban areas. However, beyond its use in electricity generation, natural gas remains uncompetitive for the industrial sector. Excess capacity in gas transportation tends to inflate the cost of the fuel. 37 Telecommunications 2.33 Colombia's telecommunications reform helped to expand and improve services, however competition remains weak and public enterprises continue to play a dominant role. Sector reforms during the 1990s have accelerated the modernization of the telecommunications networks, and facilitated the introduction of new services, such as mobile telephony and internet. The reform process in the telecommunications sector focused on liberalization rather than privatization, introducing competition into all services. However, market structures continue to concentrate power in a handful of dominant players (Table 2.5). In each of the four major services (local, national, international, and mobile), the two largest players account for 70 to 100 percent of the market, and the Herfindahl index lies in the 0.3 to 0.5 range. This situation i s somewhat attenuated by competition between different types of telecommunications services, although many of the same players reappear across the different segments of the market. Considering the sector as a whole, the two largest players (Colombia Telecommunications and Bogota Telecommunications Utility -ETB-) still account for almost half of sector turnover, and the Herfindahl index stands at 0.2. Moreover, actual competition i s geographically concentrated in the three largest cities, with four main operators in Cali, and three in Bogota. Thus, public enterprises at both national and municipal levels continue to dominate, and together still control around 90 percent of the assets in local and long distance services. Moreover, public enterprises accounted for 70 percent of the total US$ 7,500 invested in the sector during the last decade, and control 64 percent of sector turnover. Table 2.5: Market concentration intelecommunication services I, Local National Intemational Mobile Value added Others' Herfindahlindex 30 48 34 51 20 A I Market share of two largest players 71 86 71 100 42 49 Source: Adapted from Mendoza, 2004 Note: In the last year there has been a third entrant to the mobile market whose exact market share is not yet known. * Including trunking and paging. 2.34 Fixed line telephony is heavily concentrated in the largest cities, and significant waiting lists continue to exist. Local telephony service i s provided by 29 operators. B y far the largest operator i s the public enterprise Colombia Telecommunications accounting for 45 percent of the market, followed by the two largest municipal operators (ETB and EPM) together accounting for a further 45 percent of the market.There are 9.3 million fixed telephone lines inColombia, of which 7.5 million are in service, resulting in average teledensity of 17 percent. There is strong geographic concentration, with 71 percent of the fixed lines located in the largest urban centers where 29 percent population lives. Teledensity in the three largest cities (Bogota, Cali and Medellin) ranges from 25 percent to almost 40 percent. However, other important cities, like Cartagena and Barranquilla, still have teledensities below the national average. 38 Although waiting lists have been declining from the one million level in 1999, as of 2002 there were still half a million requests pending for fixed line service. The pace of expansion of fixed line services has slowed considerably following the introduction of mobile telephony (Figure 2.4 (a)). Table 2.6: Government ownership of telecommunications operators j Total state participation 193.3 89.1 2.2 43.2 j1.9 Source: Adapted from Mendoza, 2004 2.35 Long distance services operate as a triopoly, but competition from mobile and internet services is strong, leading to an overall decline in demand. Long distance services were opened-up to competition in 1998, allowing two new operators to compete with the public incumbent Colombia Telecommunications. The entrants, Orbitel (private) and ETB (municipal), paid a total of US$lSO million to obtain their licenses. Colombia Telecommunications continues to hold a dominant position in the market, although competition for international traffic is somewhat stronger than for national traffic (Table 2.6). Overall, demand for long distance telephony has been falling due to the rise of substitutes such as mobile telephony and internet (Figure 2.4). 2.36 Mobile penetration has been growing rapidly, particularly since the entrance of a third market player. Mobile service was introduced to Colombia in 1994, under a regionalized duopoly regime that later became consolidated at the national level (Table 2.6). The two private operators, Comcel and Bell South, paid a total of US$1,200 million in license fees. The rate of expansion accelerated markedly following the announcement of a third license in 2000 (Figure 2.4). The license was awarded to Colombia Movil in 2003, ajoint venture of the two largest municipal telecommunications operators. The new entrant has adopted an aggressive marketing strategy, acquiring 600,000 subscribers in its first quarter of operation, leading to temporary network congestion as demand outpaced the expansion of network capacity. Given recent growth rates, mobile density has been rapidly catching-up with fixed line density (Figure 2.4), and according to the most recent statistics surpassed fixed line density as o f June 2004. As with fixed lines, mobile services are geographically concentrated in large urban areas and along some strategic road corridors. 39 Figure 2.3: Breakdown of telecommunications sector turnover 100% telephony b 90% b 80% Cellular 5 70% telephony 60% 50% 21% 43% 5g P 40% WLom 30% distance 20% a 10% oValue 0% ad&d Local Long Cellular Value Otter Total telephony distance tdeljlony ad8d OOtkr OTelecom BET8 EIEPM DCancel OBell South mothers (a) Turnover by segment (b) Turnover by provider Source: Adapted from Mendoza, 2004 2.37 Internet penetration in Colombia is still relatively low, and broadband services are only available infour major cities. Carriers have developed two extensive national optic fiber networks (one of them over power transmission lines). Moreover, following the liberalization of international calls, three submarine wire and satellite connections were established providing multiple links to international networks. Value added services are on the rise, including corporate services, data transmission and trunking. However, Colombia still has only 2.7 million internet users, implying an internet density of 4.2 that i s well behind that of Chile (24), Argentina (12) and Peru (10). Moreover, 85 percent of Colombian internet users are concentrated in the five largest cities, while broadband services are readily available in Bogota at present (and to a lesser degree in Medellin, Cali and Bucaramanga) but only incipient in the other large cities, due to the presence of infrastructure bottlenecks. A key constraint is that local telephony operators have exclusive rights to provide broadband services over the exercise. Figure 2.4: Evolutionof fixed and mobile telephony Y E 7 20 6 4 5 x 2 4 - Y---- -z53 - 2 5 1 0 am a301 ma? +Fwed telephony Cellulartelephony -Long dstance (national) -+Long distance(internationa1) (a) Suscribers (b) Traffic inMinutes Source: Adapted from Mendoza, 2004 2.38 Sector reform has helped to accelerate technological modernization, leading to improved quality of service, and declining prices in some segments at least. As the result of a process of technological modernization and capacity expansion, Colombia's 40 telecommunications service i s almost completely digital. Some 96 percent of exchanges had been digitalized by 2001 and the proportion i s climbing rapidly towards 100 percent. Operators are in the process of replacing some old transmission lines, or leaving them as back up for new modern lines. Overall, technology and asset condition do not represent a serious constraint on continued service growth. As a result, quality of service has been improving, with the number of call failures per 100 inhabitants falling from 84 to 42 over the last five years. Tariffs have fallen significantly with the growth of competition for international long distance, mobile telephony and internet services. However, tariffs for long distance national services have remained stagnant, while the cost of local telephony services has increased substantially due to the removal o f historic cross-subsidies between local and long distance calls. In addition, local telephony tariffs also fall within the national cross-subsidy framework, which will be described in some detail in Chapter 3 below. Figure 2.5: Evolution of telecommunications tariffs - q n x n n L c c a l L4 v d u ] telepbny 7 . i A - 2 0 0 Intemet - 1 5 0s F access +Cellular telept-any Source: Adapted fromMendoza, 2004 2.39 Summary. The telecommunications sector reform introduced across the board liberalization, and opened the way for limited private participation in some segments of the market. However, the outcome has largely been one of competition between public enterprises, with market power heavily concentrated in the hands of three major national and municipal enterprises. Quality of service has improved, although significant waiting lists remain for local services. Prices have fallen in some segments of the market, but risen in others. Internet penetration is still comparatively low, and broadband services are only available inthe capital city. Water 2.40 The water sector in Colombia has been fully decentralized since the late 1980s, with services provided by more than 1,300 utilities. There are approximately 1,300 water utilities operating in Colombia, the vast majority of them municipal utilities, although there are also a handful of departmental utilities. However, the market i s geographically concentrated with the 40 largest municipal utilities serving 70 percent of 41 the urban population, equivalent to 55 percent of the national population. The extreme atomization o f the remaining service providers has led to substantial loss of scale economies, and contributes to the weak management and precarious financial viability of the smaller operators. 2.41 Only half of installed waste water treatment capacity is operational, treating less than 10percent of the country's waste water. The last complete inventory of water sector assets was conducted in 1996. The water trunk transportation and local distribution network amounts to 47,000 kilometers in length. Although it is predominantly made of PVC, almost 40 percent of the network was reported to be made o f asbestos. About 70 percent of water supply utilities provide sewerage service, using 29,000 kilometers of sewers, of which almost 60 percent are combined with storm water drainage. About 40 percent of this infrastructure was found to be less than 10 years old. There are 327 treatment plants, located in 325 municipalities, with a total installed capacity of 13 cubic meters per second, equivalent to 16 percent of current waste water volumes. However, only about half of this capacity i s currently being used, due partly to deficient maintenance that leaves plants inoperative, and partly to the fact that some plants are located in municipalities that still lack the necessary sewerage infrastructure for waste water collection. 2.42 Private sector participation is not uncommon among the largest utilities, and takes place within the framework of a distinctive mixed enterprise model. There has been a considerable diversity of public private partnerships, with more than half of the 26 largest utilities, equivalent to 13 percent of the urban market, operating with some degree of private sector involvement (Figure 2.6). Colombia has pioneered the mixed enterprise model for private sector participation in the water sector, whereby: the municipal government remains a significant (usually controlling) share of the utility; the private sector i s represented on the Board and takes a leading role in the operation of the company; and investments are co-financed between the public and private sector to varying degrees. Figure 2.6: Institutional breakdown of water sector 9% OState owed enterprises oLimited liability public companies Minotity pride share mixed companies BMajotity pride share mixed companies OPrivate companies Source: Adapted from Fernandez, 2004 42 2.43 Overall performance indicators for the 59 largest utilities show reasonable levels of efficiency, with similar performance across public and private operators (Table 2.7). Average levels of operational performance are moderate with 42 percent unaccounted for water. However, commercial efficiency i s good with meter coverage of 94 percent, collection efficiency of 87 percent and average payment delay of 81 days. Among the public utilities, performance in the two largest cities (Bogota and Medellin) i s substantially better than in the smaller cities. The utilities with private sector participation, including Barranquilla and Cartagena, show performance levels beneath those of Bogota and Medellin, but are broadly comparable to those of the other public utilities. I Public Private Average I 1 Bogota Medellin Others Barranquilla 1Cartagena Others Operational efficiency - Unaccounted for water (percent) 35 1 - Labor costs (US$/'000m3) 1 68 94 42 65 38 17 46 64 42 63 42 64 -Commercial efficiency :: Meter coverage (percent) 98 98 88 83 98 181 94 - Collection efficiency ( percent) 95 98 84 85 ' - Average payment delay (days) - - 58 115 I 2.44 Water tariffs more than doubled during the 1990s, leading to a major contraction in residential demand that is postponing the need for major investments. During the 1990s, sustained efforts were made to raise water tariffs closer to cost recovery levels, lifting the average domestic tariff from US$0.33 per cubic meter in 1990 to US$0.78 per cubic meter in 2001. Given that 88 percent of households in Colombia are metered, this large tariff increase led to a strong demand reaction, reducing average household consumption from 34 to 19 cubic meters per month over the same period (which implies a price elasticity of demand around -0.3.). Moreover, water tariffs also fall within the national cross-subsidy framework, described in Chapter 3 below. Due to significant adjustments in the cross-subsidy framework over time, price trends for specific customer groups may vary significantly from the overall average. An important consequence of rising prices has been the postponement o f major investments in the expansion of water supply capacity. 2.45 Quality of service remains deficient outside the major urban centers, with lack of disinfection of drinking water being a major concern. A recent survey found that 99 percent of the population in the 23 largest cities was receiving water of potable quality. However, in the rest of the country, almost 60 percent o f samples fail to pass the minimumquality standard. One of the main reasons for this is that almost a third of the smaller municipalities, although equipped with water treatment plants, do not make the necessary purchases o f chlorine for water disinfection. This in turn i s attributable to the weak technical capacity of the smaller utilities, as well as cash flow problems that restrict 43 the availability of resources to purchase basic production inputs. Regarding continuity of service, two thirds of households report access to a 24-hour water service. 2.46 Summary. The water sector has been fully decentralized, and i s highly atomized as a result, with more than 1,300 utilities. The larger utilities present moderate operational performance and solid commercial performance. A significant number of them have some form of private participation, often following the mixed enterprise model. The smaller utilities present very weak operational and financial indicators, and provide water of deficient quality, in many cases without any disinfection. Tariffs have more than doubled over the last decade, leading to a substantial decline in demand, and hence significant postponement of investments inexpansion of water resources. Transport 2.47 Road is the prevailing means of transport for freight and passengersflows; .whilemaritime transportation dominates international trade. The overall modal split inColombia's freight transportation shows a clear dominance of roadtransportation, with almost 80 percent of the country's ton-kilometers hauled by truck, covering all types of cargo (Figure 2.7 (a)). Railroads account for 15 percent o f ton-kilometers, and are used almost exclusively to transport coal from mines to maritime ports for export. Inland navigation flows are concentrated on the Rio Magdalena, which i s mainly used to transport oil and its derivatives. Almost all of Colombia's international trade i s channeled through maritime ports, which predominantly handle massive movements of bulk commodities (Figure 2.7 (b)). Truck flows through border crossings with Venezuela and Ecuador take comparatively small volumes, while air transportation i s only relevant for the small fraction of higher value products. The vast majority of inter-city passengers using public transportation rely on bus services, although air transportation has a particular relevance for some remote regions (Figure 2.7 (c)). Motorization levels in Colombia are relatively low by international standards, at 55 vehicles per 1,000 inhabitants (40 vehicles per 1,000 inhabitants outside Bogota) versus the Latin American average of 108 vehicles per 1,000 inhabitants. The incidence of freight in product costs i s relatively high in Colombia. According to recent estimates, it represents well over 8% of the value of the product (against a world average of around 6%). In the primary sector, the proportion i s much higher, ranging from 20% to 60% according to the product (Pizano, 2004). 44 Figure 2.7: Transportation by mode I t396 I I 147 4% I t I 85% eRoad ElRailroad OWaterway OPort OAirport gLand borders E0,s W L - a BAir 1 2.48 Approximately 15 percent of Colombia's 166,000 kilometers of road network are paved. Colombia's road network i s thought to be approximately a 166,223 kilometers in length, although precise estimates of the sub-national network are difficult to make (Table 2.8). Following a decentralization process, secondary roads were partially devolved to departmental management and tertiary roads to municipal management. However, owing to lack of local resources and capacities for road maintenance, the decentralization process could not be fully completed and therefore significant portions of the tertiary road network remain with INVIAS under nationaljurisdiction. Some 2,300 kilometers of the national primary network have been given in concession in three successive waves since 1994, covering just about all the corridors with sufficient traffic to support toll-based concessions. About 70 percent of the primary network, and 15 percent of the overall road network i s paved; with approximately half of the paved segments in the primary and half in the secondary network. Moreover, only 400 kilometers of the primary network are multi-lane highways. The quality of the road network i s relatively poor, except on the access routes to major cities. Colombia's difficult topography creates problems of variable width and frequent absence of sidings, while steep slopes and close curvatures significantly constrain the speed of circulation. In addition, severe climatic events lead to rapid erosion and frequent landslides that further jeopardize the quality of the road infrastructure service. Table 2.8: Road network by category i-- __ I 1 % 1 ~~ PuZZction Length Paved percent 1 I ` 1I - National 70.8 (underconcession) (2,300) (100.0) Secondary Departmental 66,082 2.2 Tertiary National 36,136 - Local footpaths Municipal 34,285 12,556 /Total. -1 Source: Ministry of Transport, 2002 2.49 Road traffic has been stagnant in recent years, and remains highly concentrated in a few critical corridors. Road traffic grew steadily reaching 2,3 17 45 vehicles per day in the primary network on average in 1997. Since then, traffic has declined and remained stable until 2002, climbing back to 2,300 vehicles per day only in 2003. The factors behind this unusual contraction in traffic growth appear to have been the economic recession of the late 1990s, together with the armed conflict and the resulting constraints imposed on nocturnal circulation. Fortunately, attacks on road infrastructure have been falling since 2002. Most o f the road traffic is concentrated in a few corridors, and along the access routes to large .cities, ports and border crossings. Only 4,500 kilometers of roads have traffic averaging more than 2,500 vehicles per day, of which about half have already been given in concession. These links represent a quarter of the primary network, but carry approximately two thirds of the total vehicle-kilometers on the system. Congestion i s rarely a problem, and where it does exist owes more to poor road specification and deficient maintenance than to high traffic levels. Vehicle circulation has been growing consistently on low traffic roads, surpassing 300 vehicles per day on many links, which i s considered to be the threshold level above which paving becomes economically viable. 2.50 Although the condition of the paved road network is still fair, it has deteriorated consistently in recent years, accumulating a sizeable liability for rehabilitation. The general condition of the paved road network i s still fair, but the quality of the publicly managed portions of the primary network i s deteriorating rapidly. Whereas in 1998 22 percent of the paved road was in bad or regular condition, this percentage rose to 29 percent by 2003. The highways agency has seen its budget frozen in nominal terms during the last decade, which is equivalent to a decline of about one third of its real original value. As a result, it has been forced to focus its attention on the most critical segments, repairing frequent damage from natural disasters, and neglecting routine preventive maintenance of the network as a whole. This is an expensive strategy in the medium term, given that rehabilitation costs twice as much as preventive maintenance in present value terms, and 3.5 times as much as reconstruction. Official estimates indicate that the present backlog, or cost of rehabilitating the network back to a pristine standard, i s in the order of US$380 million. However, if current practices continue, the percentage of the network in bad or regular condition i s projected to increase to 46 percent of the total by 2010, raising the cost o f rehabilitation to US$800 million. The segment of the primary road network that is under concession, on the other hand, presents an acceptable level of serviceability. However, the fiscal costs associated with these concessions has significantly reduced the volume of resources available for preventive maintenance on the rest of the network. There are two reasons for this. First, the traffic guarantees provided under the first generation o f concession contracts were triggered by lower than anticipated traffic flows, resulting in significant payments to the concessionaires (see Chapter 5 for further details). Second, it has been necessary to undertake additional investment work on these roads, that was not foreseen in the original concession contracts, and therefore had to be funded directly b y the state. 46 Figure 2.8: Concentrationof vehicle flows 100% 90% 80% 70% 6 60% g 50% 40% 30% 20% 10% 0% 0% 20% 40% 60% 80% 100% Kmd mads Source: Adapted from Ospina, 2004 2.51 The trucking industry, which serves the road haulage segment of the domestic freight transportation market, operates at low levels of efficiency. Although the trucking industry i s by far the most important mode for surface freight transportation, relatively little i s known about the sector. Trucking i s largely outsourced, with only 25 percent of truck freight carried on shippers' own fleets. However, long term contracts are unusual. The performance of the industry is poor, with an average vehicle mileage of no more than 70,000 kilometers per year, and a utilization rate of only 70 percent. Some 60 percent of the vehicles themselves are more than 20 years old, and are typically powered b y gasoline rather than diesel. Indeed, 79 percent of the 144,000 trucks have only two axles, and only 14 percent are modern tractor units with more than five axles. Modem practices such as separation of tractors from trailers, and rotation of drivers, are relatively unusual. Moreover, vehicle cycles register long delays for loading and unloading, as well as between trips. Security problems add further to the inefficiency by effectively restricting night circulation. Trucks are usually overloaded further contributing to the deterioration of the road network. Regarding international haulage to neighboring Ecuador and Venezuela, although free circulation between Andean Countries i s theoretically permitted by regulation, in practice vehicle transfers must be made at border crossings. 2.52 The organizational structure of the industry is one of the underlying causes of this low operational efficiency. The trucking industry follows a two tier structure, whereby shippers contract with formal transport firms usually lacking a vehicle fleet of their own who themselves in-turn then sub-contract the transportation services to informal owner operators. Owner-operators are highly atomized, with 82 percent owning no more than a single vehicle. There i s major tension between transport firms and operators, prompting government regulations that set mandatory rates for different types of vehicles in different regions, which bear little relation to market determined rates. Informality i s the norm, and i s used to conceal not just transport expenditures, but often the entire commercial transaction. 47 Box 2.2: Evaluation of development of logistics 2.53 The railroad network is not yet making its full potential contribution to freight haulage, but should begin to do so once rehabilitation and integration are complete. Colombia's railroad network comprises 3,300 kilometers of narrow gauge, operated by two major concessions (Atlantic0 with 1,480 kilometers, and Pacific0 with 490 kilometers), two short private lines devoted to transferring coal to sea ports (180 kilometers), and 1,170 kilometers that are currently inactive. Freight volumes increased substantially during the 1990s, due largely to the movement of coal for export, and are among the highest in South America. Network condition i s relatively poor, its 48 rehabilitation being one of the main objectives of the concessions launched in 1998, and supported b y public funds. The original rehabilitation schedule was delayed in 2000, but should be completed by 2008. The two main systems may eventually be linked, pending the results of a technical study. Thereafter, they would be better able to diversify their freight, moving diverse products to Buenaventura port, and linking Bogota with the Atlantic coast. There i s considerable potential demand for railroad services, mostly for dry bulk products and containers, which i s expected to start materializing in 2005. A greater participation of railroads in the future i s expected to reduce both the private and social costs associated with freight transportation. 2.54 Colombia's ports improved considerably during the 1990s, but now need substantial investment to improve access and accommodate increasedtrade under the Free Trade Agreement (FTA). Colombia has several major maritimeport terminals in the Atlantic coast, and one in the Pacific coast. Volumes figures are highly influenced by the exports of coal and oil, which account for three quarters of trading volumes. There are two main type of port facilities: private ports and privately operated public access ports. Private ports are usually dedicated facilities operated by large traders specializing inthe movement of specific primaryproducts such as coal, banana, coffee, oil, and grain. General cargo, typically moved in containers, i s concentrated in public ports operated by five major regional port societies (SPR). These ports were given in concession to the private sector during the 1990s; while other cargo services are provided by multiple stevedoring firms. Following this reform, the SPRs substantially improved their efficiency in loading and unloading of containers. Cartagena, for example, shows remarkable performance indicators, participating in the container transfer business and developing logistic activities around the port. Ports currently face two main constraints. The first i s restricted sea access to both Barranquilla and Buenaventura, and restricted land access to Buenaventura. The second i s the need to increase capacity and improve efficiency to accommodate potential growth in trade volumes under the imminent FTA with the United States. It will therefore be important to assess whether the current concession model, based on a multi-operator scheme, provides adequate incentives for the necessaryimprovements inthe service. 2.55 The Rio Magdalena is potentially an important waterway for freight transportation, but requires dredging works to allow all-season navigability. There are 7,000 kilometers of waterways with permanent navigation in Colombia. The Magdalena river i s by far the most important of these, traversing the country from north to south, and carrying an annual freight volume of approximately 3 million tons. This river presents some seasonal depth constraints of 6' to Barrancabermeja, 4.5' to Puerto Berrio and 3' to Puerto Salgar. However, if navigability could be improved and inter- modal facilities developed, the Magdalena offers the potential to connect river ports close to Bogota with the main Atlantic sea terminals. However, this would be in competition with new roads currently under construction, as well as the rail network in process of rehabilitation, so that the expected demand for three parallel modes would need to be carefully evaluated. There are a number of other waterways that are experiencing growing traffic; primarily the Atrato river, as well as several tributaries in the Amazon and Orinoco basins. Although these waterways carry only modest freight volumes, they 49 provide a vital thoroughfare for isolated regions of the interior, but will probably require dredging works to allow all-season navigability. 2.56 The bulk of commercial air transportation is concentrated in a few large airports, whose main shortcoming is the absence of infrastructure for freight management. Colombia has 74 airports under nationaljurisdiction, and 513 municipal or privately owned installations scattered across the country. Most of the freight and passenger flows are heavily concentrated in a few of the larger airports, while the smaller ones perform a primarily social function in isolated regions. Some of the largest facilities have been given in concession, including Barranquilla, Cali and Cartagena, as well as a second runaway in Bogota. Publicly controlled air traffic control systems are generally adequate, but many airports lack the infrastructure needed for efficient management of freight. This is particularly the case of Bogota, where cargo (mostly flowers) are transshipped directly from truck to aircraft due to the lack of adequate warehouse facilities. 2.57 Multi-modal transportation is still incipient in Colombia. Freight modal interchange and true multi-modal transportation i s still at an initial stage, but i s increasingly attracting government attention. At present, most containers continue to consolidate and de-consolidate in the immediate vicinity of the ports. There i s a lack of development of multi-modal operators, due to some regulatory constraints, the limitations implied by recurrent controls, and the imbalance of container freight between ports and major cities. Security conditions make it more difficult for multi-modal operators to obtain complete cargo liability. Moreover, ports are ill-equipped to handle a multi-modal transportation paradigm. In general, Colombian firms are still in the early stages of adopting modem logistics practices (Box 2.2). 2.58 Summary. Both freight and passenger transportation are almost entirely road based, but traffic levels have been stagnant in recent years. The condition of the paved network i s fair, but deteriorating quite rapidly storing-up future liabilities for rehabilitation. The potential for toll-based concessions on the primary network i s now all but exhausted. A key problem i s the trucking industry, which suffers from structural atomization that prevents the most efficient use of the available stock; meriting closer attention from policy makers. Capacity for rail and river freight will expand in the next few years, and. will provide a valuable alternative to road freight. Ports improved substantially during the 1990s, but now face key bottlenecks in sea and land access, as well as the need to upgrade facilities to handle larger volumes of trade and increase quality of service. Multi-modal transportation and modern logistics services require further development. A VIEW FROMTHE PRODUCTIVE SECTORS 2.59 As a complement to the supply side view of the infrastructure sectors presented above, this section presentsthe viewpoint of large industrial consumers of 50 infrastructure services. This helps to provide a more complete viewpoint of where services currently stand, as well as a diagnosis of key areas where improvements are needed. The survey covers a sample of 61 firms drawn from a variety of productive sectors and geographical locations. Although the original intention was to cover a larger sample, many firms were unwilling to provide quantitative information about the contribution of infrastructure services to their production process. As a result, the survey i s only representative at the national level, but not at the level of individual regions or productive sectors. The findings basically confirm some of the shortcomings identified in the supply side overview, adding a number of new insights, and helping to weight the relative importance of the various issues that arise from the perspective of the productive sector. The importance of infrastructure for the productive sectors 2.60 According to the national accounts, infrastructure services represent at least 16 percent of the total production costs of Colombian firms. The national input-output matrix provides one approach for estimating the weight of infrastructure services in the cost structure of Colombian firms. According to this method, infrastructure services account for about 16 percent of the total production costs of Colombian firms, with energy, transportation and telecommunications each representing close to 5 percent of the total. However, there i s enormous variance across sectors (Table 2.9). Thus, the total weight of infrastructure services in the cost structure ranges from 4 percent (construction) to 34 percent (financial services). Telecommunications i s the most significant infrastructure service for firms in the financial and personal services sector, while transportation weighs most heavily for the agricultural and hospitality sectors and energy carries the greatest weight for manufacturing. Nevertheless, it i s worth noting that these figures are likely to represent an under-estimate given that the input-output matrix only reflects the infrastructure services purchased from dedicated infrastructure providers, overlooking in-house provision of infrastructure services within the productive sector. This can be substantial as, for example, when firms maintain their own transportation fleets, auto-generate or co-generate power, or operate private wells for water supply. Table 2.9: Infrastructure intensity by productive sector ____ percent of total costs Power Telecomn Water ai Total and gas Transport unications sanitatioi -~ and mail Social, community and personal services 7.2 5.2 9.9 1.2 23.5 Financial, insurance, real estate and other 0.3 34.0 services 11.1 8.9 13.7 Agriculture, cattle, forestry, fishing 1.7 9.9 5.9 0.1 17.6 Manufacturing 5.8 2.2 1.o 0.3 9.3 Trade, restaurants, hotels 3.8 16.0 6.9 1.4 28.1 Construction 0.5 2.1 1.o 0.4 4.0 ____ Average 16.0 Source: National Input-Output Matrix, DANE2001 51 2.61 Although the manufacturing sector as a whole is not very infrastructure intensive (9 percent), in a number of important sectors infrastructure exceeds 15 percent of total costs. According to the national accounts, infrastructure services costs represent 9.3 percent of the total costs in the manufacturing sector with energy (5.8 percent) and transport (2.2 percent) being the two most significant services. However, there i s wide variation across different areas of manufacturing, and a number of sectors present a strong reliance on infrastructure, typically on one of the services in particular (Table 2.10). Key examples are glass, metals, textiles, rubber and printed materials with 13-26 percent of costs going on infrastructure services, and the bulk of these on energy, or beverages with 18 percent of costs going on infrastructure services, and more than half of these on transport. This can be contrasted with sectors such as coffee and leather, which barely devote 4 percent of their costs to infrastructure services. Table 2.10: Top 10 infrastructure intensive manufacturing activities Percentof total costs Power Telecommunications Water and Total and gas Transport and mail sanitation Glass and non-metal products 18.4 4.3 2.5 0.3 25.5 Commonmetals 14.0 2.1 2.5 0.2 18.8 Beverages 6.0 9.5 1.1 1.5 18.1 Textiles 14.1 1.4 1.7 0.5 17.6 Rubber and plastics 9.4 2.0 1.7 0.3 13.5 Printedmaterials 7.7 3.2 1.3 0.6 12.7 Other machinery and electric 0.3 11.6 provision 7.4 2.1 I.8 Sugar 7.3 2.7 0.8 0.4 11.2 Refinedfuel products 7.6 2.3 1.o 0.1 11.0 Machinerv 5.3 2.6 1.5 0.4 9.7 Source:NationalInput-OutputMatrix, DANE2001 2.62 However, survey evidence suggests that infrastructure services account for 16 percent of the costs of the manufacturing sector, with 25 percent of services provided in-house. In order to obtain a more direct measure of the weight of infrastructure services in the cost structure of a number of key manufacturing sectors, an ad hoc survey of firms was conducted. The survey took place from March through May of 2004 and covered 61 firms operating in 14 productive sectors spread across Colombia's four largest productive centers (Barranquilla, Bogota, Cali, and Medellin). This survey found that infrastructure services accounted for 16 percent of the costs of manufacturing firms on average, compared with the result of 9 percent for the manufacturing sector from the input-output matrix (Table 2.11). Consistent with the evidence from the national accounts, energy and transport are the most significant services for the manufacturing sector, each accounting for 36 percent o f total costs. The results of the survey suggest that about 25 percent of infrastructure services by value are vertically integrated within the manufacturing firms, and hence do not show-up in the input-output matrix. 52 Table 2.11: Infrastructure intensity for representative productive activities As a percentof costs Energy 5.8 Transport 5.8 1 Total I 16.1 Source: Maldonado and Sardi, 2004 2.63 Electricity is by far the dominant energy source for the productive sector; however, deficiencies inpublic supply lead 40 percent of firms to maintaintheir own facilities. Electric power i s the dominant source of energy in the productive sector, accounting for more than 80 percent of total energy consumption in more than 60 percent of the firms surveyed. More than 60 percent of firms surveyed complain that electricity prices are high, and 30 percent find deficiencies in the quality of service, with half of them experiencing more than one supply interruption per month (Figure 2.9(a)). As a result, almost 40 percent of firms surveyed have some sort of own generation facility. About three quarters of these rely primarily on their own production of power (mainly through co-generation), while the remainder maintain backup facilities as a contingency against failures in public supply. O f those firms that have their own generation capacity, almost two thirds state the need for more reliable energy supply as the primary motivation (Figure 2.9(b)). Moreover, although 60 percent of firms surveyed complain about the cost of public electricity supply, only 20 percent of those with their own generating plant are motivated by the desire to reduce costs. This indicates that quality rather than price i s the driving factor in prompting firms to invest in their own generating plant. Finally, about 80 percent of firms surveyed consider that quality of service improved following the entry of private operators, with fewer interruptions and improved client service. Figure 2.9: Perception of energy services To% , 70%7 80% - 60% 50% 6 40% $W I : 30% a 20% 10% 7Electricrty UPPhl 0% More reliable Cheaper Easier access Other BGS (b) Reasonsfor own generation 2.64 Firms see little use for natural gas. Very few of the firms surveyed see natural gas as a relevant energy source, finding coal supplies to be much more competitive for industrial purposes. About 40 percent of firms surveyed do not have access to the fuel, while 30 percent consider the fuel too expensive, and a further 20 percent too unreliable. 53 2.65 Around 80 percent of firms outsource road haulage to traditional transport operators, and are primarily concerned about infrastructure quality and security on the roads. There i s a remarkable lack of detailed information about transportation practices in the firms surveyed. As indicated by sector statistics, road haulage i s dominant, with 65 percent of firms moving more than 80 percent of their freight by truck. About 21 percent of the firms utilize their own fleets, and the remainder outsource to transport operators. There i s an incipient trend towards outsourcing to logistic operators, both to reduce costs (58 percent) and to allow firms to concentrate on their core activities (33 percent). However, the usage of multi-modal transportation i s low, affecting only 24 percent of firms surveyed. Delays are frequent and relatively long, with 50 percent of firms experiencing delays of more than four days in international cargo, and 27 percent experiencing problems in ports and airports. When asked to compare transportation services with those of neighboring countries, around 70 percent o f firms surveyed cite security and the quality of road infrastructure as worse in Colombia. The majority see most other aspects of transport as very similar to neighboring countries, with 20 percent rating the customs system as better than neighboring countries. Figure 2.10: Perception of transport servicesin Colombia relative to other countries 8096 e E 70% 80% : 50% 40% c 5 30% 20% a 10% 0% Source: Maldonado and Sardi, 2004 2.66 Firms are generally satisfied with telecommunications services, and particularly appreciate the ability to switch between providers. However, delays for new service remain long. The assessment of telecommunications service quality i s generally good, but i s significantly better for local and long distance services than for mobile and data transmission services (Figure 2.1l(a)). Almost half of the firms surveyed complain about service interruptions on fixed line telephony services. Firms are taking considerable advantage of the choice offered by competition in the telecommunications sector, with 24 percent having recently switched service provider for fixed line telephony, and 50 percent having switched internet service provider. About three quarters of these switches are primarily motivated by the desire to reduce costs. Another frequent complaint relates to the delays in obtaining new fixed line telephone service, with three quarters of firms surveyed waiting more than 60 days to be connected. 54 2.67 Water and sewerage services do not seem to represent a critical factor of productionfor most firms. The majority of the firms surveyed find water and sanitation services to be of very good quality. However, half of them complain that water and sewerage services are too expensive (Figure 2.11(b)). Two thirds of firms surveyed mention that water quality i s important for their business. Figure 2.11: Firmsatisfaction with services 50% OLocal 45% OLD national i35% E 4)% 6 33% I L D internationa 5E 0 25% mModerate 20% DCellular a 15% OGood 10% 5% O V e y goo< Cmt Customer Qualty 0% seruce c m t Qualrty I (a) Telecommunications (b) Water Source: Maldonado and Sardi, 2004 The incidence of infrastructure inselected value chains 2.68 Further insight can be gained by analyzing the situation of specific productive chains with respect to infrastructure services: textiles, paper and sugar. To provide a deeper understanding of the ways in which infrastructure services are presently contributing or restricting the development of productive activities in Colombia, three value chains are analyzed in greater detail. These three chains are textiles, paper and sugar. They were selected on the basis of contribution to GDP, , strategic importance, utilization of different infrastructure services, and geographical distribution. Table 2.12: Comparison of the three value chains SY Locationof 1 GDP chain Textiles ----E ~ - Fiber Dispersed percent, via 6.0 4.0 - Textiles and Bogota, Atlantic ports II II garments ___- Medellin 1 Paper 4 1 . 2 Valle del By road - Paper About 20 Cauca percent,via 7.0 12.0 19.0 I ~ - Printed Cundinama Buenaventura 5.0 9.0 16.0 materials II rca, Boyaca L-__.-- Sugar Cauca and i - Sugar refining 24 Valle del 4.0 3.0 -preserves I Cauca 5.0 II2.5 L__.--L_ _ ~ - 55 2.69 These three value chains, which together account for 5 percent of GDP, present very different geographical patterns of production. In aggregate, these three value chains account for almost 5 percent of Colombia's GDP. The textile value chain includes fiber production, spinning mills, textile production, and garments. Whereas fiber production i s widely dispersed, the latter stages of the chain are highly vertically integrated and concentrated in Bogota and Medellin, with substantial outsourcing of garment production. The paper value chain includes fiber production, pulp processing, production of paper and cardboard and printing. The early stages o f the chain are concentrated in the Valle del Cauca, and involve highly dispersed producers, entailing significant production costs for road haulage for processing. Curiously, paper producers pay pulp mills in coal, which they are responsible for hauling from Cundinamarca and Boyaca. The sugar value chain includes sugar cane production, sugar refining and production of preserves, and i s highly concentrated in Cauca and Valle del Cauca making this arelatively simple chain with highvolumes and low margins. 2.70 All three chains are export-focused and intensive in the use of energy and transport infrastructure. Paper production i s the only chain to rely significantly on imports, with 20 percent of raw materials (wood and recycled paper) imported through the Pacific port of Buenaventura. A substantial proportion of the final output of the three chains i s exported, with textiles and sugar relying respectively on Atlantic and Pacific ports, and paper products usingroad haulage to neighboring countries. The infrastructure intensity of these value chains ranges from 8 percent (sugar) to 19 percent (paper production), with sugar and textiles being particularly reliant on energy and paper particularly reliant on transport (Table 2.12). It i s important to note that the percentages reported in Table 2.12 do not coincide with those reported for the same industry in Table 2.10; the reason i s that the latter presents comprehensive national statistics, while the former presents the results of a non-representative sample selected with the primary purpose of providing in-depth qualitative information. 2.71 Firms are concerned about the price of electricity, and the frequency of service interruptions. Sugar, paper and textile production are particularly intensive in the use of energy. Most firms interviewed thought that electricity prices compare reasonably to those in competing neighboring countries (such as Brazil, Mexico and Central America), although prices in Brazil and Mexico are expected to fall in the future. Nevertheless, electricity prices are still thought to be high in absolute terms, due to the 20 percent surcharge on industrial users. Moreover, there are significant concerns about service interruptions linked to terrorist attacks, particularly among sugar industry interviewees. In the textile sector, small and medium-sized enterprises lack protection from voltage fluctuations and can be adversely affected. 2.72 As a result, many firms are looking to expand auto-generation but find themselves subject to regulatory restrictions on this activity. For all o f these reasons, across the three value chains, firms are looking to increase self-generation and co- generation. In the sugar industry, where co-generation by sugar m i l l s is already very significant, there are complaints about regulations that prevent the sale of excess power over the grid, as well as the transfer of energy between production processes for vertically integrated firms. Inthe textile industry, firms cited financial constraints on self- 56 generation, and expressed ambivalence about entering into activities that distract from their core business. 2.73 There is unanimous rejection of natural gas, which is seen to be costly and uncertain in supply. Natural gas i s universally regarded as too expensive to be relevant, particularly to firms located on the Pacific coast who are affected by distance related transportation tolls. Furthermore, those interviewed perceived uncertainties about security of supply in the medium term, given uncertain prospects for the sector and the difficulty of signing long term contracts. 2.74 In the early stages of the production chains, the central concern is the high cost of transportation including security concerns. Whereas sugar mills tend to rely on their own vehicle fleets, paper and textile producers outsource road haulage and complain about high regulated truck fares. Truck rates for imports are particularly high, with loads from the Atlantic coast to the centers of production charging 50 percent more than the reverse fare due to imbalances inthe pattern of trade. The security situation inthe country adds substantially to the costs of transportation; by requiring security escorts for vehicles, inflating insurance costs, and restricting nocturnal circulation of trucks. However, this has fortunately been improving in recent years. Both paper and sugar producers interviewed lament the absence of a railroad service, that could potentially reduce transportation costs by as much as a half for wood, bagasse and paper. 2.75 In the later stages of the production chain, quality becomes the over-riding concern with transportation services. Printed materials are particularly sensitive to travel time, due to demand variations across seasons and markets. Garment exports also require high quality logistics given the need to coordinate with large numbers of satellite producers, and to meet the short lead times of very demanding clients. Furthermore, Colombia is disadvantaged by being locatedfurther from the end market than competitors in Central America. As a result, some of the firms interviewed are beginning to develop their own logistics operators. Part of the problem is that use of containers i s still incipient, and confined to a few of the larger firms in the sugar and paper industry. This leads to frequentlosses of cargo, prompting complaints by foreign customers. 2.76 As far as exports are concerned, there are major concerns about access to the Pacific port of Buenaventura. Buenaventura port on the Pacific Coast accounts for 9 percent of Colombia's sea trade volume, but 44 percent o f the trade volume channeled through public access (as opposed to private) ports. The use o f Buenaventura as a gateway for exports presents a whole range of serious difficulties for those firms interviewed, even prompting some sugar producers to export via Cartagena on the other side of the country. Interviewees explained that the problems begin with the road links. Security on the critical stretch from Cali to Buenaventura i s particularly bad, with trucks facing threats as diverse as piracy and falling rocks. Furthermore, lack o f road capacity on the entrance to the port creates serious congestion. Once it arrives, cargo faces major delays in entering into the port. For example, in a typical garment export from Medellin through Buenaventura, the garment spends 22 hours reaching the port entrance, and a further 43 hours until it i s cleared for entering the port terminal. It i s important to clarify that these procedures are not related to the management o f the port facility itself, but 57 rather to infrastructure access and control procedures. Lengthy inspection processes linked to drug control are largely responsible, and particularly penalize small and medium enterprises. Finally, low frequency of maritime services from Chile to Buenaventura affect the import of pulp by paper producers. B y contrast, few problems are reported with roadexports to Venezuela and Ecuador. 2.77 Regarding telecommunications, the main concern is the absence of service in rural areas making it difficult to coordinate across different stages in the chain. Although telecommunications services do not carry significant weight in the cost structures of the interviewed firms, all three industries considered these services to be o f strategic importance. Given that all of these chains rely on highly dispersed rural production for their raw materials, a common complaint was the absence of telephone coverage in rural areas to enable adequate coordination between suppliers and downstream producers, and the high cost o f satellite connections. Sugar producers contrast this with the situation in neighboring Brazil, where there is more reasonably priced satellite access in rural areas. This finding suggests that there i s an untapped commercial demand for telecommunications in rural areas, which may require a combination of local fixed line service, extension of cellular networks and improved satellite service. The government's rural and social telephony programs (described extensively in Chapter 3) have so far overlooked this dimension of telecommunications in rural areas. 2.78 Water and sanitation services have limited impact on production, and self- supply is widespread. Water and sanitation services have a negligible incidence in the cost of production for the three value chains considered, even in the case of paper production, which is quite water intensive. Self-supply through private boreholes i s typical in paper and sugar production. Water quality was only cited as relevant for the dyeing process in textile production. 2.79 Summary. Infrastructure services represent at least 16 percent of the total production costs of firms in Colombia. There i s considerable variance across productive sectors. Energy and transport are the predominant costs for manufacturing producers, while telecommunications are particularly significant for firms in the service sectors. Concerns about the efficiency andreliability of infrastructure services lead manufacturing firms surveyed to self-supply about 25 percent of their infrastructure services by value; with 40 percent of firms running their own electricity generators, and 20 percent operating their own fleet of trucks. Surveyed firms complain about power supply interruptions, lengthy waiting lists for fixed line telephones, and considerable delays and security risks on road haulage. The opportunity to switch between competing telecommunications providers i s greatly appreciated and frequently exercised. Moreover, firms surveyed report greater satisfaction with services provided b y private versus public operators. 2.80 Surveyed firms wishing to circumvent public power supplies find themselves restricted by regulations that prevent sales of excess power over the grid, or even transfer of power across different stages of production by vertically integrated firms. Road haulage services are found to be very costly in the early stages of production, in 58 part due to restrictions on nocturnal circulation. In the later stages of production, the concern shifts more to quality considerations, given that road haulage services are not sufficiently agile to respond to real time customer demands. Exporters surveyed complain bitterly about the considerable costs, risks and delays associated with the port of Buenaventura road access. Firms reliant on agricultural raw materials, find themselves limited by scarce and costly telecommunications services inrural areas. TOWARDS AN INFRASTRUCTURE FOR GROWTH AGENDA 2.81 In defining the infrastructure investment agenda, attention in this chapter focuses on so-called `productive infrastructure' defined in Chapter 1 to include electricity generation capacity, telecommunications infrastructure, and paved roads, as well as sea and air ports. Chapter 3 will provide a detailed analysis of investment needs in `social infrastructure' defined to include household access to basic utility services, as well as ruralroads and urbantransportation services. A growth based vision of infrastructureneeds 2.82 Colombia's projected economic growth will be underpinned by the expansion of some critical productive sectors. Macroeconomic analysis predicts economic growth at an annual rate of 3 percent for the next five years. This growth projection i s under- pinnedby the expansion of three areas of productive activity in particular. First, a number of non-traditional sectors that have shown a growing participation in the country's exports in recent years and where significant investments have already been made. Examples include textiles and publications. Second, Colombia's traditional agricultural export such as bananas, coffee and flowers are expected to continue playing an important role. Third, a number of other sectors, whose international competitiveness i s not yet confirmed, but which will continue to play an important role in the domestic economy. These include glass, metals and plastics. 59 Table 2.13: Infrastructure intensity inthe primary growth-leading sectors I Energy tnfrastructure Traditional sectors -- Bananas Flowers - Sugar I * *** - Coffee - Oil and minerals ** *** I ~ Non-traditional sectors - Chemicals ** * *** **** ** *** 4 - Vehicles - Plasticsand rubber **** **** ~ - Glass **** - Metallic products *** **** **** I Source: Maldonado and Sardi, 2004 2.83 The non-traditional growth leading sectors are more intensive in their use of infrastructure than the country's traditional export activities. The infrastructure intensity of Colombia's key growth leading sectors can be measured using official statistics, adjusted on the basis of survey results to correct for under-reporting of self- supplied infrastructure services. The results suggest that non-traditional export sectors and other strategic productive activities are more infrastructure intensive than traditional mining and agricultural exports (Table 2.13). This suggests that the marginal infrastructure intensity of production may be higher than historical average levels. Many of the newer productive activities are particularly reliant on energy services, requiring both quantity and reliability of energy, due to the nature of the production processes involved as well as the demanding schedules of the international markets. While transportation services do not represent such a high proportion of costs for these higher value goods, they do place substantial demands on the quality of transportation services, requiringsophisticated logistics to ensure the efficient operation of the supply chain. This i s particularly true in those sectors that are heavily reliant on imported inputs, such as automotive, chemicals, metals, and textiles. In addition, metals and textiles also require good telecommunications, while food and textiles are sensitive to water quality issues. 2.84 The imminent FTA, as well as other ongoingtrade negotiations,will alter the composition and routes of existing trade flows composition and routes in uncertain ways. Colombia i s expecting to sign a Free Trade Agreement with the United States in early 2005, while several other regional and extra-regional trade agreements are currently under discussion. The primary impact of these trade agreements on infrastructure requirements are likely to be felt in the transport sector, impacting mostly on traffic through international gateways (ports, airports, border crossings), through variations in trade volume, composition and routes. However, it i s difficult to estimate those impacts 60 with any great degree of precision. Although a number of modeling efforts exist, these suffer from a number of methodological problems. First,the models are usually static and assume constant returns to scale, thereby failing to capture important aspects of reality. Second, the available trade data i s highly aggregated across products and regions, and available only for years whose trade characteristics differ substantially from the present ones. Third, the baseline scenario i s impossible to predict beyond the year 2006, when the present United State's Andean Trade Promotion and Drug Eradication Act (APTDEA) trading rules expire. Fourth, the free trading scenarios are themselves uncertain due to the fact that many of the key parameters have yet to be decided (scope of products to be included in the FTA, changes to existing agreements with EUand Mercosur). 2.85 Current estimates show significant increases in Colombia's overall trade volume, however with a consistently larger impact on imports than on exports. Existing analyses (Light and Rutherford, 2003, and DEE-DDE, 2003) anticipate significant changes in trade volume. Imports are expected to grow b y between 12-24 percent under different scenarios; while exports are predicted to growth at rates ranging between 2-8 percent. This asymmetry i s attributed to the fact that Colombia's existing tariffs on import tariffs tend to be higher than the trade barriers currently faced by Colombia in the U S and other regional markets (Light and Rutherford, 2003). It i s worth noting that the general equilibrium models commonly utilized may have a bias towards the underestimation of the overall trade variations under tariffs liberalization. As regards the composition of the trade flow changes, Table 2.14 summarizes those sectors in which significant increase or decrease i s expected, in proportions that may vary greatly with the scenario considered. Table 2.14: Trade flows expectedto increase or decrease after F T A Likely to increase Likely to decrease Exports Wearing apparel Textiles Cerealand grains sugar Motor vehicle andparts Coffee, vegetableoil, dairy productsL - Imports Cereal andgrains Meet Textiles Manufacturedgoods __ _______ ~ Source: Own elaboration, based on Light and Rutherford, 2003, and DEE-DDE, 2003 2.86 The FTA will imply a considerable and abrupt incremental pressure on infrastructure, mainly on transport gateways linked to international freight trade. According to the present estimates, the expected trade growth will be the equivalent to less than one year's average annual historic growth rate for exports (which was 6.7 percent per year from 1994 to 2003), but three to five years' average annual historic growth rate for imports (which was 4.4 percent in the same period). The projected increase will probably be concentrated in a relatively short period of time, namely two or three years. The increase will involve products with an average unit value (US$ per ton) higher than the present one, as a result of the growing incidence of imported manufactured goods, textiles and meat. The most impacted infrastructure services will be the international transport gateways, mainly ports and airports, and eventually border 61 crossings. Road traffic i s expected to increase somewhat (in proportion to the GDP impact of the FTA), however incremental traffic flows will be concentrated in specific segments linking the international gateways with the major centers o f production and consumption. The trucking industry i s thus expected to face growing volumes; while there will be heightened demand for high quality logistics services, particularly as regards container movements, freight facilities and multi-modal coordination. Both firms' supply chain organization and third party logistics operators will be challenged by a more competitive environment. 2.87 In the longer term, the FTA may create incentives for new industries to locate in different regions of the country. The geographical distance between Colombia's centers of economic production and its international gateways, combined with the country's challenging topography; make transport issues particularly important from the perspective of international trade. Inthe longer term, this may create economic incentives for new industries to locate closer to Atlantic ports. However, this will only be possible to the extent that other infrastructure and business conditions on the Atlantic coast are also favorable. Such considerations point to the need to incorporate a regional policy dimension into longer term strategic planning for the transport network. Strategic priorities by infrastructure sector 2.88 The foregoing analysis provides a solid basis for identifying priority measures that will enhance the ability of infrastructure services to support the country's economic growth. B y combining the supply-side diagnostic, with the perspective of the industrial consumer, and the identification o f the productive sectors likely to make the greatest contribution to the growth of the economy, it i s possible to identify the highest priority measures that need to be taken to ensure that the infrastructure services make their necessary contribution to the development of the country. The key areas for action in each sector are identified below, and classified according to whether they are primarily investment issues or require other types of interventions including policy and regulatory changes, or must combine investment with policy and regulatory reforms (Table 2.15). The two are complementary in that the regulatory framework has a strong influence on investment incentives for the private sector. 62 Table 2.15: Key areas for action on infrastructure services I Sector j 1Key actions Intervention Electricity I Safeguard security of energy supply Investmentplus I Reduce vulnerability of transmissionnetwork Investment only ~ Liberalize regulationson co-generation. Other measures Reduce industrial electricity tariffs. Other measures 1 Gas Ensure supply of natural gas to new thermal plants Investmentplus Transport ' Modemize road infrastructure Investmentplus Reform the trucking industry Investmentplus Promote better logistics practices Investmentplus Prepareport terminals for l T A Investmentplus 1 Develop alternativetransport modes (rail, river, airports) Investmentplus I I I Investmentplus Investmentplus I Improve access to telecommunications in primary (rural) Investmentonly stages of production I Water Improve water quality in smaller towns Investmentplus i Reduce industrial water tariffs Other measures I Clarify environmental standards Investmentpius Source: Own elaboration 2.89 In the electricity sector, the emphasis should be on improving reliability of service and reducing costs to industrial users. Electricity was shown to be the most intensively used of the infrastructure services by the productive sector, particularly in those industries thought to be strategic in terms o f the country's medium term economic expansion. In order to support economic growth and competitiveness, the priority should be to ensure adequate quantity and quality of supply (particularly in terms of reliability), as well as reducing costs to industrial users. The following measures would appear to be particularly important. (a) Safeguard security of energy supply. Investments in new generation capacity will need to maintain the prudent balance between thermal and hydro sources that have evolved in the last decade. This will undoubtedly require the creation of a more stable regulatory environment for the electricity sector so that incentives for private investment are recovered. (b) Reduce vulnerability in transmission. The transmission network needs investments in a number of bottleneck areas, as well as reinforcements on some existing segments to reduce the vulnerability entailed by the radial topology of the network. Both o f the above measures, by improving reliability of supply, should help to reduce the need for firms to maintain their own back-up generation facilities. (c) Liberalize regulations on co-generation. Measures should be taken to liberalize the regulatory regime for co-generation that currently prevents firms from selling excess power to the national grid or transferring it across stages in the production chain, making this economically efficient 63 form of power generation more costly to firms in financial terms. This will require a balancing of competitiveness considerations against broader equity issues. (d) Reduce industrial electricity tariffs. The surcharges applied to industrial electricity tariffs as part o f the national cross-subsidy framework are having a significant impact on production costs, and may need to be reduced as part of a broader reform to the system (see Chapter 3 for further discussion). 2.90 In the natural gas sector, the main concern is to assure supplies to gas-fired electricity generators. Natural gas i s far from presenting a competitive fuel for direct industrial use; however its role as an input to thermal electricity generation i s indirectly critical to ensuring the reliability of electricity supply to industrial users. Hence, the main strategic concern i s to assure supplies of the fuel to gas-fired electricity generators, in coordination with expansion plans for the generation sector. Furthermore, any future decision to go ahead with exports to neighboring countries will necessitate major investment ingas transportation infrastructure. 2.91 In the transport sector, coordinated interventions are needed both to improve the road network and the efficiency of haulage services provided over the network. For maximum impact on productive sectors, government interventions need to address deficiencies both with the road network itself, but also with the trucking industry that provides services over that network. The latter i s not a traditional focus of government transport policy, but i s critical to ensuring its ultimate success. The main areas for action include the following. (a) Modernize road infrastructure. The key priority here i s to reduce the maintenance backlog, and create a stable source o f finance for on-going essential maintenance activities. Although some major road links are likely to be necessary, upgrading existing facilities along critical corridors appears as the main priority. First, standards on trunk corridors should continue to improve in order to reduce vehicle operating costs and circulation times. Second, there i s scope for paving some low density links that are experiencing significant growth in traffic volumes. Third, connections with neighboring countries need to be up-graded. Fourth, links between productive rural areas and economic centers should be improved. (b) Reform the trucking industry. There i s a need to rethink public policy towards the trucking industry, in such a way as to improve the efficiency and quality o f service, and bringing it into the formal economy, while carefully managing the potential social impacts. Many specific issues need to be tackled including the rate-setting framework, incentives for fleet modernization, technical regulations and their enforcement, and defining the appropriate roles of different industry stakeholders. There is also a need to improve security on key road links, both inurban and rural areas. 64 Promote better logistics practices. Primary responsibility for improving logistics practices rests with the private sector, both with producers who need to streamline their supply chains and with carriers who need to transform themselves into modern logistics operators. However, government also has an important role to play in adapting regulations (such as those governing container movements), training small and medium enterprises, and investing in strategically located inter-modal transfer facilities and logistics centers (where traditional logistic functions are consolidated and complemented with other value added functions, like packaging and inventory management). Prepare port terminals for FTA. The FTA will make major additional demands on existing port infrastructure capacity and operational efficiency, demanding considerable new investments. On the,government side, there i s the need to improve land and sea accesses to the major ports. On the private sector side, sea terminal facilities need to be improved and up-graded. However, this will require a reform of the regulatory framework to strengthen the investment incentives of existing operators. Develop alternative transport modes. On the strategic corridor connecting Bogota to the Atlantic ports, railroad and river navigation provide potentially attractive substitutes to road transportation, but will require significant investments, some of which are already underway. However, it i s not clear that all three modes could be supported along this route, so that an integrated view needs to be taken on investment priorities across these modes. There i s a need to increase capacity in airports, in order to accommodate increased freight and passenger flows associated with growing international trade and tourism, with particular focus on Bogota's freight facilities. Furthermore there i s a scope for diversifying railroad freight from its current concentration inthe coal sector. 2.92 In the telecommunications sector, the key challenges are the expansion of broadband in urban areas, and the improvement of telecommunications for rural production units. The following strategic priorities have been identified for the telecommunications sector. (a) Improve quality of telephony service. There i s a need to redress some significant deficiencies in the quality o f telephony service, in particular waiting lists for new lines and call completion rates. (b) Expand access to broadband internet. Internet is becoming a major support to some key business activities, like the management of coordinated supply chains, inventory control, procurement management (through business to business electronic transactions), and logistics operations. The provision of broadband internet should be ensured in all major productive centers and gateways (including ports, airports, border 65 crossings, free trade zones or logistics centers), which will entail a significant broadening of access over current levels. (c) Improve access to telecommunications in primary production. Some value chains involve primary productive centers scattered across rural areas. Communication services, such as reasonably priced mobile telephony, microwaves or satellite connections, are essential to improve coordination of product flows with downstream links inthe supply chain. 2.93 In the water sector, concerns center on water quality in smaller towns, as well as the high level of industrial water charges. The key challenges for the water sector are as follows. (a) Improve water quality in smaller towns. Although water quality is quite good in the largest cities, it should be improved in other intermediate and minor urban centers. (b) Reduce industrial water tariffs. The surcharges paid by industrial water users in the national cross-subsidy system are extremely high, amounting to 45 percent over true costs of service provision on average. They also constitute a significant incentive for firms to seek their own independent supply of water, reducing the revenue base for the utility sector. However, any changes to these surcharges need to be considered in the late of more general reforms to the cross-subsidy system (see Chapter 3 below). (c) Clarify environmental standards. The provision of water from own wells and the treatment of residual water, if not properly undertaken, may cause severe environmental damage. In order to ensure sustainable development, these externalities should be properly consjdered. National and local governments should establish and enforce the suitable standards. 2.94 Finally, the potential for regional infrastructure integration has only been partially achieved, leaving significant room for progress. Colombia's pivotal location between South and Central America, makes it well placed to act as an integrating link between sub-regional infrastructure networks. The current status o f integration in each of the sectors can be described as follows. Electricity. Integration of electricity systems brings numerous benefits including the reduction of generation and transmission risks, as well as the compensation of local consumption peaks. Colombia already has connections with Venezuela (three links with a combined capacity of 380 megawatts) and Ecuador (two links with a combined capacity of 250 megawatts), and i s behaving mostly as an energy exporter (CIER, 2003); there i s no link with Panama or Brazil. Gas. The government has recently announced its intention to enter into a gas export project with Venezuela, entailing the need to develop dedicated 66 transportation infrastructure. This project i s regarded as the first phase of a regional integration project that could eventually link Central and South America. (c) Telecom. The industry has made substantial advances inthe integration of regional fiber optic networks. However, significant opportunities remain inimproving regional coordination of satellite facilities. (d) Transport. Integration facilitates trade and border areas development; connections with Venezuela and Ecuador do exist and are widely used, though concentrated in few links. Road connections with Panama and Brazil are more difficult due to sensitive environmental characteristics. However, the full benefits o f road integration are not achieved until trucking industry regulations are harmonized. Although the Andean countries have signedtreaties allowing for full operational integration, this i s not currently taking place due to opposition from local operators. Thus freight traveling from Colombia to Venezuela, for example, must be transferred from one vehicle to another at the frontier. 2.95 A number of priority projects have been identified under the Initiative for the Regional Integration of Infrastructure in South America (IIRSA). There are many advantageous opportunities for regional integration with neighbor countries. These have been identified by IIRSA during the last three years, and include improvement of road, river and gas transportation links with neighboring countries, as well as greater harmonization of trucking regulations (Table 2.16) (IIRSA, 2003). It i s also important to note that projects aimed at improving regional integration, also contribute to greater territorial cohesion within Colombia, and thereby bring an additional social dimension of benefits. 67 Table 2.16: Regional integration priorities Gas , Development of fields and gas pipeline to allow for export to western Venezuela. Roads Improvementof existing roadlinks and border crossings with Venezuela and Ecuador along a number of key comdors (some already underway): I @ Bogota- Buenaventura,Pasto - Mocoa; La Espriella - Rio Mataje. River Navigation Improvementof navigability of the Rio Metaand development of border facilities. j i of freight transfer in border crossings I Source: Own elaboration Infrastructure investmentplan 2.96 The strategic priorities identified above point to a specific investment agenda to support the growth and competitiveness of the Colombian economy. This i s developed below, while the strategic regulatory issues will be further developed in Chapter 5. Attention focuses on the electricity and transport sectors, where the bulk of necessary investments inproductive infrastructure are concentrated. Electricity 2.97 Electricity demand is projected to grow somewhat faster than GDP, reflecting growing exports and more energy intensive industries. Colombian GDP i s projectedto grow at 3 percent per year over the medium term. Although it i s known that electricity demand has GDP elasticity close to one, a slightly higher electricity demand growth scenario of 4.1 percent per year i s used. This i s to account for growth in electricity exports as well as the shift towards more energy-intensive sectors of production, as noted above. Projected investment requirements are based on the official expansion plan for generation andtransmission up untilthe year 2011. 2.98 This will entail investments in generation capacity of US$258 million per year, in order to add roughly equal amounts of hydro and thermal capacity. The expansion path involves adding 1,500 mega-watts of new capacity by 2011 (Table 2.17). Two major hydro projects are anticipated, in addition to 51 mega-watts of hydro and eolic capacity already under way. The first i s Amoya, a 78 mega-watt hydro plant that would come on line in year 2008. Since it has been designed without a dam, it qualifies to sell carbon bonds within the Kyoto Protocol mechanisms. The second i s Porce 111, a large 660 mega-watt scheme under development by EPM in Antioquia and expected to start producing by 2011. The remaining 750 mega-watts of the requirement would be met by combined cycle gas-fired plants on the Atlantic coast. The implementation of this expansion path i s expected to slightly reduce the share of hydro power in the generating portfolio from 67 percent to 65 percent, thereby helping to safeguard security of supply. 68 111 Projects Type i'11Capacity (MW) 1 I Completiondate ~ Jepirachi,LaVuelta, La Herradura I'Hydro, eolic 51 2004 AmoyA I 8 12008 3 CCGTplants (Atlantic coast) $;I%I 750 2008,2009,2010 1 i Porce 111 I Hydro 1 660 12011 2.99 Inthe transmissionsegment, some linkswill need to be reinforced, and new ones built, entailing average annual investment of US$43 million per year for the next five years. The transmission network needs to develop in a manner consistent with the projected level and spatial distribution of energy generation and consumption. Planned projects include the reinforcement of congested links, and the addition of new substations and links on the national transmission network. New circuits and substations for the regional transmission network are also considered. The two largest projects, UPME 01 and UPME 02 are relatively urgent, needing to be completed by 2007 so as to avoid congestion in the eastern central region of the country. This will lead to a concentration of investment requirements during the next three years. 2.100 Inthe distribution sector, a rehabilitation programof US$43 million per year i s needed to halve distribution losses over the next decade. In order to address the productive sector's key concern about the quality and reliability of electricity service, an investment program has been designed with the aim of reducing distribution losses from their current average level of 25 percent to 12 percent over the next ten years. The total cost of the program i s US$43 million per year, which would be spread across 24 distribution utilities. However, 25 percent of the investments would need to be concentrated on the Atlantic coast systems, where network quality i s particularly deficient. Table2.18: Summarypower sector investment needs 2006 2010 Total Average annual -- - ~- ._____ Maintenance 310 310 1,860 Rehabilitation - 1 1T O2005 43 43 43 258 PPAs 113 113 678 Generation 331 190 1,545 258 Transmission 85 0 256 - - Total 1 634 882 656 4591 - L .~ - ~ ~ m c e : Adapted from Chahin and Rojas, 2004 2.101 In summary, the overall investment needs in the electricity generation, transmission and distribution sectors total US$767 million per year. Electricity investment needs to back economic growth and competitiveness average US$767 million per year, though with a heavy concentration of investment over the period 2006/07, when needs of the order of US$S50 million per year are expected. About 60 percent of this relates to maintenance obligations and payment o f Power Purchase Agreement (PPA) 69 guarantees, and the remaining 40 percent to new investment in generation and transmission. These investment needs are entirely related to the SIN. Investments associated with the ZNI, and with expansion of coverage within the S I N will be estimated inChapter 3. Transport 2.102 Colombia faces major investment needs in transport to enhance the productivity and competitiveness of its productive sectors. The investment needs of the transport sector are by far the largest of the infrastructure sectors, and include both maintenance and rehabilitation of deficient road networks, as well as continued upgrading of road network capacity (which i s the backbone of the country's freight flows), and improvement of air and maritime gateways (which are increasingly critical for international trade). Although significant private sector participation i s foreseen, many of the transport requirements are public goods that will necessarily entail substantial government funding. 2.103 Road maintenance expenditures of US$460 million are needed to avoid costly investments in rehabilitation in the medium term. Although not strictly speaking investment, maintenance expenditures present a critical financing requirement, given that their deferral has a severe impact on future investment needs inroads rehabilitation. Total annual maintenance needs for the entire road network are estimated at US$460 million per year, of which 20 percent i s associated with the primary network (excluding privately maintained concessions), 52 percent i s associated with the departmental secondary road network, and 28 percent with the tertiary road network. Furthermore, on those sections o f the primary road network under concession to the private sector, minimum traffic guarantees of US$42 million per year on average have been triggered, absorbing resources otherwise destined for maintenance and thereby contributing to the burgeoning liability for rehabilitation. 2.104 Investments of US$519 million per year are needed in the primary road network to modernize strategic corridors and improve internal and regional integration. Total investment needs in the primary roads between 2005 and 2010 are estimated at US$3,100 million, equivalent to US$520 million per year (Table 2.19). Two thirds of this investment will need to come directly from the public sector, while the remaining third would be susceptible to public-private co-financing. Thus the public sector i s likely to be responsible for financing 80 percent of the total. These investments comprise two categories: e Rehabilitation of corridors deteriorated by accumulated deferred maintenance, widening of some roads, paving of some high traffic unpaved corridors, and a special bridge program. This group of investments totals US$l,l86 million over the years 2005-2010, amounting to an annual average of US$198 million. 70 e Construction of new links to improve the internal and international integration, shorten existing circuits, reduce gradients, incorporate some 300 kilometers of additional lanes, and improve access to key gateways such as ports, airports and border crossings. These works include U S $ l l 6 million for the primary road network under Plan 2500, an initiative aimed at improving accessibility of remote regions, with its primary focus on the secondary and tertiary networks. Some of these projects will be given under concession, in a number of cases with the contribution of public funds. Investment in this type of works for years 2005-2010 averages an annual expenditure of approximately $320 million. Rehabilitation of existing network 1,186 198 New publicly managedprojects 919 New projects 587 Plan 2500 116 153 Additional lanes 164 Works to be completed 52 Roadconcessions 1,006 169 L Public co-financing j306 j 53 ( Total 13,111 I519 Source: Adapted from Ospina, 2004 Notes: Split between public and private financing is a preliminary estimate 2.105 Investments of US$167 million per year are needed to rehabilitate the secondary road networks; over and above existing commitments under Plan 2500. The secondary and tertiary road networks include links that are key to reducing transportation delays in the early stages of production chains dependent on primary products. The very limited information available on the quality of these networks complicates the process of investment planning. Nonetheless, it i s known that rehabilitation represents the most significant investment requirement on these networks, estimated to cost US$743 million over the period 2005/10, equivalent to US$124 million per year on average. This is over and above investments already contemplated for 2005/06 under Plan 2500, which comprise a mixture of upgrading and rehabilitation works totaling US$260 million, with a 60:40 split between the secondary and tertiary networks. 2.106 Ports face investment needs averaging US$50 million per year, shared between private upgrading of facilities and complementary public improvements in access. Colombian ports face major challenges due to the expected increase in international trade following the FTA, and the opportunity to increase their share in container transfer activities. Investment needs in the sector divide into two clear categories. First, it will be necessary to increase capacity and improve performance of port facilities through investments in cranes, yards, racks, warehouses, information systems, and logistics centers. These investments can be expected to be undertaken by the private sector, subject to suitable adjustments in the regulatory framework. The cost i s 71 estimated at US$240 million over the period 2005/10, equivalent to US$40 million per year. However, there i s currently some controversy over the extent to which performance improvements demand new investments or can be achieved by more efficient operation of existing infrastructure. Second, complementary public investments are required to address severe bottlenecks in sea and road access to port facilities. This will require some US$60 million over the period 2005/10, or US$10 million annually, of which half will go towards deepening maritimeaccess to Barranquilla port alone. 2.107 Investment needs in the rail sector are estimated at US$10 million per year, focusingon expansion of coal transport capacity, and funded entirely by the private sector. Rail networks will require investment to accommodate larger volumes of coal, as well as to expand their activities to other freight. A second track will be needed along the trunk route from Cesar to the Atlantic ports, some missing links will be completed, and significant improvements of loading/unloading facilities and rolling stock are expected. Investment needs are expected to be of the order of US$60 million between 2005/10, and will be entirely assumedby private rail operators. 2.108 Investments inother transport sectors could demand a further US$70 million per year, including upgrading of airports, dredging of rivers, and multi-modal facilities. e Airports will need to modernize, to expand capacity and improve service quality for the increasing flows of international freight and passengers. The two key initiatives are Phase Iof the Bogota Master Plan and the improvement of three secondary airports (San Andre's, Rionegro and Palestina) respectively estimated to cost US$120 million and US$54 million in2005/10. In addition, the modernization of the equipment for air traffic control and navigation aids will require US$53 million over the same period. The private operators of airports already under concession are expected to contribute US$60 million towards these requirements. e Dredging works to improve river navigation will be needed on the Magdalena river (the Yuma project), and on the Meta river (connecting Colombia with Venezuela through the Orinoco), among others. The total cost of these works i s estimated at US$21 million annually over the period 2005/10. e Multi-modal transportation and logistics development are strategically important, and a national policy initiative on this issue i s currently under being developed. Some public investments in multi-modal terminals and logistics complexes will be needed to complement changes in managerial practices inthe private sector. The associated cost is difficult to estimate. 72 Table 2.20: Summaryof transport investmentneeds US$m Investmer needs 2005/1 Average an la1 investmentneeds Public Private Total Public Maintenance 2,760 - 2,760 460 Primary roads 2,411 700 3,111 402 Secondary and tertiary 1,003 - 1,003 roads 167 Ports and sea-side accesses 60 240 300 10 Railroads - 60 60 - Airports 227 60 287 38 Waterways 127 - 127 21 - 21 Total 3,828 1,060 4,888 638 177 1,163 Source: Own elaboration. 2.109 In summary, the overall investment needs in transportation infrastructure total US$SlS million per year of which almost 80 percent will need to be publicly financed. Aggregating the investment needs of each of the infrastructure sectors leads to an estimated total requirement of US$638 million per year, o f which the roads sector accounts for about 85 percent, and the air and sea ports for a further 6 percent each. It i s estimated that the private sector could finance about 20 percent of the total requirement. Other sectors 2.110 Other infrastructure sectors will also request significant investment; they will be by and large paid back through user charges. The investment needs detailed above focus on some of the largest and most strategic investment requirements in the electricity and transportation sectors. However, other infrastructure activities will also require investments to support economic growth. In most cases, these can be expected to be financed directly by the private sector based on tariff revenues, without direct government support. Telecommunications investments in particular are deemed to be significant, with the expansion of the mobile network expected to cost US$439 million per year, coming to a total of US$469 million per year if investments in long distance and value added services are also included. In addition, investments needed to export natural gas to Venezuela, are expected to amount an average of US$67 million during the next five years. 2.111 Summary. Those productive sectors expected to underpin Colombia's future economic expansion are more intensive in the use of infrastructure services, than the country's traditional agricultural and mineral exports. In particular, they will require a significant volume of reliable electricity supply, as well as cheaper and better quality transport services. Overall investment needs to provide the infrastructure needed to enhance the productivity and competitiveness o f the Colombian economy are estimated at US$2,600 million per year, which i s equivalent to 3.2 percent o f GDP (Table 2.21). 73 2.112 I t is important to note that about 46 percent of the total (or 1.5 percent of GDP) corresponds to maintenance and rehabilitation expenditures, and 48 percent (or 1.6 percent of GDP) to investment in new infrastructure assets. The transport sector makes the largest demand on resources, accounting for 50 percent of the total (or 1.6 percent of GDP), followed by 30 percent (or 1 percent o f GDP) for the electricity sector. The share of these resources that could be financed by the private sector varies substantially across sub-sectors, ranging from 18 percent in the transport sector to 100 percent in the natural gas sector, however the overall share stands at 31 percent (or 1.0 percent of GDP). Table 2.21: Overall summary of investment needs, priorities and financing sources US$m Public finance Private finance Electricity Maintenance 200 110 310 Rehabilitation 22 21 43 PPAs 113 0 Generation (SIN) 129 129 258 Transmission (SIN) 43 0 : Transport 1057 237 Maintenance 460 Rehabilitation Traffic guarantees 365 53 Road construction 297 Port upgrading 50 Airport upgrading 38 Railroad upgrading 21 10 Dredging I - Natural gas 0 67 67 Export 0 61 67 Telecom 469 Mobile 439 Long Distance 3 Value Added 27 Total 1,784 814 2,597 Maintenance 660 110 710 Rehabilitation 381 21 408 Guarantees 166 0 166 Investment 571 683 1,253 Total (as percent GDP) 2.2 1.0 3.2 Maintenance 0.8 0.1 1.o Rehabilitation 0.5 0.0 0.5 Guarantees 0.2 0.0 0.2 Investment 0.7 0.9 1.6 - . ~ Source: Own elaboration 2.113 With regard to prioritization, as a general principle, maintenance and rehabilitation expenditures should always be given the highest priority due to the need to conserve and recover existing assets. However, it should be possible to prioritize between new investments based on their relative rates o f return. In areas driven 74 primarily by private investment, i t i s assumed that prioritization will occur automatically through the market. In the electricity sector, prioritization i s difficult due to the inter- relatedness of the system and the non-discretionary nature of many investments. However, the burden of public investment could be reduced to the tune of US$239 million per year, if the process of privatization of generation and distribution assets was to be taken further. In the transport sector, some investments (notably, improvement of international gateways, leveraging of private finance, and removal of critical bottlenecks on the existing road network) would appear to be of higher priority than others. CONCLUSIONS AND R E C O M M E N D A T I O N S 2.114 Infrastructure services are not ends in themselves, but represent important intermediate inputs into the production process. The analysis of this chapter has vividly illustrated how supply side problems in the infrastructure sectors themselves, reverberate along the production chain, generating additional coping costs and compromising quality of output inthe productive sectors. 2.115 Therefore, the central recommendation is that infrastructure policies and their implementation should permanently incorporate the perspective of the productive sectors. On the basis of the analysis presented in this chapter, a number of key recommendations can be summarized, many of them with links to issues that will be covered in later chapters. The central thrust of all of these recommendations is the need to incorporate the perspective of the productive sectors into all aspects of infrastructure policy formulation, which has too often been absent to date. This approach not only introduces new trade-offs into a number of traditional policy areas (such as cross- subsidies and power sector liberalization), but also invites government attention in areas that have traditionally been overlooked (road haulage, multi-modal transportation, rural mobile telephony), and a more holistic view o f critical infrastructure corridors (such as those surrounding Buenaventura port). (a) Government needs to prioritize strategic investments to improve land and sea access to major ports, and improve freight facilities at airports. Although the quality and efficiency of Colombian ports has improved substantially during the 1990s, their efficacy i s still compromised by bottlenecks in land and sea access, which are a government responsibility. The most dramatic example i s Buenaventura port, which accounts for 44 percent of the trade handled through the public port system, and currently presents major problems for exporters in the west of the country. Problems range from tortuous and insecure road access, to slow rates o f throughput, and deficient maritime channels. Bogota's airport also presents deficiencies in terms of freight facilities provided to exporters. An integral package of measures i s needed to address each of these bottlenecks, with the aim of improving the overall efficiency of exports through these international gateways. 75 (b) Government policy needs to take a more holistic view of transportation, incorporating an explicit strategy on road haulage and logistics facilities. What ultimately matters for competitiveness are not roads, but transportation services, encompassing roads, road haulage and logistics facilities. Around the world, governments have tended to focus their attention solely on road infrastructure, overlooking the other key aspects of this nexus. Colombia i s no exception. While the complementary services offered over the road network are primarily the responsibility of the private sector, government policy and regulation can make a critical contribution to the efficiency and quality of these services. The major deficiencies of the trucking industry in Colombia clearly require some strategic thinking at the policy level. Moreover, some o f the investments needed to facilitate integrated logistics and greater use o f multi-modal transportation have public good characteristics, and hence may require small strategic investments from the public sector to provide inter-modal transport facilities and logistics centers. (c) Measures to improve security of energy supply are needed, to relieve industry of the extra financial burden of own and back-up generation facilities. The deficient quality o f electricity service provided by local distribution utilities i s prompting many industrial producers to install back-up generators, or switch to own generation facilities. Except in cases of co-generation, this involves a significant increase in costs of energy supply, and therefore may have implications for competitiveness. The key measures to improve this situation are cross-links in the radial structure of the transmission network, and investments to rehabilitate distribution networks inthe most dysfunctional utilities. (d) Incentive problems impeding construction of gas transportation infrastructure to serve new thermal generation plant need to be addressed. At present, there i s little incentive to build the gas transportation infrastructure needed to serve the country's next generation of thermal electricity generation plants, due to a mismatch between the fixed costs of transportation infrastructure and the unpredictable revenues of thermal generators. New contractual mechanisms need to be found to bridge this gap, and will probably need to be underpinned by changes in the regulatory framework that recognize gas transportation infrastructure as part of the fixed costs of thermal generation. (e), Any reform of Colombia's national cross-subsidy framework, should take into account the burden imposed on industry to finance the necessary surcharges. Discussion of the national utilities cross-subsidy framework tends to center on its efficacy as a social policy instrument, as well as its implications for the financial sustainability o f utilities (see Chapter 3 for details). However, the industrial surcharges used to finance the scheme are regarded as significant and material b y the productive 76 sector, and may be contributing to the competitiveness of Colombian firms, particularly with regard to energy costs. To the extent that higher costs motivate self-supply by industrial customers, they may be counter- productive by reducing the overall base of customers across which the fixed costs of electricity service can be spread. Evidently, a delicate balance between efficiency and equity considerations needs to be found. (f) The regulations governing own generation by large industrial users need to be carefully reviewed. The inability to trade or barter surplus power from own and co-generation facilities i s contributing significantly to the energy costs of large industrialists in a number of energy-intensive sectors. While these regulations may have their own justification in terms of maximizing scale economies in the inter-connected system, it would be advisable to do a cost-benefit analysis o f further liberalizing this regime for industrial producers of energy. It would appear to be in the national interest for firms with co-generation potential to take full advantage of the power produced, while firms maintaining their own generation plants will only be motivated to return to the public system if there are adequate improvements inthe reliability of supply. (g) Colombia's rural telecommunications policy should take into account the needs of the productive as well as the domestic sector. Colombia has a cutting-edge rural telecommunications policy. However, programs are primarily designed to address residential and social telecommunications needs in rural areas, including public telephones, community telecenters, and school connectivity (see Chapter 3 for details). This overlooks the constraints faced by the productive sector in coordinating the flow of primary products along supply chains between rural and urban areas. Addressing these needs i s likely to require different types of interventions than those considered in rural areas to date; such as expansion of fixed line telephony, extension of mobile coverage along critical production corridors, and/or policy measures to reduce the cost of access to satellite telephony. These could be accommodated within the current methodology developed for rural telecommunications namely, identifying the need, setting service targets and prices, and letting operators compete for licenses and subsidies if needed, leaving the technological decisions open to the private sector. (h) Increased financing to support improved road networks needs to be given greater priority; but location policies should also be considered in the longer term. As far as the infrastructure sectors are concerned, Colombia's most significant competitive disadvantage lies in its roads The problems relate both to road quality and to travel security, which restricts nocturnal circulation and inflates insurance costs. Security issues are relatively intractable; however road quality problems can be addressed primarily by scaling-up maintenance and rehabilitation efforts, which i s ultimately an issue of assuring the necessary finance. In the longer term, it 77 will be important to evaluate the spatial distribution of economic activity within the country relative to transport costs and the location of international gateways. 78 3. THE ROLE OFINFRASTRUCTUREINTHE ACHIEVEMENT OF SOCIAL OBJECTIVES INTRODUCTION 3.1 Infrastructure services are known to have a significant impact on quality of life at the household level. The links between infrastructure and quality of life are numerous: water and sanitation are key to public health; electricity extends the day for work and study; natural gas provides for safe and efficient cooking and heating; telephony reduces social isolation; while transport opens access to economic opportunities and improves national integration. This chapter therefore examines the extent to which infrastructure in Colombia i s available to all citizens on an equitable basis. For this purpose, attention i s limited to so-called `social infrastructure', defined in Chapter 1 to include household coverage of water, sanitation, electricity and telephony services, as well as access to communal telecommunications facilities. On the transport side, social infrastructurei s defined to include urban mass transit systems, as well as rural roads and other modalities that improve the accessibility of remote and marginalized areas. 3.2 This reality has longbeen recognized in Colombia, which is a regional leader in the use of infrastructure social policies. Chapter 1 already showed how Colombia compares very favorably with other higher income countries in the region, in terms of its ability to deliver social infrastructure services to the broadest range of its population. The importance attached to social infrastructure i s evident at the highest level in the Colombian constitution, which defines access to water and sanitation as a fundamental right of citizens, and stipulates that utility tariffs be based on considerations of social solidarity. In response to this mandate, the Colombian state has developed an extensive array of sophisticated initiatives to promote access and safeguard affordability. There are major investment programs to expand rural access injust about every sector, including a particularly well-designed scheme for rural telecommunications. Moreover, Colombia has a longstanding national cross-subsidy framework for the utilities, which i s an ambitious attempt to provide a comprehensive social safety net for low-income families. 3.3 The first half of this chapter examines progress towards the goal of universal access to infrastructure services. Historic coverage trends are first scrutinized to assess whether past levels of effort are adequate to ensure that universal access can be reached within a reasonable time period. Thereafter, the government's current policies for promoting universal access in both urban and rural areas are described and evaluated in 79 some detail, on a sector-by-sector basis. Furthermore, the feasibility of meeting the government's policy objectives is assessed by comparing future investment needs against historic levels of expenditure. In summary, rural water and sanitation stand out as being the only sector where (due to its decentralized structure) there i s no major explicit program for promoting rural access. In the other sectors, where interesting programs exist, a key issue i s finance. The resources historically available for rural investments in electricity and roads appear to be small in relation to the magnitude of unmet needs, while in electricity and telecommunications a substantial percentage of those resources available have increasinglybeen subject to fiscal restrictions on their disbursement. 3.4 The second half of the chapter assesses the affordability of infrastructure services in Colombia. An analysis of household expenditure patterns i s first used to clarify the financial burden that utility services represent for low-income households. Attention then focuses on the national utilities cross-subsidy system, which i s the government's central policy instrument for addressing this issue. The cross-subsidy system i s evaluated in terms of its effect on service affordability, its ability to target resources towards the poorest, its impact on the financial sustainability of the utilities sector, and the ways in which it distorts the incentives of the key actors. Moreover, the varying relevance of the cross-subsidy system to the different utility sectors i s also assessed. To summarize, the national cross-subsidy system, though apparently necessary to maintain affordability of basic water and electricity services, suffers from a number of major design flaws. In particular, only around 40 percent of subsidy beneficiaries come from the bottom two quintiles. At the same time, the system creates serious and deteriorating structural deficits that are threatening the financial sustainability of some sectors, particularly water. Furthermore, it i s questionable whether a cross-subsidy system of this kind i s still appropriate or necessary for the natural gas and telephony services. ACCESS TO INFRASTRUCTURESERVICES 3.5 This section begins by reviewing the current levels of access to infrastructure services in Colombia, and the rate of progress over time. It goes on to describe the universal access initiatives that exist in each sector, for both urban and rural areas. Finally, it examines the extent to which historic resource flows to these programs are enough to ensure that national policy objectives can be met within a reasonable time frame. Current state of access 3.6 Colombia has reached universal urban access to basic services (broadly defined), and continues to sustain significant coverage expansion in rural areas. One important source of data on household service coverage are the household surveys occasionally produced by DANE, and available for the years 1993, 1997, and 2003 80 (Table 3.1). According to this source, coverage of electricity has become virtually universal in urban areas of Colombia during the last decade. The same i s true for access to safe water and sanitation services, when these are broadly defined according to United Nations guidelines to include private wells in the case of water supply and septic tanks and latrines in the case of sanitation. However, if strict definitions of household connections to water and sewerage are used, the current level of access i s 94 percent for water and 86 percent for sewerage. In rural areas, electricity coverage lags behind at 83 percent, although substantial progress has been made during the last decade. Coverage of rural water and sanitation (broadly defined) stands at 66 percent and 72 percent ' respectively, with substantial progress over recent years. These levels fall to 38 percent and 13 percent respectively when the definition i s restricted to household water and sewerage connections. However, it i s questionable whether this level of service i s always appropriate in rural areas. 3.7 Coverage levels reported in household surveys are substantially higher than those registered in official statistics for electricity, water, and sewerage. It i s important to note that with respect to electricity, water and sewerage, the household survey evidence indicates coverage rates for 2003 that are of the order of 10 percentage points higher than what can be inferred from the utility cadastres. Thus, for example, UPME reports coverage of 87 percent for electricity as against 96 percent from the household survey. However, when the household survey evidence for 1993 i s compared with the results of the last national census conducted in the same year, the national coverage rates from the two sources are not significantly different, thereby indicating the reliability of the household survey evidence. Although this differential could be attributable to many factors, such as differences of definition, a potentially important explanation are clandestine service connections, which are picked-up in household visits but are absent from utility cadastres. Indeed, for the case of electricity, water and telephony, the household surveys ask separately about whether households have access to the service and whether they pay for the service. The discrepancy between these two measures, which provides some indication of the extent of illegal connections, amounts to 14.1 percent for water, 8.3 percent for electricity, and 1.6 percent for telephony. Table 3.1: Evolution of service coverage inurban and rural areas over time Nation; 1997 12003 1997 2003 1993 1997 200 - ~- 3 Electricity 77.2 83.1 85.0 93.9 95.8 Natural gas - - - - - Water 98.4 51.8 65.6 92.1 86.0 90.2 Sanitation 99.1 64.6 72.2 87.5 89.9 92.2 Fixedtelephone 64.8 5.0 10.7 40.1 48.8 55.0 Mobile telephone - 21.8 - 4.6 - - 17.7 I Source: DANE:EncuestaCalidadde Vida, 1993, 1997, 2003. 3.8 Coverage gains have been particularly large in wo new iervic :natur 1 gas and mobile telephony whose penetration has jumped some 20 percentage points in six years (Table 3.1). Following the arrival of natural gas networks in urban areas, household connections have grown by an average o f 3.5 percent per year throughout the 81 last decade, reaching 55 percent coverage in urban areas overall. Moreover, this coverage rate rises to 77 percent coverage when attention i s limitedto those municipalities that are sufficiently close to the transportation network to be able to contemplate providing the service. Finally, fixed telephone coverage expanded at a rate of 2.7 percent per year, with more than 90 percent of new connections taking place in urban areas. According to household survey evidence, mobile telephony has come from nowhere to reach a household coverage of 17.7 percent in 2003. It should be noted that this figure for household coverage i s somewhat higher than the official mobile teledensity of 13.9 percent for the same period, projected to rise to 16.9 percent as o f 2004. Mobile telephony i s primarily used as a complement to, rather than a substitute for, fixed line telephony, with 94 percent of mobile telephones representing second telephones for households that already have a fixed line connection. Although rural coverage of mobile telephony appears low, at under 5 percent, coming from zero it has already reached the same level that fixed telephony had in rural areas only six years before. 3.9 There is a clear geographical bias in access, with Regions 5-7 providing significantly higher coverage than Regions 1,4, and 8 (Figure 3.1). All regions of the country present very high levels of electricity coverage, ranging from 85-100 percent. However, for the other services, there i s a wide dispersion in access across regions, with coverage rates several times higher in the best-served regions relative to the worse served regions. In general, the best-served areas tend to be Regions 5 (Bogota), 7 (Valle del Cauca), and 6 (Antioquia), which correspond to the zones surrounding the country's three largest cities. On the other hand, the worst served areas are Regions 8 (San Andres Islands), 4 (Pacific Coast) and 1(Atlantic Coast). Figure 3.1: Service coverage by region Electricity Water Sanitation Talephone Source: DANE: EncuestaCalidadde Vida, 2003. Notes: Region 1 Atlantic Coast; Region2 Central (excludingBogota); Region 3 Oriental; Region4 PacificCoast (excludingValle del Cauca); Region 5 Bogota; Region 6 Antioquia; Region7 Valle del Cauca; Region 8 San Andres Islands; Region9 Amazonia and Orinoquia. 3.10 There is also a socioeconomic bias in access, although this appears to be stronger in urban areas than in rural areas (Figure 3.2). Access to electricity, water and natural gas i s relatively equitable across the income distribution, while access to 82 natural gas and telephony i s strongly associated with higher income levels. In urban areas, access deficits are concentrated among lower income groups, while in rural areas there are substantial deficits across the board. 3.11 Moreover, coverage statistics conceal significant differences in the quality of services provided to different segments of the population. Table 3.2 provides household survey evidence on the percentage of customers with service who express satisfaction with the quality of services they receive, and shows that around 75 percent of customers are satisfied with water and electricity services, as against 86 percent satisfied with fixed telephone services. The results also show significant disparities in quality of service between urban and rural areas, as well as significant inequities in terms of the quality of service received by different socio-economic strata. Nevertheless, it does not , necessarily make economic sense to provide the same level of quality to all consumers. Indeed, it may be appropriate to allow quality to vary to reflect differences in the cost of service provision in different settings, as well as differences in ability to pay of different customer groups. Figure 3.2: Access to infrastructure services acrossincome quintiles 100% E lectncfiy go% e880% 70% W a t e r 60% 50% $ 40% -cSanRation 30% 2 0 1 -Telephone (fixed) 10% 0% 1 1 3 4 5 &Telephone (cellulai hmrewide L I hcom quimle (a) Urban (b) Rural Source:DANE:EncuestaCalidadde Vida, 2003 3.12 Around four million people lack access to safe water and sanitation, and around two million lack access to electricity; all of them in rural areas. Around four million people (or 800,000 households) lack access to some form of safe water and sanitation, virtually all of them in rural areas. This deficit rises to seven million if access i s restricted to a household water connection, and as much as 13 million if access i s restricted to a household sewerage connection. In addition, some two million people (or 400,000 households) lack access to electricity, again all of them in rural areas. Only in the case of telephony, does the unserved population split almost equally between urban and rural areas. 83 Table 3.2: Percentage of customerswith service expressingsatisfaction with quality % Electricity Water and sewerage F :d telephc Urban Rural Total Urban Rural 1 Total Urban Rural Total Quintile 1 71.7 '1 52.3 64.9 11 75.1 1I 60.0 71.1 87.2 77.9 86.5 1 I Quintile 2 74.4 54.1 67.7 75.7 62.9 72.5 85.0 75.0 84.2 Quintile 3 78.3 54.0 72.9 78.7 64.4 76.5 86.6 82.6 86.3 ~ Quintile 4 79.9 53.4 76.6 82.7 i 63.6 81.0 87.5 86.6 87.5 ~ Quintile 5 ' 82.1 53.2 80.6 I 85.5 65.1 84.8 86.5 75.0 86.3 Total 77.9 53.4 72.7 I 80.3 62.5 77.6 86.7 79.8 86.4 ~ Source: DANE: Encuesta Calidad de Vida, 1993, 1997, 2003. 3.13 The population without access are heavily concentrated both geographically, in Regions 1-4, and socio-economically, in the lower income groups (Table 3.3). About 80 percent of the unserved population i s concentrated in the poorest Regions 1-4 (comprising the Atlantic and Pacific coasts, as well as the central and coffee-growing regions), even though these regions account for less than 40 percent of the national population. Moreover, the bottom 40 percent of the income distribution represents 55 percent of the unserved population. Across areas Across income groups Urban 1 Across regions Rural Regions 1-4 Regions 5-9 income '' Electricity 4.3 95.7 83.2 , 16.8 66.2 I Water 13.5 86.5 84.1 15.9 61.8 Sanitation 13.4 86.6 89.7 10.3 72.9 I Telephone 51.7 48.3 77.4 22.6 58.2 41.8 3.14 There is apparently no evidence to suggest that Colombia's violent conflict has significantly impeded access to services in rural areas. The violent civil conflict afflicting many o f Colombia's rural areas might be expected to be responsible for low service coverage in those areas. However, recent research suggests that looking across rural areas, and controlling for a wide variety of explanatory factors, violence does not seem to affect the probability that a given rural household will have access to infrastructure services (Box 3.1). This may reflect substantial policy efforts to improve access to services inregions affected by violence. 84 Box 3.1: Impact of civil conflict on infrastructure coverage 3.15 Current rates of expansion of basic services in rural areas are not high enough to reach universal access objectives within a reasonable time horizon. Assuming a continuation of the historic trend, Colombia has already reached universal urban access to electricity, and would achieve the same goal for water and sanitation by 2006. However, given the significant access deficit that exists in rural areas, i t i s important to evaluate to what extent sustaining historic levels of effort in service expansion would enable universal access goals to be reached within a reasonable time horizon (for example by 2020), taking into account demographic growth. Under the same continuation of historic trends, universal rural access to basic services remains a very distant goal that would not be reached for another 30 years. In order to bring forward this date to the window 2010 to 2020, the average annual expansion rate in rural areas would need to increase to 2 percent for electricity, 2.6 percent for sanitation, and 3 percent for water (Table 3.4). This represents a more than doubling effort for rural electricity and sanitation, and a 50 percent increase in effort for rural water. Nevertheless, if current trends continue, it appears likely that Colombia will reach the Millennium Development Goals for water and sanitation (Box 3.2). 3.16 Although coverage deficits in rural areas are higher, reaching universal access will require an even larger effort in urban areas. The fact that initial coverage deficits are much larger in rural or urban areas may give the impression that the bulk of coverage expansion effort needs to focus on the rural side. However, this overlooks the fact that Colombia i s 77 percent urbanized and that urban areas are experiencing demographic growth of 2 percent per year as against virtually zero demographic growth inrural areas. As a result, attaining universal coverage inurban areas will require twice as many connections as in rural areas in the case of water, and five times as many in the case of electricity. 85 Electricity Water Sanitation Urban 1 Rural I'Urban 'Rural 1Urban Rural Annual growthrate I I Il 0 Historic trend I :Year of universalaccess Historic trend I2003 'j2033 I 2006 2024 i2006 Increasedeffort - 2013 - 2018 - I ~ Source: Own calculations basedon DANE: Encuesta Calidad de Vida, 2003. Box 3.2: Will Colombia achieve the MDGsfor water and sanitation? 86 3.17 About one third of the rural population (some four million people) does not have ready access to the road network. In Colombia, the average rural household lives 2.5 kilometers from an all season road. This average ranges widely across regions, reaching 4.5 kilometers on the Pacific Coast (Figure 3.3(a)). The World Bank uses a zone of influence approach to define sustainable access to rural transport as the percentage of households that live within two kilometers of an all-season road (equivalent to a 20 minute walk). An all-season road i s defined as one that i s accessible to motor vehicles all the year round, by the prevailing means of rural transport. According to this definition, 63 percent of rural households have sustainable access to rural transport infrastructure. This result i s close to the average for low-income countries of 64 percent, but well below the average for middle-income countries of 94 percent. Of the four million people without sustainable access to rural transport, 43 percent live within one kilometer of an all-season road, and a further 46 percent between two to five kilometers distance (Figure 3.3(b)). The largest absolute numbers of isolated households are concentrated in Region 2 (Central) and Region 4 (Pacific Coast). It i s important to note that these results understate the true extent of isolation given that data for the Amazonian region, which i s the most isolated of all. were not available. Figure 3.3: Distance of householdsfrom all-seasonroad --..-- I 4 5 40 E 35 3 0 hi gE 2 5 20 1 5 I O 0 5 0 0 R1 R3 R2 R7 R6 R4 NEtiml Dktance to alksexon road(kms) I (a) Average distance (b) Frequency distribution Source: DANE:EncuestaCalidad de Vida, 2003. Notes: Region 1Atlantic Coast; Region2 Central (excludingBogota); Region 3 Oriental; Region4 Pacific Coast (excludingValle del Cauca); Region6 Antioquia; Region7 Valle del Cauca. 3.18 In the Amazonian region, river and air provide the most relevant modes of transportation, given the environmental constraints on the development of the road network. The Amazonian region with a population of 2.5 million has the lowest population density of the country (less than 20 persons per square kilometer). Accordingly, the Colombian road network i s heavily skewed towards the most populated northwestern and central regions of the country, and barely touches the extensive Amazonian region (Figure 3.4(a)). This is not surprising given the environmental sensitivity of the region, and its inclusion of extensive national parks and indigenous reserves, all of which necessarily restrict the development o f the road network. The absence of roads i s to some degree compensated by the presence of river highways, and airports, providing a much higher level o f coverage when all modes are considered together (Figure 3.4(b)). Nonetheless, most o f the waterways are not passable all the year 87 round, while air transport i s too expensive to be relevant to all but a small minority of the population. Hence, the challenge in this region i s to ensure that river and air modes of transportation provide a well functioning and accessible network, providing reasonable levels of mobility to the local population. Figure 3.4: Access to transportation infrastructure (a) Access to road networks (b) Accessto all transport modes Source: Ospina, 2004 3.19 Moreover more than half of smaller rural towns are not accessible by an all- season road. The national Electoral Register paints a more detailed picture of the degree of isolation faced by rural towns. Whereas the vast majority of municipal capitals can be reached by an all-season road, less than half o f the smaller towns enjoy such access (Figure 33a)). Moreover, a third of municipal capitals are more than four hours journey from their respective departmental capitals. Indeed, in some parts of the Pacific Coast and MagdalenaMedio region the journey can take longer than a day (Figure 3.5(b)). Figure 3.5: Accessibility of rural towns ~ 100% 80% r 80% 70% 00% 2 80% 4pi eE r 50% 40% se a m rn 40% 30% 20% 20% 10% 0% 0% Qto2hcurs 2to4horrs OLer4hwrs Road Track Footpath Rmr Airport MuniclDal CaDrtal QMunlclDal subcapltal UPollce statlo mMunicipal capital BMunlclpalsub-caprtal OPolice station (a) Modes of access (b) Time to administrative center Source: Electoral Register, 2004 Adapted from Ospina, 2004 88 3.20 Summary. Colombia has sustained substantial coverage improvements over the last decade, and i s very close to reaching universal access to electricity, water and (broadly defined) sanitation in urban areas. However, significant inequities persist in access and quality of services, both across income groups and between urban and rural areas. Coverage deficits are heavily concentrated in rural areas, as well as the more depressed Regions 1-4. To reach universal access in a reasonable time frame, the rate of coverage expansion for ruralelectricity, water, and sanitation services needs to accelerate significantly. Nevertheless, Colombia i s well placed to meet the Millenium Development Goals, at least at the national level, if not at the rural level. More than a third of the rural population, and about half of small rural towns, are not reached by the tertiary road network. Inthe Amazon region, river and air links provide the most appropriate means of transportation. Policies to promote access 3.21 This section reviewsand evaluates the wide variety of policy instrumentsthat have been developed to promote universal access to infrastructure services in both urban and rural areas. Given the diversity of practices, separate discussion i s provided on the electricity, gas, telecommunications, and water and transport sectors. Electricity 3.22 Two separate policy frameworks for universal access have been developed in Colombia: one for the inter-connected system (SIN) and the other for the non-inter- connected zones (ZNI). The S I N primarily covers urban areas and some rural areas in between the main urban areas, where more than 80 percent of the rural population i s concentrated. The ZNI covers the remaining 20 percent of the rural population characterized by highly dispersed settlements. According to official UPME figures, national electricity coverage stands at 87 percent (Table 3.5). This is substantially lower than the estimate of 96 percent that emerges from DANE'Shousehold survey data; although this estimate falls to 87 percent if only those households that pay for their electricity service are included, showing greater consistency with the UPME figures. According to the UPME estimates, there are 5.4 unserved households o f which two thirds are in rural areas. Although rural coverage rates are higher in the S I N than the ZNI, the absolute size of the unservedpopulation remains substantially higher in the S I N than the ZNI, at 2.6 versus 1.2 million. 3.23 Within the SIN, there is very little economic incentive for distribution utilities to expand coverage. Responsibility for service expansion within the S I N rests with the distribution utilities, and i s almost entirely funded through the tariff base. A key problem i s that utilities lack commercial incentives to extend coverage. On the one hand, distribution tariffs are based on average costs and do not take into account the rising marginal costs of expanding the network into more remote areas. On the other hand, unserved customers tend to be concentrated in the lower socio-economic strata and hence yield revenues that are well below even average costs. It would therefore be important to 89 explore the extent to which the regulatory framework can help to restore some of these incentives, for example, by establishing different pricing zones inrural parts of the grid. : Table 3.5: Overview of electricity coveragepatterns I Total population 1Official UPME 1 Unservedpopulation (m) coverage rates % 1 (m) ~ i ~ SIN 42.7 90 I 4.4 - Urban 34.2 92 2.1 - Rural 8.5 69 2.6 ZNI ' 1.8 35 11.2 ~ I I Total 44.5 87 ~~~ 5.6 -- Urban I 34.2 92 2.7 Rural 10.3 63 3.8 L I 1 1 I Source: Adapted from UPME, 2004 3.24 Moreover, the boundary between the SIN and the ZNI appears to be largely a political one, without full regard for economic viability. Historically there have been incentives for municipalities on the extreme fringes of the network to have themselves classified as part of the SIN, in the hope that they would eventually receive grid electricity, even though the costs of grid extension in those areas might be prohibitively high.It may therefore be relevant to revisit the boundary betweenthe SIN and the ZNI on the basis of an objective cost-based threshold for establishing the economic limits of the grid. 3.25 A new funding mechanism, the Rural Electrification Fund (FAER),has been established to subsidize investments in rural areas of the SIN. In 2003, a special fund (known as FAER) was created within the Ministry of Mines and Energy to finance the expansion of services to rural areas covered by the SIN. The fund collects resources from a surcharge of US0.40 per megawatt-hour o f electricity sold in the wholesale market, which yields approximate revenue of US$18 million per year. The funds are administered by a government committee. Projects are presented to FAER by the local government authorities. To be eligible, they must form part of the local development plan, and the investment plan of the corresponding distribution utility, and must also pass through the national project screening and evaluation system (see Chapter 5). The resulting infrastructure remains the property of the national government but i s operated by the local distribution utility under a management contract. In its first year, the fund financed six projects benefiting 535 households in the department of Boyaca at an average cost of US$1,645 each. 3.26 The ZNI presents particularly challenging conditions 'for rural electrification, as well as major inadequacies in current service provision. The ZNI covers 66 percent of the national territory, yet comprises only 4 percent of the national population, with an average population density of three persons per square kilometer. The ZNI has an installed generating capacity of 136 megawatts, which is almost exclusively diesel-based, and suffers from major diseconomies of scale given that 80 percent of 90 capacity i s in plants below the 100 kilowatts threshold. Service coverage currently stands at 34 percent, and continuity of service has been declining from 8.7 hours per day in 2000 to 5.3 hours per day in 2003. Electricity service in the ZNI i s provided by 42 formally constituted utilities in the larger agglomerations, and directly by 847 smaller municipalities elsewhere. 3.27 Investment subsidies are available through the national Non-Interconnected Zones Fund (FAZNI), to finance project proposals submitted by local authorities. Since 2000, a special fund (known as FAZNI) has been the primary source of funding for the electrification of the ZNI. The FAZNI i s funded from a surcharge of US$0.40 per megawatt-hour of electricity sold in the wholesale market, which yields approximate revenues o f US$18 million per year. The FAZNI operates under very similar administrative procedures to the FAER. Project proposals must include specific investment and operational targets, environmental and social impact assessments, as well as an institutional plan that ensures operational sustainability of the infrastructure. To ensure greater efficiency of investments, FAZNI has grouped most of the target communities into 11 clusters that present some degree of internal homogeneity. The delivery of services i s contracted out to these communities on a minimum subsidy basis, either to private operators in areas with greater commercial potential, or to community organizations in areas with more modest levels of demand. 3.28 The strategy for expansion inthe ZNI will needto pay increasing attention to renewable energy sources and isolated systems. Past interventions in the ZNI have focused almost exclusively on the provision of diesel-powered conventional mini-grid services. However, this seems inappropriate for two reasons. First, the very limited transportation infrastructure of the ZNI complicates the distribution of diesel to generating plants, leading to service outages. At the same time, the ZNI presents significant renewable energy resources, particularly in the form of solar radiation and potential for small-scale hydro-electric schemes. Second, three quarters o f the population of the ZNI live outside of municipal capitals in highly dispersed settlements, suggesting that isolated solutions (such as solar home systems) will need to play an increasing role in the expansion of access. All of these factors suggest a growing role for renewable energy sources inthe ZNI. 3.29 A major concern with both rural funds has been the imposition of fiscal restrictions that prevent them from making full use of the resources at their disposal. The philosophy behind the creation of both the FAER and the FAZNI was to create a reliable source of investment funding for rural electrification projects, financed directly from the resources of the sector. Nonetheless, under Colombian law, these funds are consolidated into the public sector accounts and count against fiscal targets. Due to the recent fiscal austerity programs, both funds have consequently had their budgets capped at levels well below their revenue collections (Table 3.6). As a result, in 2004, both funds are authorized to spend only a fraction of their potential annual revenue collection from the sector, even though the pipeline of investment projects that already have full approval far exceeds the budgeted amount in each case. Inconsequence, the two funds have an accumulated unused balance of over US$75 million. Clearly, this situation undermines the whole rationale for the creation of this kindof fundingmechanism. 91 2004 Potential annual Accumulated ' Budget 12003 Project Budget 1 Project demand revenue collection balance (31/3/04) 1 ~ demand I I FAER 0.1 6.1 3.5 5.4 18.0 12.2 FAZNI , 5.8 10.5 7.1 41.2 18.0 53.3 ~ ~ I Total 6.5 16.6 10.6 46.6 36.0 65.5 ~ 3.30 Summary. Colombia has established a solid framework for financing rural electrification that has yet to deliver major results. This i s largely due to fiscal constraints that have seriously constrained the execution of the fund resources. Longer-term concerns include the problem of sustainability of investments made in the ZNI, as well as the need to make greater use of decentralized renewable energy sources. Moreover, beyond the funds themselves, there are a number of important regulatory issues affecting the SIN that warrant some attention. In particular, the way in which expansion of rural services by distribution utilities i s rewarded through the tariff formula, and the principles used to establish the boundaries of the national grid. Table 3.7: Summary of universal access policy for electricity Objective Urban policy Rural policy Access to a household connection 11Universal access 11- Distribution utilities urimarily responsible for the SIN. ~~ to national or local electricity already achieved. - Two national funds subsidize investments in service grid. expansion for SIN(FAER) and ZNI (FAZNI). ~- 1 Natural gas 3.31 Colombia has achieved a large and rapid expansion in natural gas penetration, bringing major economic and social benefits to the urban population. Following expansion of the natural gas transportation network during the second half of the 1990s, coverage of natural gas in urban areas nationwide rose from 26 percent to 47 percent between 1997 and 2003, and as high as 77 percent if attention i s limited to those 327 municipalities that are close enough to the transportation network. The arrival of natural gas has brought substantial benefits to the urban population. First, natural gas i s a far more economic fuel for domestic cooking and heating than LPG or any of the other commonly used alternatives, meeting basic household needs at about half of the equivalent cost for LPG or electricity. Second, natural gas i s much safer than highly flammable alternatives, such as gasoline and cocinol, which were in widespread use among lower income groups at the time; although this i s no longer the case. However, the key barrier that often prevents households reaping these benefits are the high switching costs, amounting to US$400 (of which 40 percent relates to the connection charge and the remaining 60 percent to the internal investments). 92 3.32 Moreover, a number of policy measures were taken to ensure that low- income groups would not fail to benefit from the arrival of natural gas service. Cognizant of these barriers to low income take-up, the government adopted a number of policies to explicitly facilitate access b y lower income groups. First, the decision was taken to incorporate the cost of branches from the transportation network up unto the `city gate' into the transportation tariff, thereby introducing an element of cross-subsidy between transportation and distribution. Second, distribution network expansion was planned in such a way as to prioritize low-income areas; as in Bogota, for example, where the lower income southern zone of the city was the first to receive service. Third, in 126 smaller municipalities exclusive service areas were created incorporating explicit connection targets for the lower socio-economic strata. Fourth, the natural gas service was included in the national cross-subsidy framework that applies to other public services, so that lower strata customers would benefit from tariff discounts. As a result, 86 percent of natural gas customers are concentrated in the three lower socio-economic strata, which correspond almost exactly to the share of those three strata in the total urban population. 3.33 In 1997, ECOGAS established a special fund aimed at subsidizing the extension of natural gas services inlow-income areas. This fund, which i s collecting a 1.5 percent surcharge on transportation tariffs, has only succeeded in committing 2 percent of its funds to concrete projects due to delays in the development of the regulatory framework governing the use of its resources. As a result, it currently has an accumulated balance of US$10.5 million, representing a wasted opportunity to further accelerate the rate of service expansion. 3.34 However, use of service subsidies does not appear to be a very efficient or effective policy for encouraging lower income households to switch to natural gas. Given the significant cost advantage that natural gas already presents for domestic uses, it makes little sense to subsidize the use of natural gas as i s currently done through the national cross-subsidy framework. Instead, subsidy resources are best concentrated on helping lower income households to make the capital-intensive switch into this fuel. To see this, consider a household taking a typical five-year loan from a natural gas distribution utility to cover US$400 of switching costs at a 2 percent monthly (or 30 percent annual) rate. Under the status quo, a lower strata household taking such a loan faces a benefit-cost ratio of 3.83. However, the payback period is six years, during which time household energy costs will be 7 percent higher than they were before, due to the need to amortize the credit for the switching investment (Table 3.8). This scenario may not be sufficiently attractive to motivate a low-income household in making the connection. Moreover, in the case of higher strata paying full cost tariffs for natural gas, the payback period would be longer than 20 years, makingthe switch unattractive unless the investment could be at least partially funded from household savings. 3.35 Indeed, granting connection subsidies instead of use of service subsidies would substantially enhance switching incentives at a fraction of the subsidy cost. Consider the alternative scenario, where all subsidies for the use of gas are phased out, and replaced by a partial connection subsidy targeted to lower strata households, with the balance of switching costs continuing to be financed by a loan (Table 3.8). Any 93 connection subsidy above US$96 makes the switching process more attractive than was the case with the subsidized tariff for use of the service. Indeed, with a subsidy of US$123, the household i s able to amortize the credit without experiencing any increase in its monthly energy bill, while with higher connection subsidies such as US$168, the household starts to experience an appreciable reduction in household energy costs from the first month of the credit. Moreover, the total subsidy cost of achieving this result with connection subsidies i s far lower, particularly when one considers that the connection subsidy only has to be paid once to the switching household, whereas the use of service subsidy i s paid indefinitely to all households including those that switched long ago. In addition, a reform of this kind would strengthen commercial incentives for expansion of service into lower strata areas, by eliminating the tariff discount associated with these customers and thereby making them more attractive to serve. Table 3.8: Comparison of alternative gas subsidieson incentive to connect 1 Naturalgas tariff Connection charge Differencebill PV of PV of BCR Payback PV of (yrs 1-5) costs benefits penodcubsidy Status quo Strata 1-2: subsidy c Unsubsidized 1+7% 41 158 I Strata4-6: unsubsidized Unsubsidized +27% 159 130 0.82 20+ 0 ~ Unsubsidized Subsidy of US$96 +6% 34 130 Unsubsidized Subsidy of US$123 0 0 131 Unsubsidized 0 130 Source: Own calculations based on Yepes, 2004. Note: *This represents the value of the subsidy paid to the connecting household, however (in this case) the same subsidy must also be paid to all households already connected, thereby increasing the subsidy cost per connected household several times over, with the ratio depending on the number of existing customers per connecting household. 3.36 Moreover, ample resources are already available within the sector to finance a large scale connection subsidy program of the kind described above. At a connection subsidy cost of US$168 for each lower strata household waiting to be connected, universal access to natural gas in the relevant areas could potentially be financed within a five-year period, given the current revenues of US$15 million that are collected annually from surcharges on the higher strata. An alternative cost would be to roll connection costs into the cost base for the use of service tariff, so that no explicit connection charge i s made. Nevertheless, any switch from consumption to connection subsidies for natural gas should take particular care of the need to avoid harming that segment of the population that recently connected to the service and i s still amortizing its investment costs. 94 Natural gas LPG Kerosene, Electricity Wood Total gasoline, cocinol, alcohol, coal Urban 51.9 32.4 1.5 13.2 1.1 100.0 Rural 1.6 35.3 2.1 5.5 55.3 100.0 Total 42.1 33.0 1.6 11.7 11.6 100.0 3.37 Outside of the sphere of influenceof the naturalgas infrastructure,afurther 614 municipalities have access to LPG. In the rural areas, wood continues to be the dominant fuel for cooking and heating purposes (Table 3.9). In 1997, ECOPETROL launched a special program to promote the use of LPG inrural areas: Gas para el Campo. This program focuses directly on connection subsidies by providing a US$80 grant to cover LPG start-up costs, including the cost of a stove, a cylinder and three complementary refills. In its first phase, the program reached 100,000 households out of the 750,000 candidates identified. Further development of the program has been stalled due to restrictions inthe availability of LPG. Table 3.10: Summary of universal access policy for gas Objective 1 Urban policy Natural gas: access to a household 1 - Cross-subsidybetween gas connection for those living in the transportation and distribution. Ecopetrol, provides 80% subsidy 327 municipalities that are viable - Network reacheslow-income (USS80) for LPG start-up costs. for network coverage (US$200/hh.) neighborhoodsfirst. - Exclusive service areas with Presently covers 100,000 people, and ~ aims to cover a further 650,000 LPG: access to a distribution point connection targets for Strata 1-11. people. for the population living in the 614 - Cross-subsidyin tariffs to municipalities that are viable for compensate switching costs. distributionof LPG. I 1 I 1 I Source: Own elaboration. 3.38 Summary. Colombia has made tremendous progress in scaling-up natural gas service to households in urban areas. The wide range of social policies that formed part of the expansion strategy have borne fruit interms of a highlevel of penetration among low income households. However, the substantial financial costs associated with the switching process remain the key obstacle to further expansion. In this sense, it would make sense to consider replacing the existing use of service subsidy with a connection subsidy that absorbs around 40 percent of total switching costs. This would allow households to experience an immediate benefit from natural gas, as opposed to the current delay of more than five years before any saving in energy costs i s perceived. Telecommunications 3.39 Colombia has made major strides in movingtowards a modernframework for telecommunicationssocial policy. Given the growing convergence and intensifying 95 competition between different telecommunications services, current international best practice in the area of telecommunications social policy consists of defining clear social objectives, establishing non-distortionary sector levies as a basis for funding, and contracting out the delivery of social objectives on a competitive basis. In Colombia, social policy objectives for rural areas have been defined with exceptional clarity, and the vast majority of social projects are competitively contracted with finance from sector levies. However, a number of anomalies persist, in particular the system of levies discriminates significantly between different segments of the market, and the fund itself has been subject to fiscal restrictions on the disbursement o f its resources. Furthermore, the urban social policy framework i s out of line with the new realities of the market. 3.40 In urban areas, the policy framework places a major emphasis on universal household service. Household connections to fixed line telephony are substantially higher in Colombia than in many neighboring countries with much higher income levels. This is largely attributable to the existence of a national cross-subsidy scheme (described in greater detail in the next section) that substantially reduces the cost of telephone ownership for low-income households. It i s further reinforced by a regulatory requirement that all entrants to the local telephony market maintain a similar socio- economic profile of customers to that of the incumbent providers, with a view to avoiding cream-skimming of more lucrative clients. 3.41 However, the traditional cross-subsidy policy no longer appears to be so relevant as a means of promoting access to telecommunications. Although the cross- subsidy policy has been very effective in raising the level of fixed line telephone coverage in the past, there are a number o f reasons to think that such a policy may no longer be so appropriate in the new market environment. First, there has been considerable growth in the availability of public telephony solutions in urban areas, so that a ready alternative to a domestic fixed line connection now exists. Second, there has been a proliferation of creative tariff plans for fixed line service that allow consumers to select the most attractive package depending on their level of use, providing greater choice and flexibility regarding the financial terms of service. Third, there has been a rapid expansion of mobile telephony services, offered under prepayment packages, which provide low-income households with what may be an attractive alternative to owning a fixed line. Given current prices and consumption patterns, substitution of fixed line telephony by prepaid mobile telephony only becomes attractive for mobile charges below US$0.04 per minute, whereas current rates are closer to US$0.12 per minute. However, as fixed line prices complete their convergence to cost-reflective levels, and competition intensifies further in the mobile market, it is possible that this switching point may be reached. Indeed, the new mobile provider `OLA!' i s already providing a concessional prepaid package that offers calls at US$O.Ol per minute under certain restrictions. 3.42 In rural areas, the Communications Fund provides a clear mechanism for financing subsidies to key social investments, benefiting 14 million people to date. The Communications Fund established in 1996 i s used to finance the projects needed to meet the goals of the national universal access plan. The fund has a budget of the order of US$40 million per year from sector levies. T o date, the main rural access programs 96 supported by the fund have invested more than US$184 million and benefited a very high percentage of rural households (Table 3.11). Table 3.11: Results of universal access programs Beneficiaryentities Beneficiary Total cost population (US$) telephony Phase I 1999 7.415 towns (>250popn.) =3.7m 36.5m 4,929 I Phase I1 2002 3,000 towns (>150 popn.) =1.2m 14.3m 4,159 ~ Phase 111 j1 2005 4.000 towns (e150popn.) =0.3m 36.9m 9,225 I (estimated) Community I I1 telecenters I Phase I 2000 940 municipalcapitals 8.4m 8,910 Phase I1 2002 500 towns (including 42.1m 84,160 ~ I 191municipalcapitals) I1 Public sector connectivity Phase I(estimated) 2004 3.744 public institutions 46.1m 12,208 184.3m I Source: Ministry of Communications, 2004 Adapted from Cardenas and Melendez, 2004 3.43 The Compartel program for rural public telephony has been extremely successful, and is close to reaching its goal of universal access in rural areas. The first major initiative financed by the fund was the Compartel program designed to bring universal public telephony to rural localities (Table 3.11). The first two phases of the program have benefited 10,415 rural towns with more than 150 inhabitants, amounting to a total population of around five million people and reaching 74 percent of hitherto unserved rural communities. A third phase of the program, currently underway, aims to reach the remaining 16 percent of communities. Once completed, this will put Colombia alongside Chile as one of the only country in Latin America to have achieved universal access to public telephony in rural areas. Compartel works by requiring potential operators to bid for the minimum subsidy they require to install the public telephone infrastructure and provide service during a 10-year period. The biddingprocess i s open to both public and private operators, with no restriction on the technology that can be used. As a result of competition, unit costs of US$4,880 were achieved in the first two phases; these are expected to increase to US$9,225 in the third and final phase as service i s taken to smaller and more remote communities. 3.44 Furthermore, Colombia is now also close to reaching the goal of universal access to a community telecenter in each municipal capital. Followingthe approval of the national Connectivity Agenda in 2000, the fund began to finance the provision of community telecenters in the capital of each municipality (Table 3.11). Each telecenter comprises six computer terminals, and up to six telephone booths. To date 1,440 telecenters have been installed benefiting some nine million people, and the overall access goal should be met during the current year. Thereafter, the main policy objective 97 will be to achieve universal connectivity of public institutions, in particular municipalities, public schools and some key health facilities. At present, 58 percent of municipalities have internet access, and only 5 percent of public schools. The first phase of this program i s already underway and will benefit 3,744 public institutions. 3.45 A key concern is the sustainability of these facilities once the 10-year contractual period expires. The subsidies currently paid to telephone and telecenter operators are allowed to cover operating losses during the first 10 years of operation. Thereafter, it was originally expected that these facilities would be able to become self- sustaining without further subsidy support. However, it is becoming apparent that some of these facilities will continue to be loss-making from a commercial perspective, even beyond the 10 year horizon. It i s estimated that the necessary subsidies could amount to US$1.2 million per year. Hence, it will be necessary to find a mechanism for bidding out operating subsidies in subsequent years. In addition, the development of new technologies in the intervening period may present opportunities for lower cost alternatives to service provision that will also need to be considered. 3.46 The surcharges used to finance the Communications Fund are relatively high, and distort competition between different services. The fund collects resources from special surcharges on different segments of the telecommunications market that correspond to license fees and charges for the use of the electromagnetic spectrum. These surcharges discriminate substantially across different types of service ranging from 3 percent on value added services to 5 percent for long distance and mobile telephony services. Local telephony i s exempt from contributing to the Communications Fund, even though it represents more than 40 percent of sector turnover. The system of surcharges has been criticized on two grounds. First, it applies surcharges that are relatively high in comparison to what i s practiced in other countries, where surcharges of around 1percent are more typical. Second, by applying differential treatment across services the fund distorts the relative cost o f three competing forms of telecommunications: local fixed line, long distance, and mobile telephony. 3.47 Moreover, a growing share of the Communication Fund's resources are being transferred to the general budget. Due to severe fiscal constraints, the Ministry of Finance began to place budgetary restrictions on the execution of resources from the Communication Fund. In the early years of the fund (1996/98) these affected no more than 5-10 percent of resources, however by 2002 the fund was allowed to execute no more than half of its available resources. While these resources remain within the fund, accumulating interest payments, the restriction on their immediate use substantially retards the achievement of the fund's investment objectives. 3.48 Although originally conceived as an autonomous unit reporting to the Ministry of Communications, in practice the fund has been subsumed within the ministry. The Communications Fund was never able to hire a staff of its own, and thus all of its functions are carried out by different units within the ministry, while the administration of resources and contracts i s outsourced to another government agency, the National Fund for Development Projects (FONADE). From 1999, the Ministry of Communications established a dedicated team of consultants to be responsible for design, 98 , planning, implementation and evaluation of the Compartel program. While this arrangement has functioned effectively, it i s not sustainable in budgetary terms, and leaves the ministry itself devoid of the key human resources, and the fund more vulnerable to political interference. Table 3.12: Summary of universal access policy for telecommunications Ob'ective Urbanpolicy Ruralpolicy - To providepublic telephone - Local telephonyentrants: required CommunicationsFundbasedon access in all rural communities -(prioritized to maintainsame Strata distribution sector surcharges funds competitively To providebypublicofintemetcommunity). size access as incumbents. awardedsubsidiesto private - Other entrants: requiredto operators who build andoperate rural in each municipalcenter. contributetowards community systems. 1 1 - To provideintemet access in all telecommunicationsfacilities. municipaloffices, public schools, ' andmajor healthcenters. i Source: Own elaboration. 3.49 Summary. Major progress has been made in expanding access to telecommunications both in urban and rural areas. Overall, a very high percentage of the rural population has benefited from rural telecommunications programs, with universal coverage of rural public telephones and community telecenters now close to being a reality. The key challenges in the medium term are to take internet connectivity into public schools, and to safeguard the sustainability of existing rural facilities beyond the 10-year contractual period already established, with a probable need for continued subsidization at some level. A number of reforms are needed in the financing scheme used to fund these social policies. The cross-subsidy system used for local telephony i s now close to being obsolete, while the surcharges used to finance the Communications Fund are distorting competition between services, and will need to be reassessedin the context of the FTA. One alternative would be to gradually phase them out in favor of a lower, but uniform, sectoral surcharge. In addition, the fiscal restrictions placed on the execution of the funds resources i s slowing down the achievement of investment targets. Water 3.50 While the achievement of universal access to water and sanitation is a municipal responsibility, the national government helps to provide an enabling framework. Given the highly decentralized nature of the water sector in Colombia, strategies for reaching universal access are largely left to the discretion of each municipality. Nevertheless, the legal and fiscal framework established at the national level, makes a substantial indirect contribution towards the achievement of universal access goals. 3.51 First, the national judicial system is increasingly being used to enforce citizens' rights of access to water and sanitation. The Colombian Constitution identifies access to water and basic sanitation as a fundamental right of citizens. This provision has prompted a growing number o f judicial claims raised by unserved 99 communities against their respective municipalities, which are helping to make local authorities increasingly accountable for their actions inthis area. 3.52 Second, regulatory measures have been taken to reduce the barrier associated with up-front connection charges. Since 1997, the Regulatory Commission for Water and Sanitation (CRA) prohibited the then widespread practice of levying capital charges on newly connected customers. Water utilities are now only authorized to charge customers the direct costs entailed in making their connection, and are moreover required to provide financing facilities for customers to spread these repayments over a three-year period. 3.53 Third, the framework of fiscal transfers to municipalities ensures adequate, and equitably distributed, investment resources for coverage expansion. Fiscal transfers from central government to the municipalities, primarily in the framework of Law 715/01, account for about two thirds of the available investment resources in the sector. Moreover, the formula used to distribute these funds provides higher per capita grants to smaller and poorer municipalities. Nonetheless, overall per capita investments tend to be substantially higher in the larger municipalities, given that the greater financial strength of their municipal utilities more than offsets the differential in fiscal transfers (see Chapter 5 for further details). In addition to these regular transfers, the current administration's National Development Plan set aside US$250 million for high priority regional projects. Following a consultative process of public audiences, it was decided that US$95 million (about 37 percent of the total) will go to high priority local water and sanitation projects. 3.54 Fourth, national targets for coverage in urban areas are published as part of the National Development Plan. While this practice helps to give the issue of universal access some political visibility, the targets in question are simply an aggregation of the investment plans of the major municipal operators. Moreover, the plan does not set any explicit coverage objectives for rural areas. 3.55 Finally, the government operates a number of programs to support smaller water utilities. One example i s the Community Micro-Enterprise Program that provides technical assistance for the formation of public or private micro-enterprises to operate water and sanitation services at the local level. In addition, the resources set aside under the National Development Plan will finance a range o f critical investment projects, and institutional strengthening activities, aimed at improving service provision at the local level. 3.56 In practice, most of these instruments have more relevance for urban than rural areas, with the latter lacking any specific central initiative. In contrast to the electricity, telecommunications and transport sectors, there i s no major explicit rural investment programs to accelerate the achievement of universal access at the local level, but rather a wide range of different measures that affect rural service provision to varying degrees. Indeed,at the central level, there i s very little information of how rural water and sanitation services are progressing. 100 3.57 Policy goals are couched almost exclusively in terms of household connections, with little consideration given to alternative services and technologies. Household connections to water and sewerage service remain the central focus of universal access policy in Colombia. Indeed, the low prevalence of alternative lower cost technologies i s quite striking, with less than 1percent of the population served b y public taps, and less than 2 percent of by latrines. In addition, the national technical regulations governing investments in water and sanitation adhere quite rigidly to traditional engineering designs, thereby restricting the development of more appropriate low cost technologies for reaching unserved communities. There i s therefore a need to consider the potential role for alternative services and technologies within the national strategy, given that in many situations they often represent a more appropriate and cost-effective solution than the conventional alternative. ' Objective Urbanpolicy ' Ruralpolicy Access to a household connection - Constitutionalright of access - No specific national policy for rural to water and sewerage. - Elimination of capitalcharges water and sanitation. - Fiscaltransfers to municipalities - Rural municipalities receive higher - Coveragetargetspublishedin per capita fiscal transfers. NationalDevelopmentPlan. 3.58 Summary. In contrast to the other infrastructure sectors, there i s no major single universal access initiative for rural areas, but a range of measures that affect rural utilities to varying degrees. Instead, the rate of progress therefore depends primarily on municipal efforts, and as noted above has so far been inadequate to make any significant headway towards the MDGs. To its credit, the national government takes a wide range of legal, financial, regulatory, and political measures in support of the universal access goal. However, the one area where national policy does not appear to advance the achievement of universal access i s with regard to technical norms and standards. These adhere rigidly to conventional engineering designs, and place almost exclusive emphasis on household water and sewerage connections as the only acceptable solution for water and sanitation, adding unnecessarily to the costs of service expansion. Transport 3.59 The government's current policy on public transport in urban areas is outlined in CONPES 3167 of 2002. The policy emphasizes a more client centered approach aimed at improving the mobility o f urban residents by providing services that are well adjusted to their transportation needs. The main elements o f this policy include promoting a more efficient use of the car in urban areas, and developing exclusive lanes for the use of buses. The policy also marks a shift away from the traditional reaction of adding additional bus capacity, when the real problems may lie elsewhere. 3.60 In urban areas, the performance of public transport systems is a key determinant of the mobility of low-income households, affecting access to economic 101 opportunities. Detailed data on travel behavior in urban Colombia is only available for commuting. Travel patterns vary considerably across the income distribution (Figure 3.6(a)). Travel to work on foot i s disproportionately concentrated at the bottom of the income distribution, with more than half of the first quintile usingthis mode. Commuting by public transport (almost exclusively bus) is predominant among the middle classes, reaching around 50 percent for the third and fourth quintiles. Travel by private means i s almost twice as common in the top income quintile as elsewhere in the income distribution. 3.61 The average commute of almost 30 minutes is fairly constant across income levels, although controlling for mode poor commuters travel significantly longer. It i s estimated that people live, on average, about 5.7 kilometers from their place of work, with the bottom quintile living slightly closer (6.6 kilometers) than the top quintile (8.0 kilometers). The average journey to work takes 28 minutes, being substantially higher in Bogota (39 minutes) than in smaller cities (24 minutes). Average journey time remains virtually constant across income quintiles (Figure 3.6(b)). However, controlling by mode, first quintile households tend to spend 15-25 percent longer traveling by public and private transport, and walk more than twice as far as top quintile households. Moreover, all those traveling on public transport, on average, take 2.7 times as long to reach the workplace as those using other modes. The average speed of commuting i s estimated to be fairly constant across the income distribution at around 15 kilometers per hour, with the top quintilehaving an edge of four kilometers per hour over the bottom quintile. Figure3.6: Commutingbehavior acrossquintiles I 2 3 4 1 2 3 4 5 Income quintile Income quintile (a)Modalpatterns (b) By quintile Source: DANE, EncuestaCalidadde Vida, 2003 3.62 Bogota is beginning to make some headway with its longstanding urban transportation problems. Bogota has long suffered from serious urban congestion problems, which have reduced vehicle speeds to less than 10 kilometers per hour along key arteries during peak periods. Two important policy measures have been taken at the city level. First, regulations issued in 1998 introduced a rotating prohibition from circulation of 20 percent of the stock o f private vehicles each working day during peak hours based on license numbers (known as pic0 y placa). Second, the city embarked on a major new mass transit system. Although a metro-rail system had long been seen as the only real solution for the city, the very high cost of the metro built inMedellin duringthe 102 1990s, together with the accumulation of positive experiences with innovative bus systems from neighboring Brazil and Ecuador, motivated the search for an alternative approach. 3.63 A key element of the strategy is Transmilenio: an innovative and cost- effective Bus Rapid Transit system. Transmilenio relies on the widening of existing avenues and bridges to incorporate a central dedicated bus corridor, enabling vehicles to reach much higher average speeds. The exclusive use of the corridors i s given to a private operator under a concession arrangement, while revenue collection i s sub-contracted to a third party. A specialized government agency is responsible for planning and supervising the system, as well as representing local government interests in the associated contracts. The main Transmilenio corridors are complemented by feeder buses that bring in passengers from neighboring areas. The overall routing of the network i s designed to connect the main commercial centers of the city in the north and east, with the lower income neighborhoods inthe south and west, which are shaded darkest on the map below (Figure 3.7(a)). Inaugurated in 2002, Transmilenio currently transports 35,000 passengers per hour per direction during rush hour, representing about 7.6 percent of commuter journeys in the city, as opposed to 52 percent that are still made on traditional buses. However, the system i s currently operating at only 10 percent of its overall planned length, due to rise to 21 percent by the end of this year when a second 40 kilometer section i s scheduled to open. As a result of the dedicated corridor, Transmilenio buses reach average speeds of 26 kilometers per hour, compared with five to seven kilometers per hour for conventional buses that move along congested urban arteries. Figure 3.7: Geographic distribution of benefits from Transmilenio as of Dec. `03 (a) Socio-economicmap of Bogota (b) Time savingsfrom Transmilenio Source: Yepes and Hidalgo, 2004 Notes: Darker color indicates (a) poorer neighborhoods and (b) greater time savings. 3.64 These measures resulted in substantial time savings, especially for the poorest. The combined effect of Transmilenio and the regulations restricting circulation of private vehicles has been to reduce commuting times in the city by between 10 to 20 minutes each way on average journey times of 45 to 65 minutes. The largest time savings are concentrated inlower income neighborhoods (Figure 3.7(b)). 3.65 The aim is now to try and replicate the success of Bogota's Transmilenio in secondary urban centers, backed by significant public subsidy. The first phase of . Transmilenio cost US$350 million; which at US$4.3 million per kilometer compares 103 favorably with an equivalent metro-rail system. Nevertheless, an investment of this magnitude was only possible due to a major turnaround in Bogota's city finances in the decade before the project was launched. The government now aims to replicate the Transmilenio model in a further five secondary cities. Given the magnitude of the investments involved and the weaker financial situation of the smaller cities, this will necessarily require significant financial support from the central government. 3.66 Inruralareas, the flagship programfor improving accessto transportation is the `Roads for Peace' initiative, which forms part of the Plan for Colombia. The central objective i s to improve the integration of isolated areas affected by civil conflict and increase the presence of the state. An important secondary objective is to create local employment in road construction and maintenance. The program i s ajoint initiative of the Ministry of Transport, the Investment Fundfor Peace, the National Road Agency, and the National Fund for Tertiary Roads, working in close association with private firms, municipalities and local communities. A total of 397 municipalities have benefited from this initiative, which will conclude in 2005. A total of US$450 million is being invested over the period 2000 to 2005, including a significant multilateral commitment. Resources are primarily targeted at the Macizo Colombiano, the Magdalena Medio, and the southeast of the country. Projects are developed at the local level, and then submitted to the Ministry of Transport to be screened on technical and environmental grounds. The volume of pre-approved projects already in the pipeline significantly exceed the available finance. As o f June 2003,247 kilometers had been added to the primary road network, an increment of about 1.5 percent to the network. Moreover, maintenance activities had improved the quality of 34,356 kilometers on 180 stretches of the tertiary road network, which accounts for no less than 40 percent o f the overall tertiary road network. River navigability has also been improvedthrough a number of dredging projects. 3.67 The `Roads for Peace' initiative also includes a significant component for community airports. The objective of the `Community Airports' Program i s to build and improve runways at airports in isolated regions that are heavily affected by the civil conflict, and where air travel may be the only feasible mode of transportation. The program i s expected to complete investments at 15 local airports by the end of 2004. 3.68 In addition to `Roads for Peace' a number of other smaller initiatives exist. The `Alliance Program', established in 1999, provides 80 percent national co-financing from the National Fund for Tertiary Roads for maintenance activities on rural roads, and has invested an average of US$54 million during the last eight years, although with highly variable annual budget allocations. Further national co-financing for rural transportation i s also available from the National Royalties Fundfor rural roads, and from the National Development Plan budget for up grading o f airports in isolated regions. In addition, agricultural producers associations also actively support rehabilitation and maintenance of rural roads, primarily in the coffee-growing and banana-producing regions. Finally, the National Development Plan includes a commitment of national support for the rehabilitation of 2,500 kilometers o f priority primary, secondary and tertiary roads, entailing an investment of US$389 million over the period 2004/06. This i s a major commitment inrelation to historical funding levels, and it i s not yet clear how the program will be funded. 104 3.69 Summary. In urban areas, workers spend on average 30 minutes commuting to work. The poorest households are most likely to walk to work, or face the longest journeys on public transport. Commuting times are coming down in Bogota following vehicle circulation restrictions, and the completion of the first phase of the Bus Rapid Transit system. The government plans to extend the latter to secondary cities, though in contrast to Bogota, this will probably entail significant central government subsidy. In rural areas, the `Roads for Peace' initiative has had a significant impact on the maintenance of the tertiary road network, but i s now coming to a close necessitating efforts to ensure its sustainability. A new central government program to fund 2,500 kilometers of secondary and tertiary roads has been announced, although the details of its financing are not yet clear. Objective i - Urban policy Rural policy Urban: Reduce time taken to - Developmentof innovative rapid - `Roads for Peace' initiative is commute on public transport, and bus systems across the country. investingUS$90mpa. inrural road improve efficiency of car usage. - Regulationslimiting circulationof construction, rehabilitation and - Land Develop rural roadnetwork private vehicles in Bogota. maintenance. inconflict areas. - National co-financingfor local -Air: Improve airport facilities in transport projects available from isolatedareas. FNCV and FNR,including 2,500 - Water: Improve navigabilityof kilometer initiative. rivers Magdalena and Meta. Source: Own elaboration. Cost of reachinguniversal access 3.70 Building on analysis of historic access trends, and evaluation of the policy instruments in place, this section concludes by estimating the cost of reaching the government's universal access goals, and assessing the feasibility o f financing them under current arrangements. 3.71 Inelectricity, historic investmentsincoverage expansionfall short of what is needed to make reasonable progress towards universal access in rural areas. In contrast to the other infrastructure sectors, there i s no clearly defined timetable for reaching universal access in electricity. Therefore, an illustrative target i s used, that would entail raising rural coverage from 83 percent to 95 percent by 2015, while sustaining universal access in urban areas. Assuming an average cost o f US$l,OOO per new connection, this objective necessitates annual investments o f the order of US$173 million, of which almost 90 percent is required to keep pace with population growth in urban areas. Given that FAER and FAZNI as currently constituted can at best raise no more than US$36 million per year (assuming that they are allowed to spend 100percent of their collections), the remaining US$137 million would fall to the electricity distribution utilities. 3.72 In natural gas, historic investment would allow universal access to be reached by 2008, however these have been cut back in the latest tariff determination. The historical average investment cost for a new natural gas connection 105 has been US$186. In the past few years, the natural gas distribution utilities have been investing US$65.3 million per year in coverage expansion. Ifthis rate of investment were to continue, universal household access in the relevant areas would be reached as soon as 2007. However, in its five year tariff determination for the period 2003/08 the Regulatory Commission for Electricity and Gas (CREG) sets tariffs based on a substantially lower annual rate of investment of US$30.6 million per year. This level of effort will allow penetration of the service to increase from 77 percent in 2003 to 88 percent in 2008. However, if similar rates of investment persist beyond 2008, universal coverage will not be reached before 2016. Table 3.15: Cost of reaching universal access acrosssectors Goal Projected annual Historic average Historic investment needs annualinvestment resource availability Electricity No official goal exists. An -Distribution: 137 .Distribution:Na. -Funds: 36 illustrative target would be to reach -Rural:36 -Rural: 15 -All energy: 95% rural coverage by 2015. - Total: 173 160 Natural gas To reach 88% coverage inrelevant 31 65 - areas by 2008 Telephone To reach 100% access by 2020 to - Villages: 3 31 62 - rural public telephony subsidies - subsidies for telecenters - Schools: 28 in - Total: 31 -municipal for connectivity capitals subsidies inpublic institutions Transport - Urban: to extend Bus Rapid BRT 262 303 Transit systems to 5 secondary --Tertiary: 180 cities. - 2,500km: 130 - Rural: to rehabilitate 2500 kms o f - Total:572 secondary road, maintain the tertiary road network, and improve community airports. 1 i Water and To reach MDGtarget of 83% for - Urban: 192 190 sanitation water and 76% for sewerage by 2015 -Rural: 64 I L - Total:256 3.73 Intelecommunications, historic investment levels are adequate to reach rural access goals, but not to make rapid progress towards universal connectivity of schools. The government has defined its universal access objectives as 100 percent coverage of rural communities with public telephones, 100percent coverage of municipal capitals with telecenters, and 100 percent connectivity in municipalities, hospitals and public schools. It i s estimated that this will cost US$500 million in government subsidies to achieve, with 77 percent of that cost being associated with connectivity of public schools. If investments continue at the historic annual average o f US$31million, it would take four years to meet all o f these objectives except for that of universal public school connectivity, which would take a further 12 years to achieve. If, however, the full value of resources collected by the Communications Fund could be retained within the sector, the overall objectives could be reached within eight years. 106 3.74 For water and sewerage, the cost of reaching the Millenium Development Goals is broadly in line with historic investments inthe sector. It i s estimated that the cost of reaching the MilleniumDevelopment Goals, narrowly defined to mean household connections to water and sewerage, would ascend to US$190 million per year over the next decade. O f this total, 46 percent relates to investments in the water service and the remaining 54 percent to the sewerage service, while 77 percent corresponds to investments in urban areas and the remaining 23 percent to rural areas. It i s estimated that this i s broadly equivalent to what utilities have historically been investing in coverage expansion, and in any case significantly less than the US$250 million of central government transfers to municipalities for water sector investments under Law 715/01. However, even though the overall investment aggregate looks manageable, given low historic rates of expansion in rural access to water and sanitation, it seems likely that rural utilities may not be allocating sufficient resources for this purpose. Finally, if the definition of access i s broadened to include lower cost services and technologies, it i s likely that some cost savings could be achieved. For example, if the unit cost of providing rural access could be halved in this way, the annual investment requirement would fall to around US$170 million. 3.75 In transport, the cost of reaching universal access in transport is difficult to bound, but based on current government initiatives would cost US$572 million per year. In contrast to the other infrastructure sectors, universal access objectives for transportation are not so sharply defined, nor easily bounded given the enormous scale o f the problem. Nevertheless, based on the government's current objectives of introducing Bus Rapid Transit systems in five secondary cities, rehabilitating 2,500 kilometers of secondary and tertiary roads, and maintaining the accessibility of the existing tertiary road network, an annual investment of US$572 million would be required, split roughly equally between the urban and rural transport initiatives. 3.76 Beyond specific universal access initiatives, a number of other socially oriented investments totaling US$481 million per year can be identified. O f these, the largest item i s the expansion of the fixed telephony network expected to require US$271 million per year and to be funded directly from user charges. In addition, maintenance needs for existing water and sewerage infrastructure will amount to US$85 million per year, plus a further U S 9 0 million per year to cover investments in water resources, waste water treatment, and storm water drainage. Finally, the introduction of natural compressed gas as a vehicles fuel will need around US$30 million per year on average. Natural gas transportation investments required to support expansion of the distribution networks are estimated in US$9 million per year. 3.77 Summary. The water sector as a whole receives enough resources to meet the Millenium Development Goals, and even exceed them, particularly if lower cost technologies were more widely used. However, it i s not clear whether these resources are reaching rural municipalities in sufficient quantity to accelerate historically slow rates of service expansion. The natural gas and telecommunications sectors both have the financial capacity to raise the resources needed to reach universal access targets significantly more rapidly than currently planned. However, their investment capacity has been circumscribed by fiscal and regulatory decisions. Finally, the situation i s most 107 critical in the electricity and transport sectors where historic levels of investment seem to fall significantly short of what i s needed to meet reasonable policy objectives, and where alternative sources of finance are far from clear. 3.78 Overall investment requirements to provide the infrastructure needed to meet social and universal access objectives in Colombia are estimated at around US$1,500 million per year, which is equivalent to 1.8 percent of GDP (Table 3.16). About 50 percent of this investment (or 0.9 percent of GDP) relates to expansion of service coverage in urban areas, while 44 percent (or 0.8 percent of GDP) relates to expansion in rural areas. The transport sector makes the largest demand on resources, accounting for 39 percent of the total (or 0.7 percent of GDP), followed by 23 percent (or 0.4 percent of GDP) for the water sector. The share of these resources that could be financed by the private sector i s relatively low standing at 18 percent overall, and relates primarily to expansion of fixed line telephony. Table 3.16: Overall summary of investment needs, priorities, and financing sources US$m ! Public finance Private finance Total Electricitv 105 69 173 Distribution (SIN) 69 69 137 ZNI 36 0 36 Natural gas 0 Transportation 0 Distribution 0 Vehicles 0 71 302 71 271 0 31 Water 516 Maintenance I70 MilleniumDevelopment Goals 256 Water resources 7 Waste water treatment 48 Sewerage systems 17 Storm water drainage 18 - -~- _ ~ _ _ ~ Transport 572 Urban transport 262 r- Secondary and tertiary roads 310 ___.. -~ . ___ Total 1,424 210 1,633 Maintenance 170 0 170 Investment (urban) 590 132 I 722 Investment (rural) 574 69 642 _____ - Total (as % GDP) 1.6 Maintenance 0.2 Investment (urban) 0.7 Investment (rural) 0.7 0.1 0.8 ~ource: Own elaboration 108 AFFORDABILITY OF INFRASTRUCTURE SERVICES 3.79 Obtaining access to infrastructure services is of limitedrelevance if tariffs are too highto permit for continuing use of the service. This section describes and evaluates the substantial policy measures that exist to safeguard affordability of infrastructure services in Colombia. For the utilities, the discussion centers on the national cross-subsidy framework. Attention then turns to subsidies for urban transportation services. Utility services Evaluation of cross-subsidy scheme 3.80 Colombia has a unique and longstanding cross-subsidy scheme applying a common national framework to the electricity, natural gas, telephone and water services. The system dates back to the 1960s when it was introduced by the National Tariff Board. Thereafter, the principle of solidarity as a principle of tariff setting for public services was enshrined in the 1991 Constitution. The system underwent a major reform in 1994, when the Public Services Law (Law 142/94) introduced new limits on the extent of cross-subsidization that shouldbe taking place. 3.81 The cross-subsidy system is characterized by a number of interesting features that distinguish it from cross-subsidies more commonly applied in other countries. First, it employs a zonal poverty criterion based on a six level social stratification of all neighborhoods throughout the country, according to the physical quality of their housing and general amenities. Second, the subsidies and surcharges are both explicit and transparent, being reported directly in customer bills. Third, the scheme i s based on a principle of income redistribution at the national level, so that some utilities have a surplus of cross-subsidy resources that are supposed to be,transferred to other utilities that face a deficit situation. Fourth, it provides for sectoral cross-subsidy resources to be complemented by public transfers from all tiers of government. 3.82 The existing cross-subsidy scheme can be evaluated along a number of dimensions. The first and most fundamental question is whether the system i s really needed, inthe sense of whether it responds to genuine affordability problems faced by the population within each sector. The second question i s whether the system i s accurate in targeting resources to the most needy households. The third question i s whether the system delivers subsidies without undermining the financial sustainability o f the sectors. The fourth question i s whether the system introduces incentives for perverse behavior. Each of these questions i s explored in some detail below. (a) Impact on affordability 3.83 Absolute utility expenditures rise substantially across the income spectrum, butthe share of income absorbed is much higher inthe lowest income group. Utility services on average absorb around 4.5 percent of householdincome, equivalent to around US$20 per month in absolute terms (Figure 3.8). The income share falls dramatically 109 across the income distribution, from over 9 percent in the bottom quintile to less than 4 percent inthe top quintile. However, the absolute expenditure inthe top quintile i s around four times that of the bottom quintile. Across the income spectrum, electricity i s the largest expenditure accounting for 40-50 percent of total spending on utilities, followed by telephony and water with around 25 percent each. 3.84 Subsidies represent a substantial proportion of household income among the poorest, but their absolute value rises markedly in the upper quintiles. Total utility subsidies from all sectors represent 3.6 percent of household income on average, which i s equivalent to US$15 per month. The importance o f subsidies as a share of household income falls substantially across the income distribution, from almost 8 percent in the first quintile, to 5-6 percent inthe middle quintiles, to barely 2 percent inthe top quintile. However, the absolute value of subsidies raises substantially, with the top quintile receiving around US$24 per month, almost three times as much as the bottom quintile receives. The water sector accounts for 82 percent of the value of subsidies that households receive, compared to 16 percent for electricity and 2 percent for telephones. Figure 3.8: Average expenditure on householdutilities by income quintile 10% 9% 40 8% 7% c 35 30 6% 5E 5% :25 20 4% 15 3% = 10 2% 5 1% a 0 0% UTeleph EWater Income quintile CIWater Income quintile IOElectficiivI 0 Electriat,t (a) As percentage income (b) Inabsolute terms Source:DANE, EncuestaCalidad de Vida, 2003 Adapted from MelCndez, 2004 Figure 3.9: Average utility subsidy received by income quintile I s = 8% 4 7% 846 2 5% % 4% E$ 3% 2% 146 0% h o m e quintilr wWacer Bemicity (a) As percentage income (b) Inabsolute terms Source: DANE, EncuestaCalidadde Vida, 2003 Adapted fromMelCndez,2004 110 3.85 Hence, without the subsidy, the lowest quintiles would spend 17 percent of income on utility services. In the absence o f cross-subsidies, the share of household income going to utility services would rise to 8 percent on average. However, the impact differs markedly across quintiles. For the first quintile the expenditure share would rise to around 17 percent of income, for the middle quintiles to around 12 percent of income, and for the top quintile to around 6 percent of income. While there i s no theoretical basis for constructing reasonable thresholds for the percentage of household income that should be devoted to utilities, international experience suggests that values of 10-15 percent are typical for the first quintile. This suggests that the problem of affordability of utilities in Colombia is exclusively confined to the first quintile of the income distribution, and that the largest part of that affordability problem i s associated with the water and sanitation sector. Indeed, eliminating the subsidy for local telephony still leaves first quintile expenditure within the affordable range. Figure 3.10: Weight of utility subsidy by incomequintile E 18% .- E2 16% 14% 12% 10% c 2 4 0 8% 6% E 4% e 2% W a 0% W Income quintile Source: DANE, EncuestaCalidadde Vida, 2003 Adaptedfrom MelCndez,2004 3.86 Moreover, the real cost of a subsistence basket of utility services has doubled over the last decade, but is still comparatively affordable in absolute terms (Figure 3.9). The results quoted above are based on actual expenditure, and could be distorted by excessively highconsumption. One way of controlling for this is to calculate the cost o f a subsistence basket of utility services. In this case, the subsistence basket is defined with respect to the monthly `basic consumption' thresholds already defined by regulators for each of the sectors, notably: 20 cubic meters of water, 200 kilowatt-hours of electricity, and 200 pulses of local telephony (equivalent to approximately 600 minutes of local calls). The results show that the cost of this subsistence basket has more than doubled in real local currency terms. This increase is due, in approximately equal measure, to increases in the basic price of the service and a reduction in the extent of cross-subsidies. The current cost of the subsistence basket, equivalent to US$30 per month, represents around 8 percent of household income on average. 111 Figure 3.11: Evolution of expenditure of subsistence basket of utilities 10.0%T- 100,000 90% " 0 h r- 80% so" 70% 70,000 E 2 60% 0" 2 50% " 0 ; c 3 40% 43" g 30% 20" 52 2 0 % m.mo 10% 10" 4 0 0% 1893 1887 2033 -Absolute expenditure (wrth subsidy) -Absolute expenditure (without subs idy) +Expenditure cs percentage income(with 5 uhsidy) -@-Expenditure a5 percentage income(without s ut6 idy) Source: DANE,EncuestaCalidadde Vida, 2003 Adapted from MelCndez, 2004 (b) Accuracy of targeting 3.87 The current cross-subsidy scheme relies on two targeting instruments: stratification and consumption. The stratification system i s used to determine which households are eligible for subsidy, and relies on the assumption that poverty i s spatially clustered in neighborhoods, and linked to the physical characteristics of urbanization. In addition, the subsistence consumption threshold acts as a secondary filter, ensuring that subsidies to eligible households are not channeled towards non-essential consumption. 3.88 The stratification system is based on national guidelines, but leaves considerable discretion at the local level. The stratification system was first introduced in 1984, but the underlying methodology was only standardized in 1994. The stratification process i s conducted at the municipal level by mayors and special committees, in accordance with the national guidelines established by National Planning Department (DNP). This methodology provides for different approaches according to the heterogeneity of the setting. However, broadly speaking, the methodology i s based on an evaluation of a number of characteristics of the dwellings (garden, garage, faqade, roof, floor, width of lot, doors, windows, service connections), and neighborhoods (roads, pavements, garbage, service availability). Although the evaluation process looks at characteristics of individual dwellings, a single stratification is ultimately assigned to the whole cluster of dwellings. The stratification manuals specify in some detail the variables to be considered, but do not provide rigid formulae for moving from observed characteristics to specific stratifications. There i s no regular process for re-evaluating stratification from time to time, unless the authorities decide to do so, or households appeal against their categorization. 112 3.89 There is evidence that the targeting accuracy of the stratification system has deteriorated over time. Under the current stratification system, local mayors have the power as well as the political incentive to reclassify neighborhoods downward from high to low strata. The household survey evidence documents the fact that the percentage of households classified in the lower strata has increased markedly during the last decade (Table 3.17). Thus, from 1993 to 2003, the percentage classified as Stratum 1 has tripled, the percentage of those classified as Strata 1 or 2 has doubled, and the percentage classified as Strata 1, 2 or 3 (and hence eligible for subsidy) has risen from 75 percent to 90 percent. These changes are much larger than could be justified on the basis of underlying changes in the poverty rate. Moreover, the household survey evidence indicates that the targeting performance of the subsidy scheme in terms of errors of inclusion has deteriorated by five percentage points. Percentageof householdsfalling into: Stratum 1 Stratum2 Stratum3 , Error of inclusion , Error of exclusion Strata 1-3 1993 ,6.4 25.1 , 43.5 75.0 52.6 3.6 1997 12.6 42.3 32.7 87.5 56.2 1.8 2003 21.2 41.7 27.5 90.5 57.7 1.5 3.90 The variables underlying the stratification system, although associated with poverty, have relatively little statistical power to identify poor households. A recent study seeks to analyze the relationship between stratification, the underlyingcriteria, and the income level of the household (Melendez, 2004). An econometric evaluation confirms that the underlying variables on which the stratification system i s based are significantly correlated with the currently assigned strata, suggesting that at a broad level the official criteria are being applied. However, it also reveals that the variables involved in the stratification process, while statistically significant in predicting the income level of households, present combined explanatory power of less than 10percent (that i s to say that these variables are linked with poverty but that they only tell a small part of the poverty story). Moreover, if a statistical criterion i s used to derive stratification from the same set of underlying variables, it emerges that almost 70 percent of households are currently allocated to strata higher than what the statistical model would predict. 3.91 Thus, lower strata are not well aligned with lower income groups, but that higher strata do correspond to higher income groups. In order to understand the targeting properties of the stratification system, it i s helpful to illustrate how the strata map into income quintiles, and vice versa. Figure 3.10(a) shows that the beneficiary strata (1-3) are quite evenly spread across the income distribution, with close to 20 percent in each income quintile. On the other hand, the contributing strata (5-6) are very heavily concentrated in the top income quintile. Figure 3.10(b) shows that more than 90 percent of households in the first four income quintiles belong to the beneficiary strata (1-3), and even in the top income quintile as many as 70 percent of households belong to the beneficiary strata. On the other hand, less than two percent of the households in the first 113 four income quintiles belong to the contributing strata (5-6), and only 11 percent of the households inthe top quintile belong to the contributing strata. Figure 3.12: Relationship between strata quintiles 5 ow C 70% IStratum2 eStratum2 BStratum3 Stratum3 % 50% QSiratum 4 50% Stratum4 ;43% OStratum 5 0Stratum5 1 3% 3 0 % a)% 20% 5z 10% Isf 8 10% 0% 0% 1 2 3 4 5 1 2 3 4 5 lncnmeaumtile5 Income quirrtile (a) Stratuminquintile (b)Quintileinstratum Source: DANE, Encuesta Calidadde Vida, 2003 3.92 As a result, the stratification system is not very effective at channeling subsidies towards the poor. Table 3.18 reports the key targeting performance criteria for the stratification system. The results indicate that the subsidy scheme is very broad in its coverage, excluding no more than 2 percent o f the poor for services with broad coverage such as electricity, water and sanitation. However, this comes at the cost of very high leakage rates. Some 50-60 percent of subsidy beneficiaries are from the top half of the income distribution. Moreover, only 30-35 percent of subsidy resources are captured by the poor, well below the benchmark of 60-80 percent considered to be good practice in the social protection literature. Overall, the performance of the subsidy i s progressive, as indicated by the low and positive quasi-Gini coefficients. The water service presents the worst distributional performance, while the telephone service presents the best. Looking at the quasi-Gini coefficients for subsidies distributed to specific strata shows that targeting performance varies substantially, with Stratum 1 subsidies being moderately progressive in their distribution, Stratum 2 subsidies being mildly progressive, and Stratum 3 subsidiesbeingmoderately regressive 114 TargetingErrors Gross subsidy distribution Surchargedistribution Inclusion Exclusion Share to poorest I Quasi-Gini Share to poorest, 1Quasi-GiniI ~ 8.7 +0.60 - 7.9 +0.62 - I 8.7 I1 +0.61 ~ I I I 1 ! I I I 1 Source: DANE, Encuesta Calidad de Vida, 2003 Note: Poorest defined as bottom 40 percent of the income distribution. Data do not permit calculation of value of natural gas subsidies received. 3.93 However, the burden of surcharges is very progressively distributed. Although the stratification system may not be very effective at channeling subsidies to the poor, it does at least avoid burdening them with the financing of those subsidies. Less than 10 percent of subsidy resources are raised from the bottom 40 percent of the income distribution (Table 3.18). Instead, the distribution of surcharges i s strongly progressive, as indicatedby the highpositive quasi-Gini coefficients. Indeed, the payment of surcharges matches the overall distribution of income in the country quite closely, giventhat the income Gini coefficient stands at 55 percent. 3.94 The level of subsistence thresholds used to determine subsidy payments are currently under review. Law 142/94 empowered the regulators to determine the subsistence consumption threshold above which no subsidies would be provided to any stratum. These monthly thresholds were originally set to 200 kilowatt-hours per month for electricity, 20 cubic meters for both gas and water and 750 minutes for local telephony. The level of the subsistence threshold i s increasingly being questioned across the sectors, and a number of concrete reform proposals already exist, including one to reduce the threshold for local telephony to 200 minutes per month. Lowering the subsistence threshold could potentially have a significant impact on the financial balance of the sectors, and i s a change that can be made by regulation without recourse to legislation. 3.95 A review of international evidence confirms that the subsistence threshold for water and electricity is high, while for gas it is quite well positioned. The use of subsistence thresholds i s widespread throughout Latin America for the electricity and water services, but much rarer for natural gas and local telephony. In electricity, a wide range of thresholds are practiced with Bolivia, Ecuador and Paraguay in the 20-50 kilowatt-hour range, Argentina, Uruguay and Venezuela in the 100-150 kilowatt-hour range, and Brazil, Honduras and Guatemala in the 200-300 kilowatt-hour range. Based on engineering parameters i t i s possible to show that a consumption of 120 kilowatt- hours per month is sufficient to meet basic needs for lighting, refrigeration and radio or small television. In water, thresholds vary from 10-30 cubic meters per month but typically cluster around 20 cubic meters. In Chile, which launched a carefully designed water subsidy policy in 1990, a threshold of 15 cubic meters was adopted. According to the World Health Organization (WHO), 60 liters per capita per day i s a bare minimum level of water consumption (equivalent to nine cubic meters per month for a family of five), while 120 liters per day i s a comfortable allowance of water (equivalent to 18 cubic meters per month at the household level). Inthe case o f natural gas, a consumption of 20 115 cubic meters per month i s adequate for cooking and water heating, though not for extensive spatial heating. 3.96 Moreover, consumption data confirm that existing thresholds are high in relation to average householdconsumption, particularly for water. While the profile of physical consumption against income i s quite flat across the first three quintiles, it becomes much steeper thereafter so that relative to bottom quintile households the top quintile consumes twice as much electricity, three times as much water, and five times as much local telephony. Nevertheless, households in the first four quintiles still consume on average substantially less than the subsistence threshold for the corresponding service. Only the top quintile consumes on average more than the subsistence threshold, andeven then only for electricity and telephony services. Figure 3.13: Physical consumption of servicesby quintile 1 (a) Relative to Stratum 1 (b) Relative to subsistence Source: DANE,Encuesta Calidad de Vida, 2003 Adapted from MelCndez, 2004 (c) Impact on financial sustainability 3.97 A key objective of the 1994 reform was to reduce the extent of cross- subsidization in order to improve the financial sustainability of the sectors. At the time of Law 142/94, the cross-subsidy system was generating a huge financial deficit particularly in the water sector. A key objective of the law therefore became to limit the permissible extent of cross-subsidization by requiring Stratum 4 reference tariffs to rise to full cost recovery levels, and reducing the magnitude of subsidy and surcharges applied to calculate the tariffs for the remaining strata. The maximum subsidy was set at 50 percent for Stratum 1, and 40 percent for Stratum 2. Subsidization of Stratum 3 would be at the discretion of the regulator, and would be eliminated altogether for Stratum 4. Furthermore, the maximum surcharges for Strata 5 and 6, as well as for commercial and industrialcustomers, was limited to 20 percent. A two-year deadline was established for tariffs to reach compliance with all of these new requirements. However, given the magnitude of the changes involved, in practice this did not prove to be politically or socially feasible. As a result, the legal deadline has been twice extended, by five years in 1996 until 2001 (Law 286/96), and b y a further four years in 2001 until 2005 (Law 116 632/01). In the specific case of the water sector, the subsidy limit for Stratum 1 was later increased to 70 percent (Law 812/03). 3.98 Convergence to cost recovery tariffs for Stratum 4 customers has now been achieved in the energy sector, however a significant gap remains for water and telephony (Figure 3.14). The legal requirement to bring Stratum 4 tariffs in line with true cost recovery tariffs, presented a different level of challenge for each of the utility sectors. Innatural gas, given that the sector was just starting-up, Stratum 4 tariffs were set at the cost recovery level from the beginning. For electricity, the gap between cost recovery tariffs and those practiced at the time of reform was relatively modest; of the order of 25-30 percent, allowing convergence to be reached by 2001. B y contrast, in the case of water, there was a much larger gap of the order of 70 percent to be bridged, and while it has been substantially reduced over time, it has not yet been entirely eliminated. Figure 3.14: Convergence towards target tariff I 200, I Source: Own calculations based on data from Regulatory Commissions 3.99 Rebalancing of subsidies and surcharges to meet new legal limits is complete inthe energy and telephone sectors, but remains a distantgoal for the water sector. Inthe case of the energy and telephone sectors, the subsidies practiced at the time of the reform were already within or close to the limits established in Law 142/94, although the surcharges on the upper strata were in some cases substantially higher (Table 3.19). This made the rebalancing process comparatively easy, with the result that today all o f these sectors conform with the legal limits, and indeed practice subsidies that are substantially lower than the maximum allowable; with discounts offered to Stratum 1 in the 20-35 percent range well below the 50 percent limit. Inthe water sector, on the other hand, the divergence between historically practiced subsidies and surcharges and the new legal limits was huge, with subsidies to Strata 1 and 2 o f the order o f 60-80 percent, and surcharges on Strata 5 and 6 around 100 percent and 200 percent respectively (Table 3.19). As a result, the adjustment process has been extremely difficult, and current practices are still far from those legally prescribed. Greatest progress has been made in the easier area of reducing surcharges. Due to the modification of the limit for Stratum 1 subsidies to 70 percent in 2003, this i s the only Stratum that i s currently in compliance with the legal limits. 117 Electncity NaturalGas Telephone Water Limit 1Current 1 1 I I 1 ~ Initial Limt Current Initial Limit , Current Initial Lirmt Current Initial Stratum 1 -50 I -19 -45 -50 ' -35 -33 -50 1-33 -40 -70 -70 -82 I -40 -26 -32 -40 -54 , -6 1 -15 -2 -17 -15 -32 Stratum4 0 0 0 0 0 0 0 0 +58 1y20 I 0+19 1 ,+26 +20 +20 +35 +20 +31 Stratum6 +20 +85 +20 I +19 +I5 +20 I+20 +84 +20 +56 1;:5 1 1 Non- +20 ' +49 +9 i +9 i I-46 residential I<20 ' I Source: Own calculations based on data from Regulatory Commissions Note: Data for water refers to four largest cities only, and for telecommunications to the 10 major operators. 3.100 The difficulties faced by the water sector are readily explained by the magnitude and regressivepattern of the tariff increasesentailed by the law. Inorder to understand the widely varying outcomes across sectors, it i s important to understand the combined impact of moving Stratum 4 tariffs to cost recovery levels and simultaneously unwinding cross-subsidies between strata. This i s best understood by calculating the overall real percentage tariff increase required for each stratum in order to move from the pre-reform starting point to the endpoint specified in the law (Figure 3.15). For domestic customers in the energy sectors, the required tariff increases were very modest, no more than 20-30 percent and followed a reasonably progressive distributional pattern. In the telephone sector, the required increases were substantially larger at 50-70 percent, but were concentrated in the middle strata. For the water sector, the required increases were both very large, and very regressive in their incidence, entailing 40 percent tariff increases for Stratum 1 customers for example. Figure 3.15: Percentage tariff increasesrequired to reach target levels 450 400 350 300 250 200 150 100 50 0 Source: Own calculations based on data from Regulatory Commissions 3.101 In spite of progress in convergence towards the legal parameters, substantial structural deficits remain in all the cross-subsidy schemes, except for natural gas. 118 Each of the sectors presents its own particular situation with regard to the financial balance of the cross-subsidy scheme. In the natural gas sector, the structure of demand makes it possible for the cross-subsidy system to be completely self-financing within the legal parameters established. Indeed, the sector currently presents a modest surplusbetween subsidies and surcharges. 0 In the telephone sector, there is currently a deficit of US$21 million per year, equivalent to 2 percent of sector turnover. However, given that a significant gap still exists between the Stratum 4 tariff and the cost recovery level, it i s conceivable that this deficit would disappear once full convergence with legal requirements has been reached. 0 In the electricity sector, the structure of demand is such that a major structural deficit persists in the cross-subsidy scheme even though the legal parameters have been fully complied with. The deficit currently stands at US$77 million per year, equivalent to 12 percent of sector turnover. 0 In the water sector, given the unfavorable structure of demand and the fact that the sector i s still far from reaching the legally established parameters, the cross-subsidy scheme presents a large deficit of US$161 million per year, equivalent to 20 percent of sector turnover. About 60 percent of the deficit i s generated by the water service, and 40 percent by the sewerage service, given that the latter i s more prevalent in the upper strata. It i s estimated that even if convergence with the legal parameters could be fully achieved, a significant structural deficit would remain. Figure 3.16: Financial balance of cross-subsidyschemes nsurcharge OSubsidies balance QCompeEate balance Source: Own calculations based on data from Regulatory Commissions 3.102 These structural deficits can be understood in terms of the underlying imbalance between contributors and beneficiaries in the utilities market structure. The persistence of deficits in the cross-subsidy system i s easily understood by examining the relative weight of beneficiary segments (Strata 1-3) and contributing segments (Strata 119 5-6 plus non-residential customers) in the overall turnover of the utilities (Figure 3.17). These proportions vary substantially across sectors, with 60-80 percent of revenues coming from contributing segments in the energy services, compared to only 20 percent for the water sector. Figure 3.17: Relative balance between beneficiaries and contributors 100% 90% 2E 80% 70% m +8 + 80% 8 50% E 40% i- t0 3096 20% 10% 0% Gas Electricity Telephone Water 1. Beneiciaries Source: Owncalculations based on data from Regulatory Commissions 3.103 Moreover, the distribution of residential customers is heavily skewed towards the lower strata, particularly outside the major urban centers. Within the total group of contributors, residential contributors (Strata 5-6) represent a very small fraction of the total amount contributed in surcharges. The reason i s that the residential population i s very heavily skewed towards the lower strata (Figure 3.18(a)). Indeed, across all sectors, between 85-90 percent of residential customers fall into Strata 1-3. Moreover, this aggregate national picture conceals the fact that the upper strata customers are overwhelmingly concentrated in a handful of the larger cities, making the situation much worse in utilities serving smaller cities and rural areas. Thus the percentage of beneficiaries (Strata 1-3 rises from 83 percent in Bogota, to 90 percent in other cities, to 98 percent in rural areas (Figure 3.18(b)). Indeed, even Bogota has only 7 percent of its residential customers in contributing strata (5-6). Figure 3.18: Distribution of populationacross strata EE 446, -7 35% .c 3% VI ;i25% E 1546 5p 0 10% 5% Strdum Stratum Stratum Stratum Stratum Strdm 1 2 3 4 5 0 =Water 0 ElectricQ (a) By sector (b) By geographic area Source: DANE, EncuestaCalidad de Vida, 2003 120 3.104 Geographic redistribution of subsidies as foreseen in the legal framework, is only currently functioning in the electricity and gas sectors. Law 142/94 provided for National Solidarity Funds to be set-up in each of the utility sectors. Utilities with a surplus of surcharges over subsidies, would be required to transfer this into the fund, while utilities with a deficit, would be able to draw from the surplus resources. Local, departmental and national governments would be able to top-up the funds in order to compensate for any remaining structural deficit at the sectoral level. This arrangement was supposed to ensure that overall each sector would have a `compensated balance' of zero (Figure 3.16). As of today, National Solidarity Funds have been established and are functioning only inthe electricity and gas sectors. a The natural gas fund has sufficient resources to fully compensate utilities with subsidy deficits, and hence the sector has a `compensated balance' of zero. a The electricity fund receives government transfers of the order of US$77 million per year ensuring that the sector achieves a `compensated balance' of zero. It i s relevant to note that surplus surcharges transferred from the utilities to the national fund are able to cover only 20 percent of the subsidy deficit generated by the other utilities, with the government transfer absorbing the remaining 80 percent. A key concern among distribution utilities are the delays that tend to affect the payment of the government's contribution to the subsidy scheme. a Inthe telephone sector, no National Solidarity Fundwas ever established; instead Law 286/96 prescribed that the Communications Fundtake on this role. However, this i s somewhat at odds with the fund's central objective of focusing on access to rural telecommunications. Moreover, the fund has itself had half of its budget diverted to the exchequer, and i s therefore short of resources. Consequently, the fund has so far not made any contribution towards the cross-subsidy deficit, leaving utilities to left to absorb the subsidy deficits on their balance sheets. a Inthe water sector, given the extent of decentralization, the law mandated the establishment of municipal funds. However, as of 2002, only 54 percent of municipalities had established such funds, and of those only 5 percent were allocating resources to these funds. It i s estimated that these municipalities are allocating a total o f US$43 million per year, which reduces the overall sectoral deficit by about 25 percent. Notwithstanding their unwillingness to allocate resources to solidarity funds, the municipalities are receiving central government transfers worth US$250 million per year, which at least on aggregate far exceed the total subsidy deficit of US$160 million per year (see Chapter 5). However, for political reasons, municipalities prefer to allocate these resources to finance new investment projects rather than to compensate utilities for subsidy deficits. 3.105 I t is striking that the bulk of the imbalances between subsidies and surcharges are heavily concentrated in a number of geographical areas. In the case of the water sector, virtually all utilities are internally in deficit in terms of the balance 121 between subsidies and surcharges within their service area. However, almost 80 percent of the deficit can be accounted for by the utilities serving the three largest cities (Bogota, Medellin and Cali). In the case of the electricity sector, the utilities serving these three largest cities run a surplus between surcharges and subsidies, which accounts for just under 90 percent of the surplus available at the national level. However, this i s more than offset by the deficit, of which 54 percent i s related to the ZNI, and a further 33 percent to the utilities serving Meta, Caqueta and Quindio. The fact that the three largest cities have a surplus for electricity but a deficit for water, suggests that it may ultimately be possible to substantially reduce the sectoral deficit in water by concentrating actions in these three utilities. 3.106 However, the financial deficit created by the cross-subsidy system, may in practice be largely absorbed as a foregone return on equity by publicly owned utilities. Although the deficits generated by the subsidy system appear to raise serious financial concerns, in practice, given the preponderance of public enterprises in these sectors, the financial impact of the subsidies is more limited. The reason i s that the cost recovery tariffs against which the subsidies are calculated include a full market rate of return on the utilities' capital base. Since public (and particularly municipal) enterprises are not required to pay a rate of return on equity, this provides some financial space within which the subsidy deficit can be (at least partially) accommodated. To give some idea o f the potential financial space involved, the annual subsidy deficit to the water sector represents no more than 2 percent o f the total value of government equity in this sector. 3.107 The structural deficits on the various cross-subsidy schemes are likely to deteriorate further in the coming years, due to coverage expansion and new legislation. There are two reasons to think that the deficits of the cross-subsidy systems are likely to deteriorate in the medium term, even if the process of convergence to the parameters established in Law 142/94 could be fully completed. First, given that access deficits are concentrated in the lower strata, further coverage expansion will tend to exacerbate the structural deficit. Second, the recent National Development Plan Law introduced an important change to the subsidy and surcharge limits, by prohibiting any real increases in the tariffs applied to Strata 1 and 2 over and above the rate of inflation (Law 812/03). This essentially prohibits any further movement in the current level of subsidy percentages for Strata 1 and 2, creating difficulties for the telephone and water sectors where further real increases would have been required to meet the originally prescribed targets, and for the energy sector to the extent that future real tariff increases may be required to fund investment programs. Melendez (2004), estimates that for every percentage point differential between general inflation and sector specific cost increases, the measures contained in Law 812/03 will increase the annual deficit o f the cross- subsidy system by US$12 million for electricity, US$8 million for water, and US$5 million for fixed line telephony. Overall, the Regulatory Commission for Telecommunications (CRT) estimates that the National Development Plan Law could add a further US$25 million to the annual deficit up to the year 2010. Whereas in the case o f electricity, it i s estimated that the deficit could grow from its present level of US$77 million to US$363 million by 2006. 122 (d) Lmpact on incentives 3.108 The current cross-subsidy system introduces a wide range o f perverse incentives both for consumers and utilities, and most seriously for local mayors. 0 For higher strata and non-residential customers, there may be an incentive to bypass the public utility service altogether and find alternatives that avoid the payment of the surcharge. Chapter 2 presented some evidence that this indeed happening, particularly in the water sector. The same incentive may exist for higher strata residential customers in sectors where ready substitutes exist, such as telephony and gas. 0 For utilities, the current system severely reduces the incentive to expand services to unserved areas, given that these are disproportionately composed of lower strata customers that would only contribute to increase the current subsidy deficit. 0 However, the most perverse incentive of all i s that faced b y local mayors to re-classify their neighborhoods into lower strata. This incentive arises both for electoral reasons (as a popularity measure), and even more critically for fiscal reasons, to avoid the creation of surplus resources that would simply be transferred to other jurisdictions. The obvious solution to this problem would be to take the decision about strata classifications away from the local level, and to link it in a very mechanistic and transparent way to objective and observable characteristics, that could be automatically updated from time to time, for example, in line with each new census. 3.109 Summary. While utility tariffs have increased substantially in Colombia during the last decade, the absolute value of charges i s still reasonable and would appear to be affordable for most consumers. However, serious affordability problems exist for first quintile households, for whom utility subsidies represent around 9 percent of household income, and who would need to devote 17 percent of their income to utilities in the absence of subsidies. Nevertheless, the current cross-subsidy system presents a number of serious design flaws. Zonal stratifications are poorly aligned with income poverty, and their precision as a targeting proxy has been deteriorating over time. At present, more than 50-60 percent of subsidy beneficiaries come from the top three income quintiles and only around 30 percent of subsidy resources are captured by the bottom two quintiles. Indeed, the system i s mildly regressive in its overall impact on the distribution of income. Furthermore, serious and deteriorating structural deficits exist in the design o f the cross- subsidy system (equivalent to 20 percent of sector turnover for water), and are placing an increasing financial burden, in some cases on utility balance sheets, and in other cases on public expenditure. Two key underlying problems are the perverse incentives for individual jurisdictions to avoid bringingthe cross-subsidy system into balance, and the relatively discretionary nature of allocation of neighborhoods across socio-economic strata. 123 Simulation of reforms 3.110 Even within the current legal framework, a number of significant measures could be taken to improve the performance of the cross-subsidy system. The current legal framework gives regulators important areas of discretion in setting the key parameters of the cross-subsidy system, in particular the decision as to whether or not Stratum 3 households should be considered eligible for subsidy, as well as the determination of the level of the subsistence threshold beyond which no further subsidies are given. In addition, Decree 565/96 allows for the value of percentage surcharges and subsidies to be altered to the extent necessary to improve the financial balance of the sector. Finally, the regulations governing the allocation of households to specific strata could be revised to make them increasingly rigorous, objective, and transparent, as well as more difficult to change on a discretionary ad hoc basis. 3.111 Lowering of subsistence thresholds goes some way towards reducing the financial deficit, butdoes not lead to a major improvement intargeting. A simulation exercise explores the impact of separately varying a number of parameters in the design of the subsidy model, considering to what extent these serve to improve targeting and financial balance, or generate changes so large that they are unlikely to be politically acceptable. The results show that significant reductions in the subsidy deficit of around 10-20 percent simply by reducing subsistence consumption thresholds (Table 3.20). However, such a reform would have virtually no impact on the distributional incidence of cross-subsidies, with the key targeting parameters remaining largely unchanged. 3.112 Restricting water cross-subsidies to existing legal limits would appreciably reduce the financial deficit and slightly improve targeting performance. In the specific case of the water sector, it i s relevant to simulate the impact of completing the process of convergence towards the subsidy and surcharge limits stated in Law 142/94. The results indicate that the completion of this reform would of itself reduce the financial deficit associated with the water subsidy scheme by almost 20 percent. It would also give rise to modest improvements inthe targeting properties of this subsidy. 3.113 Restricting eligibility for subsidieswithin Stratum 3 would lead to significant improvements in financial sustainability and targeting. Given that Stratum 3 represents the largest, and most poorly targeted category of subsidy beneficiaries, the simulations explore the effect o f reforming eligibility for Stratum 3 subsidies. One possibility would be to eliminate the automatic subsidy entitlement for these households, but to allow them to apply for a subsidy in cases where they meet a certain number of additional criteria. This approach i s simulated using criteria related to housing characteristics, educational attainment, electricity consumption and ownership of luxury goods. The results indicate that this approach would reduce the financial deficit of the subsidy by around 30%. It would also significantly improve targeting performance with quasi-Gini coefficients turning mildly progressive. However, targeting remains unimpressive in absolute terms, with poor households capturing little more than 40% of total subsidy resources. 124 3.114 Achieving more substantial targeting improvements would thus appear to require some departure from the traditional stratification approach. There are two potential explanations for the poor targeting performance of the stratification system observed above. This first i s that the housing characteristics underlying the stratification system are not very well correlated with poverty. The second i s that they are well correlated with poverty but that they are not being very rigorously applied in practice. Melendez 2004 performs some statistical tests that help to discriminate between these two hypotheses. The study reports that many o f the variables underlying the stratification (presence of garden, presence of garage, quality of faGade, floor materials, garbage collection, availability of electricity and telephone services) are indeed significantly correlated with poverty but that they collectively account fro less than 10% of the observed variation in income across households. That i s to say that the variables are individually significant but not jointly material. At the same time, Melendez finds that a household's stratum classification i s significantly correlated with the underlying housing characteristics but that these explain less than 20% of the stratum classification. Overall, these results suggest that stratum classifications are not entirely dictated by the underlying housing characteristics, but that they would not accurately reflect poverty even if they were. The implication i s that additional non-housing poverty indicators may be needed if targeting performance i s to be improved. 3.115 One alternative to stratification might be the adoption of the SISBEN indicator used to determine eligibility for other social programs. One option might be to use eligibility criteria already developed in the context of other social programs. Colombia has a national beneficiary selection system, known as the SISBEN. However, the basis for the SISBEN was recently reformed so as to bring it more closely into line with the stratification system. As a result, the targeting performance of the new SISBEN criterion is very similar to that observed under the status quo stratification framework. Nevertheless, an illustrative simulation i s also performed with the original SISBEN criterion, and illustrates that this indicator would have substantially improved the targeting performance of the scheme, yielding a very progressive distribution of the subsidy, as indicated by the large negative quasi-Gini coefficients. At the same time, the financial deficit associated with the cross-subsidy scheme would be substantially reduced for the electricity sector, and turned into a surplus for the water and telephony sectors. However, the original SISBEN criterion i s in the process o f being phased out, and hence no longer represents an available basis for any reforms to the cross-subsidy system. 3.116 Another alternative would be the development of a new poverty indicator, independent from either SISBEN or the current stratification system. For illustrative purposes, an ad hoc criterion i s developed based on housing characteristics, educational attainment, electricity consumption and ownership of luxury goods. The simulations suggest that use o f such a criterion could reduce the financial deficit by around 40%. Targeting would also improve significantly with mildly progressive quasi-Gini coefficients, and errors of inclusion falling from around 50% to around 30%. Nevertheless, the absolute performance of such a scheme remains wanting with less than 50% of subsidy resources reaching the poor. Moreover, the administrative costs of setting-up a new and independent targeting system may be quite costly. 125 3.117 Finally, it would be highly desirable to reduce the discretion entailed by the current stratification system. In addition to considering changes to the parameters of the stratification system, an equally important area for reform would be to reduce the margin of local discretion that exists in the assignment and reassignment of households and neighborhoods to particular strata. This could be done by providing a more tightly circumscribed methodology for household classification, calibrated on the basis of census data, to ensure that a more consistent definition of the various strata is adopted across jurisdictions. In addition, it would be desirable to perform random auditing of how correctly the stratification principles are being applied inpractice. Table 3.20: Summary of impact of reforms within current legal framework Target Sustainability Accer iility I Error Error Share to QGC Percentage Percentage Average inclusion exclusion poorest reductionin who lose loss deficit (US$lm) Electricity Status quo 51.2 1.5 40.2 +0.01 - - - 3educe 51.2 1.5 41.2 +0.01 -12.4 49.6 0.89 ;ubsistence .hreshold Restricted Stratum 36.7 8.5 44.8 -0.04 -28.8 55.7 1.52 3 subsidy SISBEN (old) 9.8 30.9 76.7 -0.37 -72.6 83.9 4.10 SISBEN (new) 51.8 1.3 41.3 +0.02 -2.7 9.5 2.74 Altemativeto 28.7 14.2 47.8 -0.08 -39.6 51.3 4.31 stratification Telephone Status quo 25.6 21.5 34.8 +0.08 - - - Reduce 25.6 21.5 35.4 +0.07 -10.0 33.5 0.19 subsistence threshold SISBEN(old) 2.0 67.0 82.6 -0.47 -323.7 78.0 2.51 SISBEN(new) 24.2 22.0 37.5 +0.06 -34.5 11.5 2.08 Water Status quo 51.1 0.7 34.0 +0.11 - - - Fullconvergence 45.6 1.5 35.1 +0.08 -18.4 76.8 0.98 Law 142194 Reduce 50.7 1.2 34.5 +o.10 -14.5 51.1 0.65 subsistence threshold Restricted Stratum 23.1 17.9 41.8 -0.02 -45.2 69.0 2.47 3 subsidy SISBEN(old) 24.7 8.6 53.2 -0.13 -194.0 85.5 4.84 SISBEN(new) 49.0 0.8 34.5 +0.11 -14.1 12.6 3.58 Altemative to 29.9 10.7 40.9 0.00 -40.5 81.9 1.46 stratification - Jote: QGC - quas 3ini coef ient measures the ove I1 distr itional incidence of the subs r', positive values indicatea regressive or pro-rich distribution while negative values indicate a progressive or pro-poor distribution Source: Fedesmollo simulations based on DANE,Encuesta Calidad de Vida, 2003. 126 3.118 Summary. The current legal framework gives significant regulatory discretion over a number of key parameters of the utilities cross-subsidy framework. It is estimated, that changes to these parameters could help to reduce, though by no means eliminate, the financial deficit caused by cross-subsidies in some sectors. However, improvements in the targeting performance of the scheme would necessitate a major overhaul of the stratification system given that housing variables only explain a small proportion of variation in household income. In addition, measures should be taken to reduce the local discretion associated with the assignment and reassignment of households across strata. Urban transportation 3.119 Urban transportation expenditure absorbs as much as a quarter of the income of first quintile households, without providing an acceptable level of mobility. Urban households in Colombia spend an average of 7 percent of their income on public transportation, with the share ranging from 25 percent for the bottom quintile to 4 percent for the top quintile (Figure 3.19(a)). In absolute terms, this amounts to an average monthly expenditure of US$33, ranging between US$19 and US$42 from the bottom to the top quintile. This expenditure provides the first two quintiles with an average consumption of less than one journey per household member per working day, and overall less than would be required for two adults to commute to work (which i s to say, 82journeys per month). Figure 3.19: Householdexpenditure andconsumptionof public transport services 1 2 3 4 5 Income quirrtile Income quintile -Percentage +&solute (a) Expenditure (b) Consumption Source: Own elaboration based on EncuestaCalidad de Vida 2003 3.120 The government has two policy instruments that directly or indirectly contribute to making urbanpublic transport services more affordable. The first i s a labor regulation requiring employees of formal sector workers earning between one and two minimumwages, to give them an additional allowance that i s designed to cover the cost of a standard commute. It i s estimated that this costs employers a total of US$293 million per year, equivalent to US$18 per beneficiary or a 12 percent mark-up over the basic minimum wage. The second are a number of concessions on fuel tax and vehicle import duties that favor bus operators. These measures effectively act as an implicit subsidy with the effect of reducing the average level o f urban transport fares, benefiting all bus users in an indiscriminate manner. Although it i s difficult to estimate the full value of these measures, the diesel subsidy alone i s believed to amount to an estimated total 127 benefit of US$70 million per year for the bus industry, and consequently allow fares to be 10percent lower than would otherwise be the case. 3.121 Unfortunately, none of these policies appears to be very effective in delivering subsidy resources to poor households. Urban transport subsidies have multiple objectives, among which the most important are those of reducing congestion and improving mobility of low-income households. Performance against this second objective can be assessed by calculating subsidy incidence based on household survey data (Table 3.21). Together these two subsidies are substantial amounting to just under 1 percent o f household income, and as much for 2 percent for middle-income groups in Bogota. The results suggest that the two subsidies perform very poorly in distributional terms, in some respects worse than the utility cross-subsidies. Around 60 percent of subsidy beneficiaries are from the upper three quintiles of the income distribution, while almost 60 percent of the poorest fail to benefit from any of these schemes. Overall, only 11 percent of these subsidy resources benefit the bottom 40 percent of the income distribution, with a highly regressive quasi-Gini coefficient of +0.37. Of the two policies, the diesel subsidy i s by far the least regressive. These results essentially reflect the fact that the poorest households do not tend to be formal sector workers and hence fail to benefit from the minimum wage policy. They also illustrate the intrinsic difficulty of targeting subsidies inwidely used services like public transport. r- Targetingerrors 1Gross subsidy distribution 1 Inclusion Exclusion iShare to poorest 1Quasi-Gini Employersubsidy 87.2 91.2 17.6 1 Diesel subsidy 75.0 87.5 25.0 +O. 16 Total 57.6 59.4 11.4 Source: DANE: EncuestaCalidad de Vida, 2003. Note: Poorestdefined as bottom40 percent of the income distribution; Transmilenio cross-subsidyrelevant for Bogotaonly 3.122 Summary. Urban transportation expenditures represent some 24 percent of first quintile income, and provide these households with very limited levels of mobility. The government i s estimated to forego US$70million of revenue per year on diesel subsidy to the urban bus sector, and requires employers to spend US$293 million per year on transport subsidies to low income formal sector workers. However, only about 10percent of the beneficiaries of these policies come from the first two income quintiles, and they capture no more than 20 percent of the value o f these subsidies. CONCLUSIONS AND RECOMMENDATIONS 3.123 The Colombian state has made major efforts to expand access of infrastructure services throughout its national territory, and to ensure that these are affordable to all of its citizens. A wide range o f programs to promote universal access exist, and have helped to deliver substantial coverage improvements over time, while 128 utility and urban transport incorporate substantial subsidies and cross-subsidies that on aggregate are worth about 8 percentage points of the income of the poorest households. It i s interesting to evaluate the relative financial importance of universal access initiatives versus the cross-subsidy scheme. The evidence suggests that in recent years the country has invested some US$425 million per year on universal access initiatives, of which 20 percent corresponds to formal rural access programs, and the remaining 80 percent to expansion undertaken primarily by public utilities. B y contrast, the aggregate deficit of the cross-subsidy scheme represents a smaller financial commitment, amounting to US$259 million per year, of which only the US$77 million per year corresponding to the electricity sector i s formally covered by the government. 3.124 There is still significant roomfor improvement in the design of many of these policies. The preceding analysis leads to a number of recommendations for future policy that can be summarized as follows. (a) Greater attention should be paid to strategies for accelerating access to electricity, and particularly water and sanitation in rural areas, given that past progress has been too slow to allow universal access goals to be reached within a reasonabletime frame. Inelectricity, this will involve both scaling-up efforts through FAZNI and FAER channels, but also strengthening the regulatory incentives for distribution utilities to invest in marginal rural areas. In water and sanitation, there i s no specific central policy to promote access in rural areas, leaving this responsibility with small municipal utilities that often suffer from weak management and precarious finances. It may therefore be worth considering whether there is scope for a special program focused on expansion o f access in rural areas, buildingon some of the initiatives already in place. (b) Steps need to be taken to improve the reliability of funding sourcesfor the rural electricity and telecommunications programs. While the electricity and telecommunications sectors have established their own self- financing rural access programs, these have been unable to spend the resources at their disposal, due to overarching restrictions on public expenditure. In the case of telecommunications, sector resources have moreover been absorbed into the general budget. Further progress in rural service expansion in these sectors therefore depends heavily on the ability to provide a more secure fundingbasis for these programs. (c) More consideration should be given to the potential role of low cost appropriate technologiesin meeting needs for water and sanitation. In the water sector, the approach to universal access has tended to focus narrowly on the provision of household connections to water and sewerage as the only acceptable solution for meeting water and sanitation needs. Thus, relatively little consideration has been given to alternatives (such as public taps and on-site sanitation) that usually lead to significant cost savings, and may provide a more appropriate form of service in certain environments. 129 (d) There is pressing need to reform the design of the existing utilities cross-subsidy framework, to ensure that it targets resources effectively towards the first income quintile, without jeopardizing the financial sustainability of the utilities sectors. Many improvements could be made within the confines of the current legal framework. Phasing out Stratum 3 subsidies, and lowering the subsistence consumption thresholds, for example, would already go some way towards improving the financial sustainability of the scheme. However, in order for targeting performance to reach acceptable levels, a major overhaul of the stratification system i s required. This will entail revising the regulations governing the allocation of households to specific strata to make them increasingly rigorous, objective, and transparent, and eliminate the scope for discretionary reallocations. (e) Subsidies for the use of natural gas would best be replaced with a connection subsidy to attenuate the high costs of switching to natural gas. Given that natural gas can meet domestic cooking and heating needs at about a quarter of the cost of its typical substitutes, it i s hardto justify a subsidy for the use of this service. The high switching costs into natural gas do however represent an important barrier, and mean that low-income households must currently pay higher fuel bills for five years before the benefits of natural gas become apparent. This problem would better be addressed by substituting the existing use of service subsidy with a connection subsidy targeted at low-income households, which would be a much more efficient and effective way of raising natural gas penetration. A subsidy of the order of US$170 per connection, leaving households to finance their internal conversion, would probably be sufficient to meet this objective. (f) Subsidies for the use of local telephony services may be reaching the end of their useful life, given the increasing availability of alternative services. While cross-subsidies for local telephony have historically been successful in raising telephone ownership among low-income groups, there are reasons to think they are becoming less relevant in the modern market environment. These include the proliferation of creative tariff plans offered by fixed line providers that allow consumers to choose the most attractive financial package, the substantial growth in availability of public telephony options in urban areas, and the falling cost of prepaid mobile telephones. Moreover, the application of surcharges to local telephony to finance the cross-subsidy scheme, distorts the intensifying competition between fixed line, mobile and internet services. (g) There is a need to find new policy instruments to address the mobility problems of the urban poor. Existing urban transport policies seem to be having a significant impact on reducing congestion costs inBogota. However, they are largely by-passing the poorest households, which currently face high transportation costs and restricted mobility. Further 130 thought needs to be given about how best to reach this segment of the population. Potential options include the use of more targeted programs, such as design of Bus Rapid Transit (BRT) feeder routes around poverty maps at the city level. (h) Some thought should be given to developing a stronger national policy for rural water and sanitation. Although the national government takes many different measures to support rural services, they do not seem to amount to a coherent rural policy. This i s in contrast to the other utilities, where clearly defined rural policies exist, and i s readily explained in terms of the decentralized nature of this service. Nevertheless, the slow rate of progress with water and sanitation coverage in rural areas, suggests that it may be desirable for the national government to take a more proactive stance in this area. (i) There is a need to improve tariff incentives for service expansion in the SIN. Inaddition to the various rural electrification funds that are being developed, it would be desirable to review the tariff regulations for electricity distribution utilities operating within the S I N that currently fail to provide adequate incentives for service expansion in marginal areas. Given the shortage of fiscal resources, this would help to ensure that the distribution utilities themselves play the greatest possible role in contributing to rural electrification, without introducing excessive distortions into the tariffs paid by other customers. It would be desirable to harmonize financial contributions for rural telecommunications fund. The current structure of financing for the rural telecommunications fund, with levies ranging from 0% for local telephony to 5% for mobile telephony, according to the sub-sector does not appear to be a very equitable financing mechanism given that these different services compete against each other. It would therefore be preferable to replace them with a relatively low single levy applied indiscriminately to all telecommunications services. 131 4. LEGAL,REGULATORYAND INSTITUTIONAL FRAMEWORKFORINFRASTRUCTURE INTRODUCTION 4.1 The adoption of a new legal, regulatory and institutional framework for infrastructure services provided basis for the major reforms of the 1990s. Colombia's major infrastructure sector reforms of the 1990s, are rooted in the 1991 Constitution, and developed in the 1994 Public Utilities Law (Law 142/94). Whereas the new Constitution clearly places ultimate responsibility for service provision with the government, Law 142/94 marks an important shift by recognizing the importance of private participation in service provision, establishing regulatory frameworks, and introducing bold liberalization measures across the utilities. Promotion of competition (wherever possible) and sound economic regulation (wherever necessary) are seen as the two main instruments for assuring the improvement and expansion of infrastructure ' services. The details of the new framework are developed in subsequent legislation, usually along sector lines, although some of the more recent legislation does adopt a wider scope. 4.2 The new legal framework brought major benefits, but also incorporates serious shortcomings that prevent the reforms from fully delivering on their promises. This chapter examines the evolution of the new legal, regulatory and institutional framework for infrastructure. The central objective i s to identify the major shortcomings in the design and implementation of institutions and regulations, and recommend the changes required to correct the emergingproblems. 4.3 The first part of the chapter provides a rapid overview of the legal, regulatory and institutional framework, as well as the key features of each of the utilities sectors. The first part of the chapter will characterize the legal, institutional and regulatory framework o f each of the utilities sectors, and highlights some o f the particularities of the Colombian model, such as the strict separation between regulation and supervision, andthe constitutional foundations of the legal framework. 4.4 The second part of the chapter identifies the structural weaknessesthat limit the effectiveness of the legal and institutional framework generically across utilities sectors. As a whole, the legal and institutional framework adopted for the utilities in Colombia i s a sophisticated one and incorporates several original features, which compare favorably with the models adopted by other reforming countries in the region. However, in practice the system presents several weaknesses, gaps and even 132 contradictions that are currently eroding its credibility and effectiveness. This i s not entirely the fault of the model, but i s also attributable to the failure of some of the underlying assumption to fully materialize in practice, limiting the possible achievements of the reform (Ayala and Millan, 2003). Inparticular, private participation did not emerge on the scale anticipated, nor did it prove possible to scale back direct state involvement in service provision, thereby creating a more level playing field and freeing-up fiscal resources for other national goals. 4.5 The third part of the chapter reviews the specific challenges of applying the new legal and regulatory framework to each of the utilities sectors. While a common legal and regulatory framework exists across the utility sectors, implementation experience varies substantially, reflecting the widely differing circumstances and characteristics of each sector. It i s therefore important to complement the generic diagnosis of the legal and regulatory framework, with a sector-by-sector review of the specific difficulties that have arisen in each case. The analysis will also try to assess the relative importance of legal and regulatory shortcomings, in contributing to other sector challenges identified elsewhere in the document, such as market structure and financial sustainability. Inaddition, the transport sector which i s the only area o f infrastructure not to be covered within the framework of Law 142/94 will receive its own separate treatment. 4.6 The final part of the chapter reviews the legal, regulatory and institutional framework for the transport sectors. The transport sectors are not covered by the framework of Law 142/94. Instead, they have been evolving along separate lines, with marked differences across the different transport modes. Some quite recent reforms have introduced a framework somewhat similar to the utilities one for concessions with private operators in the road, rail and ports sectors. This comprises an Institute of Concessions (INCO) that awards and regulates concessions in these areas, as well as a Superintendence with responsibility for supervision and control across the transport sectors. However, other services such as urban transportation and tracking continue to be subject to more traditional licensing and regulation directly from the Ministry of Transport. In the air transport sector, responsibility for policy-making, regulation, and supervision, as well as provision of air traffic control services, continue to be merged within a single entity. OVERVIEW OFLEGAL, REGULATORYAND INSTITUTIONAL FRAMEWORK 4.7 The legal framework for provision of public services in Colombia adopted in the early 1990s assigns important roles both to the state and to the private sector. The Constitution of 1991, the Public Services Law 142/94 and several other laws in subsequent years have established a new legal framework for infrastructure services in Colombia. This legislation created institutions designed to allow the state to take on its new regulatory and supervisory role, while maintaining its role as a policymaker, and often as direct provider despite the shift. This section presents an overview of these laws 133 and institutions, and the remainder of the chapter will be devoted to assessing their impact on the actual performance and development of the sectors. Legal framework 4.8 The 1991 Constitution is unusually detailed and explicit about the principles governing provision of utility services. The 1991Constitution identifies public utilities as one of the core services that contribute to the well being of the population, and therefore treats this topic in considerable detail. The Constitution defines the nature of these services, and the role of the different actors in providing these services, whether directly or indirectly through the regulation and supervision of the sectors. It reiterates the ultimate responsibility of the state for ensuring the provision of these services to its citizens (Article 365, e) and its obligation to supervise and control their provision (Article 334). At the same time, the 1991 Constitution assigns an important role to the private sector by stating that these services may be provided directly by the state or delegated to the private sector or grass root community based organizations (Article 365), and b y establishing the freedom of private initiative and economic activity within the boundaries established to protect the common good (Article 333). Article 150 explains that the control of the state in these matters can be undertaken through the issuances of laws for economic intervention in a given sector. Such laws should clearly state the objectives and scope of the state's intervention, in order to preserve the legal rights of the private entrepreneurs. 4.9 Colombia is also unusual in having a comprehensive framework law for the public utilities: the Ley de Servicios Publicos (Law 142/94). In most countries around the region, infrastructure reform has been undertaken on a piecemeal basis, through sector-by-sector reform laws, or sometimes sustained only by decrees and regulations. Colombia i s unusual in having a single very detailed public utilities law that covers the electricity, natural gas, local telephony, and water and sewerage services within a consistent and unifying framework. As such, Law 142/94 i s one of the most detailed pieces of utilities legislation inLatin America. Among the key elements of the law are the creation of the Superintendence for Public Services, and the definition of the functions of the three Regulatory Commissions: Regulatory Commission for Water and Sanitation (CRA) for water, Regulatory Commission for Electricity and Gas (CREG) for electricity and gas and Regulatory Commission for Telecommunications (CRT) for telecommunications. Furthermore, the law promotes the adoption of cost recovery tariffs for the utilities, and establishes limits on the extent of cross-subsidization between customers. The law also establishes immediate liberalization of all utility services, including those traditionally regarded as natural monopolies, and explicitly exonerates service providers from any requirement to hold a concession from the conceding authority. 4.10 Regulation, control, and supervision functions are explicitly assigned to the President, who delegates them to the corresponding entities. Article 370 of the Constitution assigns direct responsibility to the President of the Republic to set the policies of the sector, and ensure adequate control and supervision of these services 134 through the Superintendence of Public Utilities (SSPD). Article 68 of Law 142194 further expands that the President may exercise these functions directly or delegate them to Regulatory Commissions. B y means of Decrees 1524/94 and 2253/94, the President did indeed delegate regulatory functions to the Regulatory Commissions (CRA, CREG and CRT), and supervision and control functions to the SSPD. 4.11 Complementary sector specific legislation is less comprehensive and shows uneven development across sectors. The legal framework for the electricity and gas sectors was completed through the issuance of Law 143/94, as well as a number of other laws and decrees. However, the water and sanitation sector does not have any sector legislation beyond Law 142/94. This legal vacuum has made it even more difficult to establish general standards and performance criteria that could apply to the large number of decentralized utilities across the country. The situation in the telecommunications sector i s more complex. The current legislation covers all dimensions of these services, but does so in a piecemeal form. Several attempts to adopt a Telecommunications Law have been unsuccessful, and the sector therefore continues to lack a unified vision. Thus basic telephony has been defined as a public utility and regulated under Law 142/94 (just like electricity and water), whereas the remaining services are regarded as ordinary commercial services, coveredby other laws and decrees. 4.12 However, there have been recent changes inlegislation that could weaken the overall legal framework in the longer term. Recently, legislation has been used as a tool to resolve disputes between the government and the Regulatory Commissions, or to pre-empt actions of the latter that could deviate from the short-term objectives of the Government. One example i s the National Development Plan Law (Law 812/03) of the current administration. This contains articles that transfer to the line ministry the authority to rule on regulatory issues such as freezing the on-going process of dismantling cross subsidies for utilities mandated by Law 142/94. 4.13 Except in telecommunications, regulations issued have focused more on pointsof detail rather than broader economic issues. The CRT has been collaborating closely with the line ministry, acting as de facto technical secretary, developing norms and regulations to support and implement the sector policies defined by the ministry. By contrast, in the case of CRA and CREG a stricter separation of roles is practiced between the Ministries and the Office of the President, whereby the latter issue the laws and regulations and the former issue regulatory resolutions. Line ministries 4.14 The line Ministries are in charge of defining sector policies, and play an active role inthe Regulatory Commissions, which they preside. The line ministries are the Ministry of Mines and Energy (MME for electricity and gas), the Ministry of Communications (MC for telecommunications and postal services), and the Ministry of Environment, Housing and Territory Development (MMAVDT for water and sanitation services). Ministries are responsible for policy formulation, as well as the granting of concessions. The ministries provide their respective Regulatory Commissions, and in the 135 case of CRA and CRT, their favorable vote i s needed to approve any decision. Furthermore, in the case of telecommunications, the Ministry i s in charge of managing and supervising the radio electric spectrum, and has the responsibility to defend competition inthe behaviors incommunications and postal services. 4.15 The advanceddegree of decentralization in Colombia places significant limits on the authority of ministries' within their respective sectors. Many aspects of the provision of utility services are undertaken by sub-national governments. Thus, water and sanitation services are a municipal responsibility, while many electricity distribution and local telephony operators operate under municipal jurisdiction. Local governments have significant authority over the services under their jurisdiction, and are relatively free to dispose of financial resources for the sector, even when these originate from the national govemment. In particular, municipal governments are responsible for guaranteeing service provision, and have the power to tax the services, define areas of service, and control public space issues. Nevertheless, the central govemment retains the responsibility of supervising the ex-post performance of all utilities nationwide, through the SSPD, and has the obligation to intervene in the management of utilities that find themselves infinancial distress. 4.16 Ministers sit on the Boards of most large national public enterprises, leading to major conflicts of interest with their roles of policy maker and regulator. As noted above, many key utilities remain under national jurisdiction, such as some of the most important electricity generation and transmission companies, the gas transportation operator, and the national telecommunications incumbent. With the recent exception of ISA and of Telecom, the line ministries represent the interest of the state in these enterprises, with the minister (or his delegate) presiding over the board of these companies. Thus, the line Minister effectively ends-up playing a triple role as policy- maker, member o f the Regulatory Commission, and member of the Board of Directors of many of the largest utilities in the country. This evidently leads to huge conflicts of interest that undermine the credibility of the whole system, and generate uncertainty among private investors and municipal operators as to whether they are being fairly treated. Regulatory Commissions 4.17 The three Regulatory Commissions are made-up of senior government representatives, together with a number of full-time technical experts. Sector regulation i s the responsibility of the CREGfor Electricity and Gas, CRA for Water (and CRT for Telecommunications. These were created in 1992 as a result of the restructuring of the earlier National Tariffs Board, although their responsibilities and capabilities were not fully established until the later passage o f Law 142/94. The three Regulatory Commissions are special administrative units within their respective sector ministries. In principle, they enjoy administrative, technical and financial autonomy, although they are fully subject to public sector budgetary procedures, and to the annual expenditure ceilings established by DNP. The Regulatory Commissions finance their activities through regulatory surcharges of up to 1percent on the operating expenditures of their respective 136 sectors, in addition to revenue from the sale of publications. The Members of the Commissions include the respective line Ministers (who are the presiding members), a number of other Ministers (Finance for CREG and Social Protection for CRA), the Director of DNP, and the Superintendent of Public Utilities (who speaks without a vote). In addition, each Regulatory Commission comprises several full-time technical experts; four of them in the case of CRA and CRT, and five in the case of CREG. These are appointed directly by the President for terms of four-year duration. These experts may have their terms renewed, and can only be replaced on grounds of resignation or term expiration. The Regulatory Commissions appoint an Executive Director, who along with the line Minister has the authority to convene the Regulatory Commission. The numbers of members required to establish a quorum varies across the three Regulatory Commissions. Inthe case of CRA and CRT, the line Minister must be present. Decisions are taken on the basis of simple majority, but in the case of CREG, the favorable vote of at least one government representative i s required. The decisions of the Regulatory Commissions can be appealed directly to each commission, and later appealed through thejudicial system, if the result of the first appeal was not satisfactory. 4.18 The main functions of the Regulatory Commissions are to promote competition (where possible) and conduct regulation of monopoly power (where necessary). The two central functions of the Regulatory Commissions are the regulation of monopoly power and the promotion of competition. For services in transition, the Regulatory Commissions have the responsibility o f determining the steps towards gradual market liberalization. They can decide when it i s appropriate to establish regulated tariffs or to allow free determination of prices inthe market place. However, all tasks relating to the award and supervision of concession contracts rest with the line Ministries, rather than the Regulatory Commissions. In principle, the tasks entrusted to the three Regulatory Commissions are similar in nature. However, in practice, there are subtle but important differences between them, some of them deriving directly from the legal framework, and others reflecting a different evolution of regulatory practice in each sector. 137 Table 4.1: Comparison of the three Regulatory Commissions Z R A ZRT ZFWG Members iexperts,lineMinister,Social 3 experts,line Minister, DNP, plus 5 experts, line Minister, Ministerof IevelopmentMinister andDNP, SSPD. 7inance andDNP, plus SSPD ]Ius SSPD. Executive Vominatedby Commissionon a Vominatedby Commissionon a Votedby Commissionbasedon Director 12-monthrotatingterm. 16-monthrotatingterm. :xperts' proposalon a rotatingterm. Quorum 5 membersincluding line 3 members, includingline Minister 5 members Minister Decision Simple majorityamong those Simple majority among those Simple majority among those present making present, in absence of unanimity present but favorable vote of DNPor minister dissentrecordedinproceedings if needed; unanimity requiredfor resolutions - Tie Vote is repeatedat next meeting, Line Minister arbitrates thereafterCommissionPresident makes final decision. Appeals Only available for specific (as Only available for specific (as Appeal possible within five days of opposedto general) opposed to general) administrative the decision administrativeprocedures procedures Remuneration Defined by the National Defined by the National President Defined by the National President President based on Law 4/92, based on Law 4/92, salary scale in based on Law 4/92, salary scale in salary scale in Decree 3535/03 Decree 353Y03 Decree3535/03 Budget Requirement to present annual Requirement to present annual Requirement to presentannualbudget budget to govemment, and abide budget to government, and abide to govemment, and abide by national by nationalexpenditure limits by nationalexpenditure limits expenditure limits I ______ Revenues Special contributionsmadeby Specialcontributionsmadeby Specialcontributionsmade by regulatedentities (Law 142/94), regulatedentities (Law 142/94), regulatedentities (Law 142/94and I plusrevenues from publications, plus revenuesfrom publications, Law 143/94), plus revenues from and retumon excessliquidity andretumon excess liquidity publications Surplus Reimbursed to contributors, or Reimbursed to contributors, or Reimbursed to contributions carriedover to next budget year, carriedover to next budget year, or carried over to next budget year, or I or transferred to the Nation if transferred to the Nation if other transferred to the Nation if other other options were not feasible optionswere not feasible options were not feasible -- - Source: Caballero et al. 2004 4.19 In the electricity and gas sector, the CREG is responsible for creating the conditions that guarantee power supply. The CREG promotes the liberalization of the market, defines methods for calculating regulated tariffs, and issues regulations governing self-generation and co-generation. It i s also in charge of developing criteria for guaranteed energy-sale commitments and of establishing regulations for implementation, planning and coordination of SIN operations. Finally, CREG regulates the functioning of the wholesale energy and gas markets. This work is undertaken with assistance from the utilities, the line Ministry, the SSPD, the Mining and Energy Planning Unit of the Ministry (UPME), the National Planning Department (DNP), and the National Dispatch Center (which i s part of the transmission utility ISA). 138 4.20 In the telecommunications sector, CRT is in charge of the economic and technical regulation of all services. These functions include tariff regulations, the definition of technical and economic conditions for interconnection, and the protection of consumers, as well as the definition of quality standards, efficiency criteria and indicators for the supervision of the operators. In contrast to other Latin American countries, it i s the line Ministry, rather than the CRT or SSPD, which i s responsible for the management and control of the radio spectrum. 4.21 In the water sector, CRA defines tariff-setting methodologies based on standard formulas and on investment plans by the operating companies and sets quality and technical standards to be followed by the utilities. Environmental regulation i s handled by Autonomous Regional Corporations (CAR). Superintendence of Public Utilities (SSPD) 4.22 The Colombian regulatory model enforces an institutional separation between regulatory and supervisory functions. The Colombian model establishes an institutional separation between these two activities, on the grounds that conflict of interest could potentially arise between them. Thus, while the Regulatory Commissions are responsible for regulation, the SSPD i s responsible for supervision and control of utilities. Another feature of the model i s that while regulation i s organized separately for the three utility sectors, supervision and control of all utilities are concentrated in a single institution with multi-sectoraljurisdiction. 4.23 The SSPD was created in 1991 by Constitutional Mandate, and its functions were specified in Law 142/94 and later elaborated in Law 689/01. The SSPD i s structured in sector divisions: water and sewerage; electricity and gas; and telecommunications. It has five regional offices spread across the national territory. The SSPD i s a technical organization with administrative, technical and financial autonomy. Unlike the Regulatory Commissions, which depend on their respective line Ministries, the SSPD i s a separate legal entity. Although empowered to manage and use its own resources, it nonetheless remains subject to the expenditure limits imposed through the national budgeting process. Like the Regulatory Commissions, SSPD finances its activities through a surcharge of up to 1 percent on the turnover of the services it supervises, as well as revenues from the sale of publications. Law 689/01, together with Article 132 in the National Development Law 812/03 provides additional resources to finance the intervention and liquidation o f public enterprises in distress. The Superintendent i s appointed by the President, and can be dismissed at any time at his discretion. 4.24 The SSP has a very broad array of responsibilities. It supervises the performance of the public services providers and monitors their compliance with service and safety standards and other regulations issued by the Regulatory Commissions. It oversees market competition, certifies stratification of residential users in the allocation of subsidies, and ensures that the subsidies reach the poor, based on this stratification. It issues opinions to the Regulatory Commissions and line Ministries regarding the 139 managerial performance of services providers and their compliance with sector laws and regulations. SSPD also investigates and requests information on irregularities that may arise within companies. It conducts inspections and tests, penalizes companies that fail to comply with the rules, and has the authority to intervene and liquidate non-performing public enterprises, where necessary. It can cancel contracts and may forbid agents from providing services for as long as 10 years. Finally, SSPD acts as an appeals body for consumer complaints against service providers and has the responsibility to guide and support community participation, especially through local Social Development and Control Committees. 4.25 The presence of publicly owned utilities in most sectors results in a multiplicity of institutions in charge of controlling their performance. In addition to the control and supervision of the SSPD, public utilities are also subject to controls from the General Audit Office (Contraloria) and the Attorney General's Office (Procuraduria).The objective of these additional controls i s to ensure compliance with official accounting procedures and proper use of public funds. The same controls apply to departmental and municipal utilities, which are supervised by the local offices of all three entities. 4.26 Anti-trust is another area where multiple agencies have received mandates of one type or another to ensure compliance with the Law. Law 689/01 clearly gave SSPD the responsibility to defend consumers against anti-competitive practices by utilities. Nevertheless, the Superintendence of Industry and Commerce (SIC) has a more general mandate to promote competition across all sectors of the economy, while the Ministry of Communications has received a similar mandate for the telecommunications sector by presidential decree (decree 2122). This multiplicity of actors in the anti-trust sphere evidently creates conflicts and ambiguities that complicate the process of promoting fair competition. 4.27 Summary. Colombia's utility sectors rest on a particularly solid legal foundation, rooted in the 1991 Constitution and developed extensively in the 1994-framework law. However, sector-specific laws have only so far been developed for the energy sector, in spite of some attempts for telecommunications. Moreover, recent changes in legislation have served to weaken the overall framework in some areas. The institutional framework comprises line Ministries that lead policy-making, while at the same time presiding Regulatory Commissions and Boards of national state-owned enterprises, with all the concomitant conflicts of interest. There are three Regulatory Commissions dealing with each of the regulated sectors. These operate by bringing together around a single table, key government representatives, such as Ministers, with technical figures, known as Expert Commissioners. The task o f regulation i s separated from that of control, with the latter being allocated to a cross-sectoral Superintendence, with responsibilities for enforcement in a broad array o f areas. The Superintendence i s also the competition authority for the utility sectors, notwithstanding the existence o f a broader anti-trust agency. 140 CROSS-CUTTING REGULATORY ISSUES FOR UTILITIES 4.28 Following ten years of experience with the new regulatory model, certain weaknesses have become apparent. The adoption of a new legal, institutional and regulatory framework during the last decade clarified the roles of government, opened-up a new role for the private sector, and providedthe basic conditions for markets to develop and expand, as well as services to improve in quality and efficiency. However, after more than ten years of experience with the new model, certain weaknesses have become apparent. In some respects, the model does not appear to have been successful in delivering the desired goals (witness the precarious financial situation of the distribution utilities). In other respects, new challenges have subsequently arisen that demand some adjustment to the model itself (witness the implications of the FTA for the telecommunications sector). 4.29 The government recognizes the need for some second-generation reforms to the legal and regulatory framework, and has initiated a formal review process. For this reason it has included a revision of the institutional framework for public utilities as part of its Programfor the Renovation of Public Administration (PRAP). Inits guidelines for this revision, the government emphasizes the importance of state intervention to defend citizen interests in imperfect markets, but at the same time recognizes that intervention should be the exception rather than the rule. Furthermore, the guidelines emphasize the need to carefully assess the benefits and costs of any state intervention and to periodically monitor their results. 4.30 This section examines the limitations of the legal framework that cut across the three utility sectors. These are sometimes attributable to design faults, imperfect implementation, or inability to adapt to changing circumstances. The four central issues identified are: first, the conflict of interests between the various roles o f the State (policy maker, regulator and provider); second, the internal structure of the Regulatory Commissions and the lack of a credible regulatory appeals channel; and the internal structure of the SSPD, the need to improve its supervisory capacity and focus on a narrower array of responsibilities; and fourth, the absence of a strong and unifiedantitrust regulation. A subsequent section performs a sector-by-sector review of the specific issues that arise in each case. Conflictsof interest within the State 4.31 Significant conflicts of interest arise between the state as policy maker, regulator, and major service provider in many key sectors. In addition to its roles as policy maker and regulator, the Colombian state continues to play a major role in the provision of utility services. Thus, as noted in Chapter 2 above, the central (and municipal) governments control around 32 percent (21 percent) of electricity generation assets, 84 percent (7 percent) of electricity transmission assets, 17 percent (20 percent) of the electricity distribution market, 39 percent (54 percent) of the local telephony market, 59 percent (30 percent) of the long distance market, and a growing share of the mobile 141 telephony market. The conflicts of interest are particularly strong in the case of national utilities, given that line Ministers are often represented both on the Boards of the utilities and on the respective sectoral Regulatory Commissions. This situation evidently introduces major conflicts of interest into the operation of the model, and makes it difficult to create a level playing field between state-owned and private investors. 4.32 A definitive resolution to this problem is likely to require further state divestment inthese areas. The clearest way of resolving this conflict of interest situation would be to undertake further privatization of some of the major state-owned utilities, particularly those that operate in competitive segments of the market. Indeed, the government has recently announced its intention to move ahead with a major privatization program for state-owned utilities. 4.33 However, it may also be helpful to consider options for a different institutional structure of state ownership. On the understanding that privatization may not always be a viable option, or may take some time to implement, it i s also relevant to consider measures that can be taken to attenuate conflicts of interest within the public sector. The Colombian government is currently considering a number of measures aimed at improving the corporate governance of the state-owned utilities, including codes of practice for good governance. While these measures are undoubtedly helpful, there i s a limit to how much can be achieved while the same actors are involved across the different institutions. To reduce the scope for conflicts of interest, it would thus be important to separate the policy making functions of the line Minister from the task of representing the particular interests of the public enterprises involved in the sector, and monitoring their performance. Privatization o f some of these public enterprises, as i s currently under consideration by the government, i s the clearest way of achieving this separation. However, in cases where privatization does not prove to be possible, financially or politically, some other mechanism for achieving a more arm's length relationship would need to be found. This could be achieved, for example, if a professionally managed Trust Fund was created to represent the Central Government as a stake holder in these enterprises (Rojas and Chahin, 2004). Regardless of the exact figure adopted to ensure representation of the State as a major shareholder in the national public enterprises, it would be important that the entity in charge of playing this role reports to, for instance the Ministry of Finance, rather than to their respective line Ministries, in order to ensure greater distance between the two. In this context, i t would be interesting to examine the Chilean experience with CORFO, a holding company that has accumulated some experience with this approach. Fuzzy boundaries between policy and regulation 4.34 There is an ongoing debate about where the boundary between policy- making and regulation should be drawn. In the water, and most particularly in electricity sectors there has been a fierce debate as to whether specific decisions fall into the policy-making or regulatory domain. This debate is by no means academic and has adversely affected the credibility o f all parties involved, and most especially the Regulatory Commissions. The debate has also had a negative impact on the perceptions of investors, who see increasing uncertainty caused by a lack of stability and clarity inthe 142 rules. For example, there was a conflict between the CREG and the line Ministry over one of the parameters used to calculate the cost of capital for the electricity distribution tariff review. This led to a blockade between the CREG and the Ministry during the ' period 2001/02, which delayed the application of the new tariff formula, and was eventually resolved by a Decree that gave the line Minister the power to arbitrate in cases of dispute within the Regulatory Commission. 4.35 One of the underlying reasons is the absence of a more detailed definition of functions, as well as the possibility for these to be transferred back to the Executive at any time. This confusion i s due in part to the lack of a detailed list of the functions and tasks that are delegated to the Regulatory Commissions. Moreover, the work of the Commission i s also weakened by the absence of rules on the transfer of functions. Thus, under Colombian law, the President or his line Minister may at any time decide to reassume any of the functions delegated to the Regulatory Commissions. 4.36 Another reason for these disputes is the perception that CREG and CRA do not have the technical capacity to fulfill their legal responsibilities. The severity of the problem of `fuzzy boundaries' differs significantly across sectors, with the problems being relatively minor in the case o f CRT and relatively severe in the case of the CREG, even though all three Regulatory Commissions operate under a common legal framework. To some extent, this variation reflects differences inthe technical capacity of the three Regulatory Commissions. CREG and CRA have developed a reputation for having weak technical capacity. As a result, line Ministries are often prompted to work through other channels to implement their policies. Moreover, technical weakness often undermines the credibility of the recommendations of these Regulatory Commissions in the face of disputes with line Ministries or regulated firms. 4.37 However, the problems are sometimes caused by inconsistent behavior on the part of policy-makers themselves. In some cases, policy-makers have chosen to diverge from the overall policy framework of the sector, leaving the Regulatory Commissions to cope with sudden major shifts in the basic parameters of their work. A pertinent example i s National Development Plan Law (812/03), which froze the rebalancing of the tariff structure to reduce the scope of cross-subsidies mandated by Law 142/94 in order to recover the financial sustainability of the utilities. The Regulatory Commissions had made sustained efforts to implement these difficult provisions over the last decade, but this process has now been interrupted until 2006. Hence, some confrontations have occurred because of disagreements between the government and the Regulatory Commission, even on issues for which the jurisdiction of the Regulatory Commissions has never been in dispute. 4.38 A final factor that softens the boundary between regulation and policy- making is the fact that the line Ministers preside the Regulatory Commissions. Inthe case of CREG, in addition, the positive vote of a government representative i s required to approve any decision. These institutional arrangements do not favor the separation between policy makingdecisions (which should take place prior to and outside the realm of the Regulatory Commissions) and regulation (which should be limited to technical specifications within the boundaries established by policy). 143 4.39 There is a voluminous literature on the definition of appropriate boundaries between policy making and regulatory functions. While it is acknowledged that the frontiers are not clear-cut, a consensus has developed on best practices. Policy-makers need to define the basic parameters within which regulators are empowered to deliver on policy goals. Regulators, as technical experts, are responsible for design of detailed norms and the implementation of specific aspects of these policies, within the defined framework (Brown, 2002). Regulators must also promote the long-term sustainability of utility services, which necessitates a certain degree of autonomy so as to be able to resist opportunistic behavior of parts of the Executive. While not so often emphasized, efficient regulation also presupposes a coherent and transparent policy-making process. Indeed, for the regulator to be able to operate within the prescribed policy parameters, those parameters must be sufficiently clear and stable. Policies can evolve and even shift, but need to change in an orderly and transparent fashion, as well as in a consistent direction. 4.40 In many countries, the delegation of Presidential responsibilities to regulatory entities is much more structured and stable than in Colombia. While it is not unusual in other countries for regulation to be defined as a Presidential responsibility that i s delegated to a regulatory entity, the delegation process i s usually governed by much more formal procedures that are more difficult and costly to reverse. Thus, in contrast to Colombia where the delegation of regulatory functions i s absolutely discretionary, the revoking of specific regulatory functions might require (for example) specific changes in the law. To the extent that delegated functions are technical in nature, such as periodic tariff revisions or the approval of interconnection agreements, such arrangements do not restrict the ability of the President or Ministers to make policy decisions. Although tensions between policy-making and regulation necessarily remain, they are forced to come out into the open and subject to broader and more objective debate. 4.41 In order to begin to draw sharper boundaries between policy-making and regulation in Colombia, it would be helpful to establish an ad-hoc Commission. Such Commission should comprise key members of both the Executive and the Regulatory Commissions. Its agenda should be to define the basic concepts that distinguish policy- making from regulatory activity, and undertake a detailed review of existing legislation to identify the responsibilities of each as currently assigned and clarify those areas where there are currently conflicts or gaps. To minimize the risks o f conflict or confusion, it would be preferable to reduce the regulator's scope of competence, but to ensure that the regulator has full freedom to operate within the scope defined. It would also be necessary to define the processes that should govern interactions between Regulatory Commissions and the Executive, as well as formal procedures for making any changes to the existing allocation of responsibilities. Finally, the ad-hoc Commission would also be a helpful forum for seeking greater consistency in the methods underlying regulatory decisions across the different utility sectors, where considerable variation of practice currently exists (for example inthe determination of the weighted average cost of capital.) 4.42 Inthe longer term, a definitive resolution of this problemis likely to require a clearer institutional separation between policy-making and regulation. It i s important to note that the proposed ad-hoc Commission would simply be a temporary 144 time-limited institution charged with providing guidance on this particular problem. The intention i s not to create an additional layer of bureaucracy. However, in the longer term, the achievement of a clearer boundary between policy-making and regulation will require a formal institutional separation between these two functions. The prior exercise of clarifying the roles and responsibilities of policy-makers and regulators will greatly facilitate this process, since it will ensure that policy-makers have the powers needed to formulate policy without needing to form part of the Regulatory Commissions. Internal structure of the Regulatory Commissions 4.43 The Regulatory Commissions are affected by serious human resource constraints, particularly at the highest levels. There i s broad consensus that the Regulatory Commissions do not have the human resources that they need to be credible players in the face of the Executive and the regulated industries. The problems are more severe for CRA and CREG, than for CRT, though all three Regulatory Commissions are affected by a number of broad structural constraints. The key issues relate to the profile, remuneration and selection of Expert Commissioners, as well as the availability of financial resources to fund the necessary level o f staffing. In addition, the line Ministers represented on each of the Regulatory Commissions do not always have adequate back- upinterms of regulatory expertise among their own staff of advisers. 4.44 The legal rules governing the profile of Expert Commissioners are overly restrictive, unnecessarily constrainingthe field of candidates. Law 142/94specifies in some detail the profile of candidates for the post of Expert Commissioner. In particular, these must be Colombian citizens with postgraduate degrees and years of worlung experience within the sector ranging from three (in the case of CRA and CRT) to six (in the case of CREG). However, the prohibition on individuals that have worked for regulated entities during the last year, necessarily rules out a substantial percentage of candidates with the necessary experience. In addition, the legal framework does not specify the desirability of maintaining a multi-disciplinary capability within the Regulatory Commissions, for example by having each of the different experts drawn from a different professional background (engineering, economics, finance, law, management). 4.45 These rules need to be revised to focus on those criteria most relevant to the job, but at the same time giving greater flexibility in some areas. It would also be important to reduce the entry safeguards so as to reduce or eliminate the restriction on prior employment in the regulated industry. Another helpful measure would be to allow Expert Commissioners for one sector to be chosen among qualified candidates from another regulated sector (for example from an electric distribution utility to CRA) or from another related sector (for example from mobile telephony to CRT). The list of qualifying professions should be widened to include lawyers in the case of CREG. One dimension where more demanding criteria would appear to be warranted is in raising the number of years of experience up to ten. 145 4.46 Remuneration levels are not high enough to attract Expert Commissioners with the necessary level of technical expertise. The lack of technical expertise i s due to a large extent to the failure to attract qualified individuals as Expert Commissioners, which play a key role in directing the work of the Regulatory Commissions. This i s attributable to the pay scales that have been set for the Regulatory Commissions, which ensure that Expert Commissioners receive lower salaries than the senior managers of even the public utilities that they regulate, let alone the managers of the private utilities. Moreover, the fact that Expert Commissioners cannot be employed by any regulated entity for a considerable period following their resignation, though understandable, severely restricts their subsequent employment opportunities and hence earnings potential. 4.47 A more competitive remuneration package could be created within the framework of the existing law, while the number of Expert Commissioners could be reduced. While the remuneration of Expert Commissioners i s necessarily bound by pay scales outlined in Law 4/92 to be no higher than that received by Ministers, they could nonetheless be raised within these limits. In particular, Expert Commissioners should receive remuneration comparable to that of managers of the large public and private utilities that they regulate, as well as comparable to that of public employees with similar levels of responsibility, such as the Board of the Central Bank. These increases could be applied gradually to newly recruited Expert Commissioners. Since it i s preferable to have a smaller number of more highly qualified individuals (say two or three rather than four or five), the budgetary impact of higher remuneration could be offset b y reducing the number of Expert Commissioners. Finally, the provision of an exit bonus to those who fully complete their term could help to offset the economic penalties entailed by the conflict of interest provisions. 4.48 The selection process of the Expert Commissioners does not in any way guarantee the necessary quality of candidates. The current recruiting system, whereby the President selects each expert unilaterally without any prior screening does not guarantee the selection of sound technical experts. While this follows typical procedures for high-level political appointments (such as Ministers), it i s not appropriate for what i s by design intended to be a high-profile technical post. In this case, it would be more appropriate to follow a formal selection process. This might, for example, entail prior selection of a shortlist of three or four qualified candidates by the corresponding line Ministry that would be presented to the President for his ultimate decision. B y establishing demanding and well-publicized criteria, and by using a shortlist mechanism, the selection process can be made more transparent and a higher quality result assured while still leaving the final selection of the Commissioners to the discretion of the President. A change of this kind could be implemented by Decree, within the framework provided by Articles 71 and 77 of Law 142/94. In addition, staggering the terms of the Expert Commissioners would help to ensure greater stability across different administrations. 4.49 In addition, lack of financial autonomy prevents the Regulatory Commissions from hiring the human resources that they need to function effectively. The resources raised b y the regulatory surcharge are generally adequate to 146 finance the necessary regulatory tasks. However although the legal framework gives the Regulatory Commissions full financial autonomy, in practice their expenditure plans are severely restricted by the expenditure limits that apply across the public sector. As a result, they are unable to secure the quantity and quality of human resources needed to provide the technical analysis supporting regulatory decisions. 4.50 A strategy is needed to provide stable enough funding to guarantee the continuity of a core group of qualified staff covering the core regulatory competencies.A possible approach would be to agree upon a core team of qualified staff within each Commission, for which a stable budget allowance i s provided, with more discretionary or specialized functions that do not require full time attention being contracted out to consultants subject to resource availability. As for the Expert Commissioners, competitive salaries would need to be established for this core group of staff. Such a strategy i s fully consistent with the financial autonomy established under Law 142/94. However, there are some difficulties of legal interpretation given that the National Budget Law seems to have de facto priority over all other legislation, thereby completely undermining the financial autonomy provisions o f Law 142/94. This i s an issue that requires more detailed legal analysis. Lack of a regulatory appeals channel 4.51 Inprinciple, regulated firms inColombia have recourseto appeal regulatory decisions through both administrative and judicial channels. From a legal perspective, the decisions of the Regulatory Commissions and the SSPD are considered to be administrative acts. Under Colombian administrative Law, individuals adversely affected by administrative acts may appeal directly to the official concerned, and thereafter to the superior of that official if the appeal i s not satisfactorily resolved in the first instance. However, this type of higher-level administrative appeal does not apply to the Regulatory Commissions nor to the SSPD. Once recourse within the same administrative branch has been exhausted, ajudicial appeal may also be raised. 4.52 However, in practice, the judicial appeal channel does not represent a viable appeal option. On the one hand, Regulatory Commissions rarely overtum their own decisions when these are directly appealed. On the other hand, the Colombian court system faces a very large backlog of cases, which greatly reduces its efficiency and its relevance as a viable channel for resolving appeals. At present, it takes between 18 months and 8 years to process a claim through the courts. Moreover, in 2002, there was a backlog of 35 percent of cases received and processed the same year, without taking into account the backlog accumulated from earlier years. The large number of cases presented to the Regulatory Commissions, in particular to the CREG, makes a significant contribution to this backlog. 4.53 Numerous possible appeal mechanisms have been under review, though many are at odds with the Colombian legal system. One possibility was to make use of a tribunal of experts to act as arbitrators of last resort. However, under Colombian law, 147 decisions involving government entities cannot be subject to decision by private parties. Another relatively straightforward option would be to create a specialized chamber within the Consejo de Estado. However, this would encounter similar problems to those faced by the standard judicial appeal route. Judges would have to be selected out of the existing pool, where specialized knowledge o f utilities i s rare. There would be no guarantee of providing a faster solution than the status quo, and it would be difficult to justify preferential treatment for this type of cases over others. 4.54 A possible solution would be to create an appeals body within the existing framework of the Regulatory Commissions. This proposition i s based on the existing principle, explained above, that those affected by an administrative act should be able to appeal to the supervisor of the official in question. Inthis case, the `superior' would need to be created and could take the form of a small cross-sectoral supra-Commission of a handful of members drawn from the Expert Commissioners of the three existing Regulatory Commissions, and could constitute the legal representative of all three (Caballero et al., 2004). The supra-Commission would be convened on an occasional basis to address appeals raised beyond the level of the Regulatory Commissions themselves. In their proposal, Caballero et al. have argued that the supra-Commission could play a variety of roles, including the formulation of cross-cutting guidelines to ensure the broader consistency of regulation. However, there i s a danger that such an arrangement would simply lead to an additional layer of bureaucracy, further blurringthe boundary between policy-making and regulation, and distracting attention from the central need to strengthen the Regulatory Commissions themselves. It would therefore appear preferable to limit the role of any supra-Commission to the processing of regulatory appeals. 4.55 This would appear to be possible within the existing legal framework, although some legal ambiguities will first need to be resolved. Such an institution could be created within the existing framework provided by Law 142/94, since this law allows the President to modify the structure and composition of the Regulatory Commissions by Decree. However, a legal interpretation would be needed to clarify Article 113 of Law 142/94, which currently excludes appeals against the decisions of the Regulatory Commissions within the Executive branch. It i s not clear whether this prohibition stems from the absence of any superior to which an appeal might be raised, or if it reflects an explicit intention to eliminate any possibility of appeal. 4.56 In order to function effectively, such an appeals mechanism will need to impose costs on those appealing in order to avoid the lodging of frivolous claims. It i s sometimes argued that creating a second instance debilitates the first one, and that instead the exclusive focus should be on strengthening the Regulatory Commissions themselves. However, international experience suggests that this argument i s not really valid and that appeal mechanisms can provide an important counter-balance in a regulatory system, as long as two basic principles are met. First, there must be a cost to raising an appeal, as for example when the losing party must bear the full costs of the appeal. Second, the appeals body should provide confirmation of the original regulatory decision in a reasonable number of cases. Under these circumstances, the appeals 148 mechanism i s less likely to be abused and the authority of the regulator i s more likely to be strengthened as a result. 4.57 Notwithstanding the significance of the appeals issue, it remains more important and urgent to resolve the other issues currently affecting the performance of the regulatory institutions, already described above. The resolution of these more fundamental problems will no doubt help to reduce the regulatory controversies that currently highlight the absence of an effective appeals mechanism. Having said that, an effective appeals mechanism remains an essential component of the regulatory system that is needed to build confidence in the fairness of the regulatory system. Hence, given the inadequacies of thejudicial appeal route, such a mechanismwill need to be developed inthe medium term. This could either take the form of a supra-commission (as described above) or an arbitration mechanism compatible with the Colombian legal framework. Problems faced by the SSPD 4.58 Problems with the performance of the SSPD had consistently been identified as one of the main problems limiting the efficacy of the reform. While this i s partly attributable to the intractable scope of SSPD's competencies, with more than 60 functions defined incorresponding legislation, it also due inpart to the incapacity of the institution (whether for internal or external reasons) to discharge its legal responsibilities. Thus, the institution has tended to under-perform in the past, and has suffered from a series of scandals involving successive Superintendents.. 4.59 In order to reverse this trend, the current administration appointed a new Superintendent with the express mandate of turning around the institution. During the last two years, the SSPD has accordingly launched a wide array of initiatives aimed at addressing each of the main problem areas. For example, the SSPD has introduced a risk classification system to act as an early warning device and developed more intensive supervisory regimes for those companies identified as being at highest risk. In the energy sector, a wholesale energy market monitoring committee i s being established to watch out for anti-competitive practices, and special measures have been taken to resolve social conflicts relating to non-payment of electricity services. Significant progress has also been made in the development and operationalization of a single national information system based on electronic capture and exchange of data on an open source technology platform. While it i s too early to assess the final results of these changes, there i s clearly a serious effort underway to overhaul the institution. The progress achieved so far already highlights the potential for a more efficient performance, and helps to identify those areas where structural limitations may require more fundamental changes. Internal structure of the SSPD 4.60 The SSPD is affected by many of the same structural weaknesses that have been identified in the case of the Regulatory Commissions. However, there i s the 149 additional aggravation that SSPD's nationwide presence in the field, as well as its direct contacts with end-users, local governments and utilities, opens the door to political interference and corruption. The importance of this problem was brought to light during the liquidation of ELECTRANTA's and the subsequent Termo Rio scandal, which led to the resignation and detention of the former Superintendent, together with a number of his aides (Ayala and Millan, 2003). In an effort to improve the technical capabilities of the institution, the SSPD, under the current administration has initiated a process of staff selection based on merits and qualifications, that has resulted in a competitive selection of the Deputy Superintendents for the three sectors and a similar selection process for other new staff. 4.61 Moreover, some of the structural problems cited in the previous section acquire additional significance. This applies for instance to the selection of the Superintendent by the President without prior screening, and (in contrast to the selection of the Expert Commissioners) without even any specific skill profile to act as a frame of reference for such a decision. While nothing prevents the President from making an excellent choice, the lack of a structured process of selection leaves too much to chance. Furthermore, limited financial autonomy also imposes particular restrictions on the SSPD, given the nature of its responsibilities: For instance, the restrictions on per diems limits the capacity of the institution to do sufficient field supervision. On the other hand, the obligation of charging all providers a supervision rate, including the smallest ones, makes the cost of recovering these revenues higher than the value of those revenues themselves. 4.62 Therefore, many of the solutions proposed to the problems of the Regulatory Commissions remain relevant for the SSPD. In particular, all the proposals related to the profile, remuneration and selection procedures for the Expert Commissioners carry over directly to the case of the Superintendent. The same can be said for the proposals to safeguard the necessary budgetary stability to allow the financing of a core group of highly qualified staff that i s critical for the discharge of the SSPD's core functions. Supervisionof utilities 4.63 SSPD's core function of supervising the performance of utilities faces many difficulties. Coordination with other public entities presents a number of serious problems. On the one hand, given that SSPD supervises rules determined by a range of other public entities, it i s obliged to adopt the diverse and often inappropriate supervision methodologies that they require. On the other hand, SSPD supervision takes place alongside that of other public sector control entities, which act in an uncoordinated and overlapping manner. Furthermore, SSPD lacks the effective sanctions and human resources to become a tough and credible supervision entity. 4.64 Although SSPD's legally assigned functions are already very broad, in practice it seems to have exceeded even those competencies established by the law. For example, while SSPD i s responsible for organizing and supporting social control and consumer participation mechanisms, in practice it has gone further and became an active 150 promoter of social control (Millan and Ayala, 2003). Given the difficulties experienced in effectively discharging even its core functions, such additional activities serve to dilute efforts and weaken the focus of the institution. 4.65 Moreover, SSPD supervises a great diversity of monitoring indicators on behalf of a multiplicity of government agencies. While SSPD i s responsible for supervising the performance of regulated firms, it does not actually design the indicators that are used for performance monitoring. Instead, SSPD receives instructions to supervise a huge diversity of criteria, indicators and models from the Regulatory Commissions. This creates two kinds of difficulties. First, the separation of the design of the supervision system from the supervision itself often leads to a lack of realism in monitoring indicators, and prevents the lessons from supervision practice being internalized in the design of the monitoring frameworks. Second, the multiplicity of parallel demands placed by different institutions leads to duplications and inconsistencies of information that increase the supervision burden both on regulated firms and on the SSPD itself. Given that the Regulatory Commissions are the origin for the bulk of the relevant performance indicators, there should be closer coordination and consultation between the Commissions and the SSPD at the moment of defining such indicators. In addition, the Regulatory Commissions should be required to justify, on a cost-benefit basis, the incorporation of any new indicators requiring additional supervision efforts on the part of the SSPD. 4.66 SSPD supervision of public enterprises partially overlaps with a range of other uncoordinated public sector controls. The supervision regime for public versus private utilities differs in so far as the former are also subject to a number of additional public sector controls including internal and external auditing. Clearly, this additional burdencannot be completely avoided given the nature of public enterprises, but it places additional administrative burdens on public utilities, and creates an inconsistent and confusing body of performance data. This system of multiple controls on public enterprises could be considerably simplified, avoiding overlaps and improving coordination of the various controlling entities. Law 142/94 gave the Regulatory Commissions the explicit responsibility for developing a unified and consistent basis for performance monitoring that would simultaneously satisfy the needs of the different control entities. Unfortunately, this work has yet to be completed but should be taken-up with renewed urgency, and would best be undertaken in consultation with a working group consisting of all the public entities that currently place control requirements on the public utilities. One possible simplification would be to exempt public utilities from external auditing and adapting their internal control to general official rules. However, this approach would not work inall cases. 4.67 SSPD lacks meaningful sanctions, such as fines, and furthermore applies those it has somewhat unevenly across regulated entities. While SSPD i s responsible for supervision, the regulatory framework does not provide for very meaningful sanctions inthe event of performance failure. Those penalties that do exist are trivial inrelation to the corresponding defects. The schedule of fines needs to be urgently revised and increased so as to represent a more meaningful penalty. Some commentators have observed that public enterprises are less likely to be sanctioned for non-performance than 151 private ones (Millan and Ayala, 2003). Moreover, the ultimate sanction of enterprise intervention and liquidation i s fraught with perverse incentives and practical difficulties (see below). 4.68 SSPD lacks the necessary human and financial resources with which to confront the large number and complexity tasks it has been assigned. Part of the problem as in the case of the Commission i s financial, since the SSPD i s subject to the same budgetary and fiscal limitations, constraining the resources available to attract staff with the necessary skills. The possibility of outsourcing certain tasks has been contemplated from the outset, but the experience with external audits (AEGR) has not so far been successful. An assessment by the SSPD (Cartagena, 2004) of the AEGR concludes that such support has not been effective and needs to be improved. The main limitations stems from the scarcity o f specialized auditors and the narrow scope of their expertise, which does not cover technical aspects. SSPD sees the need to strengthen the technical knowledge of these auditors and i s seeking to develop a roaster of auditors and mechanisms to monitor their performance. Intervention of failing enterprises 4.69 The intervention, management, and liquidation of failing public enterprises are one of SSPD's most important and difficult roles. SSPD has the power to intervene, manage, and liquidate failing public enterprises, with a view to ensuring continuity of service for the affected population. At a conceptual level, this mechanism i s intended to replicate, within the public sector, the discipline that bankruptcy proceedings impose on private companies. However, in practice, this instrument has beenfraught with a number of complications. 4.70 First, the intervention mechanism has been geared to react to poor performance, rather than to prevent it. The absence of a functional early warning system has prevented the SSPD from taking remedial measures before the situation deteriorates to a point where intervention i s the only feasible solution. This i s contrary to what happens with the SIC or with the SuperBancaria, which have preventive instruments that allow them to act before the firms reach a state that could be detrimental to consumers or competitors. As a result, problems were often left to deteriorate until a sufficiently serious situation exists to warrant direct intervention, when they might have been much more easily resolved at an earlier stage. 4.71 Second, there are a number of conflicts of interest associated with the intervention process. Due to limitedresources and lack of political will to go ahead with the liquidation of public enterprises, intervention processes have in the past often become long and drawn out. Recent legislation has made matters worse by removing the two-year deadline for SSPD to complete the intervention and liquidation process (Law 812/03). Lengthy interventions tend to create conflicts of interest with the SSPD acting both as manager and supervisor of the ailing enterprise. In addition, conflicts sometimes arise between the SSPD's responsibility to ensure continuity of service and the need to liquidate insolvent enterprises in order to satisfy the claims of creditors. 152 4.72 Third,the interventionprocessand associatedresources create moralhazard problems for public utility managers. Managers of deficient utilities do not personally face any substantial penalties. Instead, the knowledge that the SSPD i s responsible to step in and takeover in any difficult situation dilutes managerial responsibility and accountability. Furthermore, the recent approval of earmarked budgetary resources to support the liquidation process even runs the risk of creating perverse incentives for mismanagement so that enterprises can secure access to these resources. All this serves to reinforce the perception that the central government i s on stand-by to bail out insolvent enterprises, and underlines the importance of creating real sanctions for under-performing public sector managers. 4.73 Under the present administration, the SSPD has tried to expedite the liquidation of a backlog of intervention cases whose financial and operating performance had been deteriorating for some time. In 2003, the SSPD reached an agreement with the various shareholders, creditors and worker representatives to re- structure EMCALI . The SSPD also concluded the intervention of Telesantamarta, a subsidiary of Telecom, and returned it to its shareholder. Other firms that were liquidated include Electrolima SA, Empresa Distrital de Telecomunicaciones, and the Electrificadoras of El AtlBntico, Cordoba, Guajira and Sucre. Presently, the SSPD is concluding the liquidation of several other public enterprises including the Electrificadoras del Cesar, Bolivar, MaganguC, Magdalena and Choc6. Box 4.1 below develops the example of Electrolima in greater detail. Box 4.1: SSPD experience with intervention of failing companies 4.74 Furthermore, the SSPD has taken several measures to address some of the conceptual problems identified with the intervention mechanism. The SSPD has been working on the development of an early warning system incorporating remedial measures that can be taken at a much earlier stage. Some of the actions include the proposal to improve the quality and performance of external audits, the decision to unify the tasks of inspection, supervision and control to achieve a more coherent understanding 153 of the problems and adopt more integral solutions. SSPD has also developed a risk rating system in order to focus attention on the most critical cases, and has signed Improvement Agreements with several public enterprises considered to be at risk of intervention.. A case in point i s the signature of such a contract with Colombia Telecomunicaciones aimed at improving customer services and achieving a prompt resolution of complaints, thereby reducing the volume of appeals raised to the SSPD. SSPD i s also working with the distribution companies in the Atlantic Coast to improve management performance through Improvement Agreements and to reduce the severe problem of non- payment through the creation of local micro enterprises that take charge of billing and revenue collection. Finally, in order to reduce the conflicts of interest associated with the intervention process, the SSPD has been delegating the intervention process to special agents, external to the institution that are paid directly by the companies. 4.75 While the above reforms are welcome and should be continued, the real long term objective should be to make the intervention mechanism redundant by addressing the underlying structural problems that have created the need for it. The fact that there continues to be a need for the SSPD to intervene ailing enterprises i s essentially a reflection of underlying structural problems that have yet to be solved within the sectors themselves. In particular, many public enterprises suffer from a lack of positive incentives for firms to be efficiently managed, while many of the smaller utilities suffer from diseconomies of scale. The ultimate solution to the problem of intervention, therefore, i s to resume with the agenda of structural reform. In the case of the national public utilities, the government has already announced its intention to resume the privatization of some of the electricity distribution utilities. In the case of sub-national public utilities, a possible avenue that i s already being explored (for example, in the case of the San Andres water utility) i s to promote the use of management contracts in utilities with performance problems before these become too severe. Anti-trust regulation 4.76 There are a number of different anti-trust authorities with jurisdiction over the utilities sector, none of them particularly strong or well-equipped. Thus, the overall anti-trust agency i s the Superintendence of Industry and Commerce (SIC), with jurisdiction across all economic sectors, and powers to investigate anti-competitive practices as well as control over proposed mergers and acquisitions. However, in the specific case of the utilities, Law 689/01 also gives the SSPD anti-trust faculties; although some argue that this stems from an explicit Constitutional mandate. Unfortunately, SSPD has no particular competence inthis complex technical field, and as noted above, i s already overwhelmed with more responsibilities than it can realistically fulfill. Furthermore, the legal framework does not require any explicit coordination of anti-trust activities undertaken by SSPD and SIC, leading to overlapping roles and potentially conflicting jurisdictions. Finally, in the specific case of telecommunications, which i s after all the utilities sector where competition i s most relevant, the line Ministry also has a responsibility for anti-trust issues based on Decree 2122/92. 154 4.77 In order to resolve the problems entailed by the current anti-trust framework, it would appear desirable to concentrate anti-trust authority in a strengthened SIC. Efforts could then focus primarily on strengthening the technical capacity of this specialized agency, in order to be able to perform its functions more effectively. In particular, SIC would need to build-up some specialized knowledge of the utilities sectors, either by hiring its own minimum core of specialized staff or by contracting out these services to experts. 4.78 The Regulatory Commissions and SSPD would continue to play an important role by initiating cases and providing technical support to the SIC. The concentration of anti-trust authority in the SIC does not mean that the regulatory and supervisory institutions do not still have a role to play. Within such a framework, both the SSPD and the Regulatory Commissions would still have an important role to play by being able to present cases to the SIC wherever they detect competition problems. These institutions could also play an advisory role to the SIC in particularly complex technical cases involving the utilities. In order for such an arrangement to work effectively, it would be important to develop a clear definition of the roles and competences of all the agencies involved, as well as an explicit and mandatory framework for inter-agency collaboration. 4.79 Some legal changes would be required to implement the suggested reforms. The benefits of the proposed changes would be significant, both interms of strengthening anti-trust regulation, and in terms of alleviating the burden of tasks faced by the SSPD thereby enabling it to focus on its core functions. Caballero (2004) reaches a similar conclusion, but raises doubts as to the legal feasibility of such a reform given the Constitutional character of the SSPD, proposing instead an intermediate option based on unifying the currently dispersed legal framework for anti-trust regulation, but retaining a division of labor across institutions for the purposes of implementation. The problem with this recommendation i s that it i s based on the notion of collaboration between SIC and SSPD, which i s precisely what i s failing to take place at present. On the other hand, the Constitutional character of the SSPD does not per se invalidate the proposal to transfer anti-trust functions from the SSPD to the SIC. Although created in the 1991 Constitution, the functions of the SSPD were not developed in detail until Law 142/94. Therefore, modifying those functions should be possible at the level of legislation. Furthermore, there i s already a precedent of some of the SSPD's anti-trust powers being transferred to other state entities, given that Decree 2122/92 assigned the anti-trust function for the telecommunications sector to the Ministry of Communications. 4.80 Nevertheless, the competition problems identified in Colombia's utility sectors cannot be solved by anti-trust regulation alone, but are likely to require some structural measures. While it i s necessary to think about improving the framework for anti-trust regulation along the lines described, this will not be enough to undo the problems posed by extensive vertical integration and horizontal concentration of market power, as well as the difficulties o f creating a level playing field between public and private sector players. This i s evidenced by the current concentration of market power in the potentially competitive segments of electricity generation (where the four largest players control 80 percent of the market), as well as mobile and long distance 155 telephony (where the two largest players control more than 70 percent of the market). In electricity in particular, these problems are aggravated by the lack of a consistent policy for vertical separation of markets, as many public owned utilities operate in various segments of the industry, where are many of their private competitors do not. Addressing these issues will require structural changes that go beyond the immediate remit of anti- trust institutions. 4.81 Finally, ex-post anti-trust regulation needs to be complemented by a strong pro-competitive stance among the regulatory and supervision entities. There i s very little use of ex-ante regulatory measures that could more effectively prevent the accumulation of market dominance. As a result, competition in the utilities sectors is not as intense as it could be given the current legal framework. An important area of improvement would be the adoption o f ex-ante measures that foster the development of competitive markets, prevent the accumulation of market dominance, and create a level playing field for public and private firms alike. An important first step in this direction has been the ongoing development by the SSPD of a market model for surveillance of the wholesale market, along the lines of what currently takes place in California. It i s relevant to note that this initiative i s not incompatible with the proposal to transfer anti-trust responsibilities from SSPD to SIC, but i s precisely an example of where SSPD can apply its technical expertise to the identification of potential anti-competitive behavior for subsequent referral to the anti-trust authorities. 4.82 Summary. The common regulatory framework established for the public utilities faces a numberof generic problems that affect all of the different sectors to varying degrees. 0 Significant conflicts of interest arise between the state's role as policy- maker and regulator and its continued major role as service provider. These are particularly acute in the case of national utilities, where line Ministers sit both on various of the Boards of the enterprises and on the Regulatory Commissions that oversee them. This makes it very difficult to create a level playing field between public and private investors. 0 The boundary between policy and regulation i s extremely blurred, due in part to the extreme discretion with which regulatory functions may be delegated or reverted. In addition, there are deficiencies in the performance of the Regulatory Commissions, as well as difficulties posed by the lack of a consistent line from policy makers. 0 The three Regulatory Commissions face a number of human resource problems. Inparticular, the criteria, remuneration and selection process for Expert Commissioners are too discretionary to deliver a consistently high caliber of candidates. Moreover, due to practical restrictions on their supposed financial autonomy, Commissions are unable to secure an adequate base of qualified staff. 0 The judicial channel constitutes in principle appeals mechanism for regulatory decisions. However, judicial processes are extremely slow due to the backlog of cases that affects the judicial system in as well as the 156 lack of specialized knowledge of the infrastructure sectors. Hence, in practice, there i s no meaningful recourse for regulatedentities. 0 Supervision and control of public utilities i s the responsibility of the SSPD, but there have many problems with the performance of this institution. Under the current administration, there i s a major effort to turn this institution around, and to become more efficient at supervising the public enterprises. However, the SSPD still faces serious challenges, It has scarce resources and an excessive array o f responsibilities. Major coordination problems arise with those agencies on behalf of which SSPD conducts its supervision, as well as other government entities controlling the activities of public utilities. Moreover, the sanctions available to the SSPD are too small to make a significant impact on managerial incentives. A particular problem is the mechanism for intervening in failing utilities, the lack of an early warning system, and the presence of significant moral hazards and conflicts of interest. 0 There are currently a multiplicity of agencies in charge of enforcing anti trust legislation: one for the utilities (SSPD), another for telecom sector and of another for the rest of the economy (SIC). This creates conflicts and ambiguities of jurisdiction. Moreover, both agencies are technically weak and focus on ex post rather than ex ante measures. SECTOR-SPECIFIC REGULATORY ISSUES FOR U T I L I T I E S Electricity 4.83 The electricity reforms helped to improve security of supply and reduce fiscal dependency, but still arouse considerable discontent both within and beyond the sector. The electricity sector went through a major reform process in the 1990s, entailing unbundling, private sector participation, and creation of a wholesale electricity market and regulation of natural monopoly elements. This has resulted in a number of important achievements. Among these, the shift in the hydro-thermal balance of the generation portfolio from 80:20 to 60:40, with the additional thermal capacity coming almost entirely from private sector investment, must be highlighted. It i s also noteworthy that the degree of fiscal dependence of the sector has been substantially reduced, though by no means eliminated, with significant benefits for the public finances. Nevertheless, as of today, there is considerable discontent with the sector both from the utilities themselves and from society at large. 4.84 The fact that private participation did not come through on the scale originally envisaged, has introduced significant complications into the operation of the model. Key concerns are the lack of competition and investment incentives in the generation segment, and the financial distress of a considerable number of the distribution utilities. In effect, many o f the assumptions on which the reform was premised failed to come through in practice. Inparticular, the privatization process in the 157 sector was not able to go as far as originally envisaged, due (at first) to political resistance at the local level, regulatory uncertainty, and (later) to loss of interest from international investors. As a result, public enterprises continue to play a very important role in all segments of the market, leading to asymmetries in the treatment of public and private operators, intrusion of political interests at the regional and local level, and potential conflicts of interest between the different roles played by the state. Lack of competition 4.85 Vertical integration of public utilities, reinforced by lack of accounting separation and presence of multi-utilities, opens the door to anti-competitive practices. There i s considerable vertical integration in the Colombian electricity industry (recall Chapter 2). Although the legal framework requires unbundling of activities, political negotiations during the design and implementation of the reforms allowed regional public utilities to participate in various segments of the industry, even while private operators are confined to a single one. This creates asymmetries that clearly favor public operators and introduce ample scope for anti-competitive practices. Moreover, regulation of vertically integrated operators i s made more difficult b y failure to apply norms on accounting separation between different segments of production, and the existence of major operators that are also multi-utilities, adding further complexity and opacity to the cost structure. All of this creates ample scope for price discrimination between regulated and unregulated activities, particularly with regard to retail activities. 4.86 Furthermore, concentration of market power ineach segment of the industry substantially curtails the extent of real competition. In the two competitive segments of the industry generation and retail, the three largest players control around 70 percent of the market. Given that particular national or municipal government interests control multiple utilities, the extent of concentration i s in fact larger than what at first appears by examining the structure of the market. Moreover, there are two local peculiarities that further exacerbate the concentration of market power. First, the fact that thermal generation plants have limited market participation outside of extreme meteorological events. Second, the fact that frequent terrorist attacks on transmission lines have the effect of further segmentingthe market geographically, creating opportunities for further abuse of market power in isolated regions. As in the case of vertical integration, there are legal dispositions to limit the market share of any participant in any segment to no more than 25 percent of the market. However, these are not strictly applied inpractice. 4.87 Part of the solution to the problem lies in a more effective application of existing anti-trust provisions, as well as the potential development of new ones. As noted above, the remedies to many of the competition problems identified above already exist in the current legal framework. In particular, the dispositions relating to accounting separation, and limits on market shares, are of immediate relevance. In addition to the need to strengthen the technical capacity of the competition authority described above, this will also no doubt entail some political resolve to take-on the interests vested in the existing market structure. Where transmission outages create temporary pockets of market power, administrative payments to generators could be used to restrict the scope 158 for opportunistic behavior. Another potential mechanism that i s currently being explored i s the creation of a market surveillance committee for the wholesale market in order to detect and prevent abuse of market power in the wholesale electricity market. This approach has been used successfully following the power supply crisis in California. Weak incentives to invest in new plant 4.88 Under current conditions, it appears unlikely that further private capital will be forthcoming to finance the next wave of thermal generation plants. As noted above, the shift towards a more balanced thermal-hydro generation portfolio was achieved largely as a result of private investment during the mid-1990s. In order to maintain this balanced mix as overall generation capacity expands during the next decade, some 700-800megawatts each of new thermal and hydro capacity will be needed during the next decade. The current presumption i s that new hydro capacity will be installed by public utilities, while private investors will continue to provide new investment in thermal capacity. However, this latter assumption i s highly questionable, given the high levels of dissatisfaction of private generators operating inColombia today, as well as the absence of an adequate mechanism to remunerate investment in thermal plant. 4.89 International investors in electricity generation are very dissatisfied with their experience in Colombia, citing frequent regulatory changes as a central issue. A recent World Bank survey of international investors in the power sector worldwide found that about 80 percent of investors interviewed in Colombia were very dissatisfied with the results of their investments and would exit the country if possible. Argentina was the only other Latin American country to register higher levels of discontent, while the other major Latin American countries registered dissatisfaction ratios in the 10-20 percent range. Private generators, inparticular, complain that CREG makes continual and sometimes fundamental changes to the rules governing the electricity wholesale market, with an average of eight rule changes per year. They are also concerned about inequitable treatment between public and private generators. Given this experience, it i s unlikely that private investors will be willing to finance new generating capacity unless there i s a substantial and sustained improvement inthe regulatory environment. 4.90 Moreover, a durable solution still needs to be found to the problem of financing insurance investment in thermal generation. Even in the absence of the problems cited above, the fact i s that Colombia lacks an adequate system o f remuneration for thermal plants. Presently, these plants essentially provide a meteorological insurance policy, and are only rarely used at anything close to full capacity. As a result, market signals are not adequate to ensure the necessary investment. When the construction o f thermal plants was first contracted on a large scale following the 1992/93 drought, the original idea was to remunerate investment costs through a `socialized' surcharge on all power sector users. In the event, this did not prove feasible to implement and Colombia has set ad hoc capacity charges that do not provide a sufficiently clear or stable signal to promote investment. A major study of the issue commissionedby CREG in 2001 resulted in recommendations that, according to industry players, would have been too complex 159 and costly to administer. Although simpler proposals have subsequently been developed by some market commentators, both CREG and the line Ministry appear to have shelved this critical issue. Financial distress of distribution utilities 4.91 Solving the problems of the distribution utilities is likely to require regional consolidation, an improved intervention mechanism and further private sector participation. The problem of dysfunctional distribution utilities continues to plague the electricity sector, with deleterious effects on the payment chain upstream, and ultimately on the public finances. There are three elements to any solution. First, it would be desirable to consolidate many of the smaller distribution enterprises into larger regional utilities so as to reap scale economies and achieve a more solid customer base. Indeed, the consolidation of a number of small utilities into a single utility in the eastern region of the country was, in fact, mandated by the original reform legislation (Law 142/94), but has yet to be achieved. Second, the resolution of the problem with distribution utilities to a significant extent depends on the resolution of the problems with the intervention mechanism used b y the SSPD, which i s the primary channel for overhauling these enterprises. This issue was already treated in some detail above. Third, private participation in some form would appear to offer the best option for turning around these utilities. While capitalization or privatization may be desirable, it may not be feasible to attract investment into such deficient enterprises in the short term. Thus, more modest contractual arrangements such as management contracts may provide a more relevant option. A particular challenge will be to ensure that the consolidation and capitalization of the distribution utilities does not come at the cost of further vertical integration in the industryas a whole. Natural gas 4.92 The natural gas sector was not so much reformed, as newly established in line with the paradigms of the 1990s. Contrary to the other sectors where legal and institutional reforms were applied on existing industries, the gas sector was not yet an established industry and the authorities took a conscious decision to establish a new market structure, design laws and economic regulations that would promote development. Precisely because of the novelty of the sector, and its relatively incipient scale, there are a number of strategic policy questions that are still being worked out, which complicates the task of the regulator. 4.93 The key strategic policy issue inthe sector at present remains the need to find other markets that would ensure the viability of the sector, and in doing so, an important question is the role of the state in promoting its further expansion. The sector has successfully expanded to provide the gas supply needed for a more balanced generation portfolio, and achieve high rates of penetration in the relevant domestic markets.However, its overall scale remains small, and there is significant excess capacity 160 ingas transportation. All of this has prompted much debate about the extent to which the government should be proactive in promoting other uses of natural gas, such as industry, transport, and export markets. Under current economic conditions, natural gas does not appear to be an attractive option for industry or transport, suggesting only limited potential demand growth in these sectors. Clearly, the main potential for its use lies inthe generation sector, but the likelihood of this market becoming a driver for the development of a gas industry will depend of the expansion of thermal electricity generation. It will also depend in a restructuring of the payments to the providers of gas, for currently, they face a mismatch between a flow of variable income from sales to the generators with the obligations to make fixed payments for gas transportation, which represent a significant share of their costs structures. In view of the uncertainty in the thermal generation market, it i s unlikely that the producers of gas will be able to negotiate long terms PPAs, as they had been doing in the past. A second potential market, that has been recently under much discussions , are the possible exports to Venezuela, and to Central America. This appears as another possible market to be developed, although much remains to be evaluated, before being able to assess the real perspective of such markets. 4.94 The gas sector is one of the arenas where the boundary between policy making and regulation has been most hotly contested. For example, in the year 2000, a Parliamentary initiative that was originally designed to prohibit gas exports was turned with the support of the government into a plan to promote gas exports and liberalize wellhead prices. Through Decrees 3428/03 and 3429/03 the MME took back from the CREG the power to define the price of gas at the wellhead, as well as authority to solve disputes over access to international gas transportation connections. Hence, the above recommendation of an ad hoc Commission to clarify the boundaries between policy makingand regulation is particularly pertinent inthe case of natural gas. Telecommunications 4.95 The telecommunications reform has brought major benefits in developing traditional services, but falls significantly short of its potential for promoting growth of more modern services, due to problems in creating effective competition. The reform of the telecommunications sector has led to important achievements in service expansion, modernization of the networks and introduction of new services. However, there have also been some important shortcomings, that have resulted in a much slower development of modem services such as internet and broadband services. Part of this lag can be explained by the difficulties for providers to bundle services and to take advantage of the rapid convergence of technologies, due to an artificial separation of services, based on the current segmented legal framework for telecommunications. The high level of market concentration despite the emphasis on market liberalization, and the low level of private participation in a sector that has been at the international forefront of private participation have also contributed to slowing down the development of the sector. These difficulties owe much to the creation of artificial barriers to competition. Moreover, the 161 fragmentary legal framework and complex institutional framework have not helped to provide a level playing field for competition. Barriers to competition 4.96 Liberalization of long distance and mobile services was designed in such a way as to incorporate significant entry barriers. The sequencing of reforms in Colombia's telecommunications sector i s unusual with immediate liberalization followed by some private participation, as opposed to immediate privatization followed by later liberalization. As a result, entry of private operators has been difficult due to the dominant position of public incumbents. In addition, the rules governing market entry imposed substantial costs on new entrants. In the long distance market, entry was confined to firms with an existing base of 150,000 fixed line subscribers in the country, a provision evidently favoring existing incumbents. Moreover, they were required to pay a license fee of US$l50 million. This payment i s very large in relation to subsequent market conditions. However, the existing mobile providers oppose the elimination of such a payment in future license awards in order to prevent new entrants from gaining an additional advantage. 4.97 In addition, there are problems associated with asymmetric treatment of public and private operators. There have been a significant number of bailout operations for ailing public telecommunications operators. O f these, the highest profile case was the recent liquidation of the national operator Telecommunications, resulting in the creation of a new public enterprise and the transfer of substantial liabilities to the state. On a smaller scale, a number of smaller local telephony operators have been intervened by the SSPD, and consequently benefited from the injection of public resources. Finally, as noted in Chapter 3, surcharges for contribution to the Communications Fund for rural and social telephony projects are highly differentiated across segments of the industry, with local telephony being completely exempted. 4.98 There is a need to revise the existing legislative framework with a view to eliminating these various kinds of distortions. There i s therefore a need to revisit the legal framework with a view to removing the aforementioned barriers to entry, and taking measures to level the playing field between public and private operators. This i s all the more urgent given the imminence of a FTA, which based on the Chilean experience i s likely to include the liberalization of telecommunications services on the negotiation agenda. This would mean that Colombian legislation would need to comply with the principle for free competition, exposing its existing operators to increased competitive pressures. Artificial legal dichotomy 4.99 Accelerating technological convergence between telecommunicationsservices has been restricted by an artificial regulatory segmentation of the market. The 162 regulatory framework establishes separate treatment of local fixed line versus other telecommunications services. However, technological convergence means that local telephony increasingly competes for customers with mobile and internet services. This competition between services i s artificially restricted by the regulatory segmentation, preventing firms from growing by offering users multi-service packages, limiting options for expanding rural coverage via mobile rather than fixed line telephony, and tending to slow down the expansion of more sophisticated modern services such as broadband which i s still rare despite the availability of fixed infrastructure inthe principal cities. The increasing divergence between the evolution of the sector and its legal and regulatory framework therefore calls for a questioningof the status quo. 4.100 Moreover, this false dichotomy makes for a fragmentation of the legal framework governing the sector. As a result, the telecommunications sector i s governed b y a wide array of laws and regulations, from different periods and based on different paradigms and criteria. Local telephony services are under the framework of Law 142/94, and as a result subject to detailed natural monopoly regulation, in the same way as the electricity, gas, and water sectors. This includes close supervision by the SSPD. The legal and regulatory framework for the remaining services originates in Law 1900/90, and i s further developed in successive legislation. 4.101 A unifying Telecommunications Law is needed, which could be based on advanced drafts already in circulation. Efforts should be focused on adopting a single unifying law that would establish consistent treatment across competing services. A draft Telecommunications Law has already been prepared by the MC, and has been in circulation since 2002. This legislative proposal includes many of the provisions needed to provide a consistent legal framework, and provides a solid basis for developing a final draft. Inparticular, it abolishes the false dichotomy between fixed line and other services, removing telephony from the scope of Law 142/94. However, the draft will need to be revisedinorder to ensure consistency with the provisions of any future FTA Complex institutionalframework 4.102 The dichotomy in the legal framework also contributes to the creation of an unnecessarily complex institutional structure for oversight of the sector. Key problems, already noted above, are the allocation of some anti-trust and regulatory functions to the MC, and the overlapping roles of the two Superintendencies, SSPD and SIC. Moreover, in such a market-oriented sector as telecommunications, the need for heavy-handed supervision and control of the kindundertaken by the SSPD i s increasingly questionable. 4.103 The sector would be better served by a simpler and leaner regulatory framework. In principle, this could be limited to three actors. The MC should broadly retain its current role of formulating sector policies, developing social programs, and promoting the development of an information society. However, it should no longer have anti-trust responsibilities. The CRT should remain responsible for the design and enforcement of economic and technical regulation, and for the supervision and control of 163 concessions. However, the SSPD would no longer play a role in the sector, as a result of the removal of telecommunications from the `public utilities' paradigm. Finally, SIC would be responsible for the defense of competition, and for the protection to end users, placing the central focus of the sector on anti-trust regulation rather than traditional supervision of performance. Water 4.104 In contrast to some of the other sectors, the reform process in the water sector was largely confined to regulatory measures. The sector had already been fully decentralized during the late 1990s, so that the structure of the sector itself was not under consideration during the reform process. Moreover, although the water sector was liberalized along side all the others, this in practice did not make any significant difference in a sector dominated by natural monopoly characteristics. Hence the key changes in the sector relate primarily to the tariff regulation model, and (potentially) to the introduction of tighter supervision of quality parameters, although this has not worked out in practice. Given the political sensitivity of water tariffs, the substantial progress achieved to date in raising tariffs towards cross-subsidy levels i s quite remarkable, even if the process i s still far from complete. Decentralization 4.105 The sheer number of municipal utilities in the sector makes central regulation and supervision almost intractable. CRA and SSPD are together responsible for regulating and supervising the activities o f more than 1,300 utilities, many of them very small-scale operations. The practice of the last 10 years has been to apply a common framework of intensive natural monopoly regulation and supervision to all of these entities, without making any attempt to differentiate treatment according to size. The result has largely been one of paralysis, with the large fringe of smaller utilities lacking the capacity or inclination to engage in regulatory processes, and the central agencies themselves lacking the manpower to interact meaningfully with such a large number of service providers. In addition, many of the smaller utilities are in such a precarious condition, that it i s questionable whether regulation i s even an appropriate intervention, as opposed to technical assistance or outright liquidation. 4.106 As a result, there is a need to adopt a differentiated and phased approach to utility regulation and supervision, based on size groupings of utilities. Given that more than half of the Colombian population i s in any case served by only 40 utilities, significant improvements could be achieved by concentrating scarce regulatory resources on the larger utilities. Two complementary strategies warrant greater attention. First, to design different regulatory regimes for different bands of utilities according to size and sophistication, with the larger ones being subject to traditional natural monopoly regulation, the medium-sized ones to a simplified form of natural monopoly regulation focused on the most critical parameters and issues. Inthe case of the smaller utilities, it i s 164 even questionable whether regulation (as opposed to technical assistance) i s the appropriate form of central engagement. As the efficacy of regulatory efforts with the largest utilities improves, it would be feasible to begin to deepen measures in the medium-sized and smaller utilities, usinga phased approach. Tariff regulation 4.107 According to the initial vision of CRA, the regulation of the water sector would evolve according to three distinct phases. In 1995, the CRA published a strategic plan detailing its vision of water regulation and associated priorities. The plan divides into three distinct phases. During the first phase, the central emphasis would be on restoring the financial viability of water utilities through sustained tariff increases. In the second phase, which would come into force during the second five-year tariff period, attention would switch to the application of regulatory tools designed to ensure the efficiency and quality of service provision. In a final phase, of uncertain timing, the objective would be to liberalize water service provision and increase the scope for competition wherever possible. 4.108 Significant progress has been made in reaching financial sustainability, but only in the larger utilities, and over a much longer period than originally anticipated. As envisaged in the plan, regulatory attention did initially focus on raising tariffs to cost recovery levels as prescribed by Law 142/94. The necessary tariff increases were both large and regressive entailing increases of 100 percent to 300 percent on the tariffs of the lowest socioeconomic strata. Consequently, a gradualist strategy was adopted, with the initial five year deadline to 2001 established by Law 286/96, and later extended by a further five years to 2005 by Law 632/00. Inpractice, the application of the tariff increases was uneven. The largest utilities have made substantial progress, now charging reference tariffs within about 10 percent of the cost recovery benchmark, and making substantial progress towards the rebalancing of the cross-subsidy system. However, in the smaller utilities, there has been greater reluctance to embrace the goal of financial sustainability. This i s due to politicization of the Boards of the municipal utilities, which are often presided by the local mayor, and which have the ultimate power to set tariffs even if obliged to do so within the regulatory framework laid down by the CRA. 4.109 Furthermore, there has been recent political back-tracking from the goal of reaching financial sustainability in the sector. Following the social discontent caused by the recession of the late 1990s, political back-tracking over tariffs at the local level has increasingly been echoed by measures taken at the national level. In recent years, CRA itself seems to have shifted its emphasis towards reducing water tariffs, engaging in a number of high profile interventions to try and reduce tariffs in some of the major cities, most notably Bogota. Furthermore, a number of new legal measures taken in 2003 to increase the subsidy limit for Stratum 1households from 50 percent to 70 percent o f the tariff, and subsequently outlaw any future real increases in the tariffs of Strata 1 and 2 households, have effectively undermined the process of rebalancing o f the cross-subsidy framework, makingthe goal of cost recovery ever more distant. 165 4.110 There are also concerns that the current methodology for calculating cost recovery tariffs may give results that are too high in some respects and too low in others. Alongside the process of tariff convergence, there have been methodological debates as to whether the current methodology provides a correct basis for calculating the cost recovery tariff. On the one hand, the tariff formula has been criticized for simply passing on current inefficient costs to customers, without making any attempt to adjust them to efficient levels. Indeed, the only efficiency adjustment incorporatedinthe current methodology i s the assumption of 30 percent unaccounted for water, irrespective of the true value of this parameter. Moreover, there are philosophical issues about the appropriateness of the practice of providing a market rate o f return on the entire historic asset base, much of which has been funded by public subsidy. Both of these concerns would tend to suggest that current reference tariffs are too high. On the other hand, there are concerns that the tariff formula excludes important and significant new cost components faced by utilities such as taxes on corporate profits and financial transactions among others, as well as new environmental charges on water abstractions and effluent discharges. This may offset to some extent the earlier concerns that tariffs might be too low, although the net effect i s ambiguous. Efficiency, quality and competition 4.111 CRA became distracted from efforts to promote efficiency and quality, as a result of a debate about the possibility of shifting towards a radical competitive model. The original 1995 plan envisaged that once the issue of financial sustainability had been addressed, CRA would turn its attention to issues of efficiency and quality, and eventually competition. In the event, the order of this agenda became reversed when in 1999 CRA begun to explore the possibility of applying a radical new regulatory model based on vertical unbundling and open competition between water producers with interconnection agreements across networks. This model (which i s based with only minor adaptations on the electricity network) has never been applied to the water sector anywhere in the world, entailed substantial risks and certainly did little to address the underlying problems of the Colombian water sector. The idea was eventually abandoned, however in the mean time its discussion and elaboration consumed a great deal of regulatory effort, and distracted attention from more pressing tasks, such as establishing efficiency and quality parameters. 4.1 12 Colombia's water sector is inany case, already among the most liberalized in the world. Colombia's legal framework i s unusual in applying full liberalization to the provision of water services, without any use of exclusive service areas that tend to be the norm in other countries. Hence, any operator i s free to enter and provide water services in any market. Given the strong natural monopoly characteristics of the sector, this right i s rarely exercised, and usually only in the case of peri-urban communities that happen to have easy access to a low cost high quality water resource. Competitive pressures are thus effectively limited to competition for the market in those cases where concessions have been awarded for water service provision (see Chapter 5). Nevertheless, even then, 166 very few concession processes in Colombia have succeeded in attracting more than one or two bidders. 4.113 As a result, progress with the development of a regulatory framework governing the efficiency and quality of service is still well behind schedule. Ten years after the passage of Law 142/94, Colombia's water sector has still not established a complete methodology for measuring efficiency and regulating quality of service. On the efficiency side, CRA has recently adopted new regulations for the calculation of the tariffs that include some deficiency parameters and make use of Data Envelopment Analysis (a mathematical programming technique that defines efficiency frontiers) in the tariff calculations. On the quality side, some draft regulations incorporating guaranteed minimum service standards were developed by the CRA, but have not yet been finalized or formally adopted. These are central regulatory instruments that urgently need to be refined and implemented in order to finalize the basic building blocks of the regulatory framework. Moreover, some have argued that tariffs should differentiate between utilities according to the quality of service provided. Role of SSPD 4.1 14 The supervisory role of SSPD in the sector has been greatly complicated by the absence of definition on key regulatory parameters, although some progress have been made with the adoptions of new regulations that include efficiency parameters. The supervisory work of SSPD i s based entirely on parameters and methodologies supplied by the Regulatory Commissions. The recent adoptions of new regulations by CRA that include some parameters of efficiency, constitute a positive step in providing the SSPD with better tools to supervise the efficiency of the operators. A similar effort needs to be achieved in terms of quality of service parameters. To allow a better supervision of the providers, it would also be necessary to advance in the definition of of the companies investment plans, needed to provide the basis for supervision. 4.115 Furthermore, the large number of dysfunctional municipal utilities makes the intervention mechanism difficult to apply. The large number of decentralized entities, their significant politization, and well known managerial deficiencies, mean that under a strict interpretation of the intervention criteria, a substantial proportion of the sector would be eligible for intervention. The impossibility of assuming such a burden, as well as the danger of creating perverse incentives for local mayors to transfer insolvent utilities to central management, makes it difficult for the SSPD to intervene in all the cases that in theory would have required such intervention. 4.1 16 Summary. The four utilities sectors present a wide range of specific challenges of their own. These range from the promotion of competition in electricity and telecommunications, to the elusive goal of cost recovery for the water sector. a The electricity sector i s characterized by extensive vertical integration and horizontal concentration of market power, particularly among public sector players. Moreover, anti-trust regulations on market share and 167 vertical accounting separation are not being adequately applied. Given a highly unpredictable regulatory environment, as well as the absence of a clear framework for capacity charges to remunerate the meteorological insurance provided by thermal plant, it will be difficult to attract the necessary private funds to finance new gas-fired generation. Ten years after the reform, a definitive solution remains to be found to the problem of poorly managed and financially precarious distribution utilities. 0 The relative novelty of the natural gas sector makes it an area where strategic policy questions are still being worked out. After having successfully achieved a significant penetration o f gas as a substitute for electricity in households, there i s still a need to develop other markets to ensure the viability of the sector. A central question i s the extent to which the state should be involved in further promoting the growth of the industry and specifically how should the cost o f gas transportation should be allocated among the actors. As a result, the sector has been particularly susceptible to conflicts between policy-making andregulatory bodies. 0 The central issue in the telecommunications sector i s the removal of a number of significant existing barriers to competition. This will require an overhaul of the legal framework, ideally resulting in the promulgation of a new sector law that brings all telecommunications services under a common approach, and streamlines the institutional framework for sector oversight. 0 The highdegree of atomization of the water sector complicates the process of regulation and supervision from the center. The main focus of regulatory efforts has been raising tariffs to cost recovery levels, resulting in major strides being made in the larger utilities, but much more modest progress in the smaller ones where local political resistance has remained strong. However, the regulatory tools for evaluating efficiency of operations and quality of service have not yet been fully developed, leaving a major gap that restricts the efficacy of the SSPD inthis sector. TRANSPORT SECTOR FRAMEWORK Legal framework 4.117 The legal framework for the transport sector was established in the early 1990s, but has recently been substantially revised. The first major piece of reform legislation in the Colombian transport sector i s Law 1/91, which ends the state monopoly inport management, and opens the way for private participation. Thereafter, Law 105/93 establishes the legal and institutional general framework for the sector, laying down the guiding principles, setting-up the regime for concessions and user charges, and identifying the Ministry of Transport (MT) as the head of the sector. Law 336/96 builds on this base, establishing the detailed regulatory framework for service provision. In 1999 168 and 2003 important new legislation was introduced, which led to substantial changes in the institutional framework for the sector (see below). Institutional framework 4.118 Most aspects of transport are under national jurisdiction, with a few significant exceptions. Under national jurisdiction are placed, major transportation infrastructure, including primary and some tertiary roads, major sea ports, airports, and waterways; as well as most transportation services, including freight rail, trucking, inter- urban passengers, and navigation, as well as commercial air passenger and freight services. The only areas under sub-national jurisdiction are secondary roads (departmental), as well as local airports, and urban and rural passenger services (municipal). 4.1 19 For roads, rail and ports there is a careful separation of functions, while for civil aviation these are all combined in a single entity. The legal framework assigns policy-making responsibility to the Ministry of Transport (MT), although navigation i s under the Ministry of Defense's (MD) General Maritime Directorate (DGM), and the Autonomous Regional Corporation of the Magdalena Valley (CorMag) participates in the determination of policies relating to that particular waterway. Regulation of all concessions for road, rail and sea ports are the responsibility of the National Institute for Concessions (INCO), while the Superintendence of Ports and Transport (SPT) supervises and controls adherence to regulations by 5,000 road, rail and port operators. For the civil aviation sector, a specialized agency, the Civil Aviation Administration Unit (UAEAC), combines policy-making, licensing, regulatory and supervision functions, as well as being responsible for providing air traffic control services. For services under sub- national jurisdiction, most of the relevant functions are undertaken by the transport department of the local authority, and occasionally by some autonomous public agencies. 4.120 The MT has clearly defined policy-making roles, but is still developing the necessary capacity to fulfill them. The MT is the central policy-making body for the sector. Its activities are divided between two directorates, one concerned with infrastructure and the other with transportation services. Their functions include sector planning, issuing of regulations governing particular markets and services, awarding of concessions, and granting of licenses. Although the MT's responsibilities are clearly established b y law, in practice it i s still in the process of developing the capacities that i t needs to fully discharge these responsibilities. 4.121 A particular problem is the lack of an integral vision of the entire transport sector inthe formulation of policy and regulation. The MT i s weak with regard to intra and inter-modal policy decisions. It also lacks an integral transportation planning that considers all modes andjurisdictions within a unified long term framework. Furthermore, the division of regulatory functions for different sub-sectors across different institutions also makes it more difficult to ensure coherence of regulatory decisions across sub- sectors. 169 4.122 Colombia lacks the legal and institutional framework to handle multi-modal transportation, which represents a growing demand from the private sector. Although the development of multi-modal transportation in Colombia i s constrained by security considerations, there i s growing interest from the private sector in moving towards a multi-modal approach. Nevertheless, at present, there i s no legal framework covering multi-modal issues, nor any institution responsible for taking a multi-modal perspective. These deficiencies will need to be addressed before the country will be able to reap the benefits of this more holistic approach to transportation planning. 4.123 A major restructuring of the institutional framework allowed INCO to emerge as regulator, and a revamped Superintendence to focus on supervision and control. In 2003, INCO was created as an autonomous agency reporting to the MT. INCO's role i s to structure, award and administer concession contracts for the road, rail, and port sub-sectors. The creation of INCO necessitated a substantial reorganization of the earlier institutional framework, entailing the liquidation of the public enterprise, Ferrovias, as well as the restructuring o f the roads agency INVIAS (to separate out its regulatory functions) and of the then Superintendence of Ports (to narrow its role to supervision and control and broaden its scope to include road and railtransport). 4.124 The transport sector presents a huge variety of sub-sectors and diverse institutional arrangements. This reflects the existence of numerous transportation modes, and the typical institutional separation between transportation infrastructure and service provision. Institutional arrangements range from direct monopoly state provision, to heavily regulated private participation, to completely deregulated competitive markets (Table 4.2). 4.125 Except for a few significant concessions, roads are provided directly by the state; while transportation services are private and competitive, but require state licensing. Regarding road transportation, road infrastructure i s largely the subject of direct government provision whether at the national (primary and tertiary networks) or departmental level (secondary networks). The 14 toll road concessions with private operators on the highest traffic corridors of the primary network are subject to regulation by INCO. Trucking services are private and highly atomized, but subject to a licensing regime under the Ministry of Transport. However, inpractice, licensed firms sub-contract most of their services to small-scale operators, many of them operating informally. Passenger services are also subject to a licensing process, either with the Ministry of Transport (inter-urban) or the local authority (urban and rural). However, once again in practice, these services are characterized by a high degree of informality, particularly at the local level, leading to problems of congestion, pollution and poor quality services. Turning to rail transportation, there are two private coal railways and two public freight railways, the latter being operated under concession and regulated by INCO. 4.126 The main ports operate under concession to the private sector, while inland waterways are under direct public provision. Regarding water transportation, the administration of the five major national ports was assigned to five regional port societies (SPR), which are under concession to the private sector and regulated by INCO. Moreover, numerous private firms provide a variety of services within each port. A 170 number of private ports also exist, dedicated to particular commodities (coal, bananas), and operated under licenses provided by the MT. The overall authority for navigation i s the DGM, under the Ministry of Defense (MD). International and inland navigation is deregulated. However, cabotage operators require licenses issued by the MT. The principal inland waterway on the Rio Magdalena i s operated by a regional development corporation, CorMag (Corporacih del Valle del M,agdalena), while the remaining waterways are operated directly by INVIAS. 4.127 A few major airports are under concession to the private sector, but the majority are publicly operated; UECAC provides air traffic control and regulatory oversight. Regarding air transportation, Colombia has 78 airports under national jurisdiction, of which three are under concession to the private sector. A further 400 or so municipal and private airports exist to serve smaller towns, and operate under local jurisdiction. There are 27 commercial passenger airlines operating under bilateral agreements. Airfreight, provided by 15 operators, i s fully deregulated. The regulatory agency in all cases i s UECAC, which also provides air traffic control services. Table 4.2: Overview of institutionalframework for transport sector Jurisdiction Regulatory Regulatory Supervision Service provision regime agency :ntity 1 Roads Toll roads National Concession INCO SPT Private (14) Primary roads National Direct provision - - INVIAS , Secondary roads Departmental Direct provision - - Departments Tertiary roads National Direct provision - - INVIAS Trucking National License MT SPT Private (1,250+) Inter-urban passenge National License MT SPT Private (650+) Urbanpassenger Municipal License Local - Private Rural passenger Municipal License Local - Private Railroadtransport Freight rail National Concession INCO Private (2+) Water transport Inland waterways River Magdalena Regional Direct provision - CorMag Other rivers National Direct provision - INVIAS Sea ports SPR National Concession INCO Private (5+) Private ports National License MT Private Navigation Intemational National Deregulated DGM-MD Private (89) t Cabotage National License MT Private Inland _._. National __ Deregulated-. DGM-MD - Private (2+) ~~ Air transport Airports Contracted out National Concession UAEAC Private (3) National National Direct provision UAEAC MT(75) Local Municipal Direct provision - Municipalities Air transportation Passenger National Bilateral treaties UAEAC Private (36+) Freight __. National Deregulated _. - __ UAEAC __ Private (25) Source: Based on Ospina 2004 RED1Transport Sector Report 171 Regulatory framework 4.128 Many of the transport infrastructure sectors are governed by concession contracts present a modern contract-based regulatory framework administered by an agency. Under the modern paradigm, the MT provides policy guidelines, operations are conducted under concession arrangements, and independent specialized agencies regulate and control them. Infrastructures subject to this regime are the port toll roads and railroads sector, all of them under the jurisdiction of INCO. a Ports. The SPRs, as well as all subsidiary operators of port services, are bound by technical norms established by the Ministry of Transport and supervised by the SPT. The SPR's face sanctions ranging from fines to intervention or revocation of concession, should they fail to meet their obligation as established in the concession contract. The regulatory framework provides for gradual transition towards free price determination. Up to 1997, the SPT determined price floors and ceilings within which the SPRs had some flexibility to determine their own tariffs. Thereafter, tariffs can be freely determined by the SPRs, subject to regulatory approval. Although the criteria governing such approval are far from clear. While the concession contracts have provided a clear framework for the operation of the SPRs, the increasing demand for port services in the context of the forthcoming FTA i s placing new demands on port infrastructure. There i s an ongoing debate as to whether these demands are best met through the current model of intra-terminal competition between service providers, or alternatively by integrating all service provision within a given port and relying instead on inter-terminal competition. a Toll roads. Toll rates on major road concessions were established through the bidding process, and have not been subject to subsequent adjustment. However, INCO has significant power to affect the economic equilibrium of the concessions. Inthe case of first generation concessions, this i s basically through granting authorization for payment of traffic guarantees. In the case of latter generation concessions, INCO has the power to adjust the duration of the concession to compensate for fluctuations in traffic volumes. Road quality parameters are established in the concession contracts and monitored by INVIAS. a Railroads. Railroad concessionaires are free to set their own tariffs. In practice, the monopsony power of the coal industry, more than compensates for the monopoly power of the railroads. However, concession contracts do establish investment targets linked to the expansion and improvement o f services. These are supervised b y the SPT. 4.129 Many of the transportation services are subject to more traditional direct regulation by the Ministry (or local equivalent), although in practice they operate informally. Under the traditional paradigm, the MT determines policies and regulations, licenses service provision, and regulates and controls their tariffs and performance. The 172 services under this paradigm include road freight, as well as most passenger transportation, with the exception of the new Transmilenio scheme described below. a Road freight. The MT is the main regulatory interlocutor for the trucking industry, and i s responsible for setting both technical norms (supervised by the SPT) and regulating tariffs, by means of a mandatory matrix covering different routes and vehicles. The regulation of tariffs in this industry i s impractical, given the enormous heterogeneity of services provided, and unnecessary given the significant degree of competition that exists in the sector (with more than 1,000 licensed operators). In the mediumterm, it would make sense to fully liberalize the entry and pricing regime. However, this will need to be accompanied by measures to promote a more modem and efficient structure for the industry, entailing mergers of small-scale providers (possibly under cooperative arrangements), so as to be able to reap economies of scale and provide more efficient and higher quality services. a Passenger transport. Inter-urban bus services are subject to regulation by the MT, which determines fares, routes and schedules. A similar procedure pertains at the local level for urban bus services. However, in practice, arrangements depart substantially from this regulatory framework. Services are characterized by open competition among a host of informal operators, running along disorganized routes and causing serious negative externalities. 4.130 Transmilenio represents an important attempt to implant a modern regulatoryparadigmto urban passenger transportation.The Transmilenio Bus Rapid Transit (BRT) system in Bogota represents a path-breaking attempt to alter the regulatory framework for urban bus services. The system i s characterized by a trunk and feeder system that optimizes passenger movements and avoids chaotic proliferation of routes. Dedicated bus routes are provided by the city government, and bus services are provided under contract by the private sector. A government management agency oversees the operation of the whole system. A complex formula i s used to allow tariffs to adjust, almost continuously, to changing circumstances. The formula i s based on an index of passenger kilometers that prevents the operator from appropriating economies of density as passenger flows increase, but retains strong incentives for private participation. The government management agency discharges a number of regulatory responsibilities, including the authorization of tariff adjustment in line with the prescribed formula, as well as adjustment to the frequency of bus schedules. 173 Table 4.3: Overview of regulatory framework for transport sector Tariff regulation Technicalregulation Modern paradigm SPR Initially within regulatory bands, Based on concession contract now free subject to regulatory approval Toll roads Established in contract, but regulator disburses Based on concession contract traffic guarantees (1" generation), or adjusts contract duration (2ndgeneration) Freight rail Free determination Based on concession contract TransmilenioBRI Frequently adjusted based on complex formula Based on concession contract reflectingcost structure Traditional paradigm TNCking Mandatory tariff matrix reflecting different Technical norms established by MT origin-destinations and vehicle types. Inter-urban buses M T sets fares MT sets routes, frequencies andtechnical norms (but, in practice, competitivelydetermined) Urban buses Local authority sets fares. Local authorities set routes, frequencies and (but, inpractice, competitively determined) technical norms Source: Based on Ospina 2004 RED1Transport Sector Report Box 4.2: Colombian port reforms 174 4.13 1 Summary. The development of the legal, regulatory and institutional framework in the transport sector has not been uniform, and there is a marked contrast between sectors (such as ports, roads and railroads) where significant modernization has been achieved under concession arrangements, and other sectors (such as road transportation) that continue to function along traditional lines. Legislating and regulating transport i s rendered particularly complex by the huge diversity of market structures and institutional arrangements across the different modes. However, important advances have been made during the last decade. Inparticular, sectors such as ports, toll roads, and freight railroads have moved towards a modern regulatory paradigm based on concession contracts administered by regulatory entities. Results in the ports sector, in particular, have been particularly impressive, but will be increasingly challenged by the need for new investments to keep pace with the demands of the forthcoming FTA. There i s clearly scope for extending some of these lessons to transportation services, such as trucking and buses, which are still governed by a traditional regulatory framework and characterized by extreme informality. The Transmilenio experience provides an interesting model for the bus sector. The trucking industry would benefit from restructuring followed by greater liberalization of its regulatory regime. The overall diagnostic and recommendations for each of the transport sub-sectors are summarized in Table 4.4 below. 175 Table 4.4: Summaryof diagnostic and recommendationsby sub-sector Main Problem ! International Trends Recommendations for Policy and Regulation - Lack of modal integration in Sector-wide policy - Integratedvision, strategic policy and - Low capacity for policy and planning Build capacity in Ministry of Transport planning planning - Planning tools, qualified - Short term, reactive approach teams Trucking - Deregulation of markets Regulate for more efficiency industry - Low efficiency - Reductionof negative externalities Socially responsiblemodernization Inter-city - Regulationby corridor, or .Retainthe samebasic model buses - No major problems deregulation Strengthen capacity to do it better - Needfor network - Sustainable maintenance .Ensureresourcesfor maintenance Roads modemization,rehabilitation -andexpansion - Concessionof high density .Implement the concessionprogram Ensure adequate maintenance links already designed - Consolidate port hubs to Maritime - Eventual shortageof Pacific attract shipping lines transport liners - Market with low economic .Maintainderegulated regulation Inland - Investmentto improve - Complete infrastructure projects navigation - Reducedactivity navigability - Operate deregulated navigation market, attracting efficient operators Air - Difficulties in consolidating - Very difficult market - Regionally integrated - Progress with regional integration Itransport airlines cabotage - Capacity constraints Airports - Weak service for passengers - Multiple modelsof private sector participation - Proceedwith further concessions and freight - - Restricted to coal exports - Take advantage ofrailroad - Accelerate a widening in scope of economies of scale railroadusage ~~- Weak service ~ - Structured,formal, - Continue the formalizationtrend - Negative externalities integratedsystems - Focus social subsidies - Vertically integrated - Review regulatory framework Ports - Needto adjust to a terminals demandingenvironment - Public private partnerships emphasizingefficiency, and exploiting inter-terminal competition ~ ~ _ _ with local authorities- - ~ Multi- - Wide usage of multi- - Improve multi-modal regulations modalism - Low diffusion of modem modalism - Promote special logistics zones and multi-modaland logistic - Increasingly efficient suppl - Promotethe development of logistics logistics practices chain management . ~~~.___~_ qerators and SME logistics Source: C in elaboration. CONCLUSIONSAND RECOMMENDATIONS 4.132 Many of the most important challenges in the Colombian infrastructure sectors stem from underlying regulatory problems. These have a major impact on incentives to invest, and hence the possibility of financing an increasing share of infrastructure needs from the private sector. They also have important fiscal 176 ramifications, since the absence of adequate regulatory control often leads to bail out situations. The key recommendations of the chapter can be grouped into cross-cutting issues that affect all of the utilities sectors, and sector-specific recommendations for electricity, telecommunications, water and transport. Cross-cuttingrecommendations 4.133 The resolution of the following cross-cutting problems would have major benefits across all of the utility sectors. The recommendations are presented in perceived order of importance, in terms of their potential contribution to resolving the problems that currently affect these sectors. (a) Measures should be taken to reduce conflicts of interest faced by the national government. The state, particularly at the national level, faces major conflicts of interest between its roles as policy-maker and regulator, and its role as a major service provider. The ultimate solution to this problem will be for the state to divest from some o f its interests in the electricity and telecommunications sector, and indeed plans for doing so are already underway. Another measure that i s worth considering i s the separation of responsibilitiesfor sector policy-making and for representing state interests in sectoral public enterprises, which currently converge in the line Ministry. The exact mechanism for achieving this remains to be defined, however it i s likely to involve the designation of a separate entity to represent state interests in public enterprises that does not report directly to the line Minister, but rather instead, for example, to the Ministry of Finance. Such an approach has been used for some time in Chile and Peru. (b) A process for clarifying boundaries between policy-makers and regulators needs to be initiated, paving the way for a formal institutional separation of these functions. Until a common working understanding of these boundaries i s developed, it becomes very difficult for regulators to effectively discharge their responsibilities. A good starting point would be to create and ad hoc Commission representing both policy-making and regulatory interests, in order to develop principles and guidelines based on ex-post analysis of historic cases. However, the ultimate resolution of this problem will require a formal institutional separation between policy-making and regulation. The prior exercise of clarifying the roles and responsibilities of policy-makers and regulators will greatly facilitate this process, since it will ensure that policy-makers have the powers needed to formulate policy without needing to form part of the Regulatory Commissions. (c) Human resource policy for Superintendents and Expert Commissioners must be improved to ensure that high quality candidates are consistently selected. This will entail a revision of 177 procedures for profiling, selecting and remunerating these posts. Profiles should be broadened to include other disciplines, and to avoid ruling out those with recent experience in the sectors. Pay should be increased to levels commensurate with the Board o f the Central Bank, and the possibility of an exit bonus should be considered to compensate for prohibition on subsequent employment. Finally, a formal selection process i s needed whereby a short list of qualified candidates i s identified for Presidential consideration. (d) The SSPD needs to be given more meaningful sanctions, and a stronger basis for coordination with other public entities. Fines and current sanctions at the disposal of the SSPD are simply not strong enough to alter managerial behavior in the utilities. More effective tools are needed to discipline firm behavior. There i s also the need to coordinate and streamline the controls applied by the SSPD and other parallel public control bodies, as well as to require upstream regulatory agencies to consult more closely with the SSPD on the development of monitoring indicators. (e) Measures should be taken to reduce the need to resort to enterprise intervention. The SSPD i s already making progress towards the development of an early warning system that would allow problematic enterprises to be identified and remedial measures to be taken long before the need for direct intervention. These initiatives are important and should be developed further. However, it i s important to recognize that in many cases the intervention mechanism i s simply a palliative for underlying structural problems of weak incentives and limited scale that can only ultimately be resolved through structural measures. In this sense, it is essential to resume the structural reform agenda. The government has already announced its intention to do this through privatization of national public utilities, while greater use of management contracts may represent a way forward at the local level. (f) The possibility of concentrating anti-trust authority and expertise in a single national institution should be explored. The existence of two separate anti-trust agencies (SIC and SSPD), neither of which are adequately equipped to perform their functions, seriously debilitates the credibility of the anti-trust framework. It would be preferable to concentrate these functions within the SIC, and focus on improving the capacity of that institution, with a particular emphasis on the development o f ex-ante anti-trust interventions. Under such a framework, the SSPD would retain the important responsibility of being able to initiate anti-trust cases in the utilities sectors. (g) A workable appeals mechanism for regulatory decisions needs to be developed outside the judicial system, possibly in the form of a supra- 178 Commission or some kind of arbitration mechanism. The current judicial appeal system has proved intractable, while many of the alternatives identified are incompatible with the Colombian legal system. A possible solution would be to create a supra-Commission to sit over the three Regulatory Commissions and act as an appeals body for all of them. However, to avoid creating another layer of bureaucracy the supra- Commission should be limited to a small ad hoc body convened on an occasional basis to address specific appeals. Another possibility i s to devise an arbitration mechanism. However, it i s not clear whether this approach would be feasible with the broader legal framework in Colombia. Sector-specific recommendations 4.134 The widely varying situations and problems faced by each of the utilities and transport sectors leads to a number of recommendations that focus on the specificities of each. (h) Strengthen application of anti-trust regulation in electricity and telecommunications. In spite of wholesale liberalization of the electricity and telecommunications sectors, both remain characterized by vertical integration, market concentration, anti-competitive practices, and barriers to entry. Some of these problems stem from an overly timid application of existing anti-trust regulation, which can partially be resolved b y strengthening anti-trust capabilities as described above. In other cases, new legal and regulatory tools may be required to address potential anti- competitive practices. An important but difficult issue that needs to be addressed, i s the absence of a level playing field for competition between public and private operators. (i) Improve climate for private investment in thermal generation, in particular by developing capacity charges. The unpredictability of the current regulatory framework for electricity generation, combined with the absence of stable capacity payments to remunerate investments in capacity for meteorological insurance, making it increasingly difficult for private investors to finance the next generation of thermal generation plants. Both problems need to be urgently addressed. On the issue of capacity charges, a number of interesting proposals already exist, and the problem i s primarily one o f buildingconsensus. Providing greater regulatory stability i s a greater challenge that will rest on the adoption of many o f the generic recommendations described above. (i) Take definitive measures to resolve the chronic financial problems of many of the smaller distribution utilities. The precarious financial situation of many distribution utilities can no longer be allowed to 179 undermine the stability of the sector, with the resulting fiscal consequences. In order to tackle this problem a combination of regional consolidation and private sector participation will be required, even though (in the worst cases) the latter may at first only be limited to management contracts. Improvement of the SSPD intervention mechanism, along the lines described above, i s also an important element of the solution. (k) Develop a unifying legal framework for the telecommunications sector, eliminate artificial barriers to entry, and simplify its institutional structure. The telecommunications sector needs to modernize its legal framework in order to move away from the current artificial dichotomy between services that have already converged from a technological standpoint. This will entail removing local telephony services from the `public utility regime'. The existing draft Telecommunications Law provides a good starting point that will require further consensus building activities. The new law should aim to simplify the institutional framework for the sector, concentrating all regulatory functions in the CRT, doing away with traditional supervision, and strengthening the role of the SIC as anti-trust agency for the sector. (1) Regulation and supervision of the water sector should adopt a differentiated approach for different segments of the market. In order to avoid CRA and SSPD becoming paralyzed by the sheer number of small water utilities, the sector should be classified into three or four sub- groups of utilities with similar characteristics, and differentiated regulatory regimes of varying depth and complexity developed for each. This will enable regulatory efforts to concentrate where they are likely to have greatest impact. (m) The regulatory framework for quality of service parameters in the water sector should be completed. The current gaps in the regulatory framework for water relating to quality of service parameters, constitute a serious vacuum for the work of the SSPD, and also raise doubts about the definition of cost recovery tariff levels. For both of these reasons, there is an urgent need to complete these critical elements, building upon recent progress by the CRA in incorporating efficiency analysis into the new tariff-setting methodology. (n) Strengthen the capacity for integrated strategic planning and regulation in the transport sector. The transport sector functions most effectively when users are able to select and combine different transportation modes according to the optimal cost-quality balance. In many countries, the tendency to formulate policy and regulation at a narrow sub-sectoral level often overlooks the important complementarities between modes, and introduces distortions that prevent smooth multi- modal usage of the transport system from taking place. Colombia i s no 180 exception to this pattern, and it i s therefore recommended that the policy- making and regulatory institutions for the transport sector begin to adopt this broader integrated vision of the transport sector, and take measures to ensure that there i s a level playing field between competing modes of transport. (0) Adjust the legal and regulatory framework for ports to adapt it to increasing demands and to the challenges of the forthcoming FTA with the US. This will require a review of the regulatory framework to allow for intra-terminal integration of services and promoting inter- terminal competition. It will also require the reconciliation of the evolution of the regulatory framework with the acquired rights and obligations of the existing operators. (p) Reform the trucking industry to improve efficiency and pave the way for eventual liberalization. Given the tremendous heterogeneity of trucking services and the considerable degree of competition, it may be possible to gradually reduce the role of government in the determination of tariff matrices. Nevertheless, further liberalization of the industry needs to be preceded by a restructuring process that provides incentives for sector consolidation into more efficient and formalized business units, without which high quality modern services are unlikely to emerge. Any such restructuring process needs to pay careful attention to the social issues raised by the large number of one-man owner-operators of trucking services. 181 5. PUBLICAND PRIVATEFINANCINGFOR INFRASTRUCTURE INTRODUCTION 5.1 Inmeetingthe substantial investmentneeds identifiedabove, bothpublic and private financing of infrastructure will play an important and complementary role. Chapters 2 and 3 identified major investment needs to improve the competitiveness of the Colombian infrastructure sectors, as well as to meet key social obligations. Colombia already has a strong tradition of both public and private financing schemes of infrastructure projects. The purpose of this chapter i s to examine to what extent public and private financing flows can keep pace with the investments required to meet these economic and social objectives, as well as the roles most naturally assigned to each. 5.2 The first part of this chapter analyzes trends in public financing, evaluating the adequacy, efficiency and equity of past investment flows to each of the sectors. A sector by sector review of public investment in infrastructure reveals a significant decline in funding since the onset of the economic recession in the late 1990s, and the associated period of fiscal austerity. The transport sector has been most severely affected with a 50 percent decline in real expenditure on the national road network. The section also undertakes a methodological review of the budgetary mechanisms used to channel public resources to the infrastructure sectors. O f particular note i s the growing use of `future appropriations' as a means of securing stable funding for long term projects, which currently account for 40 percent o f the available fiscal space for infrastructure. Following a number of costly guarantee schemes attached to private participation in the early 1990s, Colombia has made important strides in fiscal accounting for contingent liabilities. Finally, the scope for increasing fiscal space by excluding a number of `well- run' public enterprises from the fiscal accounts is explored. 5.3 The second part of the chapter deals with trends in private financing of infrastructure, and possible mechanisms for regenerating these flows. Private investment has also declined substantially since the late 1990s, reflecting Colombia's loss of investment grade rating, as well as investor concerns about regulatory and judicial uncertainty, particularly in energy. The telecom sector remains dominated by public enterprises. The government has announced its interest in regenerating private investment, by publishing a list of infrastructure projects for private financing, and announcing a zero tolerance policy towards failing public enterprises. In order to realize these ambitions, it will be important to develop more finely targeted credit enhancement 182 mechanisms to lure international investors, as well as mobilizing the burgeoning long- term domestic resources of the pension industryfor infrastructure investments. 5.4 The third part of the chapter brings together investment needs and funding sources to assess the overall financial balance of each of the infrastructure sectors. Inorder to assess the balance between investment needs and available public and private finance, aggregated sector financial balances are presented that provide an overview of tariff revenues and fiscal transfers in each sector. The financial balances show a surplus of resources in the telecom sector, manageable deficits in the energy and water sectors, and an important deficit in the transport sector. However, the overall sectoral balances conceal substantial deficits within specific segments of the market, in particular some of the more precarious generation and distribution utilities, as well as the water companies serving the larger municipalities. PUBLIC FINANCING Review of the adequacy, efficiency and equity of public spending 5.5 Public investment fell substantially during the late 1990s, and shifted from the central to the local level, with public enterprises remaining the dominant source. Total investment averaged 17 percent of GDP during the 1990s, with 10 percent coming from the private sector and 7 percent from the public sector. However, investment levels have declined substantially since their peak of 23 percent in 1994, reaching a trough of 12 percent in the year 2000 (Figure 5.1(a)). Throughout this period, the split between private and public sector investment has maintained a relatively constant ratio of 60:40 (Figure 5.1(b)). Public sector investment i s undertaken either directly b y central or local government, or indirectly through state-owned enterprises. Indeed, just under half of public sector investment comes through national (ISA, ISAGEN, ECOPTEROL, ECOGAS and TELECOM) or municipal (such as EEB, EPM, ETB, EAAB and EMCALI) enterprises, representing about 3 percent of GDP. While central government investment slightly exceeded local government investment in 1990, this balance has shifted throughout the last decade, so that as of 2001 local governments were investing almost twice as much as the central government. Figure 5.1: Total investment (a) As percentage GDP (b) By source Source: Fainboim, 2004 RED1Fiscal Report 183 5.6 Investment in infrastructure, predominantly from the private sector, has historically averaged 3 percent of GDP, declining inrecent years. Total investment in infrastructure averaged almost 3 percent of GDP during the 1990s. However, investment levels have declined substantially since their peak of 4.4 percent in 1997, reaching only of 2.6 percent in the year 2003 (Figure 5.2).Throughout the last decade, the private sector financed about one third of infrastructure investments, the largest share being in the communications sector. Public investment in infrastructure has represented on average 19 percent of total public investment throughout the last decade, reaching a peak of 30 percent in 1997 but falling rapidly back to 20 percent by the year 2000. About 80 percent of public investment in infrastructure i s undertaken by the national and municipal state- owned enterprises. Figure 5.2: Breakdown of infrastructure investment (a) By sector (b) By financing source Source: DNP Notes: Transport includes roads, railways, ports and airports, while energy includes both electricity and gas, excludes mining and oil. 5.7 The transport sector has been most severely affected by the decline in public infrastructure investment. The simultaneous drop in both public and private infrastructure can be explained by two distinct, but coinciding, trends. Private investment has fallen as a result of declining investor confidence prompted by the loss of investment grade rating. Public investment has fallen as a result of the fiscal crisis derived from the severe economic recession of the late 1990s. As most of the investment in transport i s undertaken directly by the Central Government (as opposed to public enterprises) this sector has been particularly vulnerable to fiscal austerity measures. Thus, total investment in transport dropped from a peak of slightly above 1.0 of GDP in 1997 to below 0.5 percent of GDP in the year 2000 and still remains below 1 percent as of 2003. The energy sector has also experienced a substantial decline i s energy, with investment falling from 2.3 percent of GDP in 1996 to only 0.5 percent in 2003. 5.8 It is noteworthy that the different infrastructure sectors differ substantially as to their financing structure. Sectors such as telecommunications, natural gas, and to a large extent electricity, are able to generation almost 100% of their investment financing needs internally from user charges, whereas the water and transport sectors are 184 almost entirely reliant on fiscal transfers. The relative importance o f private finance for investment also varies significantly from close to zero in the water sector to around one third for the energy and transport sectors, to just over one half in the telecommunications sector. Nevertheless, it i s important to note that other countries in Latin America secure private finance for more than 80% of their needs inthe telecommunication sector. 5.9 The efficiency of publicinvestment ininfrastructure is largely determined by the efficiency of the public enterprises through which it is predominantly channeled. While a handful of public enterprises have succeeded in establishing operational autonomy and good governance procedures, many of them still suffer from less than optimal management procedures and a precarious financial situation. Accordingly, the government has reiterated plans to transform, merge or dismantle all together many of the public enterprises that show poor governance and results (see Box 2.1 on efficiency measures for electricity distribution companies). Figure 5.3: Composition of investment financing by sector - Interndl cash V,"I Wder ' Transport 'Energy ' Telemm'I nPnwAe .PUHlC (a) Internal versus external (b)Publicversus private Source :Own calculations based on data from DNP 5.10 Public investment is equitable in socioeconomic terms, benefiting lower income groups, but is still subject so significant geographic biases. Over the last decade, Colombia has achieved an important expansion in its infrastructure base. Electricity generation capacity increased by almost 50 percent, the number of fixed telephone lines tripled, and household connections to natural gas grew b y a factor of ten. Indeed, only the road network did not increase over the last decade. In evaluating the distribution of these new infrastructure assets, it i s relevant to consider both equity across the income distribution, and equity across geographical areas. There is ample evidence that new connections to infrastructure services were largely concentrated among lower income groups. However, notwithstanding some attempts to redirect public resources to disadvantaged areas, significant geographic disparities exist both across regions, and between urban and rural areas. 5.11 Most multilateral investment is currently directed toward the transport sector. Multilateral institutions (Andean Development Corporation -CAF-, 185 Interamerican Development Bank -1DB- and the World Bank) have a tradition of participating in infrastructure financing in Colombia. Up until the late 1980s, the largest share was directed towards financing of large electricity projects (mostly hydroelectric power). During the last five years, there has been a major shift towards financing transportation projects focused on roads, together with some institutional strengthening in the water sector. The next group o f projects in the pipeline will focus exclusively on the development of urban mass transit systems in large and intermediate cities. In addition to these investment projects, the multilateral institutions have been providing technical assistance for the structuring of projects with private participation and the World Bank has provided a limited guarantee program for toll road concessions. Interms of multilateral financing, as well as more generally, Colombia will need to pay special attention to the cost of funds. In an environment of potentially increasing interest rates, the cost of borrowing will become an important element in reducing the fiscal impact of infrastructure financing. Lender Borrower Sector Value Active 2001 CAF Central government Roads 200 2001 CAF Bogota Roads 100 2001 IBRD Central govemment Water 40 2003 IDB Cundinamarca Roads 22 2003 IBRD Bogota Urban Services 100 In process 2004 CAF Central government Urban transport- 150 Bogota 2004 IBRD Central government Urban transport- 250 Several cities 2005 IDB Central government Urban transport- 200 Cali 5.12 Taxes levied on infrastructure services are high in international comparison but no discrimination between public and private enterprises exists. Utilities have traditionally been tax exempt. However, these exceptions have gradually been abolished inrecent years, leading to a growing tax burden on the sectors. These taxes are primarily value added tax at 16 percent and corporate profit tax at around 35 percent. Taxes and other contributions vary across sectors, but have been estimated at around 20 percent of operational income for the water and electricity sectors. This i s relatively high compared with most OECD (Organization for Economic Co-operation and Development) countries where water and energy taxes typically lie in the 5-15 percent. However, it i s substantially lower than in Argentina where taxes represent around 30 percent of operational income. It i s important to note that tax policy does not discriminate between public and private utilities. However, there i s some distortionary discrimination within sectors, for example with mobile telephony facing higher Value Added Tax (VAT) than other services in the telecommunications sector. It is important to note that there i s a recent provision to exempt public utilities from VAT once again. 186 Electricity 5.13 M a n y distribution utilities suffer from poor management and precarious finances, jeopardizing the efficiency of their investments. The major public enterprises responsible for electricity generation and transmission are generally well managed and present a reasonable operating surplus. B y contrast, many of the electric distribution utilities experience major financial problems, with net losses that totaled US$77 million in 2002. There are essentially two explanations for this. The first is the deficient management of many of these utilities, that results from political interference, high costs and inefficient operation. The second is the regulatory environment, which sets tariffs according to average rather than marginal costs, and requires utilities to apply cross- subsidy schemes that do not allow them to cover the full costs of operation. Figure 5.4: Operational resultsof loss-makingdistribution utilities 10 I Source: Chahinand Rojas, 2004 RED1Electricity Sector Report 5.14 As a result, the electricity sector represents a significant fiscal burdenfor the state, although this has been declining over time. The electricity sector has historically represented a major fiscal burden, accounting for one third of the stock of public debt by the end of the 1980s. The situation has improved substantially during the last decade, with electric utilities with the share of public debt falling to around 5 percent of the total, and the sector as a whole presenting a surplus of 0.1 percent of GDP, for various reasons distribution utilities continue to represent a significant drain on the public purse. The only regular fiscal transfer to the sector is the payment to the National Solidarity Fund (FSSRI) to cover the deficit that arises between subsidies and surcharges in the national cross-subsidy system, which currently amounts to US$77 million per year. Nevertheless, there i s substantial volume of occasional expenditures associated with particular liabilities and financial crises. First, the national government has assumed historic payment obligations for projects that have proved difficult to finance from the cash flow of the respective utilities. These include the Urra hydro project at an annual cost of 187 around US$lOO million, and the Power Purchase Agreements that supported the development of the first wave of thermal generation plants at a total cost of US$500 million (over the period 1996/01). Second, the government routinely makes 'loans' to loss-making electric distribution utilities to allow them to pay off their debts to the electricity wholesale market. These amounted to a total of US$319 million over the last decade. Third, the Government has paid over US$2.5 billion (close to 3.4 percent of GDP) over the last decade to rescuebankrupt public enterprises (Box 5.1). Box 5.1: The cost of government "rescue operations" 5.15 Rural electrification is an important but manageable challenge for public investmentin the sector. The cost of reaching universal access to electricity throughout Colombia has been estimated at US$288 million and appears modest in relation with the magnitude of the subsidies that have historically been made to the electricity distribution sector. Some impression of the equity of public investment in electricity can be obtained by comparing government expenditure on subsidies to the interconnected system (SIN) (described above), with those made for electrification in the non-interconnected zones (ZNI). At present, supply plus demand-side subsidies to the SIN amount to US$7 per capita per year (or US$15 if rescue operations continue), while subsidies to the ZNI represent US$34 per capita per year. This shows that there i s at least some equity in the distribution of subsidies across geographic areas. It i s important to note that subsidies for rural electrification are entirely funded within the sector without any need for fiscal transfers. 188 Natural gas 5.16 Public investment inthe natural gas sector is concentrated inexploration and transportation, and undertaken by two major public enterprises (Ecopetrol and Ecogas). There are no fiscal transfers to the natural gas sector, so that public expenditure i s confined to those investments undertaken by the public enterprises Ecopetrol (in exploration) and Ecogas (in transportation). Distribution activities are undertaken entirely by the private sector. The public sector played a major role in promoting the expansion of the natural gas transportation system, by means of a number of BOMT contracts awarded by Ecopetrol. Moreover, 30 percent of the value paid by Ecopetrol to the private investors (as a tariff component that was used as the bid award criterion) were not passed to the later created transportation company Ecogas, effectively constituting a subsidy to the sector valued at US$l55 million. In addition, the `development surcharge' (cuotu de fomento) on the transportation tariff raises approximately US$11 million that i s used to finance public investments to expand access to low income households in rural areas adjacent to the transportation infrastructure. Telecommunications 5.17 Public investment in the telecom sector remains high, but has been declining in recent years. While there are no fiscal transfers to the telecom sector, the state continues to play an important role in the financing of the sector, with national and municipal public enterprises investing close to US$360 million per year. Nonetheless, the share of investment funded by the public sector has fallen substantially over time, from 99 percent in 1992 to 36 percent in 2002, still well above the regional average of 15 percent. The rising share of private investment i s largely attributable to the development of new mobile networks. About 75 percent of public investment in the sector was destined for fixed line local telephony services. In addition, important amounts are invested in social telecom programs, mainly through the Compartel program, as described in Chapter 3. 5.18 Moreover, the national telecom operator faces major historic liabilities, as well as potential future new ones. The national telecom incumbent, Telecom, was liquidated in 2003 leading to the creation of a successor company, Colombia Telecomunicaciones. The new operator faces major historical liabilities associated with the remaining debts and massive pension obligations of its predecessor. In addition, Telecom faces claims against guarantees provided in a number of joint venture contracts with the private sector (see below). 189 Figure 5.5: Joint venture investment by Telecom I 10004 A Source: Fainboim, 2004 REDIFiscal Report 5.19 Although Colombia's telecom sector is comparatively equitable, investments driven by commercial criteria are heavily concentrated in higher income areas. While household access to telephone services remains skewed towards higher income groups, Colombia's fixed line teledensity i s above the regional average, and coverage among lower income groups i s comparatively high. At the same time, coverage of mobile telephony lags behind when compared to countries with similar income levels. Moreover, Colombia is one of the first countries in the region to reach universal access to public telephony services in rural localities. The bulk of investment in the telecom sector i s primarily driven by commercial criteria and therefore tends to be heavily concentrated in the higher income areas of the country. Indeed, regions with higher per capita incomes, in particular Cundinamarca, Antioquia and Valle del Cauca, have historically received three times as much investment as regions with lower per capita incomes (Table 5.2). Moreover, even within these regions, investments are concentrated within the departmental capitals, which have 75 percent of the telephone lines as against only 50 percent o f the population. US$ per capita 1991/94 1994t97 1997102 High income regions 37.0 73.4 80.7 Middle income regions 20.3 44.7 33.3 Low income regions 12.4 16.3 26.2 Average 26.7 53.9 53.7 Water 5.20 Transfers from central government to the municipalities represent the main source of investment financing. About one third of public investment in water and sanitation i s funded directly by municipal enterprises, while the remaining two thirds comes from central government transfers as mandated by the 1991 Constitution. These transfers have amounted to an average annual sum of US$278 million during recent 190 years. Around one third of these resources (US$ 94 million) come through Law 715/01, being bound by exclusive use provisions for water and sanitation (Table 5.3). Further resources come from oil sector royalties through the National Royalty Fund (FNR), and a number of specific national programs. In addition, the government provides credits and co-financing for projects in the water sector through FINDETER, a specialized entity that evaluates and finances projects in the sector. Table 5.3: Publicinvestmentinwater 1998/01 I I Average annual I Percentagetotal ransfers to the Municiualities I 239.1 86.0 16.9 6.1 ational Royalty Fund (FNR) 11.2 4.0 6.2 2.2 4.0 1.4 Plan Pacific0 0.7 0.3 Total 278.1 100.0 5.21 A major concern with respect to Law 715/01 is related to the use of the resources by the municipalities. Although a share of the resources transferred to municipalities under Law 715/01 are earmarked to the water sector, in practice there i s no information about how these resources are actually spent, and hence no accountability to central government for their use. Moreover, there is widespread anecdotal evidence to suggest that these resources often fail to be transferred to the municipal water utilities, and are either diverted to other categories of expenditure or invested directly on water projects by the municipality that do not necessarily coincide with the needs and priorities of the water utility. As a result, the water sector that appears on paper to enjoy relatively large fiscal transfers, may not necessarily receive the corresponding benefit. Additional control mechanisms would therefore be required to ensure the appropriate use of these transfers. 5.22 Central government transfers are highly progressive, yet still fail to offset the greater availability of internal investment resources in larger municipalities. Central government transfers to municipalities under Law 715/01follow rules designed to ensure that additional resources reach the poorest municipalities with the highest unsatisfied basic needs. Thus, per capita fiscal transfers vary substantially from just over US$2 in the larger, richer municipalities to almost US$6 in the smaller, poorer municipalities in 2001. Notwithstanding these differences, actual investment per capita in the larger, richer municipalities (at US$27) was three times higher than in the smaller, poorer ones (at US$ 8). The essential reason for this i s that water utilities in the larger municipalities have a stronger customer base and are therefore able to raise far more internal resources for investment, overwhelming the differences in fiscal transfers. 5.23 The greater inefficiency of the smaller municipal utilities, further exacerbates the problems posed by the limited access to investment resources. Not only do the smaller municipal utilities have fewer fiscal and tariff revenues available to finance investments, but they also tend to be much less efficient than the larger 191 municipalities, therefore obtaining lesser results from a given volume of investment resources. The inefficiency of the smaller utilities reflects the diseconomies of scale in operations, as well as the scarcity of technical and managerial human resources, and widespread political interference. Despite efforts to merge companies and to promote investment of larger companies in smaller municipalities, success has been limited. Transport 5.24 Public investment in the transport sector has declined in real terms since 1995. The public sector has invested on average US$1,140 million per year in the transport sector during the last decade. However, the resources devoted to the sector have declined by 50 percent in real terms, from a peak of US$1,590 million in 1996 to US$SOO million in 2002. More than 80 percent of public investment in transport goes to the road sector, with the airport and railroad sectors making up a further 10 percent of the total (Figure 5.6(a)). Since those road corridors with sufficient traffic volume have been transferred to the private sector under toll road concessions, central government investment in roads i s concentrated on some of the lower traffic segments of the primary network. Some 60 percent of total public investment in transport i s financed by central government, with the reminder divided almost equally between departments and municipalities (Figure 5.6(b)). Figure5.6: Publicinvestmentintransport 59% I =National 8Depattmmta oMunicipl (a) By sub-sector (b)Byjurisdiction Source: Ospina, 2004 RED1Transport Sector Report 5.25 Only a small fraction of the road budget goes to maintenance, resulting in maintenance expenditures that are below prudent benchmark levels. About 62 percent of the investment in the primary network i s allocated to preventive maintenance and rehabilitation of the road network, with the remainder being absorbed by emergency works to recover roads affected by landslides, and to pay traffic guarantees on first generation toll road concessions (see below). As a result, maintenance plus rehabilitation expenditure on the national network amounts to US$24,000 per kilometer. At the same time, routine maintenance spending by departmental and municipal governments on the 192 secondary road network amounts to no more than US$2,000 per kilometer. This latter value i s low compared with international benchmarks of US$lO,OOO for routine maintenance of both paved and unpaved roads. Indeed, only four out of 32 departments come close to spending the prudent benchmark level on road maintenance (Figure 5.7). A further problem i s the tendency to engage in small scale dispersed projects improving local segments of the secondary or tertiary road network, diverting resources from more strategic investments in road corridors linking the main production centers with internal and external markets. 5.26 Transport investments are at the discretion of each localjurisdiction, leading to huge variation inthe resources assigned. As far as equity i s concerned, sub-national spending on the road network is determined by the expenditure priorities of each local jurisdiction. This stands in contrast to the water sector where there are explicit formula based transfers with exclusive use provisions. There are huge variations in transport expenditure across department, ranging from little more than US$I per capita in Cauca to almost US$95 in Casanare, with an overall average of US$8 per capita. There does not appear to be any systematic tendency to explain these geographic variations in transport expenditure. Figure 5.7: Dispersion in transport investment across departments j us$5,mo-i0,000 !US$3,0005,000 L r- 0 E us$2,0003,0o0 x k us$1,000-2,000 e Lo 3 3,600 3 5 8 100 Source: Srinivas. 2004 RED1 5.80 Two important regulatory constraints mean that this will not be straightforward to achieve. Mobilizing resources from pension funds for infrastructure i s an important policy objective, but evidently must be done in such a manner as to avoid pension fund exposure to excessive levels of risk. Two regulatory constraints precisely aimed at safeguarding prudent fund management, currently represent a constraint to pension fund financing of infrastructure projects. First, pension fund regulation currently requires that the funds invest only in bonds that are rated A minus (A-) or higher by a recognized rating agency. Many infrastructure bonds would be unable to satisfy this criterion unless either the government (or multilateral agencies) provides project guarantees or other forms of credit enhancements. Evidently, this requirement is an important one for ensuringprudent management o f pension assets. Second, pension funds are required to provide minimum returns to investors within a band around the average return of industry. Current regulation states that if the return provided by a particular pension fund is beneath this minimum level, the difference must be made-up by the pension fund manager from his own capital. However, if the rate of return i s above the upper end of this band, all the excess returns accrue to investors and none to the fund manager. This asymmetric structure creates disincentives for fund managers to actively explore alternate investment instruments that present higher risk-reward ratios. 5.81 However, there are a number of ways of getting around these risks without exposing pension funds to undue levels of risk. There are a number of alternative ways of relaxing the two regulatory constraints described above,. without exposing pension funds to undue levels of risk. One option, which has been tried in Chile, i s to permit pension schemes to offer different `funds' to different investors, and let the investors decide which funds they would like to invest in based on their own age and risk profile. Another option, which could be applied within the existing regulatory structure, i s to permit pension funds to investment in a limited volume of infrastructure securities, subject to rating criteria. This would evidently entail the need to strengthen the capabilities of local credit rating agencies to provide the necessary risk assessments. Alternatively, infrastructure projects could undertake financial engineering to improve the suitability of their bonds for pension funds. This could be achieved either by providing securitized income streams from infrastructure projects in the post-construction 220 phase after many of the major risks have passed, or by creating pooled securities that diversify risk across a variety of infrastructure sectors and projects. 5.82 Possible investment from pension funds into infrastructure. At present, pension funds operate under a limit of 30 percent for investment in equities and 20 percent for securitized bonds. Infrastructure investments by pension funds could be included under either of these categories (in keeping with their risk profile), and a smaller limit of say 10-15 percent could be established initially. Clearly, it will take time for the pension funds to get acquainted with infrastructure securities and as they gain experience, the regulator could consider increasing the limit. With the current asset stock of about $9 billion, such a structure would imply that a stock of about US$0.9 to $1.3 billion would be immediately available for infrastructure investment, with a further US$100million per year thereafter. While this seems relatively small compared to the Government's projected investment needs of about US$ 6.5 billion, it i s important to keep in mind that these resources are of the same magnitude as the entire infrastructure loan portfolio of the commercial banks. Thus, even under this relatively conservative scenario, domestic resources available for infrastructure are being doubled. In addition, if these investments turn out to be successful, a larger market for infrastructure securities is likely to be generated as the trust fund and mutual fund industry (which represents 25 per cent of the financial system) step in to take-up the new opportunities. 5.83 Summary. After a lull in private capital flows, the government has announced its intention to regenerate this source of financing for infrastructure, by publishing a list of US$6.5 billion of green field projects identified as suitable for private participation, as well as announcing its new policy o f zero tolerance towards failing public enterprises. Given Colombia's loss of investment grade rating, it seems likely that some degree o f credit enhancement will be required to attract (particularly international) investors back into infrastructure projects. However, this does not imply a return to the generous guarantees of the early 1990s. A wide range o f more sophisticated financial instruments i s now available for targeting such enhancements towards the most critical elements of risk. Furthermore, the exponential growth of pension deposits during the last decade reaching US$9 billion in 2003 and growing at a rate of US$lOO million per year into the medium term offers the potential of an attractive new source o f long-term domestic financing for infrastructure projects. However, harnessing such financing in a responsible manner, will entail a number of significant reforms in the pension industry, including the offering of multiple funds with a range of risk reward ratios, the establishment of specific limits for infrastructure investments, and the establishment of a more sophisticated risk- rating capability. The infrastructure sectors themselves will also need to take measures to make their offerings more appealing to retirement investors, including financial repackaging of securities to separate out the safer post-construction income stream, and to pool together securities across infrastructure sectors. 221 INVESTMENT PRIORITIES A N D FINANCINGNEEDS 5.84 Overall, bringing together the productive and social investment needs identified in Chapters 2 and 3, the total annual investment requirement is US$4,000, which is equivalent to 5.1 percent of GDP (Table 5.20). It i s important to note that 31 percent of the total (or 1.6 percent of GDP) corresponds purely to maintenance and rehabilitation expenditures for the water, electricity and transport sectors. The remainder corresponds to new infrastructure investment, of which 35 percent (or 1.8 percent of GDP) is productive investment and 33 percent (or 1.7 percent of GDP) is social investment. The transport sector accounts for the largest share of the total with 46 percent (or 2.3 percent of GDP), followed by electricity with 23 percent (or 1.2 percent of GDP) and telecommunications with 19 percent (or 1.0 percent o f GDP). It i s estimated that about 33 percent of these investment requirements (or 1.7 percent of GDP) would be met directly by the private sector. If the remainder of the publicly owned segments of the electricity and telecommunications industry were privatized, this would transfer an additional US$508 million of the investments into the private sphere, representing 12 percent of the total (or 0.6 percent of GDP). Table 5.20: Overview of social and productiveinvestment financing needs USSm 1Public finance Private finance Sectoral allocation I 3,208 1,024 Electricity 612 329 340 Natural gas 0 137 137 Telecommunications 45 1 321 771 Water 5 16 0 516 Transport 1,629 237 1,866 Functional allocation 3,208 1,024 1,230 Maintenanceandrehabilitation 1,307 140 1,441 Productiveinvestment 683 1,419 Social investment 1,164 201 1,364 Sectoral allocation (% GDP) 3.7 1.3 5.1 Electricity 0.8 0.4 1.2 Natural gas 0.0 0.2 0.2 Telecommunications 0.6 0.4 1.o Water 0.4 0.0 0.4 Transport 2.0 0.3 2.3 3.7 1.4 5.1 Maintenanceandrehabilitation 1.4 0.2 1.6 Productiveinvestment 0.9 0.9 1.8 Socialinvestment 1.4 0.3 Source: Own elaboration 5.85 These requirements lie well above historical investment levels, even during the peak period of the late 1990s, and hence there is a n inevitable need to prioritize among the competing alternatives. It i s necessary to confront this bottom-up desirable investment agenda, with the country's overall budget constraint for infrastructure investment. As noted in Chapter 1, Colombia has managed to sustain one of the higher 222 and more stable levels of infrastructure investment in the region. Notwithstanding, Colombia has typically managed to invest no more than 3 percent of GDP in infrastructure, of which around 2 percent of GDP from the public sector and 1 percent of GDP from the private sector. Moreover, even at the historic peak during the late 1990s, public finance for infrastructure amounted to no 4 percent of GDP, of which 2.5 percent of GDP from the public sector and 1.5 percent of GDP from the private sector. This suggests that the desirable investment agenda falls 1-2 percentage points of GDP short of the resources likely to be available, even under a best case scenario. Given that Colombia will continue to face a tight fiscal balance over the next few years, this means that the government will both need to take some difficult decisions to prioritize among competing investments, as well as take any possible measures to expand fiscal space. 5.86 The fiscal space available for public investment in new infrastructure assets could be significantly increased by further privatization of commercially orientated services. A general principle that can be applied is that maintenance and rehabilitation expenditures to preserve and restore existing assets typically bring significantly higher rates of return than investments in new assets. The estimates suggest that 1.6 percent of GDP is needed to maintain and rehabilitate assets, and that this should be given the highest priority. It i s therefore helpful to explore, under different financing scenarios, the margin that remains for new public investment once maintenance and rehabilitation obligations has been met. This exercise is performed in Table 5.22 below, which shows how the size of the margin varies as public investment levels are brought back in line with the historical peak. It also illustrates the impact of increased privatization in more commercial infrastructure services, such as electricity and telecommunications, which could potentially reduce public sector requirements by a significant 0.6 percent of GDP. 5.87 Overall, bringing together the productive and social investment needs identified in Chapters 2 and 3, the desirable annual investment agenda is US$4,000, which is equivalent to 5.1 percent of GDP (Table 5.21). The calculations show that, in the absence of further privatization, there would be a shortfall of between 0.7 percent and 1.2 percent of GDP relative to the desirable agenda of 1.6 percent of GDP of new public sector investment projects, necessitating a major prioritization exercise within the investment agenda. In the most difficult scenario, with a new investment margin o f only 0.4 percent of GDP, it would be important to include the 0.2 percent of GDP needed to remove strategic bottlenecks in the transport sector, leaving only 0.2 percent to cover some of the most sensitive social investments, such as those relating to the Millenium Development Goals for water and sanitation. 5.88 However, additional privatization of commercially oriented services could potentially create the fiscal space necessary to finance the investment agenda. However, under an increased privatization scenario, the available margin of 1.O percent for new public investment i s broadly in line with the remaining public sector investment obligations of 0.9 percent of GDP (given that 0.7 percent of GDP from the public investment agenda would have been transferred to the private sector under this scenario). 223 Yo GDP Public investment at current levels Public investment at peak levels No further privatization of 0.4 0.9 infrastructure Further privatization of electricity 1.0 1.5 and telecommunications enterprises 5.89 Aggregate financial balances constructed for each sector, are used to tie in social and productive sector investment requirements, with resource availability. This section presents the overall financial balance sheet for each of the infrastructure sectors, looking forward over the period 2004/08 (on the methodology used refer to Box 5.5). As such, it confronts the investment needs for social and productive infrastructure identified in Chapters 2 and 3, with the financial realities of each of the sectors, with the aim of evaluating the extent to which each sector has at its disposal the resources available to implement the highest priority investments. Inthis way, all o f the elements of the analysis so far are tied together within a coherent financial framework. The financial balance methodology essentially replicates the concept of a firm's balance sheet at the sectoral level, aggregating the financial balances of all the sector players. As such, it encompasses all sources of investment finance, including both the internal cash generation plus debt capacity of the public and private enterprises in each sector, and the fiscal transfers made by the state. For the purposes of projection, a sector specific demand growth scenario i s developed, and used to forecast turnover and operating costs (assuming a continuation of historic trends in efficiency improvements), leading to an estimate of net operating income. Assets are also assumed to grow in line with demand, together with liabilities maintaining a constant gearing ratio. The internal cash generation of the enterprises, together with their increase in debt finance capacity, jointly defines the space available for financing new investments. Fiscal transfers are assumed to remain constant in real terms. 224 Box 5.5: Methodology for the sectoral financial balances 5.90 With exception of the road transport sector, most sectors will have sufficient resources available to finance their investment needs. Most sectors with the prominent exception of road transportation will have sufficient overall resources available to finance their investment needs, either through operating revenues or access to debt financing. However, even when sectors are not financially constrained on aggregate, specific segments within those sectors may present a deficit situation, as for example some of the water and electricity utilities. While that i s mostly the case, some important considerations have to be taken into account when analyzing some sectors in more detail. 225 5.91 However, if pension liabilities were to be included, a much more difficult picture would emerge in terms of resource availability for infrastructure investments. The financial balances presented below do not take into account the substantial liabilities related to pension costs of the state-owned enterprises. The reason i s that no data were available on the time profiling of these liabilities. However it i s known that un-funded pension liabilities account for approximately US$834 million in the water sector, and over US$2.5 billion in the telecom sector (mostly related to the liquidated former operator Telecom). Both of these estimates are based on actuarial analysis as of 2003, and represent 84 percent and 17 percent respectively of the net asset base of the telecom and water operators to which they correspond. Pension liabilities in the electricity sector, although significant, are much more manageable, standing at US$150 million (based on actuarial analysis of 1998 expressed in 2003 US$). Evidently, the inclusion of these liabilities in the financial balances of the water and telecom sectors would seriously constrain the resources available for investment financing. Electricity Table 5.22: Financial balance for electricity sector: average yearly resources and investment for the 2004108period 1 US$ m 2003 S I N ZNI Verygood Good Bad Very bad Total SIN Resources II II I Sector resources \ : ; II - 570 50 Net resourceswithin firms 527 - I I : ; I - Sector specific funds 43 - FOES 23 - FAER 20 - FAZNI Fiscal resources FSSRI(Net subsidies) FNR 1 - Total 610 266 49 -208 717 58 Investments Maintenance 310 - New generationcapacity 258 Transmission network 43 Distribution network 137 Rehabilitation 43 - PPA guarantees 113 - Total 904 36 Balance -187 +22 Source: Basedon Chahinand Rojas, 2004 RED1Electricity Sector Report 5.92 While the electricity sector as a whole is broadly able to cover its investment needs, some groups of enterprises will remain in deficit. The electricity sector i s predominantly financed by sector resources, mainly net tariff revenues and some sector specific funds (FOES, FAER, FAZNI). Nevertheless, the sector receives a significant fiscal transfer from the National Solidarity Fund (FSSRI) to cover the deficit associated with the national cross-subsidy scheme, which amounts to 26 percent o f revenues for the distribution segment. Each of the generation, transmission and distribution segments of 226 the electricity industry has sufficient resources to cover its respective investment needs. However, within both the generation and distribution segments, there are important clusters of enterprises that face serious and persistent deficit situations, as well as others that present a very positive financial balance (Table 5.22). Inthe case of generation, those enterprises affected by the Power Purchase Agreements of the early 1990s, will continue to lack the financial resources to honor these payments necessitating continual support from the public purse. In the case of distribution, those enterprises outside of the major urban centers that lack a strong base of industrial and higher strata customers will continue to be dependent on transfers from the FSSRI, and will therefore not have an adequate financial base to finance their investment programs. These enterprises are effectively caught in a `deficit trap' since they are unable to finance the investments they need to be able to improve their operating efficiency and thereby reduce their deficit in the medium term. The deficits posed by these problematic enterprises (`very bad' category) are not trivial, amounting to 30 percent of overall sector turnover. Natural gas Residential and industrial Resources Sector resources 131 Net resources within firms 131 Sector specific funds 2 Fiscal transfers - Total 133 Investments Gas transportation 9 Gas distribution 31 \ Export 61 CNG distribution 1 30 I Balance -4 227 4.87 The natural gas sector is able to fund investment needs from tariff revenues and internal debt and equity financing, though plans to export gas will represent a substantial step-up in historic investmentlevels. The natural gas sector i s entirely self- sufficient from tariff revenues generated within the sector, without any need for fiscal transfers. Based on a conservative investment plan and steady market growth that largely continues the past trends of connecting domestic customers and providing gas for the thermoelectric expansion, incomes from the gas companies will provide ample basis for financing investment needs. However, if a more ambitious program i s to be implemented, including proposals to export gas to the neighboring countries and to expand use of natural gas in public transportation, some modest deficits may appear in the middle years of the investment period. Nevertheless, as these investments will likely be undertaken by private companies, they will be able to finance the required investments on the basis of the solid pay-back periods and returns that can be expected. Telecommunications US$ m2003 Local LD Mobile Value added Total Resources Sector resources 714 69 216 -107 892 Net resources within firms 683 69 216 -107 861 Sector specific funds (FC) 31 - - - 31 Fiscal transfers - - - - - Total 714 69 216 -107 892 Investments Service expansion 271 3 439 27 739 Social telecom 31 - - - -3 _ _ 1 Total 302 3 439 27 770 Balance I 412 I 66 I -223 tI -134 1 122 Source: Based on Mendoza, 2004 RED1Telecom Sector Report 5.93 The telecom sector is well-funded overall, although revenues from mobile and value added services are small in relation with the necessary investments. The telecom sector can finance the required investments from tariff revenues generated within the sector as well as debt and equity financing from the sector companies, without any need for fiscal transfers. The income of the telecom operators has increased substantially over the last few years, with local telephony revenues growing at 10 percent per year on average over the last decade. While the sector as a whole shows a sizeable surplus relative to investment needs, the situation varies substantially across different segments of the market. Local and long distance operators present a strong surplus inrelation to the investments needed to reach 20 percent teledensity by 2008. In the mobile telephony market, however, given the need to develop new networks, current revenues from mobile services do not appear adequate to finance the investments needed to raise mobile teledensity to 25 percent by 2008. Nevertheless, since the bulk of investments in new mobile telephony networks will be funded by operators with lucrative local telephony markets (ETB and EPM), this may not represent a significant constraint. 228 Water US$ m 2003 Large Medium Small Total Resources Sector resources 61 21 7 89 Net resources within firms 61 21 7 89 Sector specific funds - - - - Fiscal transfers 92 38 198 328 Law 715 62 27 181 270 Autonomous Regional Corporations (CAR) 27 9 16 52 National Royalty Fund (FNR) 3 1 2 6 Total 153 59 205 417 Investments Maintenance 88 30 50 168 Water resources - - 7 7 Millenium Development Goals 101 33 122 256 Waste water treatment 35 13 - 48 Sewerage systems 12 4 - 16 Storm water drainage 13 5 - 18 Total 249 85 179 513 Balance -96 -26 26 -96 Source: Based on Fernandez, 2004 RED1Water Sector Report 5.94 The water sector presents a significant deficit based on a relatively cautious investment program. In contrast to the other sectors, about 80 percent of resources for investment in the water sector derive from fiscal transfers, and only 20 percent from tariff revenues. Reliance on fiscal transfers for investment financing varies substantially according to the size of the water utility, with medium and large utilities obtaining 40-50 percent of investment finance from fiscal transfers, and small utilities obtaining as much as 90 percent o f investment finance from fiscal transfers. Investments projected in the water sector for the next five years are mainly aimed at meeting the millennium goals, the wastewater plan, and some maintenance and renovation of existent equipment. Investment needs would be substantially larger if an objective of universal access i s set, going beyond the relatively modest target entailed by the Millenium Development Goals. Furthermore, the plan incorporates relatively little investment in waste water treatment given that the total amount estimated for the whole country i s comparable to that already invested in Bogota. Even on this basis, the water sector presents a significant modest deficit with respect to the investments it needs to finance. Moreover, these calculations assume that the full resources allocated under Law 715/01 actually reach the sector, which as discussed above may not always be the case. 229 Transport US$m2003 Primary Secondary and Portsand Airports Urban Total road tertiary road Access transport network networks (Dredging) Resources Sector resources 121 - - -5 - 116 . _ Net resources within firms 121 -5 116 Sector specific funds - - - - - - Asca1 resources 141 251 0 6 - 398 Central government 141 20 5 - 166 Local governments - 231 - -1 - 232 Total 262 251 -1 - 514 Investments Upgrading 267 71- 38 262 638 Rehabilitation 198 167 - - - 365 Maintenance 460 310 - 710 Guarantees 53 - - - - 53 Total 978 477 71 38 262 1826 Balance -716 -226 -71 -37 -262 -1312 Source: Based on Ospina, 2004 RED1Transportation Sector Report 5.95 The transport sector, with its heavy dependency on fiscal transfers, presents b y far the most serious financing deficit of all the infrastructure sectors. The transport sector relies almost exclusively on fiscal transfers for investment finance; althuugh toll rev--the primay~~~twerk- c4ribttte %Freefit &total Im-e++ment resources in the sector. The resources available to the primary versus secondary and tertiary networks are broadly equal. However, sub-national governments on aggregate allocate many more resources to road finance than does the central government. Total resources available to the sector cover no more than a third of the huge maintenance, rehabilitation and investment needs, particularly in the road sector. As a result, large deficits exist across every sub-sector of the transportation system. The shortfall amounts to an annual average of slightly over US$700 million for the primary road network, and close to US$240 million for the secondary and tertiary roads. While deficits in the airport sector might exist, they are basically related to the ambitious investment program in the Bogota airport, as well as the new construction of the Palestina airport in the Manizales region. The challenge for the government will be to assure sufficient resources for secondary airports once the concession of the bigger airports, including Bogota, i s implemented. 5.96 Summary. The financial balance situation varies enormously across infrastructure sectors. While the telecom sector shows a substantial surplus of resources relative to investment requirements, the water and energy sectors present modest but manageable deficits, but the transport sector i s extremely under-funded. Even for the utility sectors that are in overall balance, important differences exist across different market segments or groups of enterprises. In particular, generators affected by early 1990s Power Purchase Agreements, electricity distribution utilities outside the major urban areas, and some of the larger water utilities, present serious and persistent deficits 230 preventing them from financing the investments that they need. In natural gas and mobile telephony, where investments are generally funded under a market paradigm, a significant step-up in investments will be required if plans to export natural gas and develop new mobile telephony networks are to be fulfilled. Moreover, the situation in the water and telecom sectors could rapidly deteriorate towards a deficit position, once the large and under-funded pension liabilities of public enterprises in both sectors start to result in significant claims. Table 5.27: Summary of overall financial balance for infrastructure sectors: average yearly resources and investment for the period 2004108 I Yearly Average Resources Investments Balance ElectricityII 775 II 940 II-165 Water 417 513 -96 Natural 133 137 -4 Gas Transport 514 1826 -1312 Telecbm 892 122 Total 1I2731 II 770 4102 II-1371 Source: Based on RED1Sectoral Reports CONCLUSIONS AND RECOMMENDATIONS 5.97 Both public and private financing of infrastructure have declined significantly, due to fiscal constraints on the one hand, and perceived increases in risk on the other. The transport sector, particularly roads, has been the most adversely affected experiencing a 50 percent decline in real fiscal transfers during the last decade. The search for stable financing for long term infrastructure needs has led to growing use of `future appropriations' and `specific funds'. Moreover, long lead times in contracting and executing major infrastructure projects often prevent them from being concluded within the annual budgetary cycle, resulting in a loss o f around 50 percent of resources allocated to transport. Private investment made an important contribution to infrastructure development in the 1990s, although the cost of the earliest schemes has proved to be quite high in terms of guarantee payments. There is currently renewed interest in private participation in infrastructure, which will probably need to be supported by carefully designed and narrowly targeted credit enhancements. It would also be desirable to harness part of the burgeoning pension fund deposits as a source of long-term local finance for infrastructure projects, although this will require some significant regulatory reforms in the pension industry, as well as financial engineering in the infrastructure sectors. 5.98 A number of policy changes would help to support significant improvements inthe flow of resources to finance infrastructure investments. Based on the preceding analysis a number of recommendations can be made over future financing policy for the infrastructure sectors. The first block of recommendations numbered (a) to (c) relate to public financing, while the second block of recommendations numbered (d) to (f) relate to private financing. 231 (a) A more stable and substantial source of funding for investmentsinthe road network is urgently needed. B y far the most serious infrastructure financing problem in Colombia arises at all levels of the road network, with important consequencesfor productivity and competitiveness, as well as future fiscal liabilities for road rehabilitation. Given the proven vulnerability of direct budgetary transfers to the road sector, an alternative funding mechanism i s needed. Motivated by similar considerations, many countries have established dedicated road funds financed from fuel taxes, that provide at least the minimum resources required for critical maintenance activities. A similar or equivalent mechanism could also be considered for Colombia. Moreover, at the sub-national level, the huge variance in discretionary allocations to the road sector suggests that there may be a case for requiring minimum investment levels compatible with critical maintenance, comparable to current practices in the water sector under Law 715/01. (b) Constraints posed by the annual budgetary cycle, may be better overcome through a multi-annual framework, than by more ad hoc approaches. Large infrastructure projects are often characterized by long lead times and sustained payment commitments over a number of years. Both of these features are difficult to accommodate within the annual budgetary cycle. In Colombia, considerable ingenuity has been exercised to get around these obstacles through `future appropriations' and `specific funds'. Although, in the road sector, the gap between budget allocations and spending capabilities within the same calendar year remains large. However, this is resulting in a significant loss of flexibility with `future appropriations' already representing 40 percent of fiscal space for infrastructure. It may therefore be more reasonable to formally handle these concerns within a multi-year budgeting framework, that allows a longer term perspective to be taken on projects that meet certain well- defined criteria. (c) A significant number of public enterprises could eventually be excluded from the fiscal accounts, on the basis of some corporate governance reforms. Public enterprises with near commercial operations continue to use up a substantial segment o f the government's fiscal space, leaving very little to finance inescapable public responsibilities, such as the maintenance of the road network. The privatization of some of these enterprises would contribute to opening-up additional fiscal space, but i s not always possible due to political and constitutional constraints. An alternative option i s to exclude `well-run' public enterprises from the fiscal accounts. The transmission company, ISA, already meets all o f the IMF criteria for a `well-run' public enterprise. A number of other companies such as ISAGEN, Ecopetrol, ETB and EPM, are not far behind, and could meet them on the basis o f sustained reforms to improve corporate governance, thereby creating significant new fiscal space within the public sector. 232 (d) Regulatory reform of the infrastructure sectors needs to be an integral part of any strategy to regenerate private investment. Private investors, banks and pension funds have repeatedly complained that high levels of regulatory and judicial uncertainty in the infrastructure sectors prevent the creation of a suitable environment for private investment. Specific problems are the continuous changes in the rules of the game, as well as the dominant presence of public enterprises leading to concerns about unfair competition. The sectors most adversely affected are electricity (particularly generation) and telecom. Therefore, a long-term view by the government, and a stable regulatory environment, as well as a judicial framework based on the rule of law and complying with international rules of arbitration, will be essential to recuperate (particularly international) private investment in infrastructure. (e) More creativity should be exercised in the design of narrowly targeted credit enhancements to support any resurgence in private investment. Some form of credit enhancements i s likely to be required for attracting both domestic and international investors. To avoid the costly payment obligations derived from these enhancements in the past, better risk distribution, pooling of projects and more creative designs o f such enhancements will be necessary. Guarantees could be gradual depending on the extent of departure from original projections and backstopped by possible contingent credit lines. An important lesson from the past i s the need for more conservative demand projections. (f) Significant long-term domestic financing for infrastructure could be attracted from pension funds, if certain regulatory changes could be made. To allow for an increased participation of domestic long-term investors such as pension funds in the funding of infrastructure projects, without compromising the security of pension deposits, some key reforms are required. These will entail strengthening the institutional capabilities of local credit rating agencies to provide more precise risk assessmentsfor infrastructure projects. In addition, some reforms to the regulations governing pension fund investments will be needed. One possibility would be to allow pension funds to offer investors the choice between a number of different funds differentiated by risk-reward profile. Alternatively, a specific limit for infrastructure investments by pension funds (subject to rating criteria) could be introduced within the current regulatory framework. In either case, the infrastructure sectors themselves will need to be creative in packaging their bonds into a form compatible with the risk requirements of pension investors. (g) It would be desirable to increase private participation in commercially viable activities. It has been shown that the fiscal space available for infrastructure investments i s relatively small relative to the desirable investment agenda that has been identified. Under such circumstances, it i s questionable that the state should continue to be 233 involved in providing services that are commercially viable, and hence could be privatized (such as some aspects of electricity and telecommunications). The state's involvement in these services effectively limits the resources available for other areas of infrastructure that can only be provided by the state (such as most of the road network, and rural access programs). (h) There is a need to improve accountability of sub-national governments for infrastructure investments. Sub-national governments have significant responsibilities for finance o f water and transport infrastructure, and to this end receive substantial transfers from the central administration. However, there is little accountability, or even information, about how these resources are being used. The limited information that these resources are not being very effectively managed to maintain and expand infrastructure assets. It would therefore be desirable to improve the accountability o f sub-national governments for the use of these resources. A particular concern i s the need to create incentives for adequate maintenance of locally managed roads. 234 6. A WAY FORWARDFORTHE INFRASTRUCTURE SECTORS INTRODUCTION 6.1 Each of the preceding chapters identified specific recommendations for ensuring that infrastructure makes the necessary contribution to meeting economic and social objectives, while loosening as far as possible the associated institutional and financial constraints. The main conclusions of the five central chapters can be summarized in the following terms. 6.2 The first chapter on international benchmarks concluded that Colombia has sustained a relatively stable level of investment in infrastructure with sustained public sector finance. As a result, Colombia's endowment of social infrastructure is beyond what might be expected for a country with comparable economic, geographic and social characteristics, however, its endowment of productive infrastructure falls short of what might be expected in comparison with its peers. This finding suggests that it may be desirable to place greater emphasis on the needs of the productive sector in devising future policies and strategies for the infrastructure sectors. 6.3 The second chapter on infrastructure and growth presented a deeper analysis of the infrastructure constraints currently faced by the productive sector in Colombia, highlighting deficiencies inaccess to key ports, quality of road networks, reliability of power supplies, and rural telecommunications. It concluded that the country needs productive infrastructure investments o f the order of US$2,600 million per year (or 3.2% of GDP) in order to keep pace with the demands of a growing economy and an increasingly open regime for international trade. 6.4 The third chapter on infrastructure and social objectives found an impressive array of policies for promoting rural access to services, but noted some financing shortfalls specifically in the rural electrification and rural roads sectors, as well as the absence of a strong national initiative for the water sector. Overall, the annual investment needs for social infrastructure are estimated at US$1,500 million (or 1.8% of GDP), of which about 60 percent relates to the transport sector. The analysis of the utility cross-subsidy system showed that the substantial financial deficits resulting from this policy could be significantly reduced by altering the design parameters o f the current framework. While cross-subsidies are effective at reducing the burden on utility costs on the budgets of low income households, they are very poorly targeted resulting in leakage equivalent to 60 percent of subsidy resources. Moreover, any substantial 235 improvement in targeting is likely to necessitate a major review of the current stratification system. Furthermore, it was found that urban transport costs weigh at least as heavily as utilities on the budgets of low income households, but do not have such a comprehensive social safety net associated with them. 6.5 The fourth chapter on the legal, regulatory and institutional framework highlighted a number of important deficiencies in the internal functioning of the Regulatory Commissions and Superintendence, including the selection procedure for leadership figures, as well as coordination problems between policy-making, regulatory and supervisory activities. The Superintendence is burdened with a huge number of onerous functions, and it is therefore questionable whether the critical anti- trust and intervention functions of the Superintendence would not better be delegated to other more specialized institutions. Furthermore, there are a number of salient sector- specific regulatory issues. In particular, the fact that the telecommunications sector is in urgent need of a modern legal framework that acknowledges the phenomenon of technological convergence. In addition, the future of private participation in the thermal generation of electricity i s strongly linked to the issue of capacity charges. 6.6 The fiith chapter on infrastructure finance found that the water and transport sectors remain heavily reliant on public finance, but that budget allocations to transport have declined sharply over the last decade. Moreover, public electricity and telecommunications enterprises, although more commercially oriented, have made repeated claims on the public budget in terms of guarantees and bailout operations. In addition to the levels o f public investment, the budgetary mechanisms through which public finance is delivered to the infrastructure sector can have a major impact on its ultimate efficacy. Particular concerns are the stability and predictability of funding for critical road maintenance activities, as well as contracting and management problems that prevent the national road agency from executing more than 50 percent of its allocated budget. Although private finance has declined substantially since the late 1990s, the government has recently announced the launch of a major new initiative for private participation in infrastructure. However, there are currently a number of significant regulatory obstacles to private finance that will need to be addressed in the context of this revival. 6.7 The purpose of this final chapter is to integrate the various individual recommendations briefly recapped above into a coherent and strategically organized agenda for infrastructure in Colombia. The chapter therefore begins by indicating how the different recommendations provided knit together into an integrated program for action, and goes on to prioritize specific recommendations within a high-low impact versus high-low difficulty framework. Thereafter, the financial implications of the different recommendations are identified, distinguishing between those with positive, neutral or negative fiscal implications, and exploring ways of reallocating fiscal space across the different recommendations. Finally, the institutional responsibilities for taking on different elements of the reform agenda are examined. 236 INTEGRATIONOF RECOMMENDATIONS 6.8 The preceding chapters generated some 40 recommendationsfor improving the infrastructure sectors in Colombia. Table 6.1 shows how the different recommendations map across the different infrastructure sectors and thematic axes. The numbering given after each recommendation provides cross-referencing to the chapter where the recommendation was originally made and i s described at greater length. Table 6.1: Overview of recommendations by sector and by theme Growth Social Finance tegulation Reformdesign of cross- Improve commercial