Report No. 39832-CO Colombia Decentralization Options and Incentives for Efficiency (In Two Volumes) Volume II: Sector Annexes August 2009 Poverty Reduction and Economic Management Colombia-Mexico Country Management Unit Latin America and the Caribbean Region Document of the World Bank This report was prepared during years 2006 and 2007 with several field visits to Colombia and close interaction with national government agencies as well as selected departmental governments and municipalities. The quantitative and econometric analysis is based on information available up to December 2006. The findings, interpretations, and conclusions expressed in this report do not necessarily reflect the views of the counterpart institutions which kindly collaborated in the elaboration and discussion process. ACKNOWLEDGEMENTS This report was prepared by a team led by Jonas Frank (LCSPS) and Christian Y. Gonzalez (LCSPE), co-task managers. Core team members were: Erik Bloom (LCSHE, education), Maria Ivanova Reyes (LCSPE, efficiency analysis), Jeffrey Rinne (human resources), Pablo Roda (consultant, roads), and Diana Pinto (Fedesarrollo, health). Inputs were also provided by Carlos David Beltrán (consultant, roads), Stephen Brushett (LCSTR), Mauricio Cuellar (LCSTR), Nicolas Estupiñan (LCSTR), Rocio Lavalle-Medina (LCSPE), Martha Laverde (LCSHE), Christine de Mariz Rozeira (LCSPS), David N. Sislen (LCSUW), Javier Suárez Pandiello (consultant), and Juan Antonio Zapata (consultant). Guidance throughout report preparation was provided by Axel van Trotsenburg (LCC1C), Laura Kullenberg (LCCCO), Miguel Lopez-Bakovic (formerly Country Manager, LCCCO), Marcelo Giugale (LCSPR), Nick Manning (LCSPS), Ron Myers (LCSPS), Guillermo Perry (formerly Chief Economist), and David Rosenblatt (LCSPE, Lead Economist). Very helpful logistical assistance was provided by: Judith Abele (LCSPS), Francisco Clavijo (LCCCO), Elsa Coy (LCSHE), Brian Gaffney (LCSPS), Michael Geller (LCSPE), Fatima Galarraga (LCCCO), Patricia Holt (LCSPE), and Patricia Mendez (LCSPS). We would like to thank Diane Stamm (consultant) for editing the report. The team appreciates the useful comments that were provided by: Jairo Arboleda (LCSSO), Rodrigo Archondo (ETWTR), Felipe Barrera (HDNED), Mauricio Cardenas (Fedesarrollo), Mauricio Santamaria (Fedesarrollo), Maria-Luisa Escobar (Brookings), Santiago Herrera (DECVP), Edgardo Mosqueira (LCSPS), Guillermo Perry (Chief Economist), Cesar Calderon (LCRCE), Kathrin Plangemann (LCSPS), Fernando Rojas (LCSPS), Eduardo Velez (LCSHE), and Raja Bentaouet Kattan (LCSHE). The peer reviewers for his report are: Jennie Litvack (LCSHD), Remy Prud'homme (Univ. Paris XII), and Eduardo Wiesner (Bogotá). We would like to thank the Colombian authorities for their cooperation in preparing the report. The main counterparts were the Ministry of Finance and the National Planning Department (DNP). In the Ministry of Finance we would like to thank Juan Pablo Zarate (Former Vice-Minister), Ana Lucia Villa, Nestor Mario Urrea, and Edgar Guio, for their support and fruitful deliberations with Bank staff. In the National Planning Department (DNP) we appreciate the continued support provided by Carolina Renteria (Former Director), Andrés Escobar (Former Sub-Director), José Fernando Arias, Felix Nates, Fernando Acosta, Oswaldo Porras, William Jimenez, Luz Stella Carrillo, Jose Lenin Galindo, and Johanna Castelblanco. This study is based on data and efficiency scores provided and developed by DNP. Detailed information about the methodology is contained in the publications "Metodología para la Medición y Análisis del Desempeño Municipal" (DNP 2005b) and "Evaluación del Desempeño Integral de los Municipios" (DNP 2008). The continued support and guidance by the Dirección de Desarrollo Territorial (Department of Subnational Governments) was critical for this analysis. In the Ministry of Transport we would like to thank Fidel Bojorquez and Edgar Carvajal. In the Ministry of Social Protection we are particularly grateful to Esperanza Giraldo. In the Ministry of Education we appreciate the support by Gloria Mercedes Alvarez and Edgar Robles. This report would not have been possible without the decisive and continued support of departmental governments and mayors from Antioquia, Bolívar, and Nariño. We would like to thank the many individuals who have contributed this effort. In Nariño, we would like to thank the Governor, Eduardo Zúñiga, for the warm welcome during the stay of the mission. We also appreciate the support and fruitful deliberations with Angela Trujillo (Secretary of Finance), Patricia Cordoba (Secretary of Education), Sonia Gomez (Secretary of Health), Gilberto Betancourt (Secretary of Infrastructure), and Fernando Viteri (INVIAS). We would also like to thank the support by Jaime Santander, Juan Carlos Vela and Cesar Ruano. Ximena Ruales provided excellent logistical support for the mission. We also would like to thank the personnel of the municipality of Pasto as well as mayors and technical personnel of the municipalities of Tablón de Gómez, San Pablo, Samaniego, El Peñol, Puerres, Ipiales, La Florida, and Yacuanquer for their participation during the workshops. In Bolívar, we would like to extend a special note of appreciation to the members of the departmental government, in particular Adolfredo Periñan (Secretary of Planning), for the decisive support that was provided. We also would like to thank the support by Carlos Alvarado (Secretary of Public Works), Augusto Salas (Secretary of Finance), Arturo Rangel Perez (Secretary of Education), and Nicolas Alvarino (Secretary of Health). We would also like to thank the support by Enrique Mazenett and Roberto Eljadue Martinez. We also thank the personnel of the Planning Department of the District of Cartagena for the discussions held during the mission. In Antioquia, we appreciate the support provided by Pedro Juan Gonzalez (Secretary of Planning), Juan Fernando Mesa (Secretary of Finance), Margarita Zuleta (Director in Secretary of Finance), Carlos Montoya (Secretary of Health), Claudia Restrepo (Secretary of Education), and Margarita Angel (Secretary of Infrastructure). The continued support of Edgar Arrubla (Advisor, Secretariat of Planning) is greatly appreciated. We also would like to thank the personnel of the Planning Department of the District of Medellín. As well we extend our appreciation to the mayors and technical personnel of municipalities in the department for their participation during the workshops. We also appreciate the contributions of the seminar participants of different workshops held in Bogotá during March 2007. In particular, Gabriel Misas, Piedad Caballero, Carolina Nieto, Fabio Sanchez, Catalina Cuervo, Jose Diaz, Jorge Fuentes, and Pedro Ibáñez. Colombia Decentralization: Options and Incentives for Efficiency Volume II ­ Sector Annexes Table of Contents Section 1: Fiscal Decentralization...........................................................................................1 A. The Subnational Tax System ................................................................................................1 B. The Evolution of the Intergovernmental Transfer System ....................................................6 C. Royalty Regime .....................................................................................................................8 D. Fiscal Responsibility ...........................................................................................................10 E. Subnational Expenditures ....................................................................................................17 Section 2: Health Sector .......................................................................................................23 A. Health Reforms and Decentralization .................................................................................23 B. Human Resources Management in the Health Sector .........................................................29 Section 3: Education Sector ..................................................................................................43 A. Background of Education Decentralization in Colombia ...................................................43 B. Human Resources Management in Education ....................................................................47 Section 4: Road Sector ...........................................................................................................51 A. History of Road Decentralization .......................................................................................51 B. Portray of the Case Studies: Challenges of the Road Sector in Antioquia, Bolivar and Nariño........................................................................................54 C. Road User Financing ...........................................................................................................56 Section 5: Water Sector .........................................................................................................61 A. Overview .............................................................................................................................61 B. Context of Decentralization ................................................................................................62 C. The Decentralization Reforms of the 1990s and Legal Framework ...................................62 D. Key Institutions ...................................................................................................................63 E. Government Strategy ...........................................................................................................68 F. Understanding Varying Approaches to Service Provision ..................................................71 Section 6: Input-Output Efficiency Analysis of Subnational Governments and Regression Analysis .....................................................................................73 A. Methodology and Model .....................................................................................................73 B. Efficiency Analysis Results ................................................................................................75 C. Regression Analysis to Determine Potential Drivers of Inefficiency .................................78 Appendix 1 Endogeneity Tests ...............................................................................................85 Appendix 2 Literature Review .................................................................................................90 Section 7: Argentina: Fiscal Decentralization Challenges and Institutional Responses ........................................................................................93 A. The Federal Organization....................................................................................................93 B. Fiscal Challenges.................................................................................................................94 C. Institutional Arrangements as a Response to the Fiscal Challenges ...................................97 Section 8: Decentralization in Spain ..................................................................................101 A. The Asymmetric Approach Towards Decentralization ....................................................101 B. Selection Criteria ...............................................................................................................102 C. Institutional Arrangements ................................................................................................103 D. Performance and Monitoring ............................................................................................104 E. Benefits ..............................................................................................................................106 F. Problems ............................................................................................................................107 G. Lessons Learned................................................................................................................107 Bibliography .........................................................................................................................109 Figures Figure 1.1A. Departmental Per Capita Tax Revenue and Fiscal Capacity ................................1 Figure 1.2A. Municipal Per Capita Tax Revenue and Fiscal Capacity .....................................3 Figure 1.1B. Overall Fiscal Balance and Debt Stock of Subnational Governments ...............11 Figure 1.1E. Department's Expenditures Per Capita ...............................................................17 Figure 1.2E. Municipalities Expenditures Per Capita ..............................................................18 Figure 1.3E. Department's Public Resources Per Capita ........................................................18 Figure 1.4E. Municipalities' Public Resources Per Capita ......................................................19 Figure 1.5E. Distribution of SGP for general purpose Per Capita ...........................................19 Figure 1.6E. Departmental Government Per Capita Expenditures in Education .....................20 Figure 1.7E. Departmental Government Per Capita Expenditures in Health ..........................20 Figure 2.1A. Change in Distribution of National Transfers to Departments and Municipalities .........................................................................28 Figure 2.1B. Percent Distribution of Total SGP Resources of Health, by Department 2005-2006 ..................................................................................39 Figure 3.1B. Average Cost per Teacher, 2006.........................................................................49 Figure 5.1A Increases in Water and Sewage Services Coverage from 1973 to 2003 .............61 Figure 5.2A Institutional Scheme for the Water and Sanitation Sector in Colombia ..............64 Figure 5.3A Water and Sewage Coverage by Department, 2003 ............................................66 Figure 6.1A Models to Measure the Level of Efficiency in Education Provision...................75 Figure 6.2A Health Sector Models to Measure the Level of Efficiency in the Provision of Health ............................................................................................76 Figure 7.1A Provincial Debt Structure, 2001 ..........................................................................95 Figure 7.2A Provincial Debt Structure, 2006 ..........................................................................96 Figure 7.3A Evolution of the Total Income and Expenditure of the Provinces ......................97 Tables Table 1.1A. The Subnational Tax System .................................................................................4 Table 1.1C. Royalty Rates .........................................................................................................8 Table 1.2C. Distribution of Royalties ........................................................................................9 Table 1.1D. Fiscal Responsibility Laws ..................................................................................12 Table 1.2D. Subnational Governments subject to Law 550 ....................................................15 Table 1.1E. Literature Review on Sub-national Response to Intergovernmental Fiscal Transfers .......................................................................21 Table 2.1A. Comparison of Health Public Sector Competences and Responsibilities Before and After Decentralization Reforms ......................24 Table 2.2A. Amount, Destination and Purpose of National Health Transfers ........................24 Table 2.3A. Recent Changes to the Health Financing and Service Delivery ..........................28 Table 2.1B. Health Resource Management Functions of Subnational Governments..............29 Table 2.2B: Human Resource Management Responsibilities in Health ..................................32 Table 2.3B: Indicators of Healthcare Facilities Intervened by Program ................................34 Table 2.4B. Responsibilities and Competences Established by Ley 715: Responsibilities in Terms of Stewardship and Management of the Health Sector in General...........35 Table 2.5B. Responsibilities and Competences Established by Ley 715: Department and Municipal Responsibilities according to SGP Purposes ...........36 Table 2.6B. Allocation Level, Percent Distribution of Resources by Level and Criteria to Determine the Amount to be Allocated, by SGP Expenditure Items ..............37 Table 2.7B. Department and Municipal Responsibilities According to SGP Purposes ..........38 Table 2.8B Department Level Resources, 2002-2006, Bolivar ...............................................39 Table 2.9B Department Level Resources, Amount and Distribution, Nariño 2004-2006 .................................................................................................40 Table 2.10B Department Level Resources 2005-2006 (millions of pesos), Antioquia 2004-2006 ..........................................................................................40 Table 2.11B Department Level Resources 2005-2006 (millions of pesos), Bolivar 2004-2006 ...............................................................................................40 Table 2.12B Department Level Expenditures, Amount and Distribution, Nariño 2004-2006 41 Table 2.13B Department Level Expenditures, Amount and Distribution, Antioquia 2004-2006 ..........................................................................................41 Table 2.14B Total Municipal Expenditures, Amount and Distribution, by Purpose 2005 ......41 Table 3.1B: Human Resource Management Responsibilities in Education ............................47 Table 3.2B: Teachers by Employment Regime & Grade .......................................................48 Table 3.3B: Teachers Monthly Base Pay, by Employment Regime & Grade ........................49 Table 4.1B. Road Indicators in Antioquia, Bolívar, and Nariño .............................................54 Table 5.1A Continuity of the Water Supply Service by Regions ............................................62 Table 5.2A Nominal Urban and Rural Coverage.....................................................................65 Table 5.3A Water Operators by Size of the Municipality .......................................................68 Table 5.4A Water and Sanitation Sector Institutional Structure..............................................71 Table 6.1A The DNP Model: Multiple Inputs and Single Outputs .........................................74 Table 6.2A Correlation between the Efficiency Scores in Education & Health Sector Models .........................................................................77 Table 6.3A Variables and Acronyms of the Regression ..........................................................78 Table 6.4A Regression Analysis for the Coverage Model ......................................................80 Table 6.5A Regression Analysis for the Quality Model ..........................................................81 Table 6.6A Regression Analysis for the Health Models..........................................................82 Table 7.1A Argentina: Assignation of Responsibilities by Government Level ......................93 Table 7.2A Provincial Fiscal Balances, 1992-2006 .................................................................94 Boxes Box 2.1A Incentive Problem Created Through Law 60 ..........................................................26 Box 2.2A The Incomplete Hospital Reform in Colombia as an Impediment to HigherAllocative Efficiency ................................................................27 Box 7.1A The Risk of Intervention Bail-outs and Fiscal Bail-outs .........................................99 Annex 1: Fiscal Decentralization ANNEX 1: FISCAL DECENTRALIZATION Annex 1A. The Subnational Tax System A1.1 Tax revenue is a function, among others, of fiscal capacity, tax rates, and tax collection practices, which are all different in each of the 1,100 municipalities and 32 departments. Assessing the efficiency of subnational governments with regards to taxation is therefore a difficult undertaking since some many important pieces of information is to date not available in Colombia. The following analysis therefore endeavors to shed light on some of the main efficiency and incentive problems, highlighting the need for further research and data development. Analysis will be focused on assessing the relative, not the absolute level of efficiency of subnational governments in raising tax revenue. A1.2 Taking into account non-oil and non-natural resource GDP,1 it becomes clear that departments raise different levels of taxes per capita at similar levels of fiscal capacity (Figure 1.1A). Different levels of fiscal capacity are reflected in the fact that in 2005 about 61 percent of departmental tax revenue was concentrated in only six departments (DNP 2006): Atlántico, Antioquia, Boyacá, Cundinamarca, Valle del Cauca, Santander.2 Figure 1.1A. Departmental Per Capita Tax Revenue and Fiscal Capacity (non-oil and non- mineral GDP per capita, in Colombian pesos), 2005 350,000 300,000 Colombian pesos per capita 250,000 200,000 150,000 100,000 50,000 0 Bo i m a Ar i la Su ía as a r s C a C At l e t a C s ar V a dí o ca Vi c a C yo G có az iño C tá Q ba N e Bo d a ia n t na ar á C re t io e i o id der te g d a r e na co és C a da ua s G er To a ge da ve i a j ir An a l l cr R ac n G na C l ív qu c u ue a a N Ma vi a al ar ho L a a nd au do Sa le d i n ti up ua l A nc Am a r ua M n au l ra e H al N P r t an an m V y ui o ch m aq ór un á tu na e is y an Pu at ov S or és dr An Fiscal capacity: from low to high n Sa Vehicle Tax Registry Liquor Tax Beer Tax Tabacco Tax Gasoline Tax Stamps Other Taxes Source: World Bank calculation based on information provided by DNP (2006) and DANE (GDP and population data from 2004). 1. Due to the importance of royalties in some departments, this measure of GDP is a more accurate mirror of fiscal capacity. Unfortunately, data on consumption are not available by departments, which would give an even better reflection of departmental fiscal capacity in the light of the fact that a large share of tax revenue are excises. 2. Departments with high fiscal capacity and efficient tax collection include Antioquia, Cundinamarca, and Meta; departments with high fiscal capacity but inefficient tax collection include Atlántico, Meta, Santander, and Valle. Departments with low fiscal capacity but relatively high efficiency in tax collection include Guaviare, and among the relatively inefficient departments are Chocó, Nariño, and Norte de Santander. 1 Annex 1: Fiscal Decentralization A1.3 As reported by Bird (2005), typical problems in departmental tax collection include, among others: high rates on the liquor tax encourage smuggling and counterfeiting (of stamps) and widespread evasion. Although the marked is dominated by departmental production monopolies (licoreras) collection costs are high and some licoreras have gone bankrupt. The yield of tobacco taxes remains low and decreasing, owing largely to the lack of control over smuggling. Lotteries are poorly administered and are not a very productive revenue source. A1.4 Taking into account non-oil and non-natural resource GDP, municipalities also exhibit different levels of efficiency levels with regards to per capita tax revenue collected. Aggregating municipalities to departments, the efficient municipalities with high fiscal capacity are located in Antioquia and Valle, in addition to the Capital District of Bogotá;3 the relatively inefficient municipalities are located, for instance, in Santander and Cundinamarca. Some of the low fiscal capacity municipalities but which exhibit relatively efficient tax administration are located in the departments of Norte de Santander, for instance; on the other hand, some of the inefficient municipalities are located in Nariño and Magdalena. These assessments need to be understood in the context of some structural conditions of tax collection: in the rural and low fiscal capacity jurisdictions the private sector traditionally plays a relatively limited role. For municipalities this is a significant constraint for the collection of the business tax, for instance: in 1998, almost 80 percent of municipalities collected less than 5 percent of their current revenues in that tax; only some thirty municipalities collected more than 40 percent of their current revenues in the business tax. A1.5 With regards to the property tax, it is estimated that about three-quarters of municipalities collect less than 80 percent of their potential under the current rates Iregui (2004:31).4 The property tax accounts for 20 percent or less of the current revenue in 63 percent of municipalities and covers less than 5 percent of personnel expenditures in 90 percent of municipalities. Several factors contribute to weak tax collection. Many cadastral values reportedly remain as low as 5 to 30 percent of market values by some estimates. Also, in most municipalities, property tax administration is faulty, as reflected in the observed gap between nominal and effective tariffs, although weak security also contributes to this gap in many cases; nominal tariffs are typically below the maximum allowed by existing legislation (see Iregui and others 2003). All in all, municipalities can still make further progress in raising more revenue through the property tax. 3. Data distortion in the case of Bogotá for the Industry and Commerce Tax, due to tax reporting of firms in this jurisdiction (the fiscal capacity might in fact be lower than what can be inferred from the data). This might also be the case in other jurisdictions with economic centers. 4. Colombia's property tax collection is above the average for developing countries (0.4 percent of GDP) and below the average for OECD countries (2.6 percent of GDP, the highest being the U.K. at 4.0 of GDP. See OECD Observer, May 4, 2001, at: www.oecdobserver.org. 2 Annex 1: Fiscal Decentralization Figure 1.2A . Municipal (Aggregated by Department) Per Capita Tax Revenue and Fiscal Capacity (non-oil and non-mineral GDP per capita, in Colombian pesos), 2005 350,000 Other Taxes 300,000 Land Tax Industry and Commerce 250,000 Colombian pesos per capita 200,000 150,000 100,000 50,000 0 un Atl e ta N re an r isa cá Su a e ór a To c a A r il a Vi ca có t e gd a re m o az ño y a nt ia és o aq r Va ío ve . C a s Q ba Bo a C á s nt na i d er as a ge C re B a lA .C G er io o tá cia ca C sa An a ll ní C ad n a t ic da ay G na ji r l im t C o lí v c é s S i oq u ld R ya N M a a via nd ue u au ov d a au ho up Am ari do La and M Sa le ua ra ua e ar di án V H Pr an ch ra m na D N Bo g e n al o ui C G tu t u Pu or C at dr An n Sa Fis cal capacity: from low to high Note: Data distortion in the case of Bogotá for the Industry and Commerce Tax, due to tax reporting of firms in this jurisdiction. This might be the case also in other jurisdictions with economic centers. Source: World Bank calculation based on information provided by DNP (2006) and DANE (GDP and population data from 2004). A1.6 As a result of the limitations in raising tax revenue, most departments and municipalities have developed significant dependence on central government transfers. In 2005, transfers constitute 46.7 percent of total resources, while tax revenue is about 27.9 percent. Although with huge variations,5 departments are per average more dependent from transfers than municipalities.6 5. There are significant differences in transfer dependence: ranging from 2.7 percent of total revenue (Casanare) to 44.8 percent of total revenue (Atlántico) in 2003. Data: World Bank Colombia PER (2005). 6. In 2003, of all municipalities, only 92.1 percent collected 30 percent or less of their total revenues in taxes. Most collected 10 percent or less. Only 12 municipalities or districts (including Bogotá and Medellin) collected more than 50 percent of their revenues through taxation. 3 Annex 1: Fiscal Decentralization Table 1.1A. The Subnational Tax System (Selected Taxes) Tax Tax base Tax rate Tax administration Tax proceeds Earmarking Liquor Tax Departments Determined nationally by law. Varying Departments. 100% Departments and 70% Health (not (delegated from according to alcoholic grade. Capital District. explicit, corresponding central government) (40% consumption tax for department; sales: to share of IVA cedido). 8% for departments = "IVA cedido"). 30% Sports Beer Tax Departments Determined nationally by law. 51% (40% Departments and 100% Departments (of Health (the share (delegated from consumption tax for department; sales: 8% for Capital District their share). corresponding to IVA central government) departments = "IVA cedido") (Bogotá). cedido) Tabacco Tax Departments Determined nationally by law. Different rates Departments and 100% Departments 16% of revenue ear- (delegated from according to sales price. DANE defines sales Capital District marked for sports. central government) price every half year. (Bogotá). Rest: to be allocated freely. Tax on Vehicle Departments and Determined nationally (Ministry of Transport). Departments 80% department No Ownership Capital District (Inspecciones de 20% municipal (Bogotá). Tránsito) or Capital (revenue-sharing). District (Bogotá). Capital District: 100%. Unidades Especiales de Desarrollo Fronterizo 100%. Gasoline Tax Tax-base sharing: Determined nationally (Ministerio de Minas y National (DIAN; To be allocated freely. 100% Road (sobretasa a la departments and Energía). Ministry of Finance). maintenance and public gasolina) municipalities. 6.5 % departments. Departments. transport. 18.5 % municipal. Municipalities. 25% Capital District. Diesel Tax National 6% National (DIAN; National revenue- Earmarked to a 100% to (ACPM) Ministry of Finance). sharing: 50% national the roads managed by level; each level of 50% departmental level. government. Property Tax Municipal Municipal: catastral valorization (defined each Municipal 100% Municipal No year): between 0.1% and 0.16% of land value (nationally defined bands). For non-urbanized land a rate of 0.33% does apply. Some municipalities (Bogotá) apply self- valuation on behalf of land owners. 4 Annex 1: Fiscal Decentralization Tax Tax base Tax rate Tax administration Tax proceeds Earmarking Business Tax Municipal Nationally defined bands: between 0.2% to Municipal 100% Municipal No (Impuesto de 0.7% of average monthly gross income Industria y (ingresos brutos) for industrial activities; Comercio) between 0.2% to 1% for commercial activities and services. Financial sector: 0.5%. Registry Tax Departmental Level defined by law. Departmental 100% Departments. 20% for pension debt. (business and In Cundinamarca 80% 80% to be allocated property; Impuesto to the department and freely. de Registro) 20% to the Bogotá District. Source: Ministry of Finance, DNP. 5 Annex 1: Fiscal Decentralization Annex 1B. The Evolution of the Intergovernmental Transfer System A1.7 The evolution of the transfer system. In Colombia, decentralization grew out of the deconcentration of national revenues to subnational administrative units. Starting in 1968 a departmental fund for education and health was financed from a fixed percentage of national revenues, and municipalities were assigned 10 percent of the then-new VAT, which was not earmarked. This was designed to solve the problem of ad hoc transfers to supplement inadequate sources of local revenue. Even after 1968 ad hoc transfers remained a problem, as mayors continued to ask the president for help to meet the cost of their new responsibilities. A major review of the system of intergovernmental transfers hardened the subnational governments' budget constraint vis-à-vis the national government and strengthened their own revenue sources (Bird 1984). A1.8 The 1991 constitution (which also made the office of governor an elected post) and Law 60 of 1993 moderately expanded the amount of revenues assigned to departments by broadening the base of the existing revenue-sharing system (the situado fiscal) to include all recurrent revenues of the government: the value added tax, customs, income tax, and special funds. They mandated a steady increase in the share of these revenues to be transferred to the departments. The share of the situado increased from 22.1 percent in 1993--net of one-time adjustments--to 23 percent in 1994, 23.5 percent in 1995, and 24.5 percent in 1996. Thereafter, the constitution committed the government to increasing the share sufficiently "to permit adequate provision of the services for which it is intended." The sharing formula was to be revised by Congress every five years. A1.9 For municipalities, the 1991 constitution broadened the base of the existing revenue- sharing system from the VAT to all government current revenues and committed the national government to increase the municipal share from 14 percent in 1993 to a minimum of 22 percent by 2002. A1.10 Thus the 1991 Constitution and Law 60 committed the national government to sharing at least 47 percent of all its current revenues with territorial governments and entities by 2002. This not only took resources away from the national government but also meant that any adjustment it tried to make on the taxation side would be substantially shared with subnational governments, weakening any intended fiscal tightening for the economy. Transfers to subnational governments increased while the central government deficit widened during 1993­2002. A1.11 The 1991­94 reforms also reduced the national government's discretion in the distribution of transfers. Prior to Law 60, the situado was paid directly to teachers and health workers under the ministries of education and health. Law 60 changed this system to one in which the situado was transferred directly to each departmental government on the basis of a formula. As a transition measure in 1994­98, each municipality was guaranteed at least the amount of VAT transferred in 1992, in constant prices. These guarantees were applicable to only a limited number of municipalities, and thus did not pose a macro-fiscal problem. A1.12 During the 1990s government spending increased dramatically. Expenditures were driven primarily by the social sectors. In particular, there were significant increases in the size of teacher payroll as local government increased hiring and at the same time, many local governments were unable to pay existing benefits leading to a growing debt on the part of local governments to the workforce. A1.13 Towards the end of the 1990s, the system entered into a crisis as the country entered a severe economic recession and subnational governments did not have enough resources to 6 Annex 1: Fiscal Decentralization serve their debt and even pay the teachers payroll. Law 60 of 1993 generated the wrong incentives and led to an unsustainable fiscal situation. A1.14 The current transfer system. The strategy of the government in 2001 was to the pass a constitutional amendment (Acto Legislativo) that set limits to the growth of total transfers. The 2001 constitutional reform temporarily de-linked transfers from national current revenues and put them on a more sustainable path, and thus limited the growth of central administration outlays (gastos generales).7 The Acto Legislativo did in fact control the growth of fiscal transfers and contributed to a reduction of the central government deficit (refer to Figure 3.1). A1.15 The 2001 reform package also included important measures regarding earmarking of transfers. The reform combined the three main transfers, which had grown and proliferated since the 1991 Constitution, into a single Sistema General de Participaciones. The Law specified the following rules for earmarking: 58.5 percent for education, 24.5 percent for health and 17 percent towards general purpose (but the bulk of this amount is earmarked towards water and sanitation). 1.1 In mid-June 2007, Congress approved a new reform to the transfer system, which can be summarized as follows: · The General System of Transfers (Sistema General de Participaciones, SGP) would grow at 4 percent in real terms during 2008­09, at 3.5 percent in 2010, and at 3 percent during 2011­16 (Box 1.2). After 2016 the increase in participaciones will equal the average growth of the central administration's current revenue during the previous four years. · In addition, subnational governments will receive an extra amount of resources earmarked for education, in particular for coverage and quality. In 2008 they will receive 1.3 percent of the SGP of 2007, in 2009 they will receive 1.3 percent of the SGP of 2008, in 2010 they will receive 1.6 percent of the SGP of 2009, and in 2011 they will receive 1.8 percent of the SGP of 2010, and will continue like that until 2016. · If real GDP growth exceeds 4 percent in any transition year, the SGP would increase proportionately, and these additional resources will be earmarked for infant care. Box 1.2 Transfer Growth and Rules SGP 2008 = SGP 2007 * ( (1+inflation)*(1+0.04)+(0.013)) SGP 2009 = SGP 2008 * ( (1+inflation)*(1+0.04)+(0.013)) SGP 2010 = SGP 2009 * ( (1+inflation)*(1+0.035)+(0.016)) SGP 2011 = SGP 2010 * ( (1+inflation)*(1+0.03)+(0.018)), continued until 2016. 7. As noted in the Introduction, in the transition, the Participaciones would grow at 2 percent in real terms during 2002­05 and at 2.5 percent during 2006­08. In 2009 subnational governments will receive the same amount of transfers they received in 2001--in other words, about 43 percent of current revenues, or about 6.9 percent of GDP. After 2010 the increase in Participaciones will equal the average growth of the central administration's current revenue during the previous four years. 7 Annex 1: Fiscal Decentralization Annex 1C. Royalty Regime Table 1.1C. Royalty rates Natural Resource Rate (% of the value of production) Coal ( extraction of more than 3 mill tons) 10% Coal (extraction of less than 3 mill tons) 5% Nickel 12% Iron and Copper 5% Gold and silver 4% Gold of alluvium 6% Platinum 5% Salt 12% Limestone 1% Radioactive minerals 10% Metallic minerals 5% Non-metallic minerals 3% Construction materials 1% Oil and other hydrocarbons (Less than 5 KBPD, thousands of barrels 8% per day) Oil and other hydrocarbons (Less than 5-125 KBPD, thousands of 8+ (production in KBPD- 5 barrels per day) KBPD)*0.1 % Oil and other hydrocarbons (Less than 125-400 KBPD, thousands of 20% barrels per day Oil and other hydrocarbons (Less than 400-600 KBPD, thousands of 20+(production in KBPD - barrels per day 400KBPD)% Oil and other hydrocarbons (more than 600 KBPD, thousands of barrels 25% per year) Source: Law 756 of 2002. 8 Annex 1: Fiscal Decentralization Table 1.2C. Distribution of Royalties (percent of value of production) Hydrocarbons Hydrocarbons Hydrocarbons Hydrocarbons (production less (production between (production between than 10,000 10-20 thousand 20-50 thousand barrels) barrels) barrels) The extra 10K The extra 30K barrels will be barrels will be distributed by this distributed by this percentage, the first percentage, the first 10K like the column 20K like the column before before. Producer Department 47.5% 52.0% 47.5% 47.5% Producer 12.5% 32.0% 25% 12.5% Municipality/District Port Municipality 8.0% 8.0% 8% 8.0% National Royalties 32.0% 8.0% 19.5% 32.0% Fund Coal (more Coal (less than 3 Nickel Iron and other than 3 million million tons) metallic minerals tons) Producer Department 42% 45.0% 55.0% 50.0% Producer 32% 45.0% 37.0% 40.0% Municipality/District Port Municipality 10.0% 10.0% 1.0% 2.0% National Royalties 16.0% 0.0% 7.0% 8.0% Fund Copper Gold, Silver and Salt Limestone Platinum Producer Department 20.0% 10.0% 20.0% 20.0% Producer 70.0% 87.0% 60.0% 67.0% Municipality/District Port Municipality 2.0% 0.0% 5.0% 3.0% National Royalties 8.0% 3.0% 15.0% 10.0% Fund Radioactive Emeralds and minerals other precious stones (follow specific percentages to municipalities and departments) Producer Department 17.0% Producer 63.0% Municipality/District Port Municipality 5.0% National Royalties 15.0% Fund Source: Law 756 of 2002 9 Annex 1: Fiscal Decentralization Annex 1.D. Fiscal Responsibility A1.16 Over the past decade, Colombia has set up a comprehensive package of fiscal responsibility rules. These rules have been complemented gradually. Several of these rules and adjustment packages do impact sensibly on resource allocation within departments and municipalities. It implies that, in the transition to a more decentralized state, Colombia still has to sacrifice more allocative efficiency for fiscal responsibility. A1.17 In the 1980s borrowing was subject to centralized control. Departments and municipalities were still appointed representatives of the central government and had no political or fiscal autonomy. This changed with decentralization process initiated as of 1986. As a consequence of higher public spending through progressive revenue-sharing, additional transfers to cover increasing costs for the education payroll, and increasing subnational expenditures, the national government experienced increasing fiscal pressures. Between 1987 and 2004, transfers have increased four times as a share of GDP.8 By the end of the 1990s Colombia had already developed into one of the most decentralized unitary countries in the Latin America region, providing nearly 45 percent of current revenue to subnational governments in the form of transfers. This limited the center's efforts at adjustment-- precisely at a time when the national deficit increased above 7 percent GDP in 1999. On the other hand, subnational governments engaged in higher borrowing but which was under- regulated. There was no effective ex-ante control of cash advances from banks, and subnational debt with the banking sector rose from 2.6 percent of GDP in 1991 to 4.6 percent of GDP in 1997. A1.18 Law 358, the so-called Ley de Semáforos (Traffic-light Law) was a response to these challenges. It brought into effect a rating system for territorial governments, based on the ratio of interest to operational savings and debt to current revenues. The rules appeared to be ambiguous first, but were later tightened.9 An additional path breaking reform was Law 549 from 1999 which regulated the pension system for subnational government employees, creating a national pension fund (FONPET) which was capitalized with tax revenue, revenue- sharing, royalties and gains from privatization. Law 550 from 1999, commonly referred to as ley de quiebras (bankruptcy law), endeavors to apply bankruptcy rules from the private sector to the public sector in general and to subnational governments more specifically. The debt restructuring plans it prescribes are intended to secure service delivery in over-indebted subnational governments (Art. 58 of the Law). A1.19 Despite these rules, however, current expenditures (gasto de funcionamiento) of the core administration of departments and municipalities--that is, excluding service delivery areas--kept growing. This process was controlled by Law 617 from 2001 which established limits for current spending10 and classified subnational governments according to their freely disposable revenue, as a more realistic proxy to qualify their capacity to pay and service debt. 8. From 1.6 percent of GDP in 1987 to about 5.8 percent of GDP in 2004. 9. The rating system allowed a fairly high degree of leveraging revenues. Territorial governments could receive a "green" rating even with 40 percent of so-called operational savings allocated to interest payments. Moreover, because some recurrent costs for human capital development are counted as investment and saving, operational savings can be considerably larger than the current account surplus (which is a more conventional measure of public sector saving) (see Dillinger and Webb 1999). 10. Law 617 does cover all subnational government expenditures. It only covers the core bureaucratic apparatus (administración central; gasto de funcionamiento) of departments, which represent 15 percent of departmental expenditures (data calculated based on DNP [2007]), Desempeño Fiscal de los Departamentos y Municipios 2005 y Comparativo 2004). The other share of subnational expenditures--the service delivery areas such as health and education--are covered under the national rules framework, including Law 819 from 2003. 10 Annex 1: Fiscal Decentralization At the same time, the Law provided new credit in order to finance their fiscal adjustment-- which is why these packages can be considered a "conditioned bailout" as the reforms were closely monitored by the central government. Law 819 from 2003 finally established comprehensive budget management and transparency rules; a medium term budget framework; and introduces market incentives for bigger municipalities and departments by making credit risk analysis mandatory prior to borrowing. A1.20 As a combined effect of these measures, there is clearly a trend of improving overall fiscal balances of both departments and municipalities11; as well, the subnational debt stock has decreased from 9.6 percent GDP in 2001 to about 5.4 percent of GDP in 2005 (Figure 1.1B). Figure 1.1B. Overall Fiscal Balance and Debt Stock of Subnational Governments (in percent GDP), 1990­2005 0.6 12 0.4 9.6 10 9.3 Subnational Government Debt Stock (in % 8.9 9.1 9.0 8.6 8.6 8.3 Overall balance departments and 8.3 7.9 7.8 municipalities (in % GDP) 0.2 7.7 8 7.4 7.4 6.0 Total debt stock GDP) 0 5.4 6 Departments Municipalities 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 -0.2 4 -0.4 2 -0.6 0 Ye a rs Source: World Bank, based on data provided by Comptroller Office, Central Bank, and Ministry of Finance. 11. This surplus is in part explained by resources from royalties (for some departments) that are still unspent. Once spent, the balance will look worse. 11 Annex 1: Fiscal Decentralization Table 1.1D. Fiscal Responsibility Laws Law Main Rules Law 358, from Art. 2: Interest payments cannot exceed 40 percent of the "operational savings" (ahorro 1997 operacional, defined as current revenue minus current expenditure and transfers on behalf of subnational government). Art. 4. If the ratio of interest/"operational savings" is more than 40 percent and below 60 percent then the subnational governments can access new credit if the debt stock of the preceding year does not increase more than the Consumer Price Index. If the debt stock exceeds this value, municipalities need explicit authorization from the departmental governor; and departments need authorization from the Ministry of Finance, based on a financing plan which ensures financial sustainability of the department or municipality. Art. 6. If the ratio of interest/"operational savings" is more than 60 percent and the ratio of the debt stock/current revenue is above 80 percent, the subnational government needs explicit authorization to borrow on behalf of the Ministry of Finance. then the subnational governments can access new Art. 9. Fiscal adjustment plans for subnational governments that have a ratio of interests/"operational savings" above 40 percent. Art. 15. For subnational governments with a ratio of interest/"operational savings" above 60 percent and a ratio of the debt stock/current revenue above 80 percent the national government establishes a transition period with the following rules: in the first year, the net debt cannot increase above 60 percent of the variation of the Consumer Price Index; in the second year, net debt cannot increase more than 40 percent of the variation of the Consumer Price Index. Law 549 from Art. 2. Funding for the Pension Fund will be provided through (i) resources from the financial 1999. Creation transaction tax which is part of the situado fiscal transfer; (ii) increases in the revenue-sharing of the Pension transfers to municipalities that become effective as of year 2000; (iii) a maximum of 7 percent Fund for of the Fondo Nacional de Regalías; (iv) 10 percent of national revenue from privatizations; (v) Subnational 10 percent of investment in entities in which the national government is the main shareholder Governments. (capitalización). Art. 15. The national government prohibits direct or indirect financial support to subnational governments which are not in line with the law. Law 550 from Art. 58 (1). The rules regarding financial restructuring applicable to the private sector are also 1999, applicable to subnational governments, both to the "administration" (sector central) as well as "Bankruptcy the "service delivery sectors" (sector descentralizado). Law". Art. 58 (2). The priority for payments is the following: a) Pension contributions. b) "Salaries" (servicios personales). c) "Payroll transfers" (transferencias de nómina). d) "General expenditures" (gastos generales). e) Other transfers. f) Interest payments. g) Amortization of debt. h) Financing of the deficit of previous years. i) Investment. The expenditure priorities can be set for a year or half a year. Art. 58 (9). Once the financial restructuring plan is signed, the subnational government is not allowed to incur other current expenditures than what is defined in the plan. Law 617 from Art. 2. Classification of subnational governments (departments; municipalities and districts) 2000. according to population and freely disposable current revenue. Art. 3. Expenditures for operation of the subnational government (gastos de funcionamiento) needs to be financed by current revenue that is freely disposable. Note: current revenue excludes the situado fiscal transfer; revenue-sharing; co-financing funds; royalties; gasoline tax; other transfers that are ear-marked, and others. Art. 3 (2). Expenditures for salaries of teachers and health personnel which are financed out of freely disposable revenue of subnational governments, and which constitute commitments for more than one year, can only be financed with freely disposable current revenue. 12 Annex 1: Fiscal Decentralization Law Main Rules Art. 4. For departments, the "operational expenditures" (gastos de funcionamiento) cannot exceed the following shares of freely disposable current revenue: · Departments in category "Especial": 50 percent. · Departments in category "Primera": 55 percent. · Departments in category "Segunda": 60 percent. · Departments in category "Tercera y Cuarta": 70 percent. Art. 6. For municipalities and districts, the "operational expenditures" (gastos de funcionamiento) cannot exceed the following shares of freely disposable current revenue: · Municipalities in category "Especial": 50 percent. · Municipalities in category "Primera": 65 percent. · Municipalities in category "Segunda y Tercera": 70 percent. · Municipalities in category "Cuarta, Quinta y Sexta": 80 percent. Art. 8. For departments, the expenditures for the Departmental Assembly and Comptroller Office cannot exceed the following shares of freely disposable revenue: · Departments in category "Especial": 1.2 percent. · Departments in category "Primera": 2.0 percent. · Departments in category "Segunda": 2.5 percent. · Departments in category "Tercera y Cuarta": 3.0 percent. Art. 10. For municipalities, the expenditures for Municipal Councils, Personerías, and Comptroller Offices cannot exceed the following shares of freely disposable revenue: · Municipalities in category "Especial": 1.6 percent. · Municipalities in category "Primera": 1.7 percent. · Municipalities in category "Segunda": 2.2 percent. · Municipalities in category "Tercera": 350 Minimum Salaries. · Municipalities in category "Cuarta": 280 Minimum Salaries. · Municipalities in category "Quinta": 190 Minimum Salaries. · Municipalities in category "Sexta": 150 percent. Limits for the Comptroller Office: · Municipalities in category "Especial": 2.8 percent. · Municipalities in category "Primera": 2.5 percent. · Municipalities in category "Segunda": 2.8 percent. Art. 12. Ear-marked revenue which is un-committed needs to be used to finance the fiscal adjustment plans. From this rule is excepted the ear-marking defined in the Constitution and Law 60 (from 1993). Art. 14. Subnational governments are prohibited to transfer resources to liquor enterprises, lotteries, and health service enterprises different to those established by law. Public enterprises which generate deficits for three consecutive years need to be closed down. Art. 20 and 26. Subnational governments which exceed the limits set forth in this law are required to execute a fiscal adjustment plan with precise performance targets Art. 62. The national government guarantees up to 100 percent of new credit that is contracted to finance the fiscal adjustment plans Art. 65. The national government sets up a fund for contingent payments to guarantee subnational debt. Art. 80. The national government is prohibited to provide direct or indirect financial assistance to subnational governments which are not fulfill the criteria set forth in this law. As well, they do not qualify for co-financing and new credit. Law 819. Art. 2. Departments and municipalities in category "especial", "one" and "two" are required to have a "primary surplus" that ensures its debt sustainability according to Law 358 (1997). Note: primary surplus in Colombia is defined as: "Se entiende por superávit primario aquel valor positivo que resulta de la diferencia entre la suma de los ingresos corrientes y los recursos de capital, diferentes a desembolsos de crédito, privatizaciones, capitalizaciones, utilidades del Banco de la República (para el caso de la Nación), y la suma de los gastos de funcionamiento, inversión y gastos de operación comercial." Art. 5. Departments, districts, and municipalities are required to have a Medium-Term Fiscal Framework which is of indicative nature. The Framework needs to contain: (i) financing plan; 13 Annex 1: Fiscal Decentralization Law Main Rules (ii) targets for primary surplus; (iii) actions that are required to reach the targets; (iv) report of previous budget execution; (v) estimate of fiscal cost regarding exemptions; (vi) an estimate of the debt that can be effectively collected and the contingent liabilities; (vii) the fiscal cost of ordenanzas and acuerdos of the previous budget year. 14 Annex 1: Fiscal Decentralization Table 1.2D. Subnational Governments subject to Law 550 (from 1999) Departmental Governments and Department Guarantee Process Municipalities 1 DEPTO DEL AMAZONAS ley 550 AMAZONAS 2 LETICIA R Ley 550 3 CAUCASIA * R Ley 550 ANTIOQUIA 4 ZARAGOZA Ley 550 5 BARRANQUILLA R Ley 550 ATLÁNTICO 6 MALAMBO Ley 550 7 ACHÍ Ley 550 8 BOLÍVAR DEPTO DE BOLIVAR R Ley 550 9 MAGANGUE R Ley 550 10 CALDAS SALAMINA R Ley 550 11 CAQUETÁ BELÉN DE LOS ANDAQUIES* R Ley 550 12 DEPTO DEL CAUCA R Ley 550 13 GUAPI Ley 550 CAUCA 14 PATIA DEL BORDO Ley 550 15 POPAYÁN R Ley 550 16 ASTREA ley 550 17 BECERRIL* ley 550 CESAR 18 CODAZZI ley 550 19 LA PAZ R Ley 550 20 BAHÍA SOLANO Ley 550 21 CONDOTO Ley 550 22 DPTO DEL CHOCO Ley 550 23 CHOCO ITSMINA R Ley 550 24 NOVITA Ley 550 25 NUQUI Ley 550 26 TADO ** R Ley 550 27 CERETE Ley 550 CÓRDOBA 28 MONTERÍA Ley 550 29 DPTO DEL GUAINIA Ley 550 GUAINIA 30 INIRIDA Ley 550 31 DISTRACCIÓN Ley 550 32 EL MOLINO Ley 550 33 FONSECA R Ley 550 GUAJIRA 34 MAICAO R Ley 550 35 RIOHACHA R Ley 550 36 VILLANUEVA R Ley 550 37 ARACATACA R Ley 550 38 DPTO DEL MAGDALENA R Ley 550 39 EL RETEN ley 550 40 MAGDALENA FUNDACIÓN R Ley 550 41 PLATO** R ley 550 42 SANTA ANA Ley 550 43 SANTA MARTA Ley 550 44 DPTO DE NARIÑO R Ley 550 45 NARIÑO SANDONÁ Ley 550 46 TUMACO Ley 550 47 NORTE DE SANTANDER PAMPLONA Ley 550 48 PUTUMAYO PUERTO ASÍS R Ley 550 49 QUINDÍO CÓRDOBA* R Ley 550 15 Annex 1: Fiscal Decentralization Departmental Governments and Department Guarantee Process Municipalities 50 DPTO DE SAN ANDRES R Ley 550 SAN ANDRÉS 51 PROVIDENCIA R Ley 550 52 COROZAL** R Ley 550 53 GALERAS Ley 550 54 GUARANDA ley 550 55 SUCRE MAJAGUAL Ley 550 56 SAN BENITO ABAD Ley 550 57 TOLÚ VIEJO R ley 550 58 TOLÚ Ley 550 59 AMBALEMA Ley 550 60 DPTO DEL TOLIMA ley 550 61 LÉRIDA R Ley 550 TOLIMA 62 LÍBANO R Ley 550 63 MARIQUITA R ley 550 64 ROVIRA Ley 550 UNIVERSIDAD DEL 65 ATLANTICO Ley 550 66 BUENAVENTURA ley 550 67 CARTAGO R Ley 550 68 VALLE DAGUA R Ley 550 69 PALMIRA R Ley 550 70 PRADERA Ley 550 71 VICHADA DPTO DEL VICHADA ley 550 Symbols: R: With a guarantee of the national government for restructured debt and for fiscal adjustment credits. * Agreements which have finalized "successfully." **Agreements which have failed due "incompliance." Source: DAF, Ministry of Finance (2007). 16 Annex 1: Fiscal Decentralization Annex 1E. Subnational Expenditures A1.21 Figure 1.1E shows departmental expenditures. As expected the largest categories are those financed by earmarked resources such as education, health and water supply and sanitation. By contrast sectors financed by non-earmarked resources such as roads are relatively small. It can be seen that those departments that have a larger share of non- earmarked resources invest more on roads. This points to the fact that in practice departments do not combine revenue from the different sources to a high degree. Own revenue is rarely used to finance SGP-funded expenditures. Figure 1.1E. Department's Expenditures Per Capita, 2005 3,500 Current Expenditures Education Health School Lunches Water and Sanitation Roads Fiscal Adjustment Program Other sectors 3,000 2,500 COP per capita 2,000 1,500 1,000 500 0 PUTUMAYO BOYACA TOLIMA META MAGDALENA AMAZONAS CUNDINAMARCA QUINDIO NATIONAL CHOCO NARIÐO ATLANTICO GUAJIRA GUAINIA BOLIVAR CESAR CAQUETA SAN ANDRES HUILA ANTIOQUIA SANTANDER SUCRE VAUPES CAUCA GUAVIARE CASANARE CALDAS RISARALDA ARAUCA NORTE DE SANTANDER VALLE DEL CAUCA Source: Bank Staff calculations based on DNP and DANE data. The departments are ranked according to the unsatisfied necessities index (NBI), and the poorest department is the one in the left-hand side. A1.22 On average, most of the municipalities spent their resources in education and health (Figure 1.2E.). These are SGP-financed sectors. Municipalities are in charge of a large number of responsibilities and since they collect very little tax revenue, they do not finance these non-earmarked responsibilities sufficiently. The one sector that most of the municipalities seem to be putting out of pocket money is water and sanitation. 17 Annex 1: Fiscal Decentralization Figure 1.2E. Municipalities Expenditures Per Capita, 2005 Current Expenditure School Lunches Education Health Transport Water and Sanitation Other Sectors 1,400 1,200 COP Thousands per capita 1,000 800 600 400 200 0 QUINDIO CHOCO NARIÐO PUTUMAYO BOYACA BOLIVAR CESAR VICHADA GUAJIRA GUAINIA AMAZONAS CORDOBA SUCRE VAUPES GUAVIARE MAGDALENA CAUCA CASANARE ARAUCA HUILA CUNDINAMARCA CALDAS RISARALDA VALLE DEL CAUCA SANTANDER BOGOTA CAQUETA TOLIMA META ATLANTICO ANTIOQUIA NORTE DE SANTANDER NATIONAL SAN ANDRES Source: DNP. The departments are ranked according the unsatisfied necessities index (NBI), and the poorest department is the one in the left-hand side Figure 1.3E. Department's Public Resources Per Capita, 2005 3,000 2,500 COP Thousand per capita 2,000 1,500 1,000 500 0 SAN ANDRES GUAVIARE CESAR BOLIVAR AMAZONAS VAUPES CALDAS Average RISARALDA VICHADA GUAJIRA GUAINIA MAGDALENA SUCRE CASANARE VALLE SANTANDER CORDOBA CAUCA NARIÑO ARAUCA HUILA CHOCO BOYACA CUNDINAMARCA QUINDIO N. DE CAQUETA TOLIMA META ANTIOQUIA ATLANTICO PUTUMAYO Ow n Revenues SGP Cof inanciacion Transf ers Royalties A verage Source: Bank staff calculations based upon DNP and DANE data. The departments are rank according the unsatisfied necessities index (NBI) and the poorest department is the one in the left hand side 18 Annex 1: Fiscal Decentralization Figure 1.4E. Municipalities' Public Resources Per Capita, 2005 3,000 COP Thousand per capita 2,500 2,000 1,500 1,000 500 0 Cundinam Putumayo Tolima La Guajira Guainia Magdalen Cauca Narino Caqueta Guaviare Casanare Norte de Meta Risaralda Quindio Choco Huila Cordoba Cesar Caldas Bogota Sucre Amazona San Boyaca Bolivar Santander Valle del Arauca Average Vichada Atlantico Antioquia Vaupes Ow n revenues SGP Transfers Royalties Other transfers Total Average Source: Bank Staff calculations based upon DNP and DANE data. The departments are rank according the unsatisfied necessities index (NBI) and the poorest department is the one in the left hand side Figure 1.5E. Distribution of SGP for General Purpose Per Capita, 2005 400 Unearmarked Water Sports Culture Earmarked for investment FONPET 350 300 COP Thousands per capita 250 200 150 100 50 0 Boyaca Guaviare Putumayo Bolivar National Choco Guainina Nariño Vichada Cordoba Sucre Magdalena Cauca Cesar Caqueta Casanare Arauca Huila Norte de Santander Meta Atlantico Santander Risaralda Quindio Valle de Cauca Bogota Antioquia Vaupes Amazonas San Andres Caldas Guajira Cundinamarca Tolima Source: Bank Staff calculations based upon DNP and DANE data. The departments are rank according the unsatisfied necessities index (NBI) and the poorest department is the one in the left hand side 19 COP Thousands per capita COP Thousands per capita 0 100 200 300 400 500 600 0 100 200 300 400 500 600 700 CHOCO CHOCO GUAJIRA GUAJIRA GUAINIA GUAINIA Source: World Bank. Source: World Bank. SUCRE SUCRE VAUPES VAUPES MAGDALENA MAGDALENA BOLIVAR BOLIVAR CAUCA CAUCA CESAR CESAR AMAZONAS AMAZONAS NARIÐO NARIÐO CAQUETA CAQUETA SAN ANDRES SAN ANDRES GUAVIARE GUAVIARE CASANARE CASANARE 20 ARAUCA ARAUCA PUTUMAYO PUTUMAYO HUILA HUILA BOYACA BOYACA NORTE DE SANTANDER NORTE DE SANTANDER TOLIMA TOLIMA META META ATLANTICO ATLANTICO ANTIOQUIA ANTIOQUIA SANTANDER SANTANDER CUNDINAMARCA CUNDINAMARCA CALDAS CALDAS RISARALDA RISARALDA QUINDIO QUINDIO VALLE DEL CAUCA VALLE DEL CAUCA Figure 1.7E. Departmental Government Per Capita Expenditures in Health, 2005 Figure 1.6E. Departmental Government Per Capita Expenditures in Education, 2005 NATIONAL NATIONAL Annex 1: Fiscal Decentralization Annex 1: Fiscal Decentralization Table 1.1E. Literature Review on Subnational Response to Intergovernmental Fiscal Transfers Reference Methodology Variables Conclusions Policy Reform options to address the issue Related to the Fiscal efforts Related to the flypaper effect hypothesis Slack and Bird Econometric analysis: Tax allowance equation (department's Increased transfers to the Design program to avoid (1983) Estimation procedures used is share of total population; income of departments have resulted in disincentive to local fiscal Ordinary Least Squares (OLS). residents of the department, number of relatively lower departmental effort as to increase total The model was estimated in students enrolled in primary education taxes: "substitution" effect (national plus local) linear form, pooled cross- and mortality rate) and expenditure expenditure on the aided section data for 22 departments allowance equation (income of residents, services by at least the amount and the national territories for tax allowance, other grants, other of the grant the years 1974 to 1977 revenues and five socioeconomic inclusive yielding 92 variables) observations in total. Sanchez Torres, Econometric analysis: Variables: Tax allowance, regional GDP, No clear relationship between The mechanisms and Gutierrez (1994) Elasticity model for 23 non taxes revenue, transfers, current transfers and fiscal efforts: conditions of the transfers must departments for the period expenditures, investment per year and per hypothesis of reduced fiscal be clearly defined to avoid Torres, Sourdis, 1980-1991 (Log-linear model) department using log specifications efforts is not supported inefficiency (discretion and Parra and simultaneous equations empirically discontinuity) (1995) model World Bank Statistical survey: Reclassified data base on municipal Local governments have Evidence support the flypaper Uselessness of the concept of (1996) Departmental and municipal revenues and expenditures, distribution generally expanded their own- effect: transfers tend to `fiscal effort' in the design of data base from 1988 to 1994 of royalties, valorization levy by source current revenues at the increase total expenditures by intergovernmental transfers municipio, property taxes, co financing same rate as their current about the amount of the transfers, departmental revenues and expenditures with the transfers because local taxes expenditures, Situado fiscal transfers increased resources made are not reduced available by larger national transfers Chaparro, Smart, Econometric analysis: Panel Variables are real own-source tax Under OLS method: small Using OLS and 2SLS transfer Make transfers more dependent Zapata (Chap. 9, data to examine finances of revenues per capita and real transfers positive effect of per capita increases are fully passed on measures of local 2005) individual municipalities under the PM per capita (transfer to transfers on per capita local through to spending and any performance and thus possibly during the period 1985-1999 municipalities), Specifications include tax revenues: transfers have reduction in tax effort is to improve fiscal incentives at using OLS and two-stage least fixed effects for municipality and year done nothing to discourage compensated for by an the local level squares (2SLS) the fiscal effort of increase in municipal Transfer formulas could be municipalities borrowing adjusted to make them simpler, Under 2SLS: negative effects: more transparent and better transfers discourage tax effort targeted to redressing differences in fiscal capacity and compensating differences in fiscal needs 21 Annex 1: Fiscal Decentralization Reference Methodology Variables Conclusions Policy Reform options to address the issue Related to the Fiscal efforts Related to the flypaper effect hypothesis Acosta and Bird Qualitative analysis of the Referring to Chap.9 great Simple generalizations are of (Chap.8, 2005) diverse problems at the expansion of transfers to little use as a guide to policy municipal level municipalities since 1993 had some adverse effects. However, differences among municipalities are great Wiesner Duran Statistical analysis: Data on transfers to sub national levels Increase of expenditures at Careful attention to local (2002) National, departmental and decentralized level is resource mobilization and municipal expenditures associated with an increase of incentives as an essential transfers at the sub national component of any successful level decentralization exercise Melo (2002) Econometric analysis: Log-linear and linear specifications are Flypaper effect is highly Analysis of the effects of used. Variables are level of expenditure, sensitive to differences in transfers on the sub-national regional income, intergovernmental fiscal structure and in public expenditure, panel data transfers, socioeconomic and institutional autonomy to collect taxes and models using three different variables to manage fiscal resources and estimators of each model: to the specification of the pooled model within groups expenditure equations. (WG) and the two steps generalized least squares (GLS) DNP (2002) Econometric analysis: Log Variables: total expenditure, current Hypothesis of `pereza fiscal' An increase of tax allowance linear model for the analysis of expenditures, tax allowance per capita is not verified empirically is associated with an increase the elasticity of expenditure in total municipal expenditure with regard to transfers and tax allowance Source: World Bank. 22 Annex 2: Health Sector ANNEX 2: HEALTH SECTOR Annex 2A. Health Reforms and Decentralization 1. The First Phase of Health Decentralization: 1990-2000 1.1. The Reforms A2.1 During the early 1990's Colombia's health system12 provided very poor services: there was very limited access even to basic health services; large parts of the population had large out-of-pocket expenses; and services were generally of poor quality (Escobar 2005). Seeking to address these problems, Colombia gradually reformed the administration, financing, organization and delivery of health services. One key piece of the reform was the creation of the National Social Health Insurance; another was the decentralization of responsibilities to subnational governments. A2.2 Following a new constitutional framework adopted in 1991, Law 60 of 1993 attempted to reorganize responsibilities13 as well as resource flow among levels of government and the different levels of care (Table 2.1A).14 The Law increased the amount of national revenues allocated to health and education transferred to departments and municipalities through a grant denominated Situado Fiscal and added another fraction of national income to be transferred to municipalities (Participaciones Municipales) out of which a percentage was earmarked to health. The decentralization laws established administrative procedures to certify local governments; if requirements were met, authority, responsibility and budgetary control of these resources were shifted to departmental and municipal levels.15 However, tight ear-marking over resources remained one of the key limiting factors for local decision-making autonomy (refer to Table 2.2A). 12. Prior to 1993 Colombia's National Health System consisted of three tiers: public provision of health care for the poor, a mandated social insurance plan for workers of the formal sector, and private insurance and provision of health services for those able to pay out of pocket. Overall health insurance coverage reached about 8 million people, or 24% of the population. 13. The central government´s main role referred to policy design, regulation and financing; departmental governments assume regional planning, management and finance, articulation of policy between local and national levels; provision of second and third level health care services; municipal governments take on policy implementation and provision of primary health care services. 14. The first level of complexity includes preventive and emergency care, basic medical, dental and diagnostic services. The second and third levels include specialized and rehabilitation care, hospitalizations and their corresponding diagnostic tests. 15. Law 10 1990, Art. 37; Law 60 1993 Arts. 14 and 16; Decreto 1770 de 1994. 23 Annex 2: Health Sector Table 2.1A. Comparison of Health Public Sector Competences and Responsibilities Before and After Decentralization Reforms Pre-decentralization Post-decentralization Decision level Decision level Function Nat Dept Mun Hosp Nat Dept Mun Hosp Financing +++ --- --- --- ++ ++ ++ + Stewardship, control and oversight +++ --- --- --- +++ + + --- Appointment of hospital directors +++ --- --- --- --- ++ ++ Planning and investment +++ --- --- --- --- ++ ++ ++ Salary rates +++ --- --- --- +++ --- --- --- Hiring and firing +++ --- --- --- --- + (*) + (*) + (*) Purchase of supplies +++ --- --- --- + ++ ++ +++ Symbols: Nat (National); Dept (Department); Mun (Municipal); Hosp(Public Hospital): +++ Full competence and responsibility ++ Shared competence and responsibility + Minimal competence and responsibility --- No competence and responsibility (*) Restricted by labor laws and union agreements Source: Londoño, Jaramillo and Uribe (2001). Table 2.2A. Amount, Destination and Purpose of National Health Transfers National Total amount Amount for health Purpose transfer Destination Situado fiscal 23% of national 25% plus 50% 45% supply subsidies 2- revenue in 1994 discretionary Department 3rd level services to 25.5% 1996 amount 5% public health 50% Certified Demand subsidies Municipalities Participación 15% National 25% plus Municipality 60% demand subsidies municipal revenue in 1994 discretionary (of this a maximum 30% to 22% in 2002 amount could be allocated to operational expenses) 40% supply subsidies for 1st and public health (5% minimum) Source: Law 60, 1993. A2.3 Although decentralization was meant to address poor service delivery, none of these measures removed financial barriers to access, improved targeting of services towards the neediest, nor rewarded providers for better performance. These issues were addressed through the health sector reform process embodied in Law 100 in 1993. This reform transformed the previous system of the National Social Health Insurance and designed two separate schemes: the Contributory (CR) and the Subsidized Regimes (SR). Both had different financing sources, target populations, and benefits.16 Both regimes are organized as a two-tiered managed competition model which is expected to improve access, contain costs, improve efficiency and upgrade quality by restructuring incentives. 16. The reason to design national insurance as two separate regimes was to allow for resources to become available to gradually increase insurance coverage for the population eligible for subsidies, with accompanying additions to the benefits package. Both regimes were expected to become equal in terms of benefits coverage in 2000, so that a single universal health insurance system could be implemented throughout the nation, but resource constraints have been an obstacle for fulfillment of this goal. 24 Annex 2: Health Sector · In the first tier, consumers have freedom of choice to enroll among a set of public or private health insurance carriers17 which offer the services included in a standard benefits packages in return for a risk adjusted premium that has a value fixed at the national level. At this level insurers are expected to engage in quality competition to attract enrollees. · In the second tier, health plans act as group purchasers for their enrollees by arranging a network of providers which they select based on the best price and quality. A2.4 In this model, the key role of the government, represented by the Ministry of Social Protection or MPS (formerly the Ministry of Health) and its affiliated regulatory bodies, is to provide information and formulate, monitor and enforce regulations directed to minimize market failures and to devise mechanisms to guarantee equitable access to healthcare services. A2.5 With the introduction of Law 100 the responsibility of delivery of personal health care services18 for the insured lies on EPS and ARS through their provider network. Departments and Municipalities maintained their service provision responsibilities, but limited to delivery of services for the uninsured or of services not covered by the benefits packages, and were also explicitly assigned the provision of a set of public health services19 denominated the Plan de Atencion Básica (PAB). Public health providers were mandated to convert to semi-public entities referred to as Empresas Sociales del Estado (ESEs), which granted them the financial and managerial autonomy necessary to prepare for competition with the private sector under the new health insurance scheme and to substitute supply side resources from national transfers with sale of services20. To finance health insurance coverage Law 100 introduced new sources of funding for each regime.21 A2.6 According to Law 100 universal health insurance coverage was to be reached by the year 2000, contingent on increased enrollment of population into the CR and transformation of supply side subsidies into demand subsidies. However the incentive framework resulting from the different laws posed difficulties for fulfillment of this goal. 17. Denominated Empresas Promotoras de Salud (EPS) and Administradoras del Régimen Subsidiado (ARS). 18. These are services consumed directly by individuals, which can be preventive, diagnostic, therapeutic or rehabilitative interventions included in the benefits packages. 19. hese comprise non-personal health services, or actions applied to collectivities or to the non-human components of the environment, such as health promotion and preventive activities, public health surveillance and risk control. Priority areas of intervention and actions and the respective responsibility of local governments were initially defined by Decretos 1938 y 1891 in 1994, and further specified by resolución 4288 de 1996. 20. The legal basis for institutional decentralization of public hospital facilities was set by the decentralization laws, through provisions mandating separation of public hospitals from administrative dependency on departments. 21. Funds for the CR were obtained by mandating wage contributions for all those with capacity to pay, and also, by increasing the amount of the contribution as percent of salary from 11 to 12%. The system's revenues were pooled in a central administrating fund called the Fondo de Seguridad Social y Garantía-FOSYGA (Social Security and Guarantees Fund) which in turn distributes these among sub-funds for different purposes. Sources of financing of the SR included resources from the national transfers mandated by Law 60, plus the following: (i) The solidarity point from the CR which was to be matched one to one by resources from the national budget; (ii) Contributions from Cajas (benefit funds); (iii) Financial returns of FOSYGA surpluses; (iv) Other miscellaneous sources such as expected revenues from taxes on oil exploitation. 25 Annex 2: Health Sector 1.2. Incentive Framework A2.7 Law 10 and Law 60 introduced levels of authority, responsibility and budgetary control and funding mechanisms so that delivery of care was based on a public network of hospitals funded by supply subsidies. On the other hand Law 100 radically changed the structure of the previous National Health System by establishing universal insurance based on an open market for public and private health insurers and providers; as well, it shifted the focus of public finance towards the provision of demand subsidies that targeted individuals rather than public providers. These different reforms however resulted in inconsistencies and contradictory incentives for local governments and service providers (Annex Box 1). Box 2.1A Incentive Problems Created through Law 60 On one hand a Municipality acted as financing agent of the SR, which required re-directing national transfers and own resources towards insurance of the poor population, but on the other hand became a provider of health services through its own public facilities, financed by the same resources. Local governments faced trade-offs between increasing insurance coverage and the opposition by public providers1 to transformation of subsidies and the policies to restructure the public network which was sensed by these as a direct threat to their survival. Source: World Bank. A2.8 The balance of forces between these opposing incentives seems to have tilted towards the supply side. This was the result of several factors: lack of clarity at the national level regarding how the transformation of supply subsidies was to be implemented; policies seeking to protect public hospitals (such as government bail out of bankruptcies); and problems in the allocation formulas for national transfers, which privileged institutional expenses versus expenditures in services to fulfill population needs. A2.9 Although important progress was made in terms of coverage by the SR reaching 9.5 million poor (22 percent of the total population) by the year 2000. However the goal of reaching universal coverage by this time was not achieved. 1.3. Lessons Learned of the First Phase A2.10 The progress and drawbacks of the past health sector decentralization and reform process have been a subject of numerous policy analyses.22 Key lessons drawn from these are: · The delegation of responsibilities and budgetary control to subnational governments needs to be more carefully matched to the potential effectiveness of the decision level and to institutional capacity to carry out these functions. Levels of decisions also need to not to overlap to avoid duplication of efforts and to promote accountability. · Minimum institutional capacities need to be in place. At the subnational government level, certification processes lacked technical assistance and the 22. Among others: Vargas and Sarmiento 1997; Sánchez et.al 1998; Sánchez and Yepes 1999; Herrera and Cortez 2000; Di Tella 2001; Santamaria 2001;Londoño, Jaramillo and Uribe 2001;Baron and Meisel 2003; Pinto 2003; MPS 2004. 26 Annex 2: Health Sector adequate financial and human resource backup to become true administrative and technical developments rather than formalities on paper. · A model designed for market competition needs a public sector that adjusts flexibly to new incentives. In Colombia, structural limitations towards greater efficiency, such as inflexible labor costs were not addressed (Box x) . · Earmarking without regional differentiation regarding helath sector needs can become a barrier to the efficient use of resources: this may lead to excessive spending in some items and deficits in others. Local autonomy is restricted and incentives are generated to comply with expenditure targets independent of benefits.23 · Allocation of resources to subnational governments needs to be linked to fulfilment of goals in terms of service coverage or good management of public resources, supported by information systems and performance indicators. Box 2.2A The Incomplete Hospital Reform in Colombia as an Impediment to Higher Allocative Efficiency Competition has not raised the efficiency of many public hospitals, which continue to operate with very low occupancy rates, while receiving hefty transfers. Hospitals have not adjusted their cost structures. Labor costs increased 40 percent from 1995, leading to increased requests for transfers. In 1998, 68 percent of ESE revenues were used to pay salaries (Escobar and Panopoulou 2003). Additional resources flowed from the government. Large payroll and wages increases took place by between 1995 and 1998 (Escobar and Panopoulou 2003). The share of personnel expenditure as a share of total expenditure increased from 40 percent to 66percent from 1994 to 2001 (but part of that was not bad). Hospitals in the first level increased personnel expenditures by 256 percent, second level by 177 percent, and third level by 103 percent (CONPES). A key explanation is that the soft-budget constraint remains. Hospitals are not wholly dependent on the market for their revenues. Source: World Bank. 2. Second Phase of Reform: Law 715 (2001) and Law 1122 (2007) A2.11 Concerns about the slow progress towards fulfillment of the expected results from decentralization and about potential fiscal destabilization led to enactment of Ley 715 in 2001 which sought to correct the weaknesses of past policies24. The overall goals of Law 715 were to clarify and distinguish functions and responsibilities at each level of government and to allocate national transfers more equitably and in accordance to verifiable criteria of performance. A2.12 Recently, changes to Law 100 which are to be implemented during 2007 were introduced by Law 1122 of 2007 (Table 2.3A). One of the major changes of this Law is the creation of a Regulatory Commission which will take over many of the functions of the National Social Security Council. Other changes that need to be highlighted due to their implications for universal health insurance coverage and efficiency are summarized below: 23. This was the case of expenditures in public health activities between 1996 and 2001, which seem to have barely met the minimum established by the law, and exposed the population to unnecessary risks. 24. The National Government had previously introduced a Constitutional Reform in 2001 (Acto Legislativo 01 de 2001) which reset the basis and growth conditions of total transfers to Departments and Municipalities until the year 2008, seeking to isolate these from macroeconomic fluctuations 27 Annex 2: Health Sector Table 2.3A. Recent Changes to the Health Financing and Service Delivery, 2007 Changes in · Sets ceiling for allocation of SGP to demand subsidies (65 percent of total Financing SGP) · Sets minimum percentage of rentas cedidas to be allocated by departments to demand subsidies (25 percent) · Increases the solidarity contribution to the SR by 0.5 percent, for a total of 1.5 percent · Allocates a minimum percentage of revenues received by territories from taxes to oil exploitation in Cupiagua y Cusiana (15 percent) · Allocates 0.02 percent of total SR municipal and district resources towards financing of external auditing and 0.04 percent towards financing of oversight activities by the Superintendencia Nacional de Salud Changes in · Restricts the amount of contracts of health plans with their own health care the providers to 30 percent of total health expenditures organization · Requires ARS to contract with the public network 60 percent of UPC of delivery · Requires ARS to privilege the public network in contracts of health promotion of services and prevention activities Source: World Bank. Figure 2.1A. Change in Distribution of National Transfers to Departments and Municipalities, 2001­04 M n of p illio s eso 2004 s Source: DNP-DDT (2005). 28 Annex 2: Health Sector Annex 2.B. Human Resources Management in the Health Sector A2.13 A key premise of Colombia's healthcare reform is that public resources can serve a larger share of the population, and be delivered with higher quality, if government shifts from financing the supply of services by healthcare institutions to instead subsidize citizen demand through expanded health insurance plans.25 In keeping with this view, Law No. 100 (1993) directed that public hospitals be transformed and that could compete on equal terms with private hospitals to sell healthcare services. These new hospitals, dubbed Social State Enterprises (ESE), operate in a marketplace regulated by the state, but are no longer to rely upon traditional budget transfers to ensure their operations. A2.14 In the marketplace, hospitals compete mainly on two dimensions: quality and cost. Healthcare personnel are the biggest cost driver for a hospital, and the principal input to service quality. Therefore, if ESE hospitals are to compete on these two dimensions, hospital autonomy must include local control over human resources management (Table 2.1B). How and to what extent has responsibility for human resources been decentralized in Colombia as part of the healthcare reform? What is the impact, to date, of these changes in authority at the level of the hospital? And finally, in light of these experiences, what additional reforms or implementation measures are presently unfolding? Table 2.1B. Health Resource Management Functions of Subnational Governments Departmental Health Only a few DHS receive SGP resources for insurance purposes for which they assume Secretariats the competences of municipalities.26 Otherwise, DHS manages funds for provision of second- and third-level individual health services for the uninsured or services not covered by the POS-S (59 percent of supply SGP), and funds for the provision of public health activities for the population in general (45 percent of public health SGP), according to departmental responsibilities. Contracts for provision of these services are to be established preferentially with public providers, although shortages of public supply allow participation of private providers. Municipal Health MHS manages the totality of funds for the provision insurance. It establishes contracts Secretariats (non- with insurers at the local level for provision of the POS-s to the eligible population. In certified) turn, insurers establish a network of providers for provision of the services covered by the POS-s. As mentioned, Law 1122 requires insurers of the RS to provide a minimum of 60 percent of the total value of the insurance contract through the public network. Health Secretariats The HS of certified Municipalitiesa are allocated 41 percent of supply SGP earmarked of certified for first-level individual health services for the uninsured, and funds for the provision Municipalities of public health activities for the population in general, within municipal responsibilities. In the case of non-certified municipalities, the DHS manages the funds. Again, by law, preference must be given to contracts with public providers. Note: a. Funds for this purpose are managed by Departments in the case of non-certified municipalities. Source: World Bank. 25. According to Law 100 (1993) workers in the formal sector and independent employees who are able to contribute to their own health insurance belong to the contributory regime (Régimen Contributivo). Their contributions are collected by the insurer of choice. Meanwhile, those who are too poor to make an insurance contribution are covered by the subsidized regime (Régimen Subsidiado). There is also a third group: individuals who are eligible for affiliation to the RS, but still uninsured. These are known as vinculados, and must rely on public hospitals for care. 26. Explanation will be provided 29 Annex 2: Health Sector 1. The Shift Towards Hospital-Based Personnel Management A2.15 Under Colombia's former National Health System all hospitals were centrally administered by the national government and Departmental Health Directorates. This meant that officials at the Department level were responsible for key human resources policy decisions, such as determining the appropriate number and balance of health professionals for each facility. They were also responsible for day-to-day operations, such as approving employee requests for leave and executing the payroll. In the Department of Antioquia, for example, the Directorate of Health was responsible for dispatching an ambulance twice each month to effect salary payments for hospital staff throughout the Department. A2.16 The changes following Law 100 have been profound. Between 1994 and 1997 the vast majority of Colombia's hospitals were transformed into ESEs.27 While legally financially autonomous, each ESE reports to a corresponding national, departmental, district, or municipal directorate of health. The vast majority are municipal ESEs. For instance, there are 136 hospitals physically located in the Department of Antioquia. Seven of these are specialized Departmental Hospitals, while four operate in "non-certified" municipalities and are thus overseen by the Departmental government rather than the municipality. Meanwhile, by August 2001, Antioquia's remaining 121 municipalities were certified to assume oversight responsibility for the remaining ESEs within their respective territories.28 A2.17 A core building block of Colombia's healthcare reform is the configuration of new municipal- and hospital-based management arrangements. Decree No. 1876 (1994) and No. 139 (1996) vest considerable authority for human resources management in the Hospital Director and Management Committee (Junta Directiva) of each newly-established ESE. It is important, then, to describe how these organizations are structured, and how persons are selected to these positions. A2.18 The Hospital Management Committee has six (or nine) members, with equal membership for three groups: i) political-administrative authorities, ii) professional-scientific members, and iii) community and business groups of the municipality. This tripartite structure is meant to ensure broad-based community representation in hospital management. A2.19 The two political-administrative members (for the common six-member Management Committee of a municipal ESE) are the Mayor or his delegate, and the Municipal Director of Health.29 The professional-scientific members consist of one representative of scientific associations active in the region served by the ESE, and one person selected by the health professionals of the ESE itself. Finally, the Hospital Management Committee includes a representative of the community (patients) served by the ESE and one representative of organizations registered with the local Chamber of Commerce (los gremios de producción).30 A2.20 With the exception of the political-administrative members of the Management Committee, each member is selected by their respective constituencies, according to the processes that the groups themselves define. Each Committee member is named to a three- 27. By 1999 more than 70 percent of second- and third-tier hospitals (categorized in terms of complexity of healthcare services) had been converted to ESEs (Escobar and Panopoulou, 2003). 28. The five are: Tarazá, Murindó, Vigía del Fuerte, Sopetrán, and La Pintada. Source: Contraloría General de la Republica, Formato No. 18 ­ A. jmejiaar@gobant.gov.co 29. For Department-level hospitals the respective representatives on the Hospital Management Committee are the Department Director of Administration or the Governor or his/her delegate, and the Departmental Director of Health or his/her delegate. 30. If these gremios de producción are not present in a given municipality, then a second community representative should be elected to the Committee. 30 Annex 2: Health Sector year term, and may be reelected. The Hospital Management Committee normally meets once every two months. Meanwhile, day-to-day operations are overseen by the Hospital Director. A2.21 A key role of the ESE Management Committee is to prepare an approved shortlist of candidates for the position of Hospital Director (Decree No. 1876, Art. 11). The criteria and process for preparing the shortlist are not precisely defined. However, once the Committee defines the process it will follow (e.g., exams, interviews, criteria for selection), the ESE must publicly invite candidates to apply. At the end of the evaluation process, the three highest-ranked candidates are presented to the Mayor (or Governor in the case of a Departmental ESE), who must select the ESE Hospital Director from that list. The Director is named for a fixed four year period.31 A2.22 The Hospital Director shall, inter alia, present to the Management Committee an Annual Institutional Development Plan, which sets out the goals and mission of the hospital and recommended steps needed to achieve those ends. The Director also presents the proposed annual budget to the Committee, along with a Staffing Plan with proposals for any changes required to improve the functioning of the hospital.32 The Management Committee, in turn, is responsible each year for analyzing, critiquing, revising and approving these key management documents. A2.23 Concerning the Staffing Plan, in particular, the Management Committee should verify the correspondence between the hospital's mission, the annual budget and the Staffing Plan, including the work hours for each employment category and the assigned salary values. If there are any changes proposed to the Staffing Plan, the Management Committee is expected to analyze and discuss the technical basis for the recommendation, the budget implications (for creation or suppression of a post), and the skill requirements appropriate to any new post. The distribution of staff also should be considered, including: · skills mix, including distribution of professional and administrative positions · distribution between career employees, provisional employees, and political appointees · number of contracted personnel and their relation to the career staff A2.24 Day-to-day personnel matters are assigned to the Logistics Branch of each ESE. However, as should be clear from the preceding summary, a number of key human resources decisions are within the remit of the Hospital Management Committee, with the participation of the Hospital Director. 2. Continued Limits to Hospital-Based Human Resources Management A2.25 Control over the Staffing Plan ­ often referred to as the "establishment list" in Anglophone countries ­ is a critical human resources management authority that formerly resided with Departmental or national government are is now entrusted to the local Hospital Management Committee and Hospital Director. The establishment list is not the only key feature of human resources management, however. Several personnel management roles are still practiced by authorities well above the facility level. 31. Law No. 1122 (2007), Art. 28 changed the term to 4 years, and includes a transitory provision to align the term of the Hospital Director with the term of the mayor or governor with the authority to name the Director. 32. See Decreto 139 de 1996, Art. 4. 31 Annex 2: Health Sector A2.26 Annex Table 4 lists these personnel management functions and delineates how they are allocated with regard to ESEs in the health sector. Table 2.2B: Human Resource Management Responsibilities in Health HR function Certified Municipality Non-certified Municipality Establishment List control ESE Management Committee (Junta ESE Management No. of posts for each specialty Directiva). Committee (Junta (e.g., doctors, nurses) and Recommended staffing guide for "core staff" Directiva), with subsequent employment type (e.g., career (planta básica) under preparation by Ministry approval required from posts, contract workers) of Social Protection; publication expected in Department Secretariat of July 2007. Health. Selection Hospital Director: ESE Management Hospital Director: Same as Committee prepares shortlist of 3 candidates certified municipality. (1/). Career posts: Same as Career posts: Examination and ranking certified municipality. carried out by the National Civil Service Commission (CNSC) or delegated by the CNSC to universities or organizations specializing in human resources and sanctioned for this role by the CNSC. Whether or not there will be a role for municipal officials in review of curricula and/or possible interviews to be determined. Provisional appointments: Hospital Director Appointment Hospital Director: Mayor appoints preferred Hospital Director: Governor candidate from shortlist. appoints preferred candidate Career posts: Hospital Director from shortlist. Provisional appointments: Justification for Career posts: Same as provisional appointment, and specific certified municipality. candidate selected by Hospital Director must Provisional appointments: be approved by CNSC. Same as certified municipality. Pay/Compensation Ministry of Social Protection (Bogotá) Same as certified defines pay range for each category of municipality. healthcare worker. Exact pay within that range is decided by the ESE Management Committee when employee is hired. Subsequent annual cost-of-living adjustments set by CONPES. Promotion Rules set by Ministry of Social Protection Same as certified decree? [To be set in new RRHH en salud municipality, with oversight law? How were rules in recent years set?] by Department Secretariat of Application of the rules is by ESE Logistics Health. Branch, with oversight by ESE Management Committee. Transfers ESE Management Committees, by mutual Same as certified agreement. municipality Evaluation Rules set by Ministry of Social Protection Same as certified decree [confirm] municipality 32 Annex 2: Health Sector HR function Certified Municipality Non-certified Municipality Dismissal Hospital staff: ESE Management Committee; Same as certified an appeal by dismissed employee referred to municipality Administrative Court (Corte Contencioso Administrativo). Hospital Director: Law 1122 (2007), Art. 27 states that the National Government will establish rules for their evaluation and potential removal for poor performance. 1/ Note: In Nariño the Association of Health Professionals prepared shortlist. Source: World Bank. 3. Limited Adjustment? The Impacts of New Management Arrangements on Human Resources in Health A2.27 Public hospitals, prior to their transformation as ESEs, were dependent upon transfers and support from the state to assure their functioning. Under the new General System of Social Security in Health (Sistema General de Seguridad Social en Salud) ESE hospitals are expected to earn revenues through the sale of their services by contracting with health insurers in the private (Empresas Promotoras de Salud ­ EPS) and publicly-subsidized regime (Administradoras del Régimen Subsidiado ­ ARS). In this way, the discipline of the market can be expected to generate improved efficiency through a decentralized healthcare delivery model. Those efficiencies would be realized, in part, through changes to the hospital's human resources, changes directed to align payroll expenditures with the type and volume of the services demanded in a given region. A2.28 To date, where human resources are concerned, there is evidence of three ESE "responses" to the new incentives and allocation of authorities: 1) Hospitals with few changes in human resources, often an increase in the wage bill but little evidence of improvement in resource utilization. 2) Hospitals that have undergone a formal restructuring process, generally with central government support Why fewer (positive) changes than expected? (group 1) A2.29 Competition has not raised the efficiency of many public hospitals, which continue to operate with very low occupancy rates, but receiving hefty transfers. Hospitals have not adjusted their cost structures. Labor costs increased 40 percent from 1995, leading to increased requests for transfers. In 1998, 68 percent of ESE revenues were used to pay salaries (Escobar and Panopoulou 2003). Additional resources flowed from government. Large payroll and wages increase took place by between 1995 and 1998 (Escobar and Panopoulou 2003). The share of personnel expenditure as a share of total expenditure increased from 40 percent to 66 percent from 1994 to 2001 (but part of that was not bad). Hospitals in the first level increased personnel expenditures by 256 percent, second level by 177 percent and third level by 103 percent (CONPES). A2.30 A key explanation is that the soft-budget constraint remains. Hospitals are not wholly dependent upon the market for their revenues. In the case of territorial bodies transfers 33 Annex 2: Health Sector should not be general or direct, but instead consistent with the quantity and value of the services effectively provided. A2.31 As the entire population is not covered by insurance, public hospitals are meant to cover those who are not affiliated. The eligible but not covered population has a right to services provided by public hospitals (or private ones, by means of contracts with regional entities). These services are covered with the so-called supply-side subsidies. In short, the shift from supply subsidies to demand subsidies is still partial. There are demand-side subsidies for enrollees in the SR, and ongoing supply-side subsidies for poor citizens not enrolled. A2.32 This continues to this day. As Law No. 1122 (2007), Art. 27 states, within 6 months the central government is to regulate the conditions and requirements for the Nation and territorial bodies to transfer resources to the Social State Enterprises when, due to market conditions in the region the ESE is not financially sustainable, even when operating efficiently. How some hospitals have restructured (group 2) A2.33 The incentives for public hospitals to restructure have not been compelling; and there are impediments, as well. This experience has underscored the need of direct policies aimed at increasing the efficiency of existing suppliers, while relying less on the idea that competition alone will take care of the problem. This fact is particularly true in a sector such as that of health care, in which, owing to political reasons public institutions operate with soft-budget constraints. A2.34 In 1999, resources from the national treasury and an Interamerican Development Bank loan were devoted to financing the restructuring of 26 second- and third-level hospitals to improve their productivity. That project brought about a 31 percent reduction in personnel costs, and a reduction in the length of hospital stays of patients. This included money to pay for severance. Table 2.3B: Indicators of Healthcare Facilities intervened by Program Indicator Year 0 Year 2 % change Service delivery Ambulatory services 2,879,560 3,583,892 24.5 Surgeries & births 281,103 335,512 19.4 Occupancy rate 74.5% 80.4% 7.9% Bed turnover 59.5% 65.1% 9.4% Expenditures Personnel (No.) 20,023 12,590 -37.1% Wage bill, regular staff (millions of $ 2004) (1/) 570,551 362,517 -36.5% Total personnel exp. (2/) 634,746 490,304 -22.8% % personnel exp./total expenditure 69.3% 61.7% -11.0% Total expenditure (millions of $ 2004) 915,451 794,472 -13.2% Revenues - Expenditures (millions of $ 2004) (8) -321,015 -48,938 -84.8% Note: "Year 0" is that in which the intervention began, while "Year 2" is after two years of the implementation process. 1/ Includes salaries and all financial contributions linked to base pay. 2/ Includes the wage bill in 1/ plus "indirect" personnel expenses for contracted employees (whether term appointments or services). Source: Ministerio de la Protección Social, as reported in "Documento CONPES Social 3415" (2006). 34 Annex 2: Health Sector Additional Information Table 2.4B. Responsibilities and Competences Established by Ley 715: Responsibilities in Terms of Stewardship and Management of the Health Sector in General Departmental level · Formulate plans, programs and projects for the development of the local health sector in accordance to national guidelines and for inclusion at the national level · Adopt, implement, execute and evaluate the health sector norms, policies, strategies, plans, programs and projects formulated by the National government · Provide technical assistance to municipalities and public providers within their jurisdiction · Supervise and control the use of resources allocated to the health sector · Supervise and control municipal compliance with National policies and norms, in conjunction with other oversight bodies · Adopt, implement, manage and coordinate the operation of a health information system · Promote social participation and exercise of citizens´ rights in relation to health and the health sector · Finance and oversee local medical and dental ethics courts · Provide the necessary care for special populations defined by the MPS Municipal level · Formulate plans, programs and projects for the development of the local health sector in accordance to national and departmental guidelines and for inclusion at the national level · .Manage the flow and execution of funds allocated to the health sector and manage the local health fund · Manage and supervise access to health care services of the population · Promote social participation and exercise of citizens´ rights in relation to health and the health sector · Adopt, implement, manage and coordinate the operation of a health information system and report the information required by the health system Source: World Bank. 35 Annex 2: Health Sector Table 2.5B. Responsibilities and Competences Established by Ley 715: Department and Municipal Responsibilities According to SGP Purposes Purpose Responsibilities Department Municipality Insurance · Management of funds for demand subsidies · Management of funds for demand subsidies (demand (only for departments created in 1991) · Cofinance enrollment in the SR subsidies · Cofinance enrollment in the SR · SR enrollment processes for the SR) · Oversight of management of municipal SR (Ley 100) funds · Oversight of compliance with health insurance regulations · Techical assistance activities to municipalities Public · Development, execution and evaluation of · Development, execution and evaluation of health (non Departmental Health Plan in accordance to Municipal Health Plan personal National guidelines · Surveillance and control of environmental health · Monitor and evaluate execution of risk factors and animal borne diseases services) municipal plans (municipalities in categories 1, 2 and 3) · Surveillance and control of environmental · Coordinate, supervise and control public risk factors and animal borne diseases (of health activities carried out by health insurers municipalities in categories 4, 5 and 6) and providers · Coordinate, supervise and control public · Surveillance and management of information health activities carried out by health insurers on local health profile and providers · Intersectoral advocacy · Surveillance and management of · Oversight and control of local sanitation information on local health profile (food production and distribution, · Intersectoral advocacy environmental conditions) and enforcement of · Techical assistance activities to norms and provisions in sanitary code municipalities and insurers · Provision of public health laboratory services (either directly or by contract) · Oversight and control of sale, distribution and safety of medications (with other national bodies) Provision of · Management of funds for the provision of · Management of funds for the provision of personal personal health services for the uninsured or first level of care personal health services health services not covered by the benefits packages (certified municipalities) services at the second and third level of care, and first · Management of funds for the provision of (supply level of care for non-certified municipalities personal health services (immunization, subsidies) · Biennial plan of public investments in cervical cancer screening, family planning, health, including infrastructure and equipment other prevention activities) · Organization and management of the public network of health service providers · Information (registry of public and private health care providers) · Oversight and control of compliance with minimum standards for infrastructure and service provision · Adopt, disseminate, implement, execute and evalute National health service provision policies Source: World Bank. 36 Annex 2: Health Sector Table 2.6B. Allocation Level, Percent Distribution of Resources by Level and Criteria to Determine the Amount to be Allocated, by SGP Expenditure Items SGP expenditure item Allocation level Criteria to determine amount to be allocated and % distribution of resources Demand subsidies Municipality % uninsured poor, weighed by municipality's share of (continuity of enrolment (100%) uninsured in relation to total uninsured in the country and new enrollees) Public health collective Department "Demand" or total population size actions (PAB) (45%) "Equity" (a weighting factor of poverty level and public health risk such risk of malaria and dengue, Municipality immunization rates, and geographic dispersion of the (55%) population) "Administrative efficiency", based on attainment of immunization coverage rates Supply subsidies Department: Total uninsured poor population adjusted for a provision of 2 and population dispersion factor and for a weighting factor 3rd level care based on the cost of services not covered by the (59%) subsidized health plan. year in the country Municipality: provision of 1st level care (41%)* Note: (*) Only certified municipalities receive these resources, otherwise the Department receives 100%. Source: World Bank, Law 715. 37 Annex 2: Health Sector Table 2.7B. Department and Municipal Responsibilities According to SGP Purposes Purpose Responsibilities Department Muninicipality Insurance · Management of funds for demand · Management of funds for demand (demand subsidies (only for departments created in subsidies subsidies 1991) · Cofinance enrollment in the SR for the SR) · Cofinance enrollment in the SR · SR enrollment processes (Ley 100) · Oversight of management of municipal SR funds · Oversight of compliance with health insurance regulations · Techical assistance activities to municipalities Public · Development, execution and evaluation of · Development, execution and evaluation of health (non Departmental Health Plan in accordance to Municipal Health Plan personal National guidelines · Surveillance and control of environmental health · Monitor and evaluate execution of risk factors and animal borne diseases services) municipal plans (municipalities in categories 1, 2 and 3) · Surveillance and control of environmental · Coordinate, supervise and control public risk factors and animal borne diseases (of health activities carried out by health municipalities in categories 4, 5 and 6) insurers and providers · Coordinate, supervise and control public · Surveillance and management of health activities carried out by health information on local health profile insurers and providers · Intersectoral advocacy · Surveillance and management of · Oversight and control of local sanitation information on local health profile (food production and distribution, · Intersectoral advocacy environmental conditions) and enforcement · Techical assistance activities to of norms and provisions in sanitary code municipalities and insurers · Provision of public health laboratory services (either directly or by contract) · Oversight and control of sale, distribution and safety of medications (with other national bodies) Provision · Management of funds for the provision of · Management of funds for the provision of of personal personal health services for the uninsured or first level of care personal health services health services not covered by the benefits (certified municipalities) services packages at the second and third level of · Management of funds for the provision of (supply care, and first level of care for non-certified personal health services (immunization, subsidies) municipalities cervical cancer screening, family planning, · Biennial plan of public investments in other prevention activities) health, including infrastructure and equipment · Organization and management of the public network of health service providers · Information (registry of public and private health care providers) · Oversight and control of compliance with minimum standards for infrastructure and service provision · Adopt, disseminate, implement, execute and evalute National health service provision policies Source: World Bank. 38 Annex 2: Health Sector Figure 2.1B. Percent Distribution of Total SGP Resources of Health, by Department 2005-2006 ANTIOQUIA BOLIVAR NARIÑO 12% 11% 8% 45% 2005 33% 55% 34% 57% 41% 1% 1% 2% 11% 11% 8% 2006 41% 33% 31% 57% 44% 58% 4% 1% 1% SGP PURPOSE SR continuity SR expansion Supply Public health Source: Authors' calculations based on data from DNP 2007 Table 2.8B Department Level Resources 2002-2006, Bolívar (2006 millions of pesos) Year SGP Supply SGP Public Health National Programs Rentas Cedidas TOTAL Amount % Amount % Amount % Amount % Total Total Total Total 2002 31,749 47.4% 8,302 12.4% 2,411 3.6% 24,575 36.7% 67,037 2003 30,570 45.4% 9,138 13.6% 1,819 2.7% 25,854 38.4% 67,381 2004 28,104 44.8% 9,725 15.5% 1,535 2.4% 23,371 37.3% 62,735 2005 31,408 48.1% 9,455 14.5% 1,772 2.7% 22,683 34.7% 65,318 2006 27,063 44.3% 9,142 15.0% 3,240 5.3% 21,632 35.4% 61,077 Source Data provided by SSD Bolivar 39 Annex 2: Health Sector Table 2.9B Department Level Resources, Amount and Distribution, Nariño 2004-2006 Year SGP (Supply And National Rentas Cedidas Sale of Services Department Resources From TOTAL Public Health) Programs Resources Balance Amount % total Amount % total Amount % total Amount % total Amount % total Amount % total 2004 43,384 54% 4,991 6% 21,140 26% 1,148 1% - - 9,337 12% 80,000 2005 42,667 49% 7,956 9% 21,783 25% 1,107 1% 1,070 1% 12,845 15% 87,428 2006 46,962 56% 4,449 5% 22,487 27% 857 1% 0% 9,796 12% 84,551 Source: IDS Nariño, Ejecuciones presupuestales Table 2.10B Department Level Resources 2005-2006, Antioquia (millions of pesos) Year SGP Supply SGP PH SGP Partial National Rentas Cedidas Department Own Resources Other Account TOTAL Subsidies Programs Transfers (Financial Balance (Regalias) Returns) Amount % Amount % Amount % Amount % Amount % Amount % Amount % Amount % Amount % total total total total total total total total total 2005 56,566 10% 18,267 3% 3,600 1% 46,031 8% 134,747 23% 40,076 7% 16,021 3% 20,347 4% 239,441 42% 575,096 2006 86,617 16% 20,511 4% 3,600 1% 17,790 3% 156,849 29% 0 0% 18,097 3% 16,697 3% 226,736 41% 546,897 Source: SS Antioquia, Ejecuciones presupuestales Table 2.11B Amount and Distribution of Total Resources, By Use, Bolívar 2002-2006 (millions of pesos) Year Demand subsidies Supply subsidies Public health DHS Total Amount % Total Amount % Total Amount % Total Amount % Total 2002 97,483 45.6% 99,048 46.4% 10,922 5.1% 6,126 2.9% 213,580 2003 99,566 56.2% 49,739 28.1% 23,684 13.4% 4,288 2.4% 177,276 2004 131,562 60.4% 60,748 27.9% 21,087 9.7% 4,587 2.1% 217,984 2005 106,870 51.3% 57,854 27.8% 17,888 8.6% 3,978 1.9% 208,471 Source: World Bank. 40 Annex 2: Health Sector Table 2.12B Department Level Expenditures, Amount and Distribution, Nariño 2004-2006 2004 2005 2006 Expenditures Value % total Value % total Value % total Operation Of IDS 7,964 11% 8,488 11% 7,075 9% Supply Subsidies 53,584 75% 56,133 71% 57,548 74% Cofinancing Demand Subsidies 691 1% 1,661 2% 1,114 1% Public Health 7,904 11% 8,695 11% 8,455 11% Institutional Strengthening 1,692 2% 3,176 4% 487 1% Special Programs 4,878 7% 8,906 11% 6,780 9% Other 2,287 3% 368 0% 3,092 4% Total 71,036 100% 78,939 100% 77,476 100% Source: IDS Nariño, Source: IDS Nariño, Ejecuciones presupuestales Table 2.13B. Department Level Expenditures, Amount and Distribution, Antioquia 2005-2006 Item 2005 2006 Amount % total Amount % total DHS expenditures Personnel 6,213 1% 5,771 1% General expenditures 2,916 1% 3,237 1% Transfers (pensions, to public sector etc) 32,174 6% 159,440 27% All other expenditures (supply and demand subsidies, public health, programs) 533,793 93% 414,819 71% Source: SS Antioquia Ejecuciones presupuestales Table 2.14B Total Municipal Expenditures, Amount and Distribution, By Purpose 2005 Department Purpose Subsidized regime Supply subsidies Public Health TOTAL Antioquia 233,288,849 72.6% 53,761,938 16.7% 34,325,878 10.7% 321,376,665 Bolívar 113,616,333 80.7% 17,178,184 12.2% 9,951,787 7.1% 140,746,304 Nariño 128,166,144 76.1% 26,116,418 15.5% 14,197,283 8.4% 168,479,845 Antioquia* 363,882,434 75.3% 53,761,938 11.1% 65,617,046 13.6% 483,261,418 Bolívar* 162,371,894 69.9% 51,650,095 22.2% 18,244,371 7.9% 232,266,360 Nariño* 153,126,764 72.9% 40,254,185 19.2% 16,549,867 7.9% 209,930,816 Note: * Calculations exclude Department capital Source: Authors' calculations based on Ejecuciones municipales, DNP 2005 41 Annex 2: Health Sector 42 Annex 3: Education Sector ANNEX 3: EDUCATION SECTOR Annex 3A. Background of Education Decentralization in Colombia A3.1 Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.2 Education has been obligatory and public since 1920 and has been essentially a national responsibility during the since the establishment of the Ministry of Education. Within the context of a centralized state, public education in Colombia has historically seen shifts between the centralized, national control of the system and a more decentralized management. At different times, the responsibility for teachers33, schools, and even curricula have been moved among the different levels of government. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.3 Much of the history of education administration in Colombia has been of the national government limiting local government's control of teachers in response to perceived misuses of local authority in hiring and contracting teachers and then expanding authority of local governments again. Teachers play a central role in education, typically accounting for 70 percent to 90 percent of expenditures. Studies have also shown that teachers are, by far, the most important factor in the "education production function" so controlling teachers is an essential function of the education system. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.4 Moves towards Centralization. In 1958, as part of the constitutional reform to restore democratic rule, the government was required to set aside at least 10 percent of its budget for education. This was the first formal attempt to "force" the national government to spend a minimum amount of resources on education and it helped propel the Ministry of Education from a relatively small entity to become one of the largest ministries in the country. It has also become a model for future reforms that established well-defined minimums for education expenditure. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.5 In the 1950s and 1960s, the provision of public education was a mixed responsibility, and the nation, departments, and municipalities all managed schools and hired and supervised teachers. Although the country itself was highly centralized, local authorities did have a great deal of flexibility in the management of institutions under their jurisdiction. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. 33. Here and through out the document, teacher is used generically for all frontline education service providers, including teachers, principals, supervisors, and the like. 43 Annex 3: Education Sector A3.6 The education system was gradually "nationalized" during the 1960s and 1970s with the national government formally taking responsibility for hiring and paying teachers. In 1960, primary teachers were made employees of the national Ministry of Education. Secondary teachers were likewise transferred during the 1970s. This was done in response to the serious delays in paying teachers by local governments and concerns about the politicization of hiring at the local level. By 1975, the public education system was almost entirely controlled by the national government. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.7 At the same time the Ministry of Education took a greater role in the supervision of education in general, including the private sector. This included establishing minimum curriculum and graduation requirements. The establishment of ICFES was part of the trend to increase control over the education system and to establish national standards. The approval of the Colombia's first unified public teachers' rules in 1979 (The Teachers' Statute) was another important element. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.8 The national government gradually allowed local governments to take on more authority in education. In 1982, local governments were allowed to hire and pay teachers through their own resources. However, this right was soon suspended (in 1986) due to concerns that local authorities were creating overly generous benefit packages for teachers. As with the initial "nationalization" of teachers, the national government became concerned about the politicization of teacher policy. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.9 While the national government tried to control the spending on the part of local governments for teachers, they also introduced or maintained a number of very generous benefits. The Teachers' Statute (Decree Law 2277 of 1979) and the various rules on teachers' benefits (consolidated and reformed in Law 91 of 1989) created a separate social security regime, including the possibility of two or even three pensions that could be earned even before the teacher retired and a number of other generous benefits. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.10 While the national government focused on controlling the administrative and financial aspects of the education sector, it did allow substantial experimentation in pedagogy and initiatives to improve quality. While there were increasing central control over the curriculum (what is taught), there was significant flexibility over how it was taught. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.11 Law 60 and the Constitution of 1991. The Constitution of 1991 represented both a consolidation of previous decentralization efforts and an increase in the move towards decentralization. In the education sector, it gave special prominence to the sector and centers the responsibility for its provision in the hands of municipalities with support from the departments. 44 Annex 3: Education Sector Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.12 To formalize the vision of the Constitution, Law 60 of 1993 set the criteria for the transfers of resources to local governments. It described how resources were to be spent and the amount of flexibility that the local government had with that spending. The law clarified the role of the Situado Fiscal and the Participacion en los Ingresos Corrientes de la Nacional (PICN), which are the main vehicles to transfer resources from the national government to local governments. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.13 PICN was the main transfer to municipalities. Under the law, municipalities must spend at least 30 percent of PICN on education, which included capital spending, educational materials, and salaries. In addition 5 percent must be spent on sports and culture, which could include some educational programs. An additional 20 percent of the participation was to be spent on the provision of potable water however municipalities that already met a certain minimum level of coverage could dedicate this money for other projects, including education. Finally, the law allowed 20 percent to be spent freely on education, health, water, housing for the poor, and sports and culture, based on the municipal budget and development plan. Thus, in theory, a municipality could spend from 30 percent to 75 percent of the PICN on education. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.14 The primary revenue sharing instruments for departments was the Situado Fiscal, which is a fixed percentage of the nation's current income. According to the Constitution, 15 percent of the Situado Fiscal is divided equally among the departments and districts, regardless of their population. This obviously favors the smaller departments over the larger ones. The remaining 85 percent was allocated based on the number of actual and potential users of public services, taking into account local efforts to raise revenues and the efficiency of public services. The Constitution specified that 60 percent of the Situado Fiscal is for spending on education, 20 percent on health and health care, an the remaining 20 percent for either of the two sectors, depending on local needs. In practice, the National Development Plans of 1994-1999 and 1998- 2002 fixed the entire percentage of the allocation of the Situado Fiscal between education and health, giving the departments no flexibility in assigning resources. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.15 Municipalities were given the responsibility for education quality and were immediately "certified" to use the resources from the PICN for education. Departments had to go through a formal certification process. Upon receiving certification, teachers were transferred and the national government started distributing the Situado Fiscal. As the Situado Fiscal was not tied to the size of the existing workforce, some jurisdictions found that they did not have enough resources to coverage their teachers. Others found that they had that they had significant new resources for education. Since the Constitution of 1886, Colombia has had some form of a 45 Annex 3: Education Sector centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.16 Both municipalities and departments could contract new teachers with their resources. National law regulated teacher salaries and benefits and all new teachers had to be contracted under the national terms. Departments could not dismiss existing teachers. Transferring teachers among schools (even in the same municipality), although possible, was difficult. In terms of human resources, local governments could increase the quantity of teachers but could not alter the cost per teacher. In many cases, increasing the number of teachers was the only real option that local governments had to increase coverage in remote areas. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.17 The 1990s saw a dramatic increase in government spending, largely driven by the social sectors. In the case of education, there were significant increases in the teacher payroll as local government increased hiring. At the same time, many local governments were unable to pay existing benefits (and in some cases salaries), leading to a growing debt on the part of local governments to the workforce. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.18 Towards the end of the 1990s, the system was in a fiscal crisis as the country entered a severe economic recession and local governments were increasing unable to pay for the education sector. The number of teachers in many jurisdictions was unsustainable and there was significant overlap in the provision of different types of services within the education sector. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.19 Lessons. Colombia's different experiments with centralization and decentralization provided the country with several important lessons that continue to influence how education policy is shaped. Perhaps the central lesson is the importance of properly managing teacher policy. Historical experience shows that both the national and local governments can be captured by teachers. In the case of the national government, teachers were given very generous pension benefits and a very generous promotion system. At the same time, local governments have a long history of offering salary supplements to teachers based on the perceived hardships that they face. While the courts continue to review the legality of hundreds of departmental schemes, the national government has completely removed local government's flexibility to set wages in any way. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. A3.20 Colombia's experiences in the past forty years have shown the national government that local governments and schools can and do innovate with positive results. Some of most regarded innovations happened during a time of very centralized control of curriculum. These innovations continue, with the explicit support of the national government. Since the Constitution of 1886, Colombia has had some form of a centralized state. Although the Constitution of 1991 created new roles and responsibilities for local governments, Colombia remains a unitary republic. 46 Annex 3: Education Sector Annex 3B Human Resource Management in Education A3.21 There are two different employment regimes in operation in Colombia, and there are important differences between them that affect human resources management. A3.22 In 1979, Decree Law No. 2277 merged a legacy of four types of teachers in Colombia · National · "Nationalized" (Law 43 (1975)), primary and secondary education offered by departments and municipalities. · Municipal · Departmental Table 3.1B: Human Resource Management Responsibilities in Education HR Function Department/Certified Municipality Establishment List Ministry of education, based upon the number of enrolled students in urban and rural Control regions. No. of posts for each Creation of additional posts contingent upon availability of resources from the General specialty Transfers System. (See Decree No. 3020 [2002], Article 7.) Selection Examination and ranking carried out by the National Civil Service Commission (CNSC) or delegated by the CNSC to universities or organizations specializing in human resources and sanctioned for this role by the CNSC. Note: Exceptions like coastal region in Nariño Appointment From the list of vacant posts, eligible teachers select, in order of merit, their position on the "list of eligibles," generally carried out in a public forum with all interested teachers present. Pay/Compensation Ministry of Education (Bogotá) negotiates pay scale. Key costs driven by the Statute. (old system experience drove costs) Annual cost-of-living adjustments by CONPES. Primas (bonuses) were granted to some groups of teachers for diverse reasons. These can no longer be created; however, some historic primas are legally valid and are still paid to those who were employed at the time the award was established. Promotion Rules established in Education Employment Law. The Ministry of Education can freeze recognition of promotion for a period of time, but not the right itself. Under new system, exam every three years in order to be promoted; 80 percent score required. Transfers Between schools managed by the Department (non-certified municipalities): According to criteria decided by the Department Secretariat of Education. (they have won the tutelas) Between schools in different departments or between certified municipality and the Department or another certified municipality: By mutual consent. Evaluation Rules set by Ministry of Education. Evaluation after first year matters; after that it really does not. Dismissal After first year: Evaluation. Appeal by dismissed employee referred to Administrative Court (Corte Contencioso Administrativo). Source: World Bank. 47 Annex 3: Education Sector Table 3.2B: Teachers by Employment Regime & Grade Decree Law No. 2277 Decree Law No. 1278 (1979) (2002) No. of No. of Grade Grade Employees Employees A 204 B 79 1 10,140 1A 7,911 2 4,064 1B 0 3 1,219 1C 0 4 3,804 1D 0 5 1,064 2A 38,364 6 3,342 2B 0 7 10,209 2C 0 8 20,253 2D 0 9 12,474 3A 168 10 30,818 3B 0 11 26,546 3C 4 12 31,897 3D 0 13 31,482 14 53,847 Other 6,976 Total 248,415 Total 46,447 Source: World Bank. 48 Annex 3: Education Sector Table 3.3B: Teachers Monthly Base Pay, by Employment Regime & Grade, 2006 Decree Law No. 2277 (1979) Decree Law No. 1278 (2002) Basic monthly pay Grade Grade Basic monthly pay (2006)** (2006)* A 475,561 B 526,818 1 590,404 1A 675,102 2 611,992 1B 918,647 3 649,442 1C 1,386,365 4 675,080 1D 1,592,802 5 717,660 2A 849,590 6 759,138 2B 1,286,988 7 849,568 2C 1,661,264 8 933,197 2D 1,793,140 9 1,033,788 Masters Doctorate 10 1,131,924 3A 1,282,012 1,558,949 11 1,292,497 3B 1,604,503 1,951,105 12 1,537,503 3C 1,826,345 2,220,869 13 1,701,900 3D 1,938,307 2,357,016 14 1,938,290 Source: * Decree No. 595, February 27, 2006. **Decree No. 596, February 27, 2006 Figure 3.1B. Average Cost per Teacher, 2006 50,000,000 Per Student Component Complement 45,000,000 40,000,000 35,000,000 C Pth u a d p r T a h r osns e ece 30,000,000 25,000,000 20,000,000 O 15,000,000 10,000,000 5,000,000 0 P tu a o u iare G ain a A n o ud a a a Q in io V d Cu a Coo m o as n q ia ag a a a o oaa o a B g ta e a ta d r u jira o oa a u ta a n re u re H ila S nA d s ic a a Cu a A ua eta R ra a a ps B liv r an d r C ld s N tio al Cs r T lim N rte d S n n e S tan e o a ea a n re u in Vu e M d len C n in m rc a a alle e a c C rd b u my V hd Byc isa ld ac ra c hc tla tic N riñ A tio u oo a n Cq e u d A az n Sc C sa a M u Ga G av o Source: Bank Staff based on DNP and DANE data. The departments are ranked according to the unsatisfied necessities index (NBI), and the poorest department is the one in the left-hand side. 49 Annex 3: Education Sector 50 Annex 4: Road Sector ANNEX 4: ROAD SECTOR Annex 4.A. History of Road Decentralization A4.1 Until the beginning of the 1990s road maintenance fell directly on overstaffed public bureaucratized organizations with deficient management of machinery and equipment assets. At the national level, the Road Fund had a specific assignment income (gasoline tax) with which it maintained regional structures (works districts) to maintain the network under its responsibility. This scheme was replicated in the departments with the public works secretariats that included the personnel responsible for maintaining the roads in their payroll. In fact, it was common to observe workers and machinery assigned to the municipal works secretariats. There was, in addition, the National Feeder Roads Fund (FNCV, in Spanish), an entity of a national nature, in charge of directly carrying out construction works and maintenance in the rural roads network. A4.2 The inefficiencies of an institutional and administrative nature were compounded by the deficient assignment of jurisdiction by network levels among territorial entities. The nation was responsible for 49,000 Km of the nearly 120,000 Km road network in the country.34 Almost three fourths of the network for which the nation was responsible corresponded to roads with local and regional impact, whose centralized management imposes inefficiencies from the standpoint of investment planning, execution, and monitoring. A4.3 The political constitution of 1991 and the laws and decrees that regulated it established the new decentralized framework for the management of the road network.35 In particular the roads were functionally defined for their assignment by levels of government. In general, the nation is responsible of maintaining and expanding the main road network connecting the department capitals among themselves and the borders and ports; The departments were responsible for the links from the municipal seats of government among themselves and from these to the main road network; and the municipalities were responsible for the roads linking the footpaths with the urban areas. A4.4 The constitution eliminated the specific assignment income for the non-social sectors and with it the allocation of the gasoline tax. Similarly, it was explicitly prohibited to assign national resources to carry out projects in sectors that have been assigned to the territorial entities. The National Road Fund was liquidated and the INVIAS was created as the agency responsible for the national main network. The INVIAS has to contract with third parties the interventions in the network and its funds come from the budget of the nation and the collection of road tolls. The law ordered the elimination of the FNCV within a period of three years and the transfer of the secondary and tertiary roads to the departments and municipalities, respectively. A4.5 Finally, the cofinancing funds were created as a temporary mechanism to support the territorial entities in financing the new responsibilities assigned under the decentralization 34. CONPES Document 2911. March 1997. The 49,000 Km for which the nation was responsible included 13,000 Km of the Main network under the responsibility of the National Road Fund, an equal number of kilometers of secondary roads and 23,000 Km of rural roads for which the FNCV was responsible. 35. Law 105 of 1993 and Decrees 2171 and 2132 of 1992. 51 Annex 4: Road Sector process, up to the year in which the transfers reach the maximum share of current revenues of the nation. The cofinancing fund for roads and urban development was created; subsequently an exclusive fund for roads was created. The Fund was administered by FINDETER which allocated non-reimbursable resources from the national budget toward projects proposed by the territorial entities. The projects had to comply with the technical guidelines of the Fund as well as have local counterparts. Furthermore, a contribution of 3 million pesos per transferred kilometer was approved to encourage territorial entities to assume greater responsibilities in terms of the network. A4.6 The decentralization of the sector has not worked out as expected nor as ordered by law. In 1995 the transfer of 13,408 kilometers of secondary network to the departments was completed, through inter-administrative agreements and with financial support from the Cofinancing Fund. However, the allocated resources were for one-time only and therefore the departments have not had a financial flow that would allow them to meet the recurrent needs of the network under optimal maintenance programs. A4.7 The case of the tertiary network is even more complex. Of the 23,000 Km that were to be transferred to the municipalities, only 7,000 had been delivered. On the one hand the municipalities resisted receiving greater responsibilities without the transfer of a source of permanent resources to maintain them. On the other, the different governments have found that the FNCV plays an important role from the political standpoint, because it opens the possibility of offering projects of a local nature into political transactions. Thus, the elimination of this entity has been postponed for more than one decade. The tertiary network in charge of INVIAS currently covers nearly 27,000 kilometers. A4.8 As part of the process, both departments and municipalities have closed down their "public work companies" and a contracting stage has begun with third parties for road works. This process has had a positive impact on efficiency and on local finances. A4.9 In 1998 and 1999, in the midst of the worst economic recession and fiscal crisis in the country in the last fifty years, the cofinancing funds were eliminated. This decision was made for three reasons. In the first place, to strengthen the finances of the central government. Secondly, because the funds had been politicized to the extreme and the allocation of resources no longer followed the technical and economic criteria with which they had initially been designed. It was not possible to ensure that these funds were being executed with efficiency and their supervision was extremely complex. This situation was due, among others, fact that to subnational governments the cofinancing funds were the budgetary part that gave them flexibility in their assignment (the transfers were to a great extent pre-assigned). In the third place, as was mentioned, cofinancing was always seen as a temporary mechanism while intergovernmental transfers were increased as provided for by the constitution. A4.10 Once the funds were eliminated, the territorial entities were left without resources to finance their greater road maintenance responsibilities. The transfers were increased, however, the part of the transfers to be freely assigned was very small which implied that road financing was subject to budgetary competition from other sectors, or depended on financing from its own small income. A 6 percent surcharge was then included in Law 488 of 1998 on Diesel consumption. The distribution of collections from this charge were 50 percent for the nation and 50 percent for the departments to finance road projects or to service the debt previously contracted by the territorial entity, as part of the fiscal reorganization program of the over- 52 Annex 4: Road Sector indebted territorial entities. Similarly, the rules of the national royalties fund were changed in such a way that regional allocations were earmarked for transportation projects (among them roads) of regional interest, targeting the flow of these capital resources toward projects that had great impact on the socioeconomic development of the regions and not scattered into small maintenance and local paving projects. However, the definition of "transportation projects" was left very ambiguous, which did not guarantee an investment in roads or in their maintenance. A4.11 At the end of the last decade, the national government invested in major road projects of lower level government responsibility. This process started with the Roads for Peace Program (Programa Vías para la Paz), that was part of the Plan Colombia. Through the program the paving of departmental roads such as Grenada­San José del Guaviare (Meta y Guaviare), Velez­ Landázuri (Santander); San Pablo y Simití (Sur de Bolívar) was begun. The resources were delivered to the municipalities for maintenance of tertiary networks and financing was provided for the construction of bridges on tertiary and secondary roads. A4.12 Under the present administration, the 2.500 program was structured which finances paving and improvement works on 2,500 kilometers of roads, for the most part in the secondary and tertiary networks. The authorization to allocate national resources to regional or local roads is obtained through the approval of the respective budget laws that include these programs. A4.13 The lessons learned from the last road decentralization process are: · A clear definition of the networks by level of government is a sine qua non condition to guarantee management efficiency. · Being a highly politicized sector, with national, regional, and municipal interests that are not always uniform, a simple financing structure is required, with clear incentives that convey the priorities of road management. · A coordinated financing policy is required between co-participation, local own revenues, and resources for roads. 53 Annex 4: Road Sector Annex 4.B. A Portray of the Case Studies: Challenges of the Road Sector in Antioquia, Bolívar and Nariño A4.14 As shown in the table below (Table 4.1B), the three departments reflect very different service delivery challenges in the road sector. Antioquia is a mountainous department with large geographical extension. The department has effectively assumed responsibility and ownership of its secondary network in the length of 4,800 kilometers. In the meantime it has performed a road survey detecting stretches of road where ownership is unclear vis-à-vis the national government and municipalities: the department rejects ownership for roughly about 20 percent of the secondary road network. In addition to the secondary network in the department, there are 1,495 kilometers of tertiary roads in charge of INVIAS and it is estimated that the municipalities respond for an additional 12,284 kilometers of tertiary roads. Table 4.1B. Road Indicators in Antioquia, Bolívar, and Narino Antioquia Bolívar Nariño Paved (Km) 1,050 86 Road surfacing (Km) 3,750 51 1,610 Dirt 3 (Km) Total (Km) 4,800 139 1,610 % Mountain 85% - 90% % poor condition 27% 49% ND 2 Area (km ) 63,612 25,978 33,268 2 Road Density (Km/Km ) 0.08 0.01 0.05 Vehicle Fleet (vehicles. 2006) 399,019 49,988 36,477 Gasoline Sales (Barrels/day) 10,576 2,844 3,423 Vehicles per Km/network 83.13 359.63 22.66 Gasoline Sales per Km/Network 2.20 20.46 2.13 Maintenance Expenditure 2006 (Millions)* 90,313 1,703 6,582 Road Expenditure per km (millions) 19 12 4 * In Antioquia the expenditure in Mega projects was excluded and the INVIAS budget in cofinanced projects Source: World Bank calculation, based on information provided by departmental governments. A4.15 Road density in Antioquia is relatively high compared to the national average. This also implies a relatively high revenue from gasoline sales per kilometer of network under its responsibility. This adds to the otherwise favorable fiscal position in terms of own revenue generation. As a result, investments in year 2006 in the secondary network in Antioquia were roughly 19 million pesos colombianos per kilometer. This is, by national average, a sufficient level of investment in order to maintain a good quality of roads; by contrast, were the network in overall good conditions, these amounts were even sub-optimally high. A4.16 The case of Bolivar is much different. This department is divided among its Northern and coastal areas, where Cartagena and other larger municipalities are located, and its South, a very poor region but which is used for agriculture. The center of the department, in turn, is covered by 54 Annex 4: Road Sector marshy lands that hinder the construction of roads. This implies that there is no direct access from the northern areas to the southern parts, and traffic must deviate into the neighboring departments of Magdalena and César first in order to reach these areas. A4.17 The decentralization process in Bolívar is incomplete in secondary and tertiary roads. Currently, the department only responds for 139 kilometers of secondary roads; more than half of them paved and, for the most part, in flat land. The only roads connecting the municipalities of Southern Bolivar and the Momposina depression belong to the national government, although they are not part of the main network that the law assigned to INVIAS. In these conditions the density of the secondary road network in Bolivar is an eighth of that of Antioquia and a fifth of that of Nariño. This means that the indicators for expenditure, traffic, and gasoline consumption per kilometer in Bolivar are relatively high. Nevertheless, almost half of this small network is classified as in poor condition, due to low investments in previous years. INVIAS has a tertiary network of 2,232 kilometers in Bolivar. Information regarding the length of the tertiary network in charge of the municipalities is not available. A4.18 Nariño, like Antioquia, concentrates most of its population in the Andean area. The length of the secondary roads are 1,610 kilometers; all roads are paved and 90 percent of the network is located in mountainous terrain. This network meets the conceptual definition of secondary network to the extent that it connects the municipalities among themselves and these with the main INVIAS network. Even though the road density of Nariño is barely two-thirds of that of Antioquia (there are no roads36 in the Pacific region), the number of vehicles per kilometer of network is very low compared to that of Antioquia and similar gasoline consumption. This factor is due to the fact that Nariño is a border department. Many trucks prefer to be supplied in Ecuador where gasoline prices are highly subsidized. Nariño, in addition has lower per capita income than Antioquia, limiting the availability of other tax revenue for roads. 36. With the exception of the Junín­Barbacoas road. 55 Annex 4: Road Sector Annex 4. C. Road User Financing 1. Direct and Indirect Cost Recovery A4.19 In regard to user financing, a distinction can be made between direct and indirect recovery of costs attributed to use of the road network. A4.20 Direct recovery.- Direct recovery is the preferred option in principle but it is generally true that tolls can only be economically collected on high volume roads (15,000-25,000 vpd for all costs and not less then 1,500 vpd for collection costs only). So this option will generally apply to a small percentage of the overall road network and even then generally only to a portion of the national road network. However improvements in tolling technology and increases in management efficiency secured through private sector participation in concessioning of roads have probably expanded this frontier over time. Additionally a number of countries have innovated financing mechanisms under which road improvement and subsequent maintenance can be funded through a blending of user recovery and government subsidy ­ see Uruguay case study further below. A4.21 In Colombia, there is limited recourse to tolling on publicly managed roads at any level of the network ­ there is by contrast a fairly extensive program of road concessions under which private sector partners are collecting tolls to at least partly finance scheduled maintenance activity. It may nonetheless be a relevant solution to the financing requirements of high volume roads on the secondary network. At least legislation should facilitate the possibility of departmental governments to be able to raise resources in this way and to be able to allocate directly to road expenditure. A4.22 Indirect recovery.- This type of funding may be of more practical relevance in countries with low traffic volumes and could also apply more readily to sub-national networks. The so called "second generation" road funds rely in principle on a proxy indicator for road use ­ normally a fuel levy per liter sometimes in combination with an access charge based on the vehicle license fee which allow for revenue to be generated in line with the deterioration caused to the road by the user. These funds have become quite widespread in many parts of the world but have not really taken off in the region outside of Central America ­ see section further below. While the funds have definitely contributed to increasing resources for maintenance, they have fallen short of a total solution to the problem. There a number of typical problems that need to be resolved if the reform is to be successful. Tow of the principle issues in Central America have been: revenue may not be stabilized if not direct channeled to the road sector; revenues may not be easily adjusted upwards if the fuel levies are set by law rather than given some degree of discretion. A4.23 Road funds could provide part of the solution for the sub-national network in Colombia with the advantage of raising a higher level of resources at the decentralized level and being able to relate what is raised directly to the maintenance needs on the network. Ideally the funding mechanism would be a newly created fuel levy expressed in cents per liter. A second best solution would be to allocate a portion of existing gas taxes and re-label this as a levy (as was done in some of the Central American cases). It may be prudent to proceed with such a reform on a pilot case-by-case basis with "willing participant" departments. It will also be necessary to 56 Annex 4: Road Sector come up with a mechanism to allocate funding between the secondary (department) and tertiary (municipal) which could be based on traffic levels or more simply by network size. A4.24 Fiscal decentralization and the road sector. As stated earlier, a successful program of roads management decentralization needs an equivalent effort to be made in regard to decentralization of funding. There are a large number of examples of the failures (in whole or in part) of decentralization programs in the road sector ­ Bolivia, Brazil, Colombia, Ecuador, Mexico ­ which can be attributed to this among other factors. Fortunately there is also one clear success story, albeit a work in progress still, and that is Peru ­ see Box 3. The system is however based on the transfer of budget resources on a permanent basis for the management of the sub- national network, rather than on a user pays model. This is not entirely satisfactory (to date) as the amount actually transferred lags critically behind what would be needs to fully fund the maintenance of the sub-national network, i.e. estimated in 2004 that whereas about 65 percent of the maintenance needs of the national network were met, the comparable figures were 42 percent and 24 percent respectively for the regional and rural networks. A4.25 The experience also illustrates the attention given to restructuring and strengthening the technical support available to the sub-national road management agencies. Additionally, while it took a while for the decentralization process to take off, what was of great help was to have the basic legal and institutional framework established ­ as part of the process of revising the Constitution in this case. A4.26 In the case of Colombia, there is the benefit of the recently approved (July 2007) CONPES document regarding the plan vial regional (PVR). This can be expected to be helpful to the extent that it provides the technical underpinning as well as the formal transfer of responsibility to the sub-national level to those roads that ought to be managed there. Notably however the PVR does not really help much in terms of the reorganization and stabilization of recurrent financing for sub-national road maintenance. Access to debt financing for sub-national road investment is mentioned but this is likely to be relevant to only a handful of departments and then generally not in any case for maintenance funding. Thus an additional CONPES would have to be prepared to authorize any reforms in regard to fiscal decentralization for the roads sector. 2. The Uruguay Experience with Road Management Reform37 A4.27 Starting in 1994, Uruguay has reformed its road management structure through a number of phases to create a more efficient and effective core management agency as well as to create new network management arrangements with strong private sector participation. A4.28 The initial impetus for reform was from the need to modernize road management and maintenance policies and to increase the efficiency in the use of limited available resources. The cornerstone of the government's strategy was to place highest priority on road maintenance over new construction ­ and to that end a priority network was defined. Core institutional reforms were seen as essential to the success of the strategy, such that the national roads directorate (DNV) was to be reorganized in order to prioritize planning, regulation and control and move out 37. Adapted from Brushett (2007). 57 Annex 4: Road Sector of execution activities. Implementation of maintenance programs would be contracted out to the private sector, with some residual force account activities largely restricted to rural districts. A4.29 DNV has since embarked on a gradual, but substantial restructuring of its organization and business processes ­ with greater orientation towards performance and results. By end 1998 about 23 percent of national road maintenance had been contracted out. Staffing had been reduced from an original 3,450 to fewer than 2,000 (subsequently to fall to 1,810 by 2003). Specific attention was paid to the creation of micro-enterprises consisting of former employees which were allocated routine maintenance contracts on some of the lower trafficked roads in the national road network. The original contracts were re-tendered in 2002 and 2003. The micro- enterprises have survived and seem to have flourished in a more competitive environment. Substantial improvements have been obtained in maintenance management and planning systems. Base unit cost parameters have been determined for maintenance works. These are set to drive a generalized reduction in maintenance costs and a reduction in the variations in cost structures of the different maintenance activities. A4.30 Increasing attention was paid to new forms of contracting out which increase private sector participation. Firstly was the use of contracts for rehabilitation and maintenance (CREMA) ­ performance based contracts usually of 4-5 years duration covering sub networks on 150-250 kilometers. These arrangements helped DNV realize cost efficiencies and also helped tie up maintenance funding over as multi-year period. Secondly was the concessioning of heavily trafficked transport corridors to the private sector, normally under 20-30 year contracts under which tolls could be charged. The initial concession dated from 1994, but the use of this mechanism has increased since 1997 with currently 542 kilometers under concession and a further 650 kilometers under consideration. Thirdly was the awarding in 2002 of the "Megaconcession" under which about 15 percent of the national network and assigned its maintenance and operation to a different entity through a concession contract. The "Megaconcession" is based on the principle of cost sharing ­ users are required to pay a toll but there is also a subsidy payable by the government. The concessionaire enjoys a minimum revenue guarantee and also receives subsidy payments based on kilometers of road and structures maintained. A4.31 The core reform of DNV has been largely accomplished, but further attention needs to be given to staff incentives retention to consolidate capacity creation in policy and regulation. Road asset preservation has been improved. Considerable progress has been made on private sector participation and diversifying sources of revenue. 42 percent of the DNV network is being managed through private sector contracts and concessions. Multi-year contracts help mitigate against budget cuts while road concessions generate incremental toll revenue. The "Megaconcession" is now a proven mechanism for accelerating commercial management and can, with suitable reforms, be envisaged as a vehicle for further revenue diversification and become more attractive for direct private investment ­ such as securitization of toll revenues; and government guaranteed bond issues. 58 Annex 4: Road Sector 3. The Central American Road Funds38 A4.32 Five Central American countries ­ Costa Rica, El Salvador, Guatemala, Honduras and Nicaragua ­ are the only ones in the region which have established national road maintenance funds. The road funds were created over the period 1993 to 2000. With the exception of Costa Rica, the road funds are dependent 100 percent on fuel levies, although in the case of Nicaragua this is only so since 2007 following a revision to the law. The fuel levies are in the range 3-8 cents per liter. Each road fund is under the supervision of a board or technical committee with mixed public and private sector participation, though in every case under the chairmanship of the minister responsible for roads. A4.33 In relation to generally accepted best practices, the Central America road funds fall short in at least some respects - though there is a good deal of divergence between the country situations. Only El Salvador and Nicaragua are separate public entities with executive boards; and only El Salvador has direct channeling of resources to the road funds, rather than through budget line allocations. Costa Rica's road fund is focused primarily on road improvement rather than on road conservation. Funding has been generally limited to all or part of the main road network - only Cost a Rica and Nicaragua formally allocate funds to the urban network. A final, important point to underline is that the fuel levies paid over are generally specified in the law and may not be considered to be road user charges in the correct sense of the word - the levies are not set, nor can they vary, in reference to cost the road user actually imposes on the road network. A4.34 The case of Honduras provides a good illustration of the qualified success of road funds to date. The law provides for paying over various fuel levies to the Road Fund ­ but these have never been transferred in full. Maintenance carried out has increased, but still falls short of meeting all assessed needs. The condition of the paved road network has improved, but that of the unpaved network which has borne the brunt of the funding shortfall has not. The Road Fund is fully staffed, but its effectiveness has been reduced by frequent turnover and changes in the government appointed Chief Executive. A4.35 Nonetheless the economies of the Central America sub-region have clearly on balance been beneficiaries of the reform, that is: amount of maintenance carried out has increased, all of which is contracted out to the private sector; resources made available have generally been utilized strategically, allocated first to a priority network; road funds have helped to build new contracting capacity, in some cases with the creation and support to micro-enterprises for routine maintenance; and the condition of the network has improved, on average 34 percent of all public roads and as high as 70 percent of main roads are assessed as in good condition. 4. The Peru Experience in Decentralization of Road Management39 A4.36 Decentralization of responsibilities for road management is an important dimension of the reform process in Peru. The policy is based on Peru's constitutionally mandated decentralization dating from 1993 further to which regional governments (26) and municipal 38. Adapted from Brushett presentation "Central American Road Funds" to the 2007 University of Birmingham Senior Road Executives Program. 39. Adapted from World Bank Report No. 36484-PE - Republic of Peru Decentralized Rural Transport Project, November 2006. 59 Annex 4: Road Sector governments (1,832 ­ of which 194 are provincial) were directly elected first in 2003. The objective is to improve service provision, including roads, by assigning responsibility closest to the user and allowing for a high level of citizen participation in decision making. A4.37 Sub-national governments in 2006 accounted for 35 percent of government expenditures and more than half of all public investment in Peru. However, with some exceptions, to date the fiscal discretion available to sub-national governments is quite limited and priorities are still set to a large extent by national priorities. High priority has been assigned by the current administration to the expansion of infrastructure in rural areas, which will a continuation of the emphasis that has been placed since 1995 on developing sustainable, user oriented approaches to rural road improvements. A4.38 Building on the decentralization reforms, management of rural roads has been progressively handed down to the municipal level, especially since 2002. A specific institutional model has been developed ­ the Provincial Road Institutes (PRIs) managed by a board comprising all mayors of a given province. Provias Rural ­ the department of the Ministry of Transport which used to be directly responsible for managing the roads ­ now provides technical assistance to develop the PRIs. As of 2006, 38 PRIs had been established and were handling all maintenance and some rehabilitation activity in their respective provinces. Inter alia the PRIs have taken on the management of the micro-enterprise program for rural roads. Since June 2006, budget resources have been transferred on a permanent basis to municipalities to finance the routine maintenance of all rehabilitated rural roads ­ amounting to 14,500 kilometers on a national scale. A4.39 A number of issues are currently being addressed in Peru to enhance the decentralization process, that is: (i) increasing funding available for sub-nationals roads ­ Peru lags other Latin American counties and has sufficient resources now only to maintain 38 percent of the total network; (ii). addressing limited management capacity at sub-national level, which has proved difficult due to fragmentation and the small size of municipalities; and (iii) enhancing coordination of planning and prioritization of infrastructure investments across various sectors ("bundling") ­ to increase the effectiveness of investment of roads and its impact on user livelihoods through reduced travel costs and increased incomes. 60 Annex 5: Water Sector ANNEX 5: WATER SECTOR Decentralization and the Provision of Water and Sanitation Services in Colombia A. Overview A5.1 Colombia, with a population of about 42 million inhabitants, has made impressive progress in expansion of water and sanitation services in urban areas in recent years. According to United Nations guidelines ­ which include private wells in the case of water supply and septic tanks and latrines in the case of sanitation ­ access to safe water and sanitation services has become virtually universal, especially in large urban areas of Colombia during the last decade. Using stricter definitions of household connections to water and sewerage, the current level of access in urban areas would be 97 percent for water and 90 percent for sewerage, which still compares impressively with higher income countries like Argentina (70 percent) and the rest of the region. Rural coverage (broadly defined) is much lower (53 percent for water and 57 percent for sanitation), with these levels shrinking to 52 percent and 15 percent respectively when the definition is restricted to household water and sewerage connections. Figure 5.1A Increases in Water and Sewage Services Coverage from 1973 to 2003 % Water Supply Sewerage Source: World Bank. A5.2 Although there has been substantial progress over the expansion of urban and rural coverage in recent years, there remain significant discrepancies in the coverage ratios in large urban areas and in peri-urban and rural areas, where efforts must be more than doubled in sanitation expansion and increased by 25 percent in water to achieve universal access by 2020. However, the increase in coverage rates masks shortcomings in the quality of service. Although the overall performance indicators for the 59 largest utilities show reasonable levels of efficiency and service levels, in smaller municipalities, water rationing and intermittent supplies are common in most water supply systems. Only slightly less than 50 percent of all drinking water outside the major urban centers is being properly treated and, as a result, the drinking water quality in many systems is substandard sufficient pressure in the water supply systems is often lacking, adding to the risk of bacterial contamination. Sewage collection systems do not have sufficient hydraulic capacity to handle wastewater flows, resulting in overflow problems and direct health and environmental impacts. The share of wastewater receiving any kind of treatment is about 10 percent, which is low for a middle income country such as Colombia. 61 Annex 5: Water Sector A5.3 Regarding continuity of service, two thirds of households report access to a 24-hour water service in large urban areas. In smaller municipalities, lack of systematized information makes it difficult to assess general service quality, but widespread anecdotal evidence shows that the continuity of the services in the many areas is far from being continuous (24 hours). Table 5.1A Continuity of the Water Supply Service by Regions Year Atlantic East Central Pacific Bogota Antioquia Valle del Total Cauca 1997 48.3% 61.7% 75.6% 65.1% 79.9% 84.1% 93.2% 71.4% 2003 56.&% 76.8% 88.3% 73.0% 97.4% 96.2% 92.7% 82.6% Source: ECV. B. Context of Decentralization A5.4 Colombia has had a highly decentralized framework for the provision of water and sanitation services since the 1986 the Decentralization Law, which transferred the responsibility for operating and managing the water and sewerage systems to municipalities, and decentralization has been a major force in shaping the water and sanitation sector's structure. Colombia now comprises 1,091 municipalities and over 1,300 registered water and sanitation service providers in urban areas, including municipal departments, utilities (public, private and mixed) and other authorized organizations and some 12,000 community based organizations providing service in rural areas. The market is geographically concentrated with the 40 largest municipal utilities serving 70 percent of the urban population, equivalent to 55 percent of the national population. C. The Decentralization Reforms of the 1990s and Legal Framework A5.5 Over the 1990s the Central Government introduced institutional reforms for the water and sanitation sector. These reforms aimed to incorporate market elements into the industry and were oriented to the enterprise development of utilities. A5.6 The 1991 National Constitution defined a more decentralized model for the management of the public services. This labor was now responsibility of private or public enterprises with the assistance of the Central Government. The new institutional structure gave the Central Government the obligation of formulating the legal and regulatory frameworks for the Water and Sanitation sector, and to municipalities and local agencies the responsibility of an efficient management of the public services. The Constitution also established general rules for private sector participation in the economy and created the Superintendency of Public Enterprises (SSPD) to control the activities of utilities. The national Constitution of 1991 confirmed the right of municipalities to provide water and sanitation services, including the power to grant concessions or other forms of private sector participation. The Constitution provided the legal framework and established the grounds to introduce more aggressive reforms in the water sector by clearly separating service provision and policy making, and by allowing PSP in the infrastructure sector. 62 Annex 5: Water Sector A5.7 In 1994, the 142 Law was enacted to promote and develop enterprise characteristics into public services. The law defined the administration regime for the domiciliary public services and the general framework for transformations in their regulation, control and vigilance and with the purpose of deepening the decentralized spirit of the Constitution; consolidation of utilities through its modernization, private participation, increases in efficiency and promotion of competition. A5.8 Law 142/1994 is one of the most complete examples of utilities legislation in the region and the water and sanitation sector is regulated by this law. Law 142 emphasizes in efficiency of service provision through the introduction of competition and the promotion of PSP. Among the key elements of the law are the creation of the Superintendency of Public Services and the definition of the functions of the Regulatory Commissions, the Comisión Reguladora del Agua in the case of the Water and Sanitation Sector. The Law promotes the 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. A5.9 The normative framework was complemented with Laws 60/1993 and 715/2001, which decided the distribution of national budget resources for the municipalities and defined specific percentages for the Water and Sanitation Sector. The responsibility of guaranteeing access of the population to public services was decentralized, but was funded by a significant amount of public resources. A5.10 The new regulatory framework established tariff-setting methodologies, which allow utilities to include the economic cost of efficient management of public services into tariff (CRA Resolutions 08 and 09/1995 and 287/2004). With the definition of these methodologies, the sector demanded reliable technical and economical information and utilities finally had a financial equilibrium in their activities. A5.11 The general framework allowed the creation and coexistence of different private participation schemes, enterprise development and different competition schemes, such as free entrance to the market (competition in the market), direct contracting with municipalities, association schemes, use of third parties in PSP processes (Bogotá) and communitarian organizations. A5.12 The Government is now contemplating the issuance of a new Water Law to enhance the policy environment as regards water resources management, particularly. The water and sanitation sector does not have any sector legislation beyond Law 142/94, and 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. D. Key Institutions A5.13 The Ministry of Environment, Housing and Regional Development (MAVDT) is now responsible for the water and sanitation sector policy at the national level. Within the MAVDT, the newly-created Viceministry of Water and Sanitation oversees the sector, formulates sector policies, and plans sector development. The Departamento Nacional de Planeación (DNP) is 63 Annex 5: Water Sector responsible for designing and monitoring the policies, plans, programs, studies, and investment projects for the water and environment sectors, including supporting decentralization and institutional reform efforts. The Water Regulatory Commission (CRA) promotes competition among service providers, controls monopolies, defines tariff-setting methodologies based on standard formulas and on investment plans submitted by the operating companies, and sets quality-of-service and technical standards to be followed by utilities. The Superintendency of Public Enterprises (SSPD) is responsible for monitoring and supervising the adequacy and efficiency of utility operations, establishing uniform accounting systems, supervising the administration of subsidies, and monitoring the general administration of public service companies. Environmental regulation is handled by Autonomous Regional Corporations (CARs). Figure 5.2A Institutional Scheme for the Water and Sanitation Sector in Colombia Sector Policy Framework A5.14 The Government's policy for the sector is very much focused on improving the institutional and financial capacity and sustainability of service providers, an important tenet of which has been the promotion of private sector participation. There has been significant experimentation with private participation, with more than half of the 26 largest utilities, equivalent to 13 percent of the urban market, operating with some degree of private sector 64 Annex 5: Water Sector involvement through various types of models, mainly mixed capital, concession and management contracts. A5.15 The performance of all these utilities has improved. 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 is represented on the Board and takes a leading role in operation of the company and investments are co-financed between the public and private sector to varying degrees. In spite of a favorable Government policy, PSP in the water sector is still not widespread and is far lower than has been expected, in part as a result of security problems which deter foreign operators and investors and in part as a result of the inherent complexities of carrying out PSP processes, which require strong political commitment at the local level, Government technical assistance and financial support in order to be successful. Table 5.2A Nominal Urban and Rural Coverage Service Census 1993 ECV 2003 New Beneficiaries 1993-2003 (Millions of Inhabitants) Urban Water Supply 94.6% 97.4% 6.7 Sewerage 81.8% 90.2% 7.7 Rural Water Supply by Network and 41.1% 66% 3.4 other sources* Sewerage + Household Solutions 50.8% 57.9% 1.3 Sewerage 14.4% 15.2% Household Solutions 36.4% 42.7% Total National Water Supply by Network and 79.7%** 90.0% 10.1 other sources* Sewerage + Household Solutions 73.2% 82.0% 9.0 Notes: * Includes wells and standpipes. ** Includes only water supply. Source: ECV 65 Annex 5: Water Sector Figure 5.3A Water and SewageCcoverage by Department, 2003 Sewerage Water Supply Source: World Bank. Sector Issues A5.16 The main outstanding sector issues in the country are the following: · Poor Utility Performance, particularly in small municipalities (less than 50,000 inhabitants) as a result of inadequate management capacity and political intervention in management, including appointment of managers. Poor utility performance is reflected in the low quality of service provided and in inadequate commercial systems, low billing and collection, and inadequate maintenance, more evident in small and medium size municipalities. · Large Regional and Urban ­ Rural Disparities in Service Coverage. Access to water and sanitation in rural areas is still a long way to go, with 53.1 percent coverage for water and 57.9 percent for sanitation, and 52 percent and 15.2 percent respectively when the definition is restricted to household water and sewerage connections. · High Investment Needs. Total investment required for investments in the services in the next five years (2004-2008) is estimated at US$ 2.1 billion or approximately US$ 420 million per year. These figures include investment in wastewater treatment and maintenance of existing assets, as well as investments needed to reach the Millennium Development Goals for the sector. 66 Annex 5: Water Sector · Limited Subsidies to the Poor. Cross subsidies are part of the design of the tariff structure and have proven to be an effective mechanism for income redistribution in large urban areas. The cross subsidy mechanism has become less effective in areas where the entire population is in the lower income strata. The current scheme has proven an insufficient tool to expand service delivery or provide sufficient financing in municipalities with predominantly poor populations, where the underlying imbalance between contributors and beneficiaries is greater. A specific mechanism, Solidarity Funds, was proposed to allow a better service to low-income groups while helping utilities achieve financial sustainability by providing an additional direct subsidy to the service provider, but since the funds are supposed to be financed with municipal resources, fiscal constraints at the local level have inhibited their rollout. As of 2002, only 54 percent of municipalities had established such funds, and of those only 5 percent were allocating resources to these funds. The Government has commissioned a detailed study of the current cross-subsidy mechanism and stratification system to define the reforms needed to solve the structural problems mentioned above · Low Wastewater Treatment Coverage. Only about 10 percent of the wastewater generated in the country undergoes any kind of treatment and only half of installed wastewater treatment capacity is operational. This is a result of deficient management, and lack of the necessary sewerage infrastructure for wastewater collection in some municipalities. The remaining 90 percent is discharged without any treatment, contaminating a significant part of the natural water resources. There is no clear mechanism for financing municipal wastewater treatment plants. · Weaknesses in the Regulatory System. The regulatory activity is costly and inefficient since the CRA and the SSPD have to oversight responsibility for over 1,300 service providers in the country. Large utilities serving over 6 million inhabitants like Bogotá are regulated in the same way as those serving 2,000 inhabitants. 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. The existing regulatory and legal frameworks do not provide sufficient comfort to potential private investors and constitute an obstacle to the efforts of consolidating the reform initiated in 1994. · Limited Coordination among Central Government Agencies. There are many institutions in the sector, some with overlapping functions in certain areas. Under such circumstances, coordination efforts are a must. Utilities frequently complain about the large amount of information requested by the government, and about the costs associated with generating it. · Lack of a national rural water policy and no explicit rural investment programs. In contrast to other infrastructure sectors, there is no explicit central policy scheme to promote rural access to water and sanitation, beyond the fiscal transfers and the general legal framework. Although Government Programs such as the Microempresas Comunitarias have proven useful, the poor performance of rural utilities in rising. 67 Annex 5: Water Sector A5.17 Despite the need for improvements, the water and sanitation sector in Colombia is relatively advanced as compared to many of its Latin American neighbors. In particular, the sector is characterized by an adequate policy framework; the Government has been a pioneer in adopting good sector policies a decentralization policy, promoting PSP and providing financing. About 12 million inhabitants in five of the largest cities have good and efficient water and sewage service. There is also a successful performance of utilities with specialized operators, which are commercially independent, financially sustainable, separated from the municipalities though owned by either the public or private sector. A number of large utilities include private sector participation within the framework of a distinctive mixed enterprise model. A5.18 There has been significant experimentation with private participation, 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. 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 is represented on the Board and takes a leading role in operation of the company; and investments are co-financed between the public and private sector to varying degrees. The experience with these utilities (including Cartagena, Barranquilla, Santa Marta, Tunja, Monteria, Palmira, Girardot, and others) is encouraging. The performance of all of them has greatly improved, customer satisfaction is high and they provide good examples for resolving the water sector problems. Table 5.3A Water Operators by Size of the Municipality Population Number of Municipalities Private ESP Mixed ESP Public ESP** Municipal Range Municipalities with Provider Registered No. % No. % No. % No. % Service Provider >50,000 125 124 101 42% 20 8% 98 41% 20 8% 25,000 to 193 173 28 14% 5 3% 107 54% 59 30% 100,000 < 25,000 800 590 53 9% 6 1% 154 27% 366 63% Total 1118 887 182 18% 31 3% 359 63% 445 44% Notes: * Does not include community organizations providing services ** Includes public share companies and industrial entities, among which are some which are out of compliance with Law 142 as they were created after January of 2008 Source: Superintendency of Public Services (April, 2006) E. Government Strategy A5.19 In recent years, the Government has implemented a modernization and reform program for the water and sanitation sector to address the issues described above. Current strategy and policy for the sector is outlined in the 2002-2006 National Development Plan (NDP), and aims to consolidate the reform undertaken and to create the financial, institutional and monitoring conditions to ensure investment in sector needs and the modernization of utilities to improve efficiency levels and performance. Government's strategy is based in the following actions: 68 Annex 5: Water Sector (i) Establishing specialized operators ­public or private ­ where technically, financially and institutionally feasible; A5.20 The benefits of PSP in the delivery of water supply and sanitation services in Colombia are well documented and PSP remains the core of the Government's approach in larger urban agglomerations. There has been significant private participation in the sector, mostly under the concession or mixed enterprise models. There are currently around 100 private operators serving 15 percent of the urban population. Private participation is also occurring through outsourcing activities in the three main cities (Bogotá, Medellín and Cali), where public enterprises remain in charge of providing the services. In Bogotá for instance, the public utility has used BOT contracts for the construction of water and sewage treatment plants, and recently awarded management contracts for the operation and billing of services in five zones of the distribution network. As part of the policy to encourage this type of reform, the Government designed a Program for Utility Modernization (Programa de Modernización Empresarial - PME), aimed at facilitating the incorporation of the private sector in the management and operation of the water and sanitation services in the utilities of participating territorial entities and at providing financial support to these utilities, while ensuring provision of services to the poor. A World Bank study40 shows that since the operator took over in Cartagena, 95 percent of new connections were in households classified as Strata 1 and 2 (i.e. the poorest). In Barranquilla, the corresponding figure was 86 percent; in Tunja 79 percent; and in another private utility, Santa Marta, over 50 percent. The implementation of this Program is being supported by the Bank through the ongoing loan Water Sector Reform Assistance Project (CO-7077) which supports key PSP transactions, targeting improvement of efficiency and expansion of water and sanitation services in medium-size cities or regional associations of municipalities (with populations of up to about 300,000 inhabitants) and small municipalities (with populations of up to about 12,000 inhabitants). (ii) Increasing Financial Support to publicly run utilities through a National Investment Program. A5.21 Achieving the financial self-sustainability of the sector is a core objective of Government's strategy. The National Development Plan 2002-2006 focuses on promoting the concept that financing of sector utilities will be mainly based on income generated by collection of bills as well as by incorporating private investors, where feasible. Municipalities can subsidize the tariff of low-income customers or subsidize part of the investments using the revenue transfers of Law 715 and existing public funds allocated to the sector. As part of the efforts to increase the available financing sources for the sector, Law 788 of 2002 additionally established a tax rebate scheme as an incentive to investment by regional utilities. Only when the aforementioned resources are proved insufficient, given affordability constraints, to cover investment needs and effectively provide services, can the Government make available to individual municipal utilities, under strict eligibility conditions, additional resources from the 40. Colombia: Expanding Services to Low-Income Areas comparing private and public water utilities, (Sotomayor, 2001) 69 Annex 5: Water Sector national budget to finance specific investments and help solve structural problems, at all instances working with the utility towards the goal of eventually achieving financial self- sustainability. A5.22 Consistent with the objectives outlined above, Law 812 of 2003, which adopted the NDP, established the 2002-2006 medium-term national financial commitments to the sector. The Government will provide capital investment grant resources to be utilized based on the outcomes of the recently conducted participatory public consultation process known as the Audiencias Públicas (August 2003). As part of this process, departmental and regional investment needs where examined and evaluated by a technical committee based on poverty- and technically- weighted criteria, resulting in fund allocation at a departmental level. Under the last NDP, approximately US$ 135 million has been allocated for investment in the water and sanitation sector under the umbrella of the Audiencias Públicas. Implementation arrangements for accessing these funds, particularly the direct oversight by the MAVDT and the development of a harmonized and publicly disseminated set of procedures, along with the use of financial intermediaries for fiduciary management, has been designed to mitigate the risk of political capture of the public consultation process to ensure the efficient use of grant transfers. The Government has designed a mechanism to centralize all publicly funded41 investments in the sector, including the allocations deriving from the Audiencias Públicas, through a Program called Ventanilla Única (to be financed under the proposed APL), managed by the newly- established Viceministry of Water. (iii) Implementing Institutional Strengthening and Technical Assistance Programs. A5.23 The Government is currently implementing a policy effort to improve the institutional development and financial and technical performance of publicly-run utilities in small- and medium-size localities (both urban and rural) where PSP is not likely to be adopted in the short- term. The Government has designed two different Programs, Fortalecimiento Institucional and Microempresa Comunitaria, which aim to bring utilities to reasonable levels of management, operational efficiency and financial viability and pave the way for private sector participation by improving the productivity of utilities and strengthening of management and administrative systems. (iv) Rationalizing the Institutional Framework and Improving Sector Coordination. A5.24 Coordination at the national level has been fostered through the strengthening of the MAVDT, consolidating the Viceministry of Water as the key coordinating institution for sector development and the creation of an umbrella Program for managing national financial support to the sector, the Ventanilla Única. The Ventanilla Única establishes a uniform approach to analyzing and financing sector investments by various involved agencies, to coordinate sector investments in a more rational way across the different institutions involved in sector investment, regulation, and management. 41. Refers to publicly provided grant resources and not credit funds provided through, for example, FINDETER. 70 Annex 5: Water Sector Small-scale providers A5.25 Small-scale service providers (both public and private) play an important role in Colombia with more than 1,700 local entities providing service in small municipalities and over 4,600 in rural areas. · Microempresas Comunitarias (community based micro-enterprises): Targeted at municipalities and rural areas of less than 5.000 inhabitants, this program aims to establish community based micro-enterprises (local or regional) that can effectively respond to consumer demands, recuperate investment costs and operate with reasonable levels of efficiency. · Fortalecimiento Institucional: Targeted at urban and rural areas of less than 12.000 inhabitants with poor management capacity, this program aims to implement a business- like approach in those service providers through a combined strategy of providing technical assistance and training. · Operator-Constructor model: Targeted at small municipalities with populations of up to about 12,000 inhabitants, which decide to incorporate a private operator through a simplified process. The differentiation with PSP processes in larger municipalities is made specifically to reflect the conditions in small municipalities (from the private sector's standpoint, the business is different in the two size ranges of municipalities, and the required type of operator is different in both) and is possibly the most significant innovation of the Colombian approach to PSP. F. Understanding Varying Approaches to Service Provision A5.26 The sector reforms have been successful. A large number of public and private group of specialized operators have been conformed (608 Public Services Enterprises ESP), with a 38 percent private participation. In larger cities and some municipalities there has been significant progress in service coverage and quality. Table 5.4A Water and Sanitation Sector Institutional Structure, 2006 Population Range Total Number of Number of Service Providers Public Service Direct Municipal Population Municipalities Service with PSP /1 Providers Service Provision 2005 Providers No. No. No. No. % No. % No. % > 500,000 14,181,984 7 26 21 81% 5 19% 0 0% 100,000 to 500,000 10,308,336 48 101 55 55% 41 41% 5 5% 25,000 to 100,000 8,715,618 203 260 78 30% 132 51% 50 19% < 25,000 8,846,941 841 710 77 11% 201 28% 432 61% Total 42,052,879 1099 1097 231 21% 379 35% 487 44% Note: /1 Of the 231 Service providers with private sector participation, 123 are principle service providers Source: World Bank. A5.27 However, in almost 500 of the municipalities direct management and operation of the utilities by the municipal administration persists, with consequential poor results in the quality of the services. In other municipalities, with already an ESP created, there are problems with management and service providing due to poorly designed contracts signed between municipalities and specialized operators. 71 Annex 5: Water Sector A5.28 The sector has developed a share scheme based in three important mechanisms: (i) cross subsidies per strata; (ii) specific destination transfers from the central government to municipalities; and, (iii) national budget resources used for financing programs and strategic investment. The sector has presented deficiencies which do not allow financial equilibrium, most importantly to problems in focalization of demand subsidies. A5.29 In recent years, policies have promoted the development of the sector by using the Departments as intermediaries between the Central Government and the municipalities. This way, the programs have achieved regional impact; investment plans are integral and the enterprise transformation has taken place throughout the country. The Government's current strategy looks for improvements in service coverage and through: (i) effective inter-institutional coordination; (ii) acceleration of utilities' modernization processes; (iii) structuring regional programs to generate economies of scale; (iv) articulating different funding sources and facilitating credit access within a context of fiscal discipline; (v) improving the control and monitoring of regulation requirements; and (vi) regionally-oriented investment plans for short and long term. 72 Annex 6: Input-Out Efficiency ANNEX 6: INPUT-OUTPUT EFFICIENCY ANALYSIS OF SUBNATIONAL GOVERNMENTS AND REGRESSION ANALYSIS This study is based on data provided by the National Planning Department (DNP). The information is from year 2006. The efficiency scores were developed by DNP and constitute the primary basis of the analysis. Detailed information about the methodology is contained in the publication "Metodología para la Medición y Análisis del Desempeño Municipal" (DNP 2005b). The continued support and guidance by the Dirección de Desarrollo Territorial (Department of Subnational Governments) in DNP was critical to conduct this analysis. A6.1 This section analyzes the levels of efficiency of subnational governments in the delivery of health and education services. The purpose of the efficiency analysis is threefold: · Understand the relative efficiency of subnational government service delivery given the incentive constraints highlighted in Section A of the report (Volume I) and the variance among them. · Provide insight into the drivers of inefficiency (which will be achieved through regression analysis). · Highlight areas of further research. A6.2 Most studies that were conducted previously in Colombia use hospitals and schools as the unit of analysis. Examples include Peñaloza (2003), Pinzon (2003), and Iregui et al (2006), among others.42 In turn, in the methodology applied by DNP (DNP, 2005b) the unit of analysis is the municipality. Although resources are not always managed autonomously at this level-- health and education services are shared responsibilities among the three levels of government-- this approach is useful as it is at the municipal level that inputs and outputs are recorded. A6.3 However, there is a methodological constraint which implies that the results of the econometric analysis need to be interpreted with caution. · Some of the inputs included in the models cannot be clearly associated with a single level of government; and not all inputs are truly managed autonomously at the local levels. This is primarily explained by the decentralization model followed in Colombia. Given this limitation, the purpose of the analysis is merely to shed light on several potential determinants of inefficiency in service delivery. A. Methodology and Model A6.4 The DNP methodology for measuring efficiency of subnational governments is based on the Data Envelopment Analysis (DEA) technique. This technique identifies "peer subnational governments" for an individual subnational government; it then estimates the efficiency of the subnational government by comparing its performance with that of the "best practice" entity chosen from its peers. This leads to an empirical approximation of the production possibility frontier, defined as the maximum attainable output of a subnational government for a given input 42. Also refer to the Literature Review in this Annex with more detailed information about efficiency studies in Colombia. 73 Annex 6: Input-Out Efficiency level; and inefficiency is defined as the distance from the observed input-output combinations to the efficiency frontier. A6.5 Efficiency analysis has two dimensions. According to Coelli et al (2005) input and output dimensions are defined as follows: (i) Output orientation approach. This approach responds to the question "by how much can output quantities be proportionally expanded without altering the input quantities used?" (ii) Input orientation approach. This approach responds to the question "by how much can input quantities be proportionally reduced without changing the output quantities produced?" A6.6 In the methodology of DNP efficiency scores are computed under the output orientation approach. Given that the DEA analysis yields relative rankings, the results can be interpreted as follows: a score of 1 implies that the municipal government is on the efficiency frontier of education coverage. In turn, an output efficiency score of less than 1, for instance a score of 0.50, indicates that the inefficient producer attains 50 percent of the education coverage output obtained by the most efficient producer with the same input intake. A6.7 The DNP model uses multiple inputs and single outputs (Table 6.1A.). This reflects the fact that in practice the production of services requires several types of resources and inputs. Table 6.1A. The DNP Model: Multiple Inputs and Single Outputs Sector Dimension Output Inputs a. Total investment in education excluding payroll and Total number of students contracts with private entities Coverage enrolled in public schools b. Total Number of Official Teachers Education c. Classroom space available in education institutions Total Number of Students d. Total expenditure in education Quality with a Score above average in e. Total Number of Teachers with grade level higher the ICFES exam than 6 or 2 on the new scale Total Number of affiliates to Coverage f. Investment in the subsidized regime the subsidized regime g. Payroll of personnel in the subsidized regime Health h. Total public investment in the immunization Number of vaccinated children Managerial program for MRM (measles, rubella Skills i. Number of vaccine doses provided by the central and mumps) government Source: DNP. 74 Annex 6: Input-Out Efficiency B. Efficiency Analysis Results B.1. Education A6.8 Figure 6.1A shows the distribution of municipalities in the coverage and quality models. The Spearman rank correlation coefficient, which measures how correlated the rankings are in each model, is approximately 0.30 indicating a low rank correlation between the coverage and quality models. As shown in the figure the scores of the coverage model are biased towards one (the most efficient municipalities), while it is the opposite in the case of the quality model. That is, most municipalities are indeed able to show good results in coverage. A possible explanation is that municipalities are able to achieve progress in enrollment with the current level of resources (expenditure, teachers and space) since enrolling students does not necessarily depend on additional inputs. In turn, obtaining good education quality seems more challenging as more than half of the municipalities are achieving less than 50 percent of the outcomes of the best comparable municipalities. Figure 6.1A: Models to Measure the Level of Efficiency in Education Provision Coverage Quality Efficiency of Education Enrollment Quality of Education Distribution of the Efficiency Scores Distribution of the Efficiency Scores 150 80 60 100 Frequency Frequency 40 50 20 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Score Score 100 Efficiency Score School Enrollment 20 40 0 60 80 0 20 40 60 80 100 Efficiency Score Quality - Number of Students with a good ICFES Source: Based on the efficiency scores provided by the DNP team. 75 Annex 6: Input-Out Efficiency B.2. Health A6.9 The figure below shows the distribution for the multiple inputs and single output models in health (Figure 6.2A). The spearman rank correlation (approximately 0.09) indicates that there is not a significant rank correlation between the rankings under both models, i.e., a municipality efficient in the affiliation to the subsidized regime is not necessarily efficient in the provision of vaccinations (and vice versa). Figure 6. 2.A Health Sector Models to Measure the Level of Efficiency in the Provision of Health Coverage Model Managerial Skills Model Efficiency of the Affiliation to the Subsidized Regime Efficiency of the Vaccination Campaign Distribution of the Efficiency Scores Distribution of the Efficiency Scores 100 80 80 60 Frequency Frequency 60 40 40 20 20 0 0 40 60 80 100 0 20 40 60 80 100 Score Score 100 80 Efficiency Score (Vaccinations) 20 40 0 60 40 60 80 100 Efficiency Score (Affiliates Subsidized Regime) Source: Based on the efficiency scores provided by the DNP team. 76 Annex 6: Input-Out Efficiency B.3. Combining the Input-Output Efficiency Analysis for Education and Health A6.10 This analysis seeks to detect basic management capacities independent of sector-specific capacities. The analysis shows that there is no correlation between the efficiency scores from education and health (Table 6.2A). In other words, municipalities located on the efficiency frontier of education are not the same as those located in the efficiency frontier of health. This result is surprising since one would expect a positive correlation, or at least that the largest and richest municipalities would be in the frontier in both sectors. This implies that a municipality that is efficient in service delivery in one sector will not necessarily be efficient in providing the service in another sector. This creates the need to continue with very specific institutional strengthening activities. Table 6.2A Correlation between the Efficiency Scores in the Education and Health Sector Models Education Health Sector Output Inputs Total Students Number of Total with ICFES Number of Vaccinated Enrollment score above Affiliates Children average Total investment Total in education, total 1.0000 Enrollment number of teachers & classroom space Total investment Education Total in education & Students with total number of ICFES score teachers with 0.2989 1.0000 above grades higher than average 6 and grade 2 (on new scale) Investment in the subsidized regime Number of & 0.0947 0.1427 1.0000 Affiliates Payroll of personnel in the subsidized regime Total public Health investment in the immunization Number of program & Vaccinated -0.0194 -0.0206 0.0787 1.0000 number of vaccine Children doses provided by the central government N = 610 Source: World Bank based on the efficiency scores provided by DNP. 77 Annex 6: Input/Output Efficiency Analysis C. Regression Analysis to Determine Potential Sources of Inefficiency A6.11 The objective of the regression analysis is to find potential sources of inefficiency that can be influenced through central government policies ("explanatory variables"), differentiating them from those factors that can hardly be changed or altered ("control variables"). The report uses a Tobit model where the dependent variable is the efficiency score obtained from the multiple inputs/single output DEA model; the "explanatory" and "control variables" are outlined in Table 6.3A. Table 6.3A Variables and Acronyms of the Regression Type Variable Health Education Dependent variable Efficiency Score Efficiency Efficiency Municipality expenditure in Municipality expenditure in 1.A. Explanatory Testing the SGP the health sub-category, education financed with SGP variable earmarking subsidized regime or PAI (current pesos) (current pesos) Municipality Municipality expenditure in Municipality expenditure in 1.B. Explanatory expenditure financed health financed with Royalties education financed with Royalties variable with Royalties (current pesos) (current pesos) Municipality Municipality expenditure in Municipality expenditure in 1.C. Explanatory expenditure financed health financed with FNR education financed with FNR variable with FNR (current pesos) (current pesos) Departmental Transfers to the Departmental Transfers to the 1.D. Explanatory Departmental Transfers municipality for health municipality for education variable expenditure in current pesos expenditure in current pesos Dcapital = 1 if Bogota, 0 otherwise; 2.A. Explanatory Dpacifico = 1 if from Pacific region, 0 otherwise DAtlantico; = 1 if Regional Dummies variable from Atlantic region, 0 otherwise; DLlanosorientales = 1 if from Llanos Orientales, 0 otherwise Dummy identifying certified 2.B Explanatory Certified Municipalities municipalities (1) and non-certified Variable in education municipalities (0) in education 2.C Instrumental Enrolled students Enrolled students in public schools Variable Population density Population divided by Population divided by municipality 2.D Control variable (population / km2 municipality area (population area (population / km2) municipalities) / km2) 2.E Control variable Tax revenue SAME SAME 2.F Control variable Constant Source: World Bank. A6.12 The variable SGP is an endogenous variable in the education sector: the efficiency scores are measured using the number of teachers as input while outputs are the number of enrolled students as well as the number of affiliates to the subsidized health regime. On the other hand, education and health SGP transfers are allocated to subnational governments according to the number of enrolled students, teachers' wages, and number of affiliates to the subsidized health regime. This required performing tests for endogeneity in the education and health sectors by using the Durbin-Wu-Haussmann Test. Only in the education sector was there an endogeneity problem which was addressed through the use of instrumental variables in the education sector regressions. 78 Annex 6: Input/Output Efficiency Analysis A6.13 A dummy variable identifying certified municipalities in education sector was used in the regression analysis. C.1 Regression Analysis for the Education Sector A6.14 Not surprisingly, the regression analysis shows that the most important determinants for municipal efficiency can be found both in structural factors--conditions which can hardly be influenced through policy--as well as in institutional and fiscal arrangements among levels of government. A6.15 Structural factors. Two important structural factors are related to (i) geographic location and (ii) population size. As shown in Tables 6.4A and 6.5A below, municipalities in different regions (Atlantic region, Pacífico, Llanos Orientales) have a different level of performance. For instance, in the education coverage model the municipalities in the Atlantic region perform better (although they do not necessarily also perform better in education quality compared to other regions). This finding indicates that regional differences matter which underscores the importance of the socio-cultural context in which services are produced and delivered. A6.16 As for population size, the regression analysis indicates that municipalities with greater population density perform on average better than those with lower population density. For instance, efficiency scores are statistically superior in Bogotá when analyzing the scores of both the coverage and quality models. The same also holds for other large municipalities, for instance Medellin. A6.17 Institutional and fiscal factors. The most important institutional factor determining the level of efficiency is the condition of being certified as municipality. On a scale from 1 to 100, the efficiency scores of certified municipalities are higher in a range from 10 to 29 on average (See Tables 6.4A and 6.5A). A6.18 Since municipalities with more than 100,000 inhabitants were automatically certified, for large part these results endorse the finding regarding population size highlighted above. In the end, it is the different institutional arrangements within municipalities that explain differences in efficiency levels. A6.19 What about the impact of intergovernmental transfers on municipal efficiency? The tables below show that there exists a marginal positive impact of SGP transfers on the level of efficiency; however, in econometric terms this impact is very low and practically negligible leading to the interpretation that SGP transfers do not have a major impact in improving the provision of education services; these transfers do not do any harm, that is, they do not lead to more inefficiency, but they also do not seem to be the main driver for more efficiency either. A6.20 More worrisome, perhaps, is the fact that in parallel to SGP transfers there are a series of revenue sources which have a negative impact on expenditure efficiency: (i) royalties (which are kept by "producing" jurisdictions); (ii) transfers from the National Royalty Fund (aimed at compensating for different levels of income from royalties); and (iii) transfers from departments to municipalities. This finding underscores the need to reassess the transfer system in Colombia in order to achieve better education services. 79 Annex 6: Input/Output Efficiency Analysis Table 6.4A Regression Analysis for the Coverage Model Dependent Variable: Efficiency Score Independent Variables (1) (2) (3) (4) (5) (6) Spending financed with 1.85e-06** 7.52e-06** 9.97e-07** .0000177** 4.22e-07* 1.33e-07** SGP transfers Spending financed with -2.07e-07 -7.17e-07* -2.41e-07 -7.32e-07** -9.05e-07 Royalties Spending financed with 1.36e-07 -.0000133** -2.92e-07 1.89e-06 -9.63e-08 FNR Spending financed with -5.65e-06** -4.61e-06* -5.33e-06** -8.49e-06** 7.06e-06 -.0005727 departmental transfers Llanos Orientales -3.27316** -2.642621* -3.328228** -3.792494** Capital 95.72898 908.5907** 54.55671 -37.63356 Pacifico -.1923042 -.2768645 -.2687817 -2.437783* Atlantico 7.735119** 5.93083** 7.793696** 3.766387** Population Density .0025852** .0063503** .0012605 .0004192 -3.91e-06 .0002257 Tax revenues -4.69e-08 -3.87e-07** -2.63e-08 1.42e-08 -1.22e-09 -2.46e-08** Certified Municipality 15.22172** 24.78182** Certified*Quality Spending financed with -.0000177** SGP Education Constant 68.74416** 66.18239** 68.71693** 65.00961** 88.74851** 86.30874** N 990 743 990 990 40 40 Log likelihood -4051.0422 -14254.376 -4038.339 -3976.1923 -135.11746 -134.33177 Chi2 141.36 169.25 166.77 291.06 10.56 12.13 Prob>chi2 0.0000 0.0000 0.0000 0.0000 0.1030 0.0164 ** Significant at 95 percent of confidence interval. *Significant at 90 percent of confidence interval. Models (1) through (5) refer to quality expenditure, since non-certified municipalities do not spend in payroll; model (6) refers to payroll expenditure. And models (5) & (6) correspond only to certified municipalities. Source: World Bank calculations based on data provided by DNP. A6.21 On the other hand, the impact of tax revenue on education efficiency resembles some of the incentive problems of subnational governments in Colombia. The regression results indicate that there is a slightly negative relationship between tax revenues used for education expenditures and efficiency: municipalities with low levels of tax revenue have a slightly higher efficiency than revenue rich municipalities. This might be explained by the fact that certified municipalities in particular do not use own revenue to finance education services in light of availability of SGP transfers. Should a future transfer system in Colombia evolve that compensates for low fiscal capacity, these findings should be taken into account since, contrary to common thought, also the jurisdictions 80 Annex 6: Input/Output Efficiency Analysis with low levels of tax revenue--most probably also implying low levels of GDP per capita as well as fiscal capacity--can be efficient with the use of resources. Table 6.5A Regression Analysis for the Quality Model Dependent Variable: Efficiency Score Independent Variables (1) (2) (3) (4) (5) Spending financed with SGP 7.33e-06** 5.65e-06** .0000155** 1.93e-06** 2.47e-07 Spending financed with -3.16e-07 -2.61e-07 -4.99e-07 -8.74e-06** Royalties Spending financed with FNR -.0000148** -.0000124** -2.92e-06 -.000012 Spending financed with -2.23e-06 -2.36e-06 -5.23e-06** .0000126 departmental transfers Llanos Orientales .3016441 .2928574 -.0001247 Capital 898.0602** 715.9064** 155.7237* Pacifico -.3514174 .0858176 -.7604163 Atlantico -12.40332** -11.84016** -13.88956** Population Density .0139788** .0126144** .0089495** -.0008785 .0071962** Tax revenues -3.79e-07** -3.00e-07** -6.13e-08* -9.53e-09 -4.58e-08 Certified Municipality 10.48483** 29.7085** Certified*Quality Spending -.0000144** financed with SGP Education Constant 24.52148** 24.73595** 22.54025** 57.77949** 42.503** N 732 732 732 27 27 Log likelihood -14076.403 -14074.137 -12980.036 -556.71738 -577.26301 Chi2 259.06 286.36 319.13 31.61 14.65 Prob>chi2 0.0000 0.0000 0.0000 0.0000 0.0021 ** Significant at 95 percent of confidence interval. *Significant at 90 percent of confidence interval. Models (1) through (4) refer to quality expenditure, since non-certified municipalities do not spend in payroll; model (5) refers to payroll expenditure. And models (4) & (5) correspond only to certified municipalities. Source: World Bank calculations based on data provided by DNP. C.2 Regression Analysis for the Health Sector A6.22 Similar to education, health outcomes are both influenced by structural as well as institutional and fiscal factors. 81 Annex 6: Input/Output Efficiency Analysis A6.23 Structural factors. Regional differences do explain health outcomes. As presented in Table 6.6A, this is true both for enrollment into the subsidized regime as well as for vaccination. However, the following differences are worthwhile noting: · The Andean (highland) region (dummy variable) has better performance with regards enrollment into the subsidized regime compared to other regions. · In providing MRM vaccinations, a municipality in the Llanos Orientales, Pacific or Atlantic region is performing better on average. Table 6.6A Regression Analysis for the Health Models Dependent Variable: Efficiency Score Independent Variables Efficiency Score Affiliation to the Efficiency Score Vaccination Campaign Subsidized Regime Model Spending financed with SGP Health -7.38e-09 -8.88e-08 Spending financed with SGP Ribereños -4.26e-06 -.0000387 Spending financed with General SGP -4.25e-06 .000017 Spending financed with Royalties 7.56e-07 3.72e-06 Spending financed with FNR -3.83e-06 1.06e-06 Spending financed with departmental transfers -8.05e-07* -.0000623 Llanos Orientales -1.803587* 5.966716** Capital -233.478** 56.03398 Pacifico -3.585577** 27.05966** Atlantico -4.41512** 7.004161** Population Density .001015 .0009273 Tax revenues 8.70e-08** -1.83e-08 Constant 68.38526** 41.84617** N 691 909 Log likelihood -2589.4924 -4023.5083 LR chi2 77.36 160.30 Prob>chi2 0.0000 0.0000 ** Significant at 95 percent of confidence interval. *Significant at 90 percent of confidence interval. Note: For the Subsidized regime model the spending refers to the accounts specifically assigned to this regime. For the vaccination model the spending refers to the accounts specifically assigned to the PAI. Source: World Bank calculations based on data provided by DNP. 82 Annex 6: Input/Output Efficiency Analysis A6.24 It is worthwhile noting that, contrary to common thought, population size does not matter much for influencing efficiency in health outcomes. This is surprising, since health services are usually affected by economies of scale. A6.25 Institutional and fiscal factors. The SGP transfers have no impact on the efficiency in the health sector. In other words, a higher increase in earmarked transfers does not lead to a higher affiliation to the subsidized regime and to a higher number of MRM immunization participants. This casts doubt on the current policy which is built on the assumption that increasing transfers will lead to more enrollment and better services. A6.26 Similar to education, departmental transfers to municipalities are problematic as they do in fact have a negative impact on enrollment into the subsidized regime. A6.27 However, there is a difference to the education sector with regards to the use of tax revenue. The amount of tax revenue allocated towards health expenditures leads on average to more enrollment into the subsidized regime (Table 6.6A). In other words, tax revenue-rich municipalities have a slightly higher efficiency than those with only low levels of tax revenue. This might be related to fiscal capacity and levels of income, but further data development is necessary to corroborate this finding. C.3. Agenda for Future Research A6.28 There are several factors which were used in the regression analysis but did not yield a statistically significant result. · Royalty revenue continues to be an increasingly important source of finance for some municipalities. A time series analysis is required to account for volatility in resource flow and impacts on efficiency. · No conclusions yet can be drawn on the optimum mix of revenue sources to achieve higher efficiency. · Further research is required on effects related to economies of scale in municipalities of different size. 83 Annex 6: Input/Output Efficiency Analysis 84 Annex 6: Input/Output Efficiency Analysis Appendix 1: Endogeneity Tests 85 Annex 6: Input/Output Efficiency Analysis Endogeneity Tests: Education Models The endogeneity problem arises because the SGP transfers to the education sector are allocated according certain specific size conditions (number of enrolled students, total population size, among others). Therefore we have the following endogeneity relation: scorei = 0 + 1 SGPi + .... + i (1) sgp i = 0 + 1enrolled i + i (2) Given the relation between equations (1) and (2), the estimators of equation (1) are not efficient ( cov(sgpi , i ) 0 ). To achieve that the estimators of equation (1) are efficient the analysis has relied on the use of the number of enrolled students as an instrument for the SGP transfers. The results for the Durbin-Wu-Hausman Endogeneity test are presented below: Step 1. Testing the endogeneity of the SGP transfers (quality expenditure). An auxiliary regression of the SGP transfers on all regressors is performed. Source | SS df MS Number of obs = 743 -------------+------------------------------ F( 10, 732) = 510.12 Model | 4.5665e+15 10 4.5665e+14 Prob > F = 0.0000 Residual | 6.5527e+14 732 8.9518e+11 R-squared = 0.8745 -------------+------------------------------ Adj R-squared = 0.8728 Total | 5.2218e+15 742 7.0375e+12 Root MSE = 9.5e+05 ------------------------------------------------------------------------------ cal_sgp | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- cal_reg | .0406191 .0255532 1.59 0.112 -.0095473 .0907854 cal_fnr | 1.941969 .1729826 11.23 0.000 1.602368 2.28157 cal_transf~t | -.2297596 .1556394 -1.48 0.140 -.5353124 .0757932 dllanosori~t | 5496.732 83541.44 0.07 0.948 -158512.7 169506.1 dcapital | -7.70e+07 4963686 -15.51 0.000 -8.67e+07 -6.72e+07 dpacifico | 257955.4 126854.2 2.03 0.042 8913.886 506996.9 datlantico | 30604.44 112504.4 0.27 0.786 -190265.4 251474.3 popdens | 157.6769 146.2309 1.08 0.281 -129.4051 444.7588 ingresostr~s | .0160348 .0024633 6.51 0.000 .0111987 .0208708 alumatri | 78.81944 4.484264 17.58 0.000 70.01588 87.62299 _cons | -113386 54799.6 -2.07 0.039 -220969.2 -5802.882 ------------------------------------------------------------------------------ Step 2. Retrieval of the residuals from the auxiliary regression. The residuals from the auxiliary regression are saved. Step 3. Estimation of augmented model (1). An augmented regression (1) is estimated now including the residuals from Step 2 as an additional variable. Since the residuals are significant there exists an endogeneity problem. 86 Annex 6: Input/Output Efficiency Analysis Tobit regression Number of obs = 743 LR chi2(11) = 206.03 Prob > chi2 = 0.0000 Log likelihood = -2979.3896 Pseudo R2 = 0.0334 ------------------------------------------------------------------------------ scorematri~a | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- cal_sgp | 7.51e-06 8.05e-07 9.34 0.000 5.93e-06 9.09e-06 cal_reg | -7.20e-07 3.67e-07 -1.96 0.050 -1.44e-06 4.74e-10 cal_fnr | -.0000133 3.13e-06 -4.25 0.000 -.0000195 -7.18e-06 cal_transf~t | -4.53e-06 2.21e-06 -2.05 0.041 -8.87e-06 -1.89e-07 dllanosori~t | -2.612625 1.182038 -2.21 0.027 -4.933213 -.2920361 dcapital | 907.6711 112.6191 8.06 0.000 686.5761 1128.766 dpacifico | -.2725786 1.823273 -0.15 0.881 -3.852046 3.306889 datlantico | 5.938086 1.597098 3.72 0.000 2.802648 9.073524 popdens | .0063533 .0020974 3.03 0.003 .0022357 .0104709 ingresostr~s | -3.87e-07 4.55e-08 -8.50 0.000 -4.76e-07 -2.98e-07 residuals | 8.68e-06 9.60e-07 9.04 0.000 6.79e-06 .0000106 _cons | 66.17887 .764037 86.62 0.000 64.6789 67.67883 -------------+---------------------------------------------------------------- /sigma | 13.38149 .3475451 12.69918 14.06379 ------------------------------------------------------------------------------ Obs. summary: 1 left-censored observation at scorematricula<=7.5599999 742 uncensored observations 0 right-censored observations Step 4. Estimation of model (1) using enrolled students as an instrument. Since the residuals in the above regression resulted statistically significant the next step is to use instrument variables to obtain efficient estimators. Regressions presented in the document use the instrumented variable technique to achieve this goal. 87 Annex 6: Input/Output Efficiency Analysis Endogeneity Tests: Health Models The endogeneity problem arises because the SGP transfers to the health sector are assigned according to certain specific size conditions (number of affiliates to the subsidized regime, total population size, among others). Therefore we have the following endogeneity relation: scorei = 0 + 1 SGPi + .... + i (1) sgp i = 0 + 1 affiliates i + i (2) Given the relation between equations (1) and (2), the estimators of equation (1) are not efficient ( cov(sgpi , i ) 0 ). To achieve that the estimators of equation (1) are efficient the analysis has relied on the use of the number of total affiliates to the subsidized regime as an instrument for the SGP transfers. The results for the Durbin-Wu-Hausman Endogeneity test are presented below: Step 1. Testing the endogeneity of the SGP transfers (quality expenditure). An auxiliary regression of the SGP transfers on all regressors is performed. Source | SS df MS Number of obs = 409 -------------+------------------------------ F( 10, 398) = 5.17 Model | 1.2787e+17 10 1.2787e+16 Prob > F = 0.0000 Residual | 9.8458e+17 398 2.4738e+15 R-squared = 0.1149 -------------+------------------------------ Adj R-squared = 0.0927 Total | 1.1125e+18 408 2.7266e+15 Root MSE = 5.0e+07 ------------------------------------------------------------------------------ regsub_sgp | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- regsub_reg | -.9991094 5.428724 -0.18 0.854 -11.67167 9.673449 regsub_fnr | -1.960306 127.8961 -0.02 0.988 -253.3967 249.4761 regsub_tra~t | -1.428952 4.653969 -0.31 0.759 -10.57839 7.720483 dllanosori~t | -5951170 6321661 -0.94 0.347 -1.84e+07 6476852 dcapital | -2.42e+07 5.93e+08 -0.04 0.968 -1.19e+09 1.14e+09 dpacifico | -5857106 7312755 -0.80 0.424 -2.02e+07 8519349 datlantico | -5750597 7753145 -0.74 0.459 -2.10e+07 9491638 popdens | -7114.09 15497.71 -0.46 0.646 -37581.69 23353.52 ingresostr~s | .0322307 .2887482 0.11 0.911 -.5354315 .5998929 afiregsub | 177.8371 271.9651 0.65 0.514 -356.8306 712.5048 _cons | 6153309 4270552 1.44 0.150 -2242350 1.45e+07 ------------------------------------------------------------------------------ Step 2. Retrieval of the residuals from the auxiliary regression. The residuals from the auxiliary regression are saved. Step 3. Estimation of augmented model (1). An augmented regression (1) is estimated now including the residuals from Step 2 as an additional variable. Since the residuals are not significant there does not exist evidence of an endogeneity problem in the health sector models. 88 Annex 6: Input/Output Efficiency Analysis Tobit regression Number of obs = 409 LR chi2(11) = 68.09 Prob > chi2 = 0.0000 Log likelihood = -1488.8326 Pseudo R2 = 0.0224 ------------------------------------------------------------------------------ scoreregsub | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- regsub_sgp | 3.34e-07 2.85e-07 1.17 0.242 -2.26e-07 8.94e-07 regsub_reg | 1.61e-06 9.03e-07 1.79 0.075 -1.60e-07 3.39e-06 regsub_fnr | -.0000225 .0000238 -0.94 0.346 -.0000693 .0000243 regsub_tra~t | -2.69e-07 6.38e-07 -0.42 0.673 -1.52e-06 9.84e-07 dllanosori~t | -3.318877 2.009476 -1.65 0.099 -7.269392 .6316378 dcapital | -96.17054 113.1253 -0.85 0.396 -318.5683 126.2272 dpacifico | -2.511013 1.984777 -1.27 0.207 -6.412971 1.390945 datlantico | -3.919897 1.986526 -1.97 0.049 -7.825293 -.0145006 popdens | .0033186 .003058 1.09 0.278 -.0026931 .0093304 ingresostr~s | -5.51e-12 6.09e-08 -0.00 1.000 -1.20e-07 1.20e-07 residuals | 3.47e-07 2.85e-07 1.22 0.224 -2.14e-07 9.08e-07 _cons | 65.89893 2.136473 30.84 0.000 61.69875 70.09911 -------------+---------------------------------------------------------------- /sigma | 9.27072 .3247645 8.632252 9.909189 ------------------------------------------------------------------------------ Obs. summary: 1 left-censored observation at scoreregsub<=46.279999 408 uncensored observations 0 right-censored observations 89 Annex 6: Input/Output Efficiency Analysis Appendix 2: Literature Review 90 Annex 6: Input/Output Efficiency Analysis i. Literature Review on Colombia Study Method Inputs Outputs Control and Explanatory Variables Health Studies Alonso Stochastic Production function: "Consultorios [---] (2005) frontier externos", hospital beds, medico general, personal asistencial, enfermera professional, personal administrativo, enfermera auxiliar. Nupia and DEA Horas medico, Horas médico Número de Sánchez especialista, horas auxiliares, horas consulta externa, (2001) administrativo, numero consultorios, número de consulta horas consultivo, número de camas y urgencia, número camillas. de egresos, número de partos, número de cirugías. Pinzon DEA Número de camas habilitadas, Número de Certificación del municipio (dummy); número de camas; número de (2003) consultorios habilitados, unidades consultas, camas al cuadrado; transferencias de la nación; costo por funcionario odontológicas, unidades actividades asociado al proceso productivo; dispersión población; mercado de ginecobstétricos. Personal (horas de odontológicas, oferta de servicios (número de instituciones privadas en municipio); personal administrativo, profesional de partos, egresos. clasificación de la institución de acuerdo a su portafolio de servicios. la salud, personal auxiliar asistencial). Peñalosa DEA Número de camas, número de salas de Número de egresos Control variables: (2003) cirugía, número de personal clínico por hospitalización, Tasa de infección nosocomial, Dummy descentralización, Dummy asistencial, número de personal número de consulta hospitales universitarios, Valor de los activos totales ($ corrientes). administrativo, nivel de tecnología, externa, porcentaje Explanatory variables: gasto de funcionamiento ($ de infección Porcentaje de transferencias, porcentaje de recursos nación/ingresos, corrientes). nosocomial. número de instituciones por Km2, Dummy institución pública, Dummy autonomía, Dummy gestión Education Studies Barrera and DEA Teacher's education, teacher-pupil Test scores "Policy variables": Gaviria ratio ( main resources present in class (ICFES; as a Total per capita investment in education by municipality during 1994 (2003) room); computer facilities and measure of and 1999; investment in education as a percentage of total spending infirmary (proxies to resources outside quality); number of in 1994 and 1999; number of public libraries; presence of a "house classroom); number of students (scale students of culture. effects). (enrollment). "Exogenous Variables": 91 Annex 6: Input/Output Efficiency Analysis Study Method Inputs Outputs Control and Explanatory Variables Percent of schools that operate in the rural area of the municipality; percent of population in the urban area; index of quality of life in 1993; index of unsatisfied needs in 1993; whether the school is located in a capital of a department; fixed effects model by departments. Melo et.al. Cobb- Personnel: relación alumno/docente ICFES test scores. Context variables: (2006) Douglas promedio; nivel de educación de los Ingreso medio de los hogares de los estudiantes de cada Production maestros (docente sin formación establecimiento; propiedad juridica del plantel (público o privado); Function universitaria; docentes con formación zona de ubicación del plantel (urbano, rural); modalidad de colegio universitaria; docentes con postgrado. (académico o no); jornada (completa, mañana, tarde, nocturna); si el Infraestructura: laboratorios de colegio organiza talleres de padres. quimica, fisica, ciencias e idiomas, biblioteca, canchas deportivas, piscina, numero de inodoros por alumno. Gaviria and Linear ICFES Score Socio-economic level of students: education and employment of Barrientos model (1999) parents; quantity and quality of interaction among parents and (2001) children (partipación de la madre en fuerza laboral). Mina Panel ICFES Score Características educativas: número de maestros por alumno; situado (2004) (2000) fiscal educación, porcentaje de escuelas privadas. Carácterísticas socioeconómicas: NBI, coeficiente Gini, número de ataques de grupos insurgentes. Variables de entorno: Altura; distancia de municipio de la capital; municipio capital de departamento (dummy). 92 Annex 7: Argentina Decentralization ANNEX 7: ARGENTINA: FISCAL DECENTRALIZATION CHALLENGES AND INSTITUTIONAL RESPONSES A. The Federal Organization 7.1 The Republic of Argentina was established in 1853 as a federal country. Each of the twenty-four provinces relinquished some of their powers but retained governmental attributes in three areas: executive, legislative, and judicial. This institutional arrangement exists up to date. 7.2 Expenditure Responsibilities. The division of responsibilities is outlined in Table 7.1A below. The administration of Buenos Aires varies from these arrangements since security, justice, and the prison system are handled by the national government. 7.3 In the 1990s, services were decentralized to a large extent. For example, responsibilities for primary and secondary education--originally held by the national government--were delegated to the provinces; national schools were managed by provinces. Similarly, social housing funding was decentralized. However, the social security system, which accounts for around 40 percent of the budget, remained in the hands of the national government; this implies that provincial governments have no responsibility over this significant expenditure item. Table 7.1A Argentina: Assignation of Responsibilities by Government Level National Government Provincial Government Municipal Government Defense Primary and Secondary Education Markets, Higher Education Economic Development Cemeteries Health: Prevention Housing Economic Development Foreign Relations Health: Hospitals Lighting, Sweeping and Foreign Trade Regulation Cleaning. Water and Sewers Federal Security and Justice Justice and Security, Solid Waste Collection and Federal Prisons Provincial Prisons Disposal Streets and Drainpipes Interprovincial Transportation Passenger and Cargo Terminals Local and Regional Roads. Land and Trade Use Planning and Control, Parks Postal Service Power and Gas Fire Control Parks Source: World Bank. 7.4 Taxes and Revenue. The provinces can determine their own taxes according to their constitutions. Most of them have defined different taxes and tax rates. Provinces have taxing powers as regards real estate and gross income. The latter is a sales tax and is of cascading nature, which ultimately undermines competitiveness. In the 1990s, an attempt was made to eliminate this tax and to replace it with a final consumer sales tax. This was based on a pact between the national government and the provinces. The federal pact foresaw that provinces eliminate gross income taxes in all phases, except for final consumption; the national government would, in turn, reduce the rate of the provisional employers' contributions levied on employment, which would render the Argentine tax revenue system more efficient and neutral. 7.5 Many provinces eliminated this tax in the productive phases within their province but not for products from other provinces. This was supposed to be the next step in a second phase of 93 Annex 7: Argentina Decentralization reforms but has never actually implemented. The gross income tax is therefore asymmetric in practice. The alignment of tax rates and the elimination of taxes in intermediate phases are pending reforms if the country wishes to improve its level of competitiveness. The gross income tax is a major source of provincial revenue. It is therefore unlikely to be eliminated. B. Fiscal Challenges 7.6 The economic cycles over the last 15 years have had an impact on Argentina's fiscal policy. As tax collection decreased as a result of adverse external shocks, public spending had to be adjusted accordingly. However, adjustment was difficult as provincial and municipal governments as well as the national government increased their expenditure level in times of prosperity.43 Therefore, the fiscal deficit and public debt continued to grow. Table 7.2A Provincial Fiscal Balances (Economic Result), 1992­2006 Province Years with Deficit Average deficit in relation to current revenue (in percent) Rio Negro 9 -6.2 Jujuy 9 -6.1 San Juan 6 -5.9 Buenos Aires 6 -5.1 Formosa 6 -2.8 Catamarca 6 -2.7 Mendoza 6 -1.9 Cordoba 5 -3.0 Chaco 5 -2.7 Entre Rios 4 -2.4 Tucuman 4 -1.3 Neuquen 3 -1.0 Misiones 3 -0.9 Corrientes 3 -0.1 Chubut 2 -0.8 La Rioja 2 -0.5 Salta 2 -0.1 C.A.B.A. 1 -0.7 Santa Fe 1 -0.5 Tierra del Fuego 1 -0.4 Santiago del Estero 1 -0.2 La Pampa 0 0.0 Santa Cruz 0 0.0 San Luis 0 0.0 Source: National Direction of Fiscal Relations with Provinces of the Finance Department of the Ministry of Economy, www.mecon.gov.ar. 43. One factor inherent to the federal tax co-participation that further deepens the economic cycles is that resources are distributed automatically between the nation and the provinces according to actual revenues collected in each month. In other words, intergovernmental transfers increase during economic booms and facilitate an increase of the provinces' public spending whereas in times of recession these transfers decrease and so the provinces have to adjust their spending level. 94 Annex 7: Argentina Decentralization 7.7 In the 1990s, tax collection at the national and provincial levels increased and in some years this revenue fell, chiefly, as a consequence of external crises.44 Fiscal problems continued for some provinces during the entire decade; on the other hand, some provinces--for example, San Luís, La Pampa, and Santa Cruz-- had balanced budgets in most of the decade. Fiscal performance did not depend on the size of the province or its level of income; rather, it was rather related to issues of governance such as the quality of the public administration. Table 2 shows the provincial current deficits in the last 15 years as well as an average of these deficits in relation to their current income. 7.8 From 1997 forward, the fiscal situation worsened. The municipalities and provinces continued to borrow with guarantees from the co-participation funds,45 and this indebtedness increased current expenditures due to interest payments. In the 1990s, the provinces' creditors were mainly the private sector and international entities. As a response, many political pacts were signed in the 1990s between the national government and the provinces to implement a series of reforms. These reforms included privatizations of provincial banks as well as public utilities in charge of electricity, water, etc., for which they received technical and financial assistance. 7.9 The crisis period after year 2000. Between 2000 and 2002, the level of public revenues declined even further because of the recession, the increasing current deficit, and unsustainable foreign debt. This crisis led to the end of convertibility with a devaluation of the Argentine currency as well as the conversion into pesos of deposits and credits in dollars. In 2001, provincial debt bonds were used as a quasi-money instrument. From 2002, the national government became the principal creditor to provinces (see Figures 7.1A and 7.2A). Figure 7.1A Provincial Debt Structure, 2001 Regional Provincial Debt 2001 Provincial Infrastructure Development Banks and Trust Fund Trust Fund Fiduciary 0.50% 14.14% Institutions 33.76% Federal Government 0.48% International Organizations 10.56% Consolidated Debt Bonds 3.28% 37.28% Source: National Direction of Fiscal Relations with Provinces of the Finance Department of the Ministry of Economy, www.mecon.gov.ar. 44. The Mexican economic crisis (tequila effect in 1994); Asian crisis (1997); Russia (1998); Brazil (1999). Argentina had to face these crises with worsened export prices and declining capital flows. 45. The creditors and bond holders demanded that the debt be secured by co-participation funds. 95 Annex 7: Argentina Decentralization Figure 7.2A Provincial Debt Structure, 2006 Provincial Debt 2006 International Federal Organizations, government, 11.49% 4.43% Regional Infrastructure Trust Fund, Consolidated 0.49% Debt, 0.81% Bonds, 18.40% Banks and Provincial Fiduciary Development Institutions, Trust Fund, 1.19% 63.19% Source: National Direction of Fiscal Relations with Provinces of the Finance Department of the Ministry of Economy, www.mecon.gov.ar. 7.10 Post-crisis period. The country's scenario changed radically after the devaluation of its currency and a new favorable international context. The country had a sustained growth rate which led to positive developments in the fiscal accounts of both the national and provincial governments (Figure 7.3A). The international price increase of Argentine export commodities had turned the deficit into a fiscal surplus from 2003 onwards. Taxes were collected directly through a new tax (retention) on the export value and indirectly through an increase of VAT revenues, the profit tax at the national level and gross income taxes (on sales) at the provincial level. With the additional revenue trust funds were created to finance public works based on agreements between the national government and the provinces. 7.11 Between 2002 and 2006 all provinces displayed a fiscal surplus. However, provincial indebtedness continued to grow. The provincial debt in 2006 totaled 83 billion pesos; the federal government is the main creditor for nearly 70 percent of all provincial debt, as opposed to only 15 percent in 2001. The federal government is practically the only source of funding the. 96 Annex 7: Argentina Decentralization Figure 7.3A Evolution of the Total Income and Expenditure of the Provinces (in million pesos, 2006) $ 90,000 $ 85,000 $ 80,000 $ 75,000 $ 70,000 $ 65,000 $ 60,000 $ 55,000 $ 50,000 $ 45,000 $ 40,000 $ 35,000 Total income Total expenditure $ 30,000 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Source: National Direction of Fiscal Relations with Provinces of the Finance Department of the Ministry of Economy, www.mecon.gov.ar. C. Institutional Arrangements as a Response to the Fiscal Challenges 7.12 Important institutional arrangements were implemented to check the fiscal challenges that endured in Argentina. These included institutional strengthening, fiscal responsibility framework, monitoring of fiscal accounts, and federal intervention. 7.13 Institutional Strengthening. The Ministry of Economy, in conjunction with the Ministry of the Interior, developed the provincial finance programs with the assistance of multilateral organizations. These programs funded investment projects in the provinces that had a current surplus. Additionally, these programs supported institutional development activities in all provinces including the ones with deficit. The institutional development programs provided funding for modernizing cadastres and tax collection budgeting systems. 7.14 Fiscal Responsibility Framework. In August 2004, a Federal Fiscal Responsibility Law was adopted46 with the purpose of achieving financial equilibrium both in the national government and in all provinces. The main objectives were: · Promoting transparency in fiscal information, budget norms and multi-annual budgets. · Establishing limits on the increase in public spending which must not to exceed nominal GDP growth. 46. We should clarify that Argentina's federal model requires that some fiscal laws must be expressly adhered to by the provinces through a ratifying provincial law. Three provinces - La Pampa, Salta and San Luis - did not ratify the Fiscal Accountability Law; some, like Salta, have however adopted their own frameworks. 97 Annex 7: Argentina Decentralization · Establishing limits on indebtedness so that in the provinces the debt service would not exceed 15 percent of the current income. · Institutional framework: the Federal Fiscal Responsibility Council made up of representatives from the national government and the provinces is the authority for monitoring and sanctions. 7.15 It is too early to judge if the Fiscal Accountability Regime can be effective during crisis situations. Given the favorable context of the post-crisis there have been no situations--to date--in which all the sanctions had to be applied. The indebtedness is one of the structural problems which may lead to weak budget constraints particularly as the provinces' dependency on the central government is very high. 7.16 Monitoring of Fiscal Accounts. Both the Ministry of Economy and the Federal Fiscal Responsibility Council are monitoring provincial fiscal accounts. The Federal Tax Commission--created by the co-participation law some decades ago--is monitoring compliance of said Law and addresses conflicts between local and federal co-participation taxes (double taxing). No full monitoring system exists within the provinces to ascertain the delivery and the quality of services. However, statistics from various ministries of the federal government, such as the National Direction of Information and Evaluation of the Educational Quality of the Ministry of Culture and Education,47 provide an incomplete picture of the situation. 7.17 Other Federal Instruments: Federal Intervention. In extreme cases of lack of governance or public order the National Government can resort to Federal Intervention, through which the provincial powers will be substituted. The 1994 constitutional reforms strengthened the role of Congress to take decisions on federal interventions. "Art. 6º.- The federal Government shall intervene in the territory of the provinces to ensure the republican government modality, to repel foreign invasions, and on the request of provincial authorities to sustain or reestablish them, in case they were overthrown by an act of sedition, or invasion by another province." "Art. 75.- Congress is authorized to: (......) Paragraph 31. Order the federal intervention in a province or the city of Buenos Aires. Approve or revoke the decreed intervention, during its recess, by the Executive Branch." "Art. 99.- The President of the Nation has the following attributions: (...) Paragraph 20. Decree the federal intervention of a province or the city of Buenos Aires during Congress' recess, which it shall convene simultaneously for considering this matter." 7.18 Once there is an intervention the following steps are taken: · Immediate removal of the governor, vice-governor and members of the Legislative Branch. Members of the Judicial Branch are declared in commission. · The national Executive Branch appoints the Federal Interventor, based on a time limit. · The Federal Interventor manages the co-participation transfers. 47. http://www.me.gov.ar/diniece/ 98 Annex 7: Argentina Decentralization · Instructions are given to ensure prompt normalization and full functioning of the government. · Extension of the intervention is possible in case of a well-founded resolution issued by the national Executive Branch ratified by Congress. · In all cases, the Federal Intervention ends when local elections are called to elect the authorities who will direct the provincial government. 7.19 In the past, reasons for interventions were fiscal issues, deficient provision of justice services or electoral functioning, and/or conflicts between the executive, legislative and judicial Branches in the provinces (Annex Box 7.1). However, there is no catalog which outlines explicit situations for an intervention. Box 7.1A The Risk of Interventions: Institutional Bail-outs and Fiscal Bail-outs The cases of Santiago del Estero in December 1993 and of Corrientes in December 1999 were related to alterations of the public order due to protests of public employees. The protests were related to delays in the payment of salaries; provincial debt service increased and it was impossible to raise sufficient resources in order to cover the cost of salaries. The Federal Intervention paid the outstanding salaries. The lesson of this cases was that an institutional bail-out led to a fiscal bail-out. The other cases - Catamarca in April 1991 and Santiago del Estero in March 2004 ­ were related to civil unrest. The population had lost confidence in the justice system and its independence as well as effectiveness. Due to the massive protests the provinces became ungovernable and an intervention was necessary to improve the deficient justice services. The Federal Intervention in Corrientes in 1992 was the result of institutional issues resulting from a Governor's election that ended in a tie in the Electoral College. Source: World Bank. 99 Annex 7: Argentina Decentralization 100 Annex 8: Decentralization in Spain ANNEX 8: DECENTRALIZATION IN SPAIN A. The Asymmetric Approach Towards Decentralization 8.1 Asymmetric decentralization--different levels and scope of autonomy for subnational governments--was a key feature of the decentralization in Spain. This process offers lessons that are worthwhile to be taken into account for any country that is redefining intergovernmental relations. In Spain technical considerations based on efficiency principles were less important than political objectives. The thrust behind this process, which developed after 1980, was essentially motivated by the political and the nationalist aspirations of certain jurisdictions-- mainly, though not exclusively, the Basque and Catalonia regions. 8.2 Two modalities were employed to transfer responsibilities. They were the "fast track" approach for the so-called historical communities and the "slow track" approach for all other territories; the latter required a minimum of five years before responsibilities were transferred. Nonetheless, from a technical standpoint, there were never set minimum requirements. Similar to Colombia, Spain--prior to transferring powers--did not need to accredit subnational governments using management capacity indicators. 8.3 On the expenditure side, asymmetric decentralization was implemented by transferring responsibilities related to education (all levels, including university); healthcare (medical care and pharmaceutical services); public order (police forces directed by the regional Basque Country and Catalonia governments); and culture (development of their idiomatic peculiarities). Most other responsibilities were transferred in a symmetric fashion.48 8.4 As regards to fiscal matters, only the Basque and Navarra regions exhibit asymmetry. In particular, these communities, whose (statutory) funding system is much more "autonomist", can collect all taxes (except for customs duties) whereas the central government collects for the rest of the country. The Basque and Navarra regions get to keep all these resources; they merely have to make a general contribution to the central government in exchange for the non-decentralized services which are provided in their territory. In the other regions, the financing system is similar but has gained more autonomy over time such as increasing the participation of the regional governments in the most important taxes (Personal Income Tax, VAT and Excise Taxes) and widening these governments' normative capacity for defining taxes or establishing fiscal benefits (the exception are the indirect taxes subject to fiscal harmonization within the European Union setting). The financing system--excluding the specific case of the territories with foral rights (the Basque Country and Navarra)--was adjusted over time through five-yearly reviews with a clear trend towards enhanced symmetry in the instruments. This has lead to all regions enjoying increased fiscal autonomy. The principal differences which remain are related to the scope of the compensation mechanisms (extent and criteria for distribution of unconditional transfers). 8.5 Aside from the transferred taxes, the autonomous communities are authorized to create taxes with the stipulation that they do not levy taxes on bases that are already taxed by the central administration or on tax bases outside their territory. Consequently, at the moment the 48. Mountains, agriculture, livestock and fishery in inland waters; urban development and housing; roads; ports and airports that do not develop any commercial activities; hydraulic structures, canals and irrigation; environmental protection; monument patrimony of the community, cultural promotion, libraries, museums; sports; and tourism promotion. 101 Annex 8: Decentralization in Spain asymmetry in tax matters is related to the different ways in which regional governments exercise this autonomous decision-making power rather than to a central government decision of treating the different regions in different ways. Some examples of asymmetrical taxes established by the autonomous communities are related to contaminating emissions (ecological taxes), bank deposits, deficient land use, certain games of chance, or trade activities on large areas (malls). Likewise, the regional governments have introduced different (asymmetrical) modifications in some aspects of the income tax (complementary deductions); the inheritance tax (successions and donations); the tax on property transfers and documented juridical acts; and the tax on retail hydrocarbons sales (which is earmarked for healthcare). 8.6 At the municipal level, the asymmetry in responsibilities is related to management capacity based on size in terms of the municipal population. Thus, legally binding responsibilities increase with population size (fewer than 5,000 inhabitants, up to 20,000, up to 50,000, up to 100,000 and more than 100,000). As regards funding, they all have access to the same taxes as they have normative capacity for the discretionary establishment of different types of taxing; the central difference is related to the transfer distribution criteria that in most cases are unconditional. 8.7 At the moment, Spain is discussing the viability of implementing a "second decentralization" process, which would cover the transfer of responsibility in certain matters from the regional to the municipal level. This transfer will be necessarily asymmetrical as the immense majority of the more than 8,000 municipalities in Spain lack the minimum critical mass in terms of population and management capacity for these tasks, and so these formulas could be tried out only in a relatively small number of municipalities. B. Selection Criteria 8.8 As mentioned, in the field of regional autonomies, the criteria for assigning different powers were not of a technical nature but rather responded to political developments. The national constitution regulates the matters that can be the object of decentralization and the institutional mechanisms for doing so. Concretely, two modalities were established for reaching the maximum ceiling of responsibilities: one that is quicker and that was used for the so-called Historical Communities and that gave rise to the asymmetry and a second - slower - one (that required a minimum of five years for implementation at the level of governments with common attributions). The Autonomy Statutes of every region (a sort of regional constitutions) approved in the last instance through laws of the national parliament, contained the matters and fields of competence in which the decentralized administration would be set. Nonetheless, they never set minimum requirements from a technical point of view and it was not necessary to accredit or assess ex-ante the management capacity required for service delivery. Therefore, political motivations were key to defining the set of responsibilities to be managed. 8.9 The financing system, excluding the specific case of the territories with foral rights (the Basque Country and Navarra), was adjusted over time through five-yearly reviews with a clear trend towards enhanced symmetry in the instruments, so that all regions have enjoyed increasing fiscal autonomy. The principal discrepancies are related to the scope of the leveling mechanisms (extent and criteria for distribution of unconditional transfers). 102 Annex 8: Decentralization in Spain C. Institutional Arrangements 8.10 The institutional relations between the different government levels in Spain are organized in a rather complex fashion. For example, once an agreement with a region has been reached to assume a new responsibility, the top-down transfer of responsibilities proceeds through a bilateral mixed commission (central government ­ regional government) to discuss the transfer rhythm as well as the technical conditions and valuation of the services for financing ends. Once the transfer is accepted by both parties, management of the transferred responsibility becomes an autonomous task of the regional government that then will make all relevant decisions as regards organization and financing matters. In the field of education, for example, the central administration reserves the capacity to establish the basic guidelines of the educational plans as well as the minimum salary for different employee categories. Regional governments, by contrast, define the final content of the curricula (respecting the minimum content defined by the central government) and establish the requirements for accessing the teachers' capabilities and for wage complements. In health, the situation is similar. The central government identifies the basic services to be provided by the national healthcare system, and the regional authorities are authorized to expand (and indeed, some do so) this set of services. There is no common salary framework for professionals in the healthcare sector who therefore receive a different compensation depending on the community in which they work. 8.11 The Fiscal and Financial Policy Council--composed of representatives of the central government and the autonomic governments--reviews and organizes the financing model. It also discusses and proposes modifications to be implemented by passing a law of the national parliament. This council is also responsible for coordinating the different government's indebtedness rules policies. Expenditures are also coordinated by intergovernmental administrative units which seek to build consensus on basic aspects of management of the principal decentralized services. Examples are the National Health Council or the University Coordination Council. In addition to these multilateral agencies, bilateral contacts between administrations exist as another mechanism. One practice consists of specific investment agreements for developing concrete programs in which the three government levels participate (central, autonomous, and municipal levels); another are the program agreements for developing specific objectives of a multi-annual nature. 8.12 For municipalities the Spanish Federation of Municipalities and Provinces (FEMP) advocates on behalf of local government interests with respect to municipal autonomies and funding. Its leaders--representatives of the municipalities--are elected democratically at the local level. They act as a basic interlocutor of the local governments. Similarly, there is a National Local Administration Commission that is a permanent agency for collaboration between the central state administration and the local administration. It too is composed of an equal number of representatives of both administrations. 8.13 Finally, the relations between the regional governments and local governments are mostly informal based on bilateral agreements and primarily for projects fomented by the regional governments with conditioned funds. Water supply, public transportation, solid waste management, and even tax collection, are shared responsibilities; they are often managed through consortiums in which regional and municipal governments participate as well as private companies. 103 Annex 8: Decentralization in Spain D. Performance and Monitoring 8.14 Albeit that the Spanish decentralization model is based on a formally unitary State, in practice the system operates much like a federal administration. The principle of autonomy enshrined in the constitution rules out strong supervisory powers of the national government vis- à-vis subnational governments. Each government has full capacity and autonomy on internal organization, setting its own priorities and managing its budget--there is no interference from the other government administrations. Because of the "political" character of the decentralization model, in practice it is unrealistic that the national government could re-centralize any of the responsibilities transferred to regional governments, even if there were hard evidence and technical criteria regarding the efficiency and/or equity in management of some of these responsibilities that would demonstrate the usefulness to proceed that way. 8.15 Indeed, the only example of a specific attempt to re-centralize in recent years is related to coordination of the deficit and debt within the context of requirements of the European Union's Stability and Growth Plan. The operational implementation of this Plan required the establishment of deficit limits and indebtedness targets for the subnational level. Budget Consolidation Scenarios were prepared in which each regional government committed not to exceed certain debt and deficit limits. However, the incompliance of these targets by some communities was never subject to any sanctions. 8.16 Still, this does not mean that the central administration does not monitor developments at the subnational level. Recently, the cabinet has approved a Royal Decree to create the State Agency for the Evaluation of Public Policies and Quality of the Services, the central mandate of which is to gather information on the effects and results of public programs, as well as to improve accountability to citizens. Some of the concrete objectives in this sense are: · To promote an evaluation culture and practices. · To elaborate information systems, methodologies, and indicators. · To improve the quality of the services as a commitment to citizens. · To analyze the activities and services State Agencies provide to citizens. 8.17 This new Agency does not replace the decision-making capacity of the public administrators. The idea is to provide independent technical assessments, with well-founded and rigorous recommendations so that the decision-makers could take their decisions in better conditions. 8.18 From the central administration, the Evaluation Agency will contribute to: · Improve public services, raising the quality and response to social demands. · Improve the design of policies and programs, analyzing the obtained results and impact. · Rationalize public spending, optimizing resources and the more efficient use thereof. · Increase accountability and the democratic quality, promoting transparency, participation of and consultation with citizens and stakeholders. · Enhance competitiveness and productivity of the Spanish economy, eliminating obstacles for entrepreneurs and promoting improvements in the regulation. 104 Annex 8: Decentralization in Spain 8.19 Though in principle, the Agency is set up to evaluate the services provided by the central administration, the central government wishes to widen its scope of activity to the autonomous communities. In this case the autonomous communities would sign a collaboration agreement with the Agency; they also appoint representatives to participate in the Directing Council, the highest governing body of the Agency, which is also made up of representatives of the Ministries of Public Administrations, Economy and Finance, the Presidency, Foreign Affairs and Cooperation, the union organizations as well as renown independent professionals.49 8.20 The auditor general and the audit offices--at the central administrative and regional level respectively--are other institutions charged with the explicit objective of controlling the legality and efficiency of public acts. These institutions have only tackled the issues of legality, and thus far their efficiency controls have been ineffective. 8.21 Two examples of situations in which decentralization has not had the expected impact in terms of efficiency are the following: the case of the universities (at the autonomic-regional level) and the case of urban planning (at the municipal level) 8.22 The Case of the Universities. Spain is a country in which most higher education is public. Even though there are private universities most students (85 percent of the total number) attend public universities. When management of the universities was transferred to the autonomous communities, different counter-efficient effects occurred, partly because of the double conflict related to autonomy. · Although there are general guidelines for drafting and designing curricula for the different studies and there are minimum payment levels for all teachers in the entire country, the autonomous communities are independent in their decision as to the number of universities operating in their jurisdiction; the studies that are being offered; and the definition of teacher salaries. · On the other hand, the Constitution guarantees the universities' own "autonomy", which implies, among others, autonomy in terms of self-organization and budget allocation. 8.23 The first result of decentralization is the increase in the number of universities and of the number of associated centers. This is not something that is either good or bad in itself--one could argue that getting the services nearer the users is good provided the quality level is maintained--but in fact this has generated a series of problems. One of them is related to the proliferation of university centers and student selection norms which, in turn, have led to the hiring of teachers with a lower profile. This inevitably deteriorates education quality and learning outcomes. 8.24 Although not the only reason for low performance, decentralization plays an important role in explaining declining service quality of universities. In the smaller jurisdictions, the regional governments often do not have the capacity to resist the pressure from professional and business interest groups to establish more and more university centers--justified on the grounds of "proximity to the users". This then challenges the capacity of universities to employ highly qualified staff and teachers in all areas. Some independent studies on efficiency of the universities, as well as some internal evaluations, have concluded that a significant number of the 49. For further details on operation and the organizational structure, please consult the Agency's webpage: http://www.map.es/ministerio/organismos/Agencia-de-Evaluacion.html 105 Annex 8: Decentralization in Spain new universities created within the framework of the decentralization are operating in clearly unsatisfactory conditions. 8.25 The Case of Urban Planning. In Spain, urban planning is a municipal responsibility although it is subject to a certain degree of supervision from the regional government. It seems logical that municipalities are responsible for territorial organization of the cities (zoning), i.e. for deciding what part of the city is reserved for residential or industrial uses, and so forth. However, regulation of urban planning is not always efficient, mainly due to political circumstances. 8.26 Recently, the media have covered a series of cases in which ambitious urban development plans were proposed and submitted by private construction companies. If implemented, these plans would lead to a duplication, triplication or even higher multiplication of the number of residents in some municipalities. Certainly, this type of proposals is attractive for any local politician as the local population will have jobs for a considerable period of time and as the new residents will increase the municipality's tax revenue. But in most cases, these urban development plans do not sufficiently account for the conditions required to deliver good public services; this has caused significant problems related to the provision of water, waste collection, and others. These problems are particularly severe in smaller municipalities--of the approximately 8,000 municipalities in Spain, almost 7,000 have less than 5,000 inhabitants-- both for reasons related to economies of scale, and for the capacity of a small municipality to resist the political pressures exercised by powerful groups. 8.27 In sum, responsibilities that could be clearly decentralized on efficiency grounds have led to service delivery problems in cases where there was insufficient regulation. E. Benefits 8.28 For Spain as a whole there does not exist an overall evaluation of the level of efficiency achieved with the asymmetric decentralization process. Clearly though, some autonomous governments have benefited from the asymmetric strategy and these are typically located in the foral territories (the Basque Country and Navarra). 8.29 As was stressed earlier however, decentralization was not promoted by the Central Administration as part of a program that focused on detecting management capacities in the regional governments that initially assumed new responsibilities; rather it was a response in politically turbulent times--at the end of a dictatorship and transition to democracy--to the demand for self-government of citizens particularly in jurisdictions with a strong regional identity. In essence, asymmetric decentralization was a political decision. Therefore, applying the Spanish decentralization model to other countries may be a risky undertaking. 8.30 Most studies that have tried to assess the level of efficiency of services provided by the subnational governments have put more emphasis on a comparison of different management forms (direct, shared or hiring of the private sector) as opposed to the type of responsible government (national versus subnational). Studies still cannot provide hard evidence that the fiscal capacity of the jurisdiction is an important determinant of the level of efficiency; by contrast, most studies corroborate that the level of competence in the operation matters most. 106 Annex 8: Decentralization in Spain 8.31 On the other hand, the analysis of the efficiency of the asymmetric decentralization model also needs to take into account the political and institutional constraints and how the incentive schemes have affected the agents' behavior. The final judgment in this matter would probably not be very favorable for the model's asymmetrical nature. 8.32 Today there seems to be widespread belief however, that the principal benefit resulting from the process consisted in showing that it is easier to respond to citizens' needs from lower levels of government. Opinion polls show that a large majority of the population supports this process and is, all in all, satisfied with the administrative organization that was selected. It is also most likely true that the asymmetrical nature of the process has helped to highlight the management capacity of pioneering regional governments which spearheaded this effort--and these then constituted examples for other jurisdictions that tried to imitate their behavior by widening the scope of responsibilities and services. F. Problems 8.33 Despite the above-mentioned benefits, it is clear that the decentralization was challenging from several perspectives. One critical element was that the model and financing framework discouraged fiscal responsibility of the regional governments. Financing was primarily based on unconditional transfers--the regional governments obtained resources but the center had to come up with the political cost. Thus, the governors of the autonomous communities were responsible for managing public services--from which, in most cases, the population benefited considerably--but they did not need to assume the political cost of collecting taxes. The regional government could then justify shortcomings in services on the grounds of insufficient funding, ultimately blaming the Central Administration. 8.34 This dilemma was addressed partially when regional governments were transferred new taxes and capacity for regulation. However, this led to a new conflict related to the trade-off between increasing local fiscal autonomy and horizontal equity. Territorial solidarity is enshrined in the Spanish Constitution, but this principle was challenged precisely through the asymmetrical strategy. 8.35 The transfer of responsibilities in education and health50 to all Autonomous Communities required solidarity mechanisms through compensatory transfers so as to ensure feasibility in the provision of these services in all jurisdictions regardless of their fiscal capacity. The unequal distribution of schools and hospitals and the regional governments' decision-making capacity on the levels of coverage of the services (beyond the guaranteed minimum levels) generate financial tensions, partly because organizational and wage differences tend to exercise upward pressure on public expenditures. G. Lessons learned 8.36 The principal problems in the asymmetrical decentralization in Spain were due to the political nature of the process. All in all however, there is hardly any doubt that overall outcome 50. In Spain, education is mandatory and free of cost until the age of 16; there is a national healthcare system that ensures generalized access to medical and hospital care. 107 Annex 8: Decentralization in Spain of the process can be considered positive, insofar as the administration of a large size of public expenditures has been decentralized, at the same time that the Welfare State was consolidated. This was achieved through broad consensus and political agreements which tried to correct shortcomings of the system as they appeared while respecting the local idiosyncrasy of the jurisdictions. 8.37 The initial asymmetry of the decentralization model later disappeared as more and more local governments adopted new responsibilities. But it was this initial element of asymmetry which clearly helped to contrast and compare different management models which, in general, have been well-accepted by the population, although this had led to certain inequalities which continue to put pressure on the financing system. In this sense, the primacy of the principle of autonomy over that of efficiency has clear political advantages in an environment where there is regional resentment due to inequality. This is particularly evident in the case of the territories with foral rights (the Basque Country and Navarra), whose privileged funding system has allowed these jurisdictions to have better conditions to access public goods and services. 8.38 If there were countries who consider to designing an asymmetric decentralization process from scratch the Spanish experience offers the following lessons. Considering there are no identity or self-management problems in the subnational governments, probably the best recommendation - from a strictly economic point of view and in light of gaining efficiency ­ would be to try to meet the following conditions: · Identify the subnational governments in which there are already or in which it is possible to create institutions that could autonomously manage the services subject to decentralization with a higher probability of efficiency. · Start a transfer process based on criteria for funding the actual cost of the transferred services. · Implement a funding system that contains relevant fiscal responsibility and accountability elements for the subnational governments, i.e. an own fiscal space with tax revenues on which the subnational governments have some degree of normative capacity. · Maintain centralized control of the process, including explicit clauses on reversibility and re-centralization of transferred services that do not attain a set of minimum efficiency objectives agreed upon ex-ante. In this sense, it would be convenient to reach agreements with subnational governments on the core principles of the process, primarily related to the desirable degrees of local autonomy and the levels of interterritorial equity. 108 Bibliography Acosta, and Bird. 2005. "The Dilemma of Decentralization in Colombia." In Richard Bird, James Poterba, and Joel Slemrod, eds., Fiscal Reform in Colombia. Cambridge: MIT Press. Ahmad, E., and K. Baer. 1997. "Colombia." In Teresa Ter-Minassian, ed., Fiscal Federalism in Theory and Practice. Washington, D.C, International Monetary Fund: 457-503. Alesina, Alberto, et al. 2000. "Decentralization in Colombia." Working Paper No. 15. Fedesarrollo, Bogotá. Baron, D. Juan and R. A Meisel. 2003. "La Descentralización y las Disparidades Económicas Regionales en Colombia en la Década de 1990." Centro de Estudios Económicos Regionales. Banco de la Republica, Cartagena. Baron, G. 2007. "Cuentas de Salud de Colombia 1993-2003." Ministerio de la Protección Social y Departamento Nacional de Planeación, Bogotá. Barrera, F., and A. Gaviria. 2003. "Efficiency of Colombian Schools." Working Paper. Fedesarrollo, Bogota. Becerra, Marcelo, S. España y A. Fiszbein. 2003. "Enfoques Sobre la Eficiencia del Gasto en Educación Básica en la Argentina." Documento de Trabajo No. 6/03. Banco Mundial, Washington, D.C. Bird, Richard, et al. 2005. "Fiscal reform in Colombia: Problems and Prospects." Cambridge: MIT Press. Bird, Richard and A. Fiszbein. 1996. "Fiscal Decentralization in Colombia: The Central Role of the Central Government." Paper prepared for Conference on Fiscal Decentralization in Developing Countries, September 19-20, Montreal. Bird, Richard, and E. Snack. 1983. "Local Response to Intergovernmental Transfers: The Case of Colombia. Public Finance, 38(3): 429-439. Bonet, Jaime. 2006. "Desequilibrios regionales en la política de descentralización en Colombia." Documentos de Trabajo sobre Economía Regional No. 20. Banco de la República, Cartagena. Burki, Shahid Javed, G. E. Perry, and W. R. Dillinger. 1999. "Beyond the Center: Decentralizing the State." World Bank, Washington DC. Cardenas, Gaviria, Meléndez. 2005. La infraestructura de transporte en Colombia. Fedesarrollo. Carrizosa, Mauricio, Christian González, and Steven B. Webb. 2006. "Taxing and Spending in Colombia: Reducing Fiscal Rigidities to Enhance Equitable Growth." Colombia Policy Note, Washington, D.C. Cano, Rodolfo, y Luis Fernando Ramírez. 2006. "Descentralización Fiscal y Eficiencia en los Servicios Sociales a Nivel Departamental y Distrital en Colombia (2001-2005)." CEPAL, Seminario Regional de Política Fiscal. Universidad La Salle. 109 CEPAL. 2005. "La institucionalidad social en Colombia. Desarrollo Humano en Búsqueda de una Descentralización con Centro y Sentido." Serie Políticas Sociales, No. 91. Santiago, Chile. Chamorro, Luz Helena. 1997. "Balance del Proceso de Conformación de Regiones en Colombia." Planeación y Desarrollo, 28(4):115-150. Chaparro, Juan Camilo, Michael Smart, and Juan Gonzalo Zapata. 2005. "Intergovernmental Transfers and Municipal Finance in Colombia." In Richard Bird, James Poterba, and Joel Slemrod, eds., Fiscal Reform in Colombia. Cambridge: MIT Press: 247-286. Coelli, Timothy et al. 2005. "An Introduction to Efficiency and Productivity Analysis." 2nd edition, New York: Springer. CONPES. 1997. "Estrategias para el fortalecimiento de la Red Secundaria y Terciaria de Carreteras." Documento 2965, Noviembre. CONPES. 1997. "Avances del Plan Nacional de Desarrollo en el Sector Transporte." Documento 2949, Septiembre. CONPES. 1997. "Avances del Programa de Descentralización vial. Red Secundaria y Terciaria." Documento 2911, Marzo. Contraloría General de la República. 2006. "Evaluación de las transferencias intergubernamentales (1994-2005)." Agosto, Bogotá. Corporación Transparencia por Colombia. 2005. "Indice de Transparencia Departamental, Resultados 2004-2005." Bogotá. Dalmau, M. and J. Puig-J. 1998. "Market Structure and Hospital Efficiency: Evaluating Potential Effects of Deregulation in a National Health Service." Review of Industrial Organization, 13(4) August: 447-66. Di, Tella R., and W.D. Savedoff . 2001. "Diagnosis Corruption: Fraud in Latin America's Public Hospitals." Inter-American Development Bank, Washington, D.C. Dillinger, William and Steven B. Webb. 1999. "Decentralization and Fiscal Management in Colombia." Policy Research Working Paper 2122, Washington, D.C. DNP. 2000. "Desempeño Fiscal de los Departamentos y Municipios Año 2005 y Comparativo 2004." Dirección Nacional de Planeación, Bogotá. DNP. 2002. "Evaluación de la Descentralización Municipal en Colombia: Balance de Una Década." Tomo I. Marco conceptual y resultados de progreso municipal, Bogota. DNP. 2003. "Resultados del Desempeño Fiscal de los Departamentos y Municipios Año 2002 y Comparativo 2000." Bogotá. DNP. 2004. "Balance del Desempeño Fiscal de los Municipios y Departamentos 2000-2003." Bogotá. DNP. 2005a. "Evaluación del Sistema General de Participaciones." Dirección de Desarrollo Territorial Sostenible, Bogotá. DNP. 2005b. "Metodología para la Medición y Análisis del Desempeño Municipal." Bogotá. 110 DNP. 2005c. "Los Municipios Colombianos hacia los Objetivos de Desarrollo del Milenio." Salud, Educación y Reducción de la Pobreza, Bogotá. DNP. 2006a. "Bases del Plan Nacional de Desarrollo." Bogotá. DNP. 2006b. "Incidencia del Gasto Público Social sobre la Distribución del Ingreso y la Reducción de la Pobreza. Misión para el Diseño de una Estrategia para la Reducción de la Pobreza y la Desigualdad." Departamento Nacional de Planeación, Bogotá. DNP. 2006b. "2019 -- Visión Colombia II Centenario." Propuesta para Discusión, Presidencia de la Republica. Dirección Nacional de Planeación, Bogotá. DNP. 2007. "Operaciones Efectivas de Caja de los Gobiernos Departamentales." Consolidada por el DNP a partir de información de la Contraloría. DNP. 2008. "Evaluación del Desempeño Integral de los Municipios 2006". Dirección Nacional de Planeación, Bogotá. Duarte, J. 1996. "La Debilidad del Ministerio de Educación y la Politización de la Educación en Colombia: Dos Problemas a enfrentar en el Plan Decenal." Coyuntura Social, No.14, May. Escobar, M. L. and P. Panopoulou. 2002. "Health chapter" In M. Giugale, O. Lafourcade, and C. Luff, eds., Colombia: The Economic Foundation of Peace. World Bank, Washington, D.C. Escobar, Maria Luisa. 2005. Health Sector Reform in Colombia. Development Outreach. May. At: ww1.worldbank.org/devoutreach/may05/article.asp?id=295 - 28k. Escobar-Lemmon, Maria C. 2003. "Political Support for Decentralization: An Analysis of the Colombian and Venezuelan Legislatures." American Journal of Political Science 47(4):683- 697. Escobar-Lemmon, Maria C. 2006. "Executives, Legislatures, and Decentralization." The Policy Studies 34(2):245-263. Fedesarrollo. 2006. "Modificación al Régimen de Transferencias: Una Reforma Inaplazable, Economía y Política." No. 23, Bogotá. García, F., C. Marcuello, G. Serrano, y O. Urbina. 1996. "Evaluación de las Eficiencia en Centro de Atención Primaria." Una aplicación del Análisis Evolvente de Datos. Revista Española de Salud Pública, No. 70: 211-220. Gaviria, A. y J. Barrientos. 2001. "Determinantes de la calidad de la educación en Colombia." Departamento Nacional de Planeación, Archivos de Economía 159, Noviembre. Gaviria, A. y C. Umaña. 2002. "Estructura salarial de los docentes públicos en Colombia." Coyuntura Social, No. 26, May. Gaviria, Alejandro, Carlos Medina, and Carolina Mejía. 2006. "Evaluating the Impact of the Health Care Reform in Colombia: From Theory to Practice." Documento CEDE, 2006-06. Giedión, U. and L. Morales. 1999. "Aproximación a la Medición de la Eficiencia Económica y Eficiencia de la Gestión en los Hospitales Públicos del Distrito Capital." Secretaría Distrital de Salud, Bogotá. 111 Giugale, Marcelo, and Webb. 2000. "Achievements and Challenges of Fiscal Decentralization, Lessons from Mexico." World Bank, Washington, D.C. Gobernación de Antioquia. 2006. "Informe de Gestión de la Inspección y Vigilancia a los Actores del Sistema General de Seguridad Social en Salud ­ Régimen Subsidiado." Dirección Seccional de Salud de Antioquia .Grupo de Inspección y Vigilancia a los Actores del SGSSS- Régimen Subsidiado. Gonzalez, Christian Y., David Rosenblatt, and Steven Webb. 2002. "Stabilizing Intergovernmental Transfers in Latin America: A Complement to National/Subnational Fiscal Rules?" Policy Research Working Paper 2869. World Bank, Washington, D.C. Granada, A., V. Estrada, and D. Pinto. 2005. "Availability of Information, A Bottleneck for Health Research and Policy Evaluation?." Revista de Salud Pública, 7(2): 201-214. GES. 2006. "Resultados Financieros del Seguro Público." Grupo de Economía de la Salud. Gutiérrez. 2007. "¿Lo que el viento se llevó? Los partidos políticos y la democracia en Colombia 1958-2002." Editorial Norma. Hanushek, E. 1986. "The Economics of Schooling: Production and Efficiency in Public Schools." Journal of Economic Literature, 24: 1141-1177. Hanushek, E. 1989. "The Impact of Differential Expenditures on School Performance." Educational Research, 18: 45-51. Heinesen, E., and B.K. Graversen. 2005. "The Effect of School Resources on Educational Attainment: Evidence from Denmark." Bulletin of Economic Research, 57(2): 109-43, April. Herrera, S. and P. Gaobo. 2005. "Efficiency of Public Spending in Developing Countries: An Efficiency Frontier Approach." World Bank, Washington, D.C. Herrera, Santiago, and Gaobo Pang. 2005. "Efficiency of Public Spending in Developing Countries: An Efficiency Frontier Approach." World Bank, Washington, D.C. Herrera, Vicente, y Ariel E. Cortez. 2000. "Análisis de la Descentralización de la Política Social y la Municipalización del Sistema General de Seguridad Social en Salud en Municipios de Categoría 4,5 y 6." ESAP-CINAP, Pág. 127. Bogotá. Hernández, German Humberto. 2004. "Impacto de las Regalías Petroleras en el Departamento del Meta, Ensayos sobre Economía Regional." Banco de la República, Villavicencio. Inventarios Viales Departamentales. 2007. "Reportes de los departamentos al Ministerio de Transporte." INVIAS. 2006. "Inventario Red Troncal y Red Terciaria a Cargo de la Nación." Bogotá. Iregui B., Ana María, Ligia Melo, and Jorge Ramos. 2006. "Evaluación y Análisis de Eficiencia de la Educación en Colombia." Bogotá. Iregui, A., L. Melo, and J. Ramos. 2006. "Evaluación y Análisis de Eficiencia de la Educación en Colombia." Banco de la República. Borradores de Economía 381, Febrero. Jaramillo, I. 2001. "Evaluación de la Descentralización en Salud en Colombia." Departamento Nacional de Planeación, Archivos de Economía 170, Deciembre. 112 Latin Source. 2006. "Tax Reform Country." August. Laverde, Martha. 2004. "Education Reform in Antioquia: A Relevant Experience for Colombia." En Breve, No. 43, March. World Bank, Washington, D.C. Marcelo, D., and N. Ariza. 2005. "Evolución de los resultados de la educación en Colombia (1997-2003)." Departamento Nacional de Planeación. Archivos de Economía 286, July. Martinez, Ana Constanza. 2006. "Determinantes del PIB per Cápita de los Departamentos Colombianos 1975-2003." Archivos de Economía, No. 318. Bogotá. McEwan, Patrick J., and Luis Benveniste. 2001. "The Politics of Rural School Reform: Escuela Nueva in Colombia." Journal of Education Policy, 16(6): 547-559. McPake, Barbara, et al. 2003. "Is the Colombian health system reform improving the performance of public hospitals in Bogotá?" Health Policy and Planning, 18(2): 182-194. Melo, Ligia. 2002. "The Flypaper Effect under Different Institutional Contexts: The Colombian Case." Public Choice, 111: 317-345. Miika, L. and U. Hakkinnen. 2000. "Determinants of Cost Efficiency of Finish Hospitals: A Comparison of DEA and SFA." National Research and Development Center for Welfare and Health, Finland. Mina, A. 2004. "Factores Asociados al Logro Educativo a Nivel Municipal." Universidad de los Andes, Documento CEDE 2004-15, March. Ministerio de Hacienda. 2006. "Informe sobre la viabilidad fiscal de los departamentos, Vigencia 2005." Bogotá. Ministerio de la Protección Social. 2006. Informe de Actividades al CNSSS 2005-2006. Moncayo, Edgard. 2002. "Las políticas regionales en Colombia." Archivos de Economía (7 Separatas). Departamento Nacional de Planeación, Bogotá. Monsalve, Alfonso and Irene Gaviria (eds.). 2004. "Colombia, Democracia y Paz." Tomo VIII. Universidad de Antioquia, Medellín. Naranjo, Garcia, Diana Maria, Beatriz Eugenia Rigol Torres, and Luis Fernando Suárez Vélez. 2001. "Manual de Inducción para los Representantes de las Juntas Directivas de las Empresas Sociales del Estado." Gobernación de Antioquia, Dirección Seccional de Salud de Antioquia. Nuñez, J., R. Steiner, X. Cadena, and R. Pardo. 2002. "¿Cuáles colegios Ofrecen Mejor Educación en Colombia?" Departamento Nacional de Planeación. Archivos de Economía 193, June. O'Neill. 1998. "Understanding the Great Power-Giveaway: Decentralization in the Andes." Paper presented to the 1998 Meeting of the Latin American Studies Association, Chicago. Parandekar, Suhas. 2006. "The Efficiency of Brazilian Municipalities." March. World Bank, Washington, D.C. Peñaloza, M.C. 2003. "Evaluación de la Eficiencia en Instituciones Hospitalarias Públicas y Privadas con Data Envelopment Analysis." Departamento Nacional de Planeación. Archivos de Economía 244, Deciembre. 113 Pérez, G.V. 2005. "La infraestructura del transporte vial y la movilización de carga en Colombia." Documentos de trabajo sobre economía regional, no. 64, Banco de la República, Cartagena de Indias. Piñeros, L. and A. Rodríguez. 1998. "Los Insumos Esocolares en la Educación Secundaria y su Efecto Sobre el Rendimiento Académico de los Estudiantes: Un estudio en Colombia." LCSHD Paper Series, No. 36, December. World Bank, Washington, D.C. Pinto, D., G. Carrasquilla, F. Gil, C. Collazos C, and J. Rincón. 2005. "Certification as an Indicator of Health Decentralization. A Look at Fulfillment of Certification Requisites in Colombian Municipalities". Revista Gerencia y Políticas de Salud, 4(8): 149-162. Pinzon, M. 2003. "Medición de Eficiencia Técnica Relativa en Hospitales Públicos de Baja Competitividad Mediante la Metodología de Data Envelopment Análisis." Departamento Nacional de Planeación. Archivos de Economía 245, December. Plaza, Beatriz, et al. 2001. "Managed Competition for the Poor or Poorly Manager Competition? Lessons from the Colombian Health Reform Experience." Health Policy and Planning, 16 (Suppl. 2): 44-51. Procuraduría General de la Nación. 2006. "El Derecho a la Educación." Bogotá. Programa Presidencial de Modernización. 2005. "Las Regalías Colombianas. Una herramienta para el fortalecimiento del control social." DNP, Bogotá. Ramírez M., J. M. Gallego, and C. Sepúlveda. 2004. "The Determinants of the Health Status in Developing Countries: Results from the Colombian Case." No. 41. Facultad de Economía, Universidad del Rosario. República de Colombia. 2002. "Autorización a la Nación para Contratar Empréstitos Externos con la Banca Multilateral hasta por US$100 Millones, o su Equivalente en Otras Monedas, Destinados a Financiar Parcialmente la Primera Fase del Proyecto de Modernización de Redes de Prestación de Servicios de Salud." Documento CONPES, Social 3175. República de Colombia. 2002. "Política de Prestación de Servicios para el Sistema de Seguridad Social en Salud y Asignación de Recursos del Presupuesto General de la Nación para la Modernización de los Hospitales Públicos." Documento CONPES, Social 3204. República de Colombia. 2006 "Concepto Favorable a la Nación para Contratar Empréstitos Externos con la Banca Multilateral hasta por US$50 Millones, con el Fin de Financiar Parcialmente el Programa de Reorganización, Rediseño y Modernización de Redes de Prestación de Servicios de Salud para las Vigencias 2006­2007." . Documento CONPES, Social 3415. Republica de Colombia. 2007. "Presupuesto 2007." Departamento de Antioquia. Restrepo, Darío, et al. 2006. "Impacto del Acto Legislativo 01 de 2001." Universidad Nacional, CID-FCE-GTZ (processed data set). Rodriguez, Oscar (ed.). 2002. "Los Avatares del Servicio de Salud para Pobres." Colección La Seguridad Social en la Encrucijada, Volumen 2. Universidad Nacional de Colombia, Bogotá. 114 Rojas, Fernando. 2003. "At the Crossroads of Decentralization, Recentralization, Federalization." In Marcelo M. Giugale, et al., eds., Colombia--The Economic Foundations of Peace, 871-896. World Bank, Washington, D.C. Rojas, Fernando, and Jonas Frank. 2003. "Assessment of the Impact of Decentralization: The Case of Colombia (1991-2001)." Public Sector Group/Decentralization Thematic Group, Funding 2002-2003. World Bank, Washington, D.C. Rosenbaum, John Sudarsky. 2001. "El Capital Social de Colombia." Departamento Nacional de Planeación. Ruiz, F. 2005. "Social Health Insurance and Access: Results from a Colombian Longitudinal Study." Saenz, L. 2001. "Modernización de la Gestión Hospitalaria Colombiana: Lecciones Aprendidas de la Transformación de los Hospitales en Empresas Sociales del Estado." LAC RSS No. 46, Febrero. Sanchez, Fabio, and Catalina Gutierrez. 1994. "La Descentralización Fiscal en América Latina: Problemas y Perspectivas. El caso de Colombia." Banco Interamericano de Desarrollo. Sánchez, L.E., and F. Yepes. 1999. "La Descentralización de la Salud en Colombia." Estudio de Casos y Controles. ASSALUD, Asociación Colombiana de la Salud. Sánchez, L.E., F. Yepes, and B. Cantor. 1998. "La Descentralización de la Salud: El Caso de Tres Municipios Colombianos." (Informe Técnico) Bogota D.C. ASSALUD, Asociación Colombiana de la Salud. Sanchez Torres, Fabio, Catalina Gutierrez Sourdis, y Juan Carlos Parra. 1995. "Transferencias Intergubernamentales y Comportamiento Fiscal de Los entes Territoriales. Una Aproximación Econométrica, Coyuntura Económica." 89-109. Sanchez, F., O. Nupia, and P. Urdinola. 2000. "Gasto Público, Eficiencia y Cobertura del Sector Salud." CEDE, Universidad de los Andes. Sánchez, Fabio, and Mario Chacón. 2005. "Conflicto, Estado y Descentralización: Del Progreso Social a la Disputa Armada por el Control Local, 1974-2002." Documento CEDE, No. 33, Bogotá. Sarmiento, A., C. Alonso, G. Duncan, and C. Garzón. 2005. "Evaluación de la Gestión de los Colegios en Concesión en Bogotá 2000-2003." Departamento Nacional de Planeación, Archivos de Economía 291, September. Sarmiento, A., W. Castellanos, A. Nieto, C. Alonso, and C. Perez. 2005. "Análisis de Eficiencia Técnica de la Red Pública de Prestadores de Servicios Dentro del Sistema General de Seguridad Social en Salud." Departamento Nacional de Planeación, Archivos de Economía 298, Deciembre. Sarmiento, A., L.P. Tovar, and C. Alam. 2002. "Situación de la Educación Básica, Media y Superior en Colombia." El Tiempo, Fundación Corona, Fundación Antonio Restrepo Barco, (second edition). 115 Slack, Enid N., and Richard M. Bird. 1983. "Local Response to Intergovernmental Fiscal Transfers: The Case of Colombia." Public Finance, 38: 429-439. Small, Kenneth A., Clifford Winston, and Carol A. Evans. 1989. "Road Work. A New Highway, Pricing & Investment Policy." The Brookings Institutions. Sudarsky, John. 2001. "El Capital Social de Colombia." Dirección Nacional de Planeación, Bogotá. Umaña, C. 2004. "Esquema de Incentivos Para la Carrera Docente." Departamento Nacional de Planeación, Archivos de Economía 270, Octubre. Uribe, Maria Camila. 2006. "Property Tax in Colombian Municipalities: Tax Base and Institucional Issues." Lincoln Institute of Land Policy, August. Vargas J.E., and A. Sarmiento. 1997. "Descentralización de los Servicios de Educación y Salud en Colombia." DNP, Casa editorial el tiempo, Bogota. Vargas, J., and A. Sarmiento. 1997. "La Descentralización de los Servicios de Educación en Colombia." Comisión Económica para América Latina y el Caribe. Serie Reformas de Política Pública 50. Vergara, C. and M. Simpson. 2001. "Evaluación de la Descentralización Municipal en Colombia. Estudio general sobre los antecedentes, diseño, avances y resultados generales del proceso de descentralización territorial en el sector educativo." Departamento Nacional de Planeación, Archivos de Economía 168, December. Wiesner, Eduardo. 1992. "Colombia: Descentralización y Federalismo Fiscal, Informe Final de la Misión para la Descentralización." Presidencia de la República, Departamento Nacional de Planeación, Bogotá. Wiesner, Eduardo. 1995. "La Descentralización, el Gasto Social y la Gobernabilidad en Colombia" Departamento Nacional de Planeación, Asociación Nacional de Instituciones Financieras, Bogotá. Wiesner, Eduardo. 2002. "Transferencias, incentivos y la Endogeneidad del Gasto Territorial." Departamento de Archivos de Economía No. 174, Bogotá. World Bank. 1987 "Colombia--A Review of Recent Fiscal Decentralization Measures." Report No. 6631-CO, Washington, D.C. World Bank. 1989. "Colombia--Decentralizing Revenues and the Provision of Services: A Review of Recent Experience." Report No. 7870-CO, Washington, D.C. World Bank. 1995. "Colombia--Local Government Capacity: Beyond Technical Assistance." Report No.14085-CO, Washington, D.C. World Bank. 1996. "Colombia -- Reforming the Decentralization Law: Incentives for an Effective Delivery of Service." Washington, D.C. World Bank. 2004. "Colombia ­ Recent Economic Developments in Infrastructure (REDI), Balancing Social and Productive Needs for Infrastructure." Report No. 30379-CO, Washington, D.C. 116 World Bank. 2005a. "Colombia Public Expenditure Review." Report No.25163-CO, Washington, D.C. World Bank. 2005b. "Country Assistance Progress Report for the Republic of Colombia." Report No. 32999-CO, Washington, D.C. World Bank. 2006. "Colombia--Contracting Education Services." Report No. 31841-CO, Washington, D.C. World Bank. 2007. "Colombia 2006-2010: A Window of Opportunity." Colombia Policy Notes 2007, Washington, D.C. World Bank. 2008. "The Quality of Education in Colombia: An Analysis and Policy Agenda", Report No. 43906-CO, Washington, D.C. 117 MAP SECTION IBRD 33388R 75°W 70°W Puerto Bolívar C a r i bb e an S e a J IR A Ríohacha UA G To Santa Marta LA Maracaibo Barranquilla Pico Cristóbal Colón AT L Á N T I C O (5,775 m) COLOMBIA A Cartegena Valledupar N LE DA 10°N Maganqué CESAR Lago de AG Maracaibo M Sincelejo PA NA El Baneo SU SU Monteria M C CR AR Acandí NORTE DE E E Ocaña A SANTANDER LÍV Turbo A D OB To R.B. DE BO ÓR Mérida Cúcuta C VENEZUELA uca To Ca Cord C d Guasdualito Bucaramanga Yarumal le na Arauca iller da Socorro ANTIOQUIA Ma g ARAUCA CÁ Atra r ER a O AND Casanare PAC IF IC SANT to YA Occi Medellin Puerto BO Carreño OCEAN d i dental Quibdo Chiquinquirá Puerto S CASANARE CHOCÓ LDA Tunja Nueva CA Manizales CUNDINA- Yopal 5°N RISARALDÁ Santa Rita 5°N Cartago Pereira MARCA BOGOTÁ VICHADA Atacavi ta Armenia Me QUINDIO tal San Pedro Gaviotas Ibaque Giradot en Villavincencio Chaviva Vich a da ri al M A a O DISTRITO Buenaventura Buga ntr VALLE DEL LI Puerto TO Palmira Inirída Ce er CAPITAL ill CAUCA Cali M E TA rd a San Juan er Co de Arama ill Neiva Guaviare GUAINÍA rd CAUCA HULA Mapiripana Tabaquén Co San José Guapí Popayan del Guaviare Brujas San Vicente Garzon del Caguán Calamar Patía Tumaco Neg Florencia G U AV I A R E NARIÑO ro San Rafael Miraflores Mitu Pasto Cagu Mocoa Vau pés án Yavarate To Ipiales PUTUM CAQUETÁ VA U P É S Puerto AY Ibarra Asis O Macujer Puerto Huitoto 0°N Puerto 0°N Leguízamo Puerto Lérida E C U A DOR Pizarro Caq u etá Puerto Santander Locas de Ptu Cahuinari ma y CO LOMB I A o La Pedrera AMAZONAS El Encanto SELECTED CITIES AND TOWNS BRA ZI L DEPARTMENT CAPITALS NATIONAL CAPITAL PERU RIVERS This map was produced by the Map Design Unit of The World Bank. The boundaries, MAIN ROADS 0 80 160 240 320 Kilometers colors, denominations and any other information shown RAILROADS on this map do not imply, on Leticia the part of The World Bank 0 40 80 120 160 200 Miles Group, any judgment on the DEPARTMENT BOUNDARIES legal status of any territory, or any endorsement or INTERNATIONAL BOUNDARIES acceptance of such 75°W 70°W boundaries. FEBRUARY 2009