51513 Local Gains from Global Opportunities: Improving Central America’s Investment Climate September 15, 2008 Finance and Private Sector Unit Poverty Reduction and Economic Management Unit Latin America and the Caribbean Region Document of the World Bank FISCAL YEAR (January 1 – December 31) ABBREVIATIONS AND ACRONYMS A&A Accounting and Auditing ADR Alternative dispute resolution AML Anti-money laundering ANSI American National Standards Institute ASTM American Society for Testing and Materials ASYCUDA Automated System for Customs Data ATM Automatic Teller Machine BCP Basel Core Principles for Effective Bank Supervision BIPM Bureau International de Poids et Mésures, France BRC British Retailer Council BS British Standards BSE Bovine Spongiform encephalopathy CA Central America CBI Caribbean Basin Initiative CEAC Electrification Council for Central America CEHM Centro Hondureño de Metrología, Honduras CENAME National Metrology Institute Guatemala CENAMEP National Metrology Institute Panama CEPAL Economic Commission for Latin America & the Caribbean (ECLAC) CFT Countering the financing of terrorism CIPM Comité international des poids et mesures CMC Calibration and Measurement Capacity CNA Colegio Nacional de Abogados de Panamá COGUANOR Comisión Guatemalteca de Normas , Guatemala COHCIT Consejo Hondureño de Ciencia y Tecnología, Honduras COMTRADE United Nations Commodity Trade Statistics Database CONACYT Consejo Nacional de Ciencia y Tecnología, El Salvador COPANIT Comisión Panameña de Normas Industriales y Técnicas, Panamá COPANT Comisión Panamericana de Normas Técnicas CPSIPS Core Principles for Systemically Important Payment Systems CRI Costa Rica DB Doing Business DG SANCO Directorate General for Health & Consumer Protection, Costa Rica DIN Deutsches Institut für Normung e.v. – German Standards Institute DR-CAFTA Dominican Republic-Central American Free Trade Agreement DTNM Dirección Técnica de Normalización y Metrología, Nicaragua DvP Delivery versus payment EAP East Asia and Pacific ECA Ente Costarricense de Acreditación, Costa Rica Accreditation Body ECA Eastern Europe and Central Asia EFQM European Federation for Quality Management EFTPOS Electronic Funds Transfer Point of Sale EN European Standards Institute ENN Formerly INTECO, Standards Institute Costa Rica EPR Empresa Propietaria de la Red ES Enterprise Survey FDI Foreign Direct Investment FOCA Foro Centroamericano de Acreditación FSAP Financial Sector Assessment Program ii FSC Forest Stewardship Council FTA Free Trade Agreement FTA Free Trade Agreement FVO Food and Veterinary Office, DG SANCO, Costa Rica GCI Global Competitiveness Index GCR Global Competitiveness Report GDP Gross Domestic Product GLOBALGAP formerly EUREPGAP, private Retailer Standard with global extension GMP Good Manufacture Practices GSM Global System for Mobile Communications GTM Guatemala GTPIR Grupo de Trabajo Planificación Indicativa Regional HACCP Hazard Critical Control Point HSBC Hong Kong Singapore Bank Hi-OECD High Income OECD HND Honduras IAAC Inter American Accreditation Cooperation IADB Inter-American Development Bank IAF International Accreditation Forum IC Investment Climate ICA Investment Climate Assessment ICR Insolvency and Creditor Rights ICT Information and communications technologies IEC International Electrotechnical Commission IFRS International Financial Reporting Standards ILAC International Laboratory Accreditation Cooperation IMF International Monetary Fund INTECO Standards Institute Costa Rica IPPC International Plant Protection Convention ISA International Standards on Auditing ISO International Standards Organization ITU International Telecommunication Union LAC Latin America and the Caribbean LACOMET National Metrology Institute Costa Rica LANAMET National Metrology Institute Nicaragua LPI Logistics Performance Index MENA Middle East and North Africa MER Mercado Eléctrico Regional MIGA Multilateral Investment Guarantee Agency MLA Mutual Recognition Arrangement MOU Memoranda of understanding MSTQ Metrology, Standardization, Testing, Quality Assurance NAFTA North American Free Trade Agreement NAFTA North American Free Trade Agreement NIC Nicaragua NMI National Metrology Institute OECD Organization for Economic Co-operation and Development OGA Oficina Guatemalteca de Acreditación, Acreditación Body Guatemala OHA Oficina Hondureña de Acreditación, Honduras OHN Organismo Hondureño de Normalización, Honduras OIML Organisation Internationale de Métrologie Légale OLADE Organización Latinoamericana de Energía ONA-MIFIC Oficina Nicaragüense de Acreditación, Nicaragua ORT Oficina de Reglamentación Técnica, Costa Rica PAN Panama PISA Program for International Student Assessment iii PvP Payment versus payment QI Quality infrastructure R&D Research and development ROSC Reports on the Observance of Standards and Codes RTGS Real Time Gross Settlement S&T Science and technology SAIDI System Average Duration Index SAIFI System Average Interruption Frequency Index SAR South Asia Region SENASA Servicio Nacional de Salud Animal, Costa Rica SIECA Secretaría de Integración Económica Centroamericana SIEPAC Sistema de Interconexión Eléctrica para América Central SIM Sistema Interamericano de Metrología SLV El Salvador SME Small and Medium Enterprise SQAM Standards, Quality Assurance, Accreditation, Metrology SSA Sub-Saharan Africa STQM Standards, Testing, Quality Assurance, Metrology TBT Technical Barriers to Trade TFP Total factor productivity TTS Transferencia telefónica segura de fondos UNCTAD United Nations Conference on Trade and Development US United States WBI World Bank Institute WDI World Development Indicators WDR World Development Report WEF World Economic Forum WITS World Integrated Trade System WTO World Trade Organization Vice President : Pamela Cox Country Director : Laura Frigenti Sector Director : Marcelo Giugale Sector Manager : Lily L. Chu Lead Economist : Humberto Lopez Task Manager : David M. Gould Co-Task Manager : Veronica Alaimo iv Local Gains from Global Opportunities: Improving Central America’s Investment Climate Table of Contents ACKNOWLEDGEMENTS ................................................................................................... IX EXECUTIVE SUMMARY .......................................................................................................X CHAPTER 1. CENTRAL AMERICA’S MACROECONOMIC ENVIRONMENT AND THE ROLE OF THE INVESTMENT CLIMATE UNDER FREE TRADE .............1 INTRODUCTION..............................................................................................................................................1 1. STRATEGIC CONTEXT......................................................................................................................1 2. REGIONAL MACROECONOMIC ENVIRONMENT ..........................................................................4 3. TRADE AND FINANCIAL FLOWS ...................................................................................................11 4. LOOKING AHEAD ..........................................................................................................................14 CHAPTER 2. BENCHMARKING THE INVESTMENT CLIMATE: WHERE DOES CENTRAL AMERICA STAND? ................................................................................ 16 INTRODUCTION............................................................................................................................................16 1. BENCHMARKING CENTRAL AMERICA .........................................................................................17 2. MICRO DATA LESSONS ..................................................................................................................25 3. INVESTMENT CLIMATE, PRODUCTIVITY AND TRADE ..............................................................28 4. SUMMARY AND CONCLUSIONS .....................................................................................................33 CHAPTER 3. GOVERNANCE AND CORRUPTION: IMPLICATIONS FOR BUSINESS INVESTMENT ...................................................................................................35 INTRODUCTION............................................................................................................................................35 1. POLITICAL SYSTEM MEASURES ....................................................................................................38 2. OVERALL GOVERNANCE INDICATORS .......................................................................................40 3. BUSINESS-RELATED MEASURES ...................................................................................................42 4. CUSTOMS REGULATIONS -- A KEY ELEMENT OF THE DR-CAFTA AGENDA .....................54 CHAPTER 4. INFRASTRUCTURE...................................................................................60 INTRODUCTION............................................................................................................................................60 1. INFRASTRUCTURE OVERVIEW ......................................................................................................61 2. TRANSPORT SECTOR ......................................................................................................................64 3. ENERGY SECTOR ............................................................................................................................74 4. CHALLENGES AND STRATEGIES ..................................................................................................84 CHAPTER 5. FINANCIAL SECTOR ................................................................................88 INTRODUCTION............................................................................................................................................88 1. CENTRAL AMERICA’S FINANCIAL SECTOR .................................................................................90 2. FINANCIAL SECTOR REGULATION AND SUPERVISON ...............................................................96 3. CREDITOR RIGHTS AND INSOLVENCY ..................................................................................... 103 4. PAYMENT SYSTEMS ..................................................................................................................... 108 5. ACCOUNTING AND AUDITING STANDARDS ........................................................................... 114 v CHAPTER 6. INNOVATION, SKILLS, AND QUALITY ...............................................119 INTRODUCTION......................................................................................................................................... 119 1. HOW CENTRAL AMERICAN COUNTRIES PERFORM................................................................ 120 2. REASONS FOR LOW INNOVATION ............................................................................................ 127 3. POLICIES TO ADDRESS LACK OF INNOVATION ...................................................................... 138 BIBLIOGRAPHY.................................................................................................................. 162 Tables Table 1.1: Per capita GDP Growth in Central America, 2000-2007 (in percent)............... 4 Table 1.2: Volatility of per Capita GDP Growth, 2000-2006 (standard deviations).......... 9 Table 1.3: Structure of Merchandise Exports, 1996-2007 (Percentage)........................... 13 Table 1.4: Estimated Size of Remittances, 2005-2006 ..................................................... 14 Table 2.1: Central America’s Position in International Rankings .................................... 18 Table 2.2: Governance indicators ..................................................................................... 23 Table 2.3: Other investment climate indicators ................................................................ 25 Table 2.4: Multivariate analysis of IC variables and firm’s characteristics ..................... 27 Table: 2.5: Firm performance and the investment climate in Central America................ 30 Table 2.6: Firm productivity, trade liberalization, and the investment climate................ 32 Table 3.1: Confidence in institutions in 2004 and 2006 (scale: 0 – 100) ......................... 39 Table 3.2: Agreement with “A market economy is the most convenient for the country� over time (% that “agree� and “highly agree�)................................................................. 40 Table 3.3: Determinants of Bribes in Central American countries................................... 53 Table 3.4: Logistic Performance in Central America ....................................................... 55 Table 4.1: Analysis of Changes in Growth....................................................................... 62 Table 4.2: Comparative Survey on the Quality of Infrastructure ..................................... 66 Table 4.3: Logistics, transportation model ....................................................................... 67 Table 4.4: Percentage of firms using each type of transportation.................................... 68 Table 4.5: Theft or breakage during transportation .......................................................... 69 Table 4.6: Cargo volumes and market shares ................................................................... 70 Table 4.7: Containerized Cargo volumes (TEU) .............................................................. 70 Table 4.8: Characteristics of Central American Energy Regulatory Authorities ............. 81 Table 4.9: Possible regional hydro projects...................................................................... 82 Table 5.1: Financial depth indicators in Latin America, 2006 ........................................ 92 Table .5.2: Business environment indicators that can influence financial depth.............. 93 Table .5.3: Indicators of branch intensity across countries, 2003..................................... 94 Table 6.1: Use of ICT, 2005 ........................................................................................... 123 Table 6.2: Share of adult population in each skill group, 2004* .................................... 126 Table 6.3: World Economic Forum Innovation Indicators............................................. 127 Table 6.4: Returns to education, pooled years................................................................ 131 Table 6.5: Most Relevant Standards for Central America.............................................. 133 Table 6.6: Organizations With or In the Process of Accreditation, 2007 ....................... 134 Table 6.7: Recommendations for an Integrated Regional Quality Infrastructure System ......................................................................................................................................... 141 vi Figures Figure 1.1: Trade between the US and Central America before and after DR-CAFTA implementation (three-month moving average).................................................................. 2 Figure 1.2: Poverty (%) and Inequality (Gini %) ............................................................... 3 Figure 1.3: Per capita growth in Central America .............................................................. 5 Figure 1.4: Per capita growth in Central America .............................................................. 6 Figure 1.5: Sectoral GDP Shares in Central America, 1990-2006 ..................................... 7 Figure 1.6: Inflation and Fiscal Deficits ........................................................................... 10 Figure 1.7: Central American Trade Flows with the US and EU, 1996-2007.................. 11 Figure 1.8: Share of Central American Exports in World Exports, 1994-2006 ............... 12 Figure 2.1: Global Competitive Index, 2007-2008........................................................... 19 Figure 2.2: Obstacles for business according to entrepreneurs’ subjective perceptions... 21 Figure 2.3: Investment Climate constraints identified among the three most serious by entrepreneurs..................................................................................................................... 22 Figure 2.4: Differences in the investment climate by firm characteristics, Central America average, 2007 .................................................................................................................... 26 Figure 3.1: Percentile ranking of Central American countries and LAC average on governance dimensions..................................................................................................... 42 Figure 3.2: Central American country rankings by Doing Business topic (2008)............ 44 Figure 3.3: Firm-reported country values and LAC average on selected and aggregated indicators........................................................................................................................... 47 Figure 3.4: Homicide Rates in Central America, 2006..................................................... 50 Figure 3.5: Firms Security Costs and Losses Due to Crime as a Percentage of Firm Sales, 2006................................................................................................................................... 51 Figure 3.6: Bribe payments in Central American countries ............................................. 52 Figure 3.7: Tons of cocaine seized in Central American countries (rounded) ................. 57 Figure 4.1: Perceptions of Transport as a Major or Severe Constraint to Growth ........... 65 Figure 4.2: Overall Network and Paved Road coverage characteristics........................... 68 Figure 4.3: Vehicle fleets in Central America .................................................................. 68 Figure 4.4: Impact of Breakage and Theft in Transit on Costa Rican Firms.................... 69 Figure 4.5: Market share by Country................................................................................ 70 Figure 4.6: Total Cost per container, including Operating and Inventory Expenses caused by Inefficiency .................................................................................................................. 71 Figure 4.8: Total Cost per Container, Including Operating and Inventory Expenses Caused by Inefficiency ..................................................................................................... 71 Figure 4.9: Liner Connectivity Index of CA and Comparator Countries, 2006 ............... 72 Figure 4.10: Projected Port Throughout by Country and Coast ....................................... 73 Figure 4.11: Per capita consumption per month ............................................................... 74 Figure 4.12: Energy intensity (BEP/K US$)..................................................................... 75 Figure 4.13: Electric and Energy intensity ....................................................................... 75 Figure 4.14: Hydroelectric potential and net generation .................................................. 76 Figure 4.15: Electrification Index ..................................................................................... 77 Figure 4.16: Electricity tariffs and prices ......................................................................... 78 Figure 4.17: Peak Load/Effective Capacity ...................................................................... 79 Figure 4.18: Imports and Exports of Electricity ............................................................... 79 vii Figure 4.19: Electricity interruptions................................................................................ 80 Figure 4.20: Electrical Losses (%).................................................................................... 84 Figure.5.1: Indicators of the Banking Sector in Central America, 2000 & 2006 ............. 91 Figure.5.2: Trends in Financial Sector Depth, 2000 - 2006.............................................. 92 Figure.5.3: Trend in branch intensity in Central America, 2003-2007............................. 95 Figure 6.1: Patents Granted to Residents Per Million Labor Force (2000-2005 Average) ......................................................................................................................................... 121 Figure 6.2: Scientific Publications as a Percent of GDP, 2005 ...................................... 121 Figure 6.3: Percent of Firms that Introduced New Products .......................................... 122 Figure 6.4: Percent of Firms that Introduced New Processes......................................... 122 Figure 6.5: Expenditure on R&D as a % of GDP, latest year available ......................... 123 Figure 6.6: Percentage of Firms That Use Technology Licensed from a Foreign Company ......................................................................................................................................... 123 Figure 6.7: Percentage of Firms That See Inadequately Educated Workforce as a Major or Severe Obstacle............................................................................................................... 124 Figure 6.8: Number of Researchers, 2005 ...................................................................... 124 Figure 6.9: Secondary School Enrollment, 2005 ............................................................ 125 Figure 6.10: Mean Years of Schooling, Workers Age 25-65 (Latest Year Available) .. 125 Figure 6.11: Percentage of Firms Offering Training Programs for Employees ............. 126 Figure 6.12: ISO 9001:2000 Certifications per US$ Billion Industrial Value-Added, 2006 ......................................................................................................................................... 127 viii ACKNOWLEDGEMENTS This report was prepared by a team lead by David Gould (TTL) and Veronica Alaimo (co-TTL), comprising Josef Loening (Macro), Alberto Leyton and Amalia Prado (Governance), Enrique Fanta and Tanya Gupta (Customs), Gabriel Demombynes and Diana Hincapie (Crime), Emmanuel James (Transport), Fernando Lecaros and Patricia Rodriguez (Energy), Aquiles Almansi (Finance, with advice from Adolfo Rouillon (Legal), Mario Guadamillas (Finance), Henri Fortin (Accounting), and the collaboration of Cecilia Brady (consultant), Rashmi Shankar (Finance), and Daniel Alonso (Legal)), Susana Sanchez (Finance) and Thomas Haven (Innovation, with assistance from Jane Hwang). Ricardo Bebczuk prepared a background paper on access to finance in Guatemala and Nicaragua, and the German Metrology Institute (Physikalisch-Technische Bundesanstalt) with authors Reinhard Schiel, Clemens Sanetra, and Alexis Valqui prepared a background paper on quality infrastructure. Maria Ivanova Reyes Peguero provided excellent research assistance. Micky O. Ananth provided valuable editorial support. Peer reviewers are J. Luis Guasch (LCSSD), Pablo Fajnzylber (LCRCE), and Daniel Lederman (DECRG). The team benefited from insights provided in meetings with Lily Chu, Humberto Lopez, and Jordan Schwartz. The findings and views expressed here are exclusively those of the World Bank. ix EXECUTIVE SUMMARY Central America has made significant progress in stabilizing its economies after decades of civil war and economic volatility that plagued the region through the 1990s. 1 The region has taken important steps to improve the business climate by enhancing macroeconomic stability, reducing red-tape, simplifying the regulatory and tax framework, strengthening the legal framework for contracting and ensuring the soundness of the financial system. Indeed, real per capita grew an average of 3 percent in Central America over the last 5 years, which was the most robust and stable the region had seen since the early 1990s. Yet, there is significant room to improve. Despite Central America’s recent expansion, when compared to the rest of Latin America, other middle income countries, or East Asia, the region is falling behind. Relative to these country groupings, Central America’s per capita growth since 2005 has been around 0.5 to 2 percentage points below its comparators. A key factor contributing to Central America’s lagging growth is its relative investment climate challenges. The investment climate refers to key factors that shape the opportunities and incentives for firms to invest productively, create jobs and grow. The region’s rate of economic growth is dependent on the economic performance of individual businesses; hence, unless firms are provided the right incentives to invest and grow, accelerating overall growth will be difficult to achieve. Improving the investment climate is particularly important in the context of the current global economic environment. The slowdown in the global economy, particularly the US economy (Central America’s main trading partner), high food and oil prices, and competition from other developing countries, calls for policies to enhance the productivity and competitiveness of Central American firms. While the economic performance of the Central American economies over the past few years has benefited from the robust global economic environment, it is unlikely that such a positive trend will continue, highlighting the need for improving the investment climate. The Dominican Republic-Central America Free Trade Agreement (DR-CAFTA) represents a unique opportunity for the region to enhance the investment climate, achieve faster growth, and reduce poverty. Nonetheless, the extent to which Central American countries take and full advantage of the opportunities offered by DR-CAFTA will depend on whether domestic markets are agile and the investment climate attractive. Thus, an accompanying bold reform agenda that addresses the many factors behind a good investment climate will maximize the benefits of DR-CAFTA. This Central America Investment Climate Assessment contributes to these efforts by presenting an evaluation of the investment climate challenges faced by the region and 1 In this report, Central America refers to the six countries of Guatemala, El Salvador, Honduras, Nicaragua, Costa Rica, and Panama x provides practical policy options that can help policy makers improve the business environment, increase competitiveness and accelerate growth. Challenges that Confront Central America For Central America to take greater advantage of the new global opportunities embodied in greater market access, it will have to tackle important obstacles at the country and regional level. Country-specific measures are, by far, the most pressing for improving the investment climate and may, in the near-term, be less difficult to pursue because they do not require additional regional coordination efforts and the difficulty of generating political consensus across national boarders. Many challenges regarding education, infrastructure investment, innovation, supervision of the financial sector, as well governance issues loom large in all of Central America, but the issues differ between countries as do the solutions. While there is much to tackle on the country level, there are also challenges that are best addressed regionally. For example, customs harmonization to facilitate and reduce the costs of goods transport is an issue that is inherently regional. Likewise, without a regional focus in energy generation and logistics investment (for example in ports), there is a risk of duplication of investments and missed opportunities to build on potential synergies. In many cases, the relatively small size of the countries in Central America necessitates regional actions to capture economies of scale, but there is a trade-off between the benefits and coordination costs. Many of the investment climate obstacles identified in this study have both country-specific and regional responses. The top challenge for Central America is governance—strengthening institutions, the rule of law, and confronting crime and violence. Governance was identified as a top constraint by businesses in the region and is confirmed as a key obstacle through empirical analysis linking the level of governance indicators to firm-level productivity in Central America. Weak governance increases firm costs directly though informal payments needed to “get things done,� the cost of regulatory and legal uncertainty, as well as security costs needed to prevent theft. Central American firms, compared to other regions in the world, spend the second highest amount on security and suffer the highest losses as a share of sales. These costs have a direct impact on reducing domestic investment as well as in dissuading foreign investment. Indeed, the analysis of this study suggests that the benefits to improving the governance framework (as well as other key factors that determine the investment climate) are amplified in the context of greater market openness. Another key constraint for the business climate is the lack of adequate physical infrastructure, including reliable energy supplies. In the case of Central America, the formal barriers to international trade, such as tariffs and quotas, are small when compared to the physical bottlenecks in the costs of production and movement of goods. But the regional challenges are not uniform across Central American countries. Logistics challenges in Panama and El Salvador, for example, are much less of a constraint than in Nicaragua. In terms of energy security and supplies, the lack of reliable and secure energy supplies has created severe competitiveness problems for Central America. Erratic supplies of fuel gives rise to price manipulation, queues at gas stations, and xi increase the cost of manufactured products. With few exceptions, fuel costs for transport are important because manufacturing centers are relatively far away from international shipping centers. Energy security and efficiency are also key issues as the region faces increasing prices and system unreliability. Central America has made progress in taking steps toward deepening capital markets and enhancing financial sector supervision and regulation, but challenges remain. The financial sector plays two important roles in creating an attractive investment climate. First, its stability reduces economic volatility and enhances the willingness of investors to make long-term capital commitments, such as building new factories and investing in productive infrastructure, the building blocks for growth. Secondly, it facilitates the allocation of scarce financial resources toward the most productive investments. If the financial sector is operating efficiently, businesses with the highest potential for profits and lowest risk should have access to financing, regardless of business size. Central America has made substantial progress over the last decade in taking steps toward deepening capital markets and enhancing financial sector supervision and regulation. Nonetheless, due to the rapid regionalization of financial services over the last few years, Central American supervisors still cannot, for a variety of legal and logistical reasons, adequately implement consolidated supervision of all financial instruments and groups within countries. Due to the lack of regional cross- country information sharing, it is extremely hard for country supervisors to enforce standard minimum capital provisioning and sometimes this is compensated for by implementing distortionary regulations. In terms of access to finance, while many firms within Central America may be underserved, some evidence from Guatemala and Nicaragua suggests there does not appear to be systematic bias against credit supply to small and medium sized business, after accounting for demand and creditworthiness. Innovation and the absorption of new technologies and production processes is a critical element for further business development and growth. Central America as a whole underperforms the rest of Latin America on measures of innovation and skills development, although there are large differences between countries with Costa Rica and Panama being ahead of other countries in this respect. Globally, governments have often taken active measures to promote the environment for innovation and technology absorption because individual firms do not necessarily reap the entire societal benefits of their private investments. For example, once a firm has shown that a certain product can be manufactured or grown successfully in a country, other firms can copy the leader and avoid risky research and development investments. These spillover effects mean that without special incentives firms invest less in R&D than would be socially optimal. Central American countries would also benefit from institutions that monitor and measure quality standards—the “quality infrastructure�. New products and process often require inputs that have exact measurements and analyses during the development and production states. Hence, reliable local quality infrastructure can promote business development and growth. xii Moving Forward: Local Gains from Global Opportunities While there is the need for clear immediate reforms, new policies should be based on a long-term strategy and not temporary stop-gap measures. Regional integration and coordination activities can serve as a vehicle to promote sustainable development at the country-level as well as to take advantage of economies of scale within the region. Due to fiscal constraints as well as vulnerabilities to natural disasters and periodic financial crisis, policy changes at the country level have often been driven by the need to “fight fires� rather than by a long-term development strategy. The uncertain policy environment has restrained investment and growth, making the cost of implementing partial or temporary reforms sometimes higher than doing nothing at all. Reforms should be strategic and not just set to meet the pressing needs of the day; but a sense of urgency can also help generate the political will for implementing difficult actions that require fiscal trade-offs. Regional trade, integration, and coordination efforts can help provide a broader political dynamic to facilitate new reforms and carry them forward. Substantial reforms are often politically costly in the near term, while the benefits may not be visible for years to come. The political cycle tends to favor quick fixes, which is why an important step forward is to depoliticize the reform process and create a long-term reform agenda with a broad-based consensus among the various segments of society and countries. To a large extent, the recently implemented DR- CAFTA and the regional agenda (customs harmonization, monetary coordination, regional disaster risk management, and energy initiatives) are an excellent beginning in this process. This Central America Regional Investment Climate Assessment will hopefully assist in this process by identifying potential areas of reform in the context of a more open economy to help achieve higher growth and reduce poverty. To be sustainable in the long run, however, these suggested reforms will have to be backed by a broad segment of the Central American society. Because of Central America’s tight fiscal constraints and the need to continue to invest in broader social programs, fiscal costs of some reforms must be carefully managed. While a large number of the reforms proposed in this report entail little or no fiscal cost, such as changes in regulations, better governance and legal framework for business, and a more efficient allocation of existing expenditures, other important reforms will have a near-term fiscal impact, such as improvements to infrastructure. The countries in Central America have seen fiscal balances improve significantly over the last few years, but the cost of reforms will have to be carefully managed so that the fiscal costs of these investments are balanced by the fiscal envelope and the potential for higher growth and a more attractive investment climate. All too often, however, short-term fiscal gaps due to unexpected natural disasters, external market turbulence, or other factors, are met cutting back on long-term investment. The vulnerability of the region to the potential external shocks suggests the need to maintain fiscal contingency funds to manage the costs of these events, rather than derail long-term investment plans. In the longer run, cutting investment may be counterproductive as future fiscal revenues may be much less if the economy and, hence, the tax base is lower. xiii Setting Country and Regional Priorities To provide some prioritization among the key recommendations in this report, a difficulty versus impact analysis is undertaken. A useful framework for understanding the tradeoffs in implementing policy involves mapping each of the recommendations into a two-dimensional space according to whether they represent high- or medium-impact measures, and whether they entail high or low levels of difficulty in their implementation. 2 Recommendations are considered to be higher impact to the extent that they produce substantial results in terms of increasing competitiveness. Recommendations are considered more difficult to implement to the extent that they represent either high fiscal costs, or legal difficulties (where new legislation may be required), or political difficulties (if building regional or within-country country consensus is difficult or vested interests are seriously affected). On this basis, it is possible to divide the investment climate recommendations of this report into four general groups (Figure 1). Group I corresponds to the high-impact, low-difficulty measures that appear in the northwest quadrant of the two-dimensional mapping; these are the highest priority measures in the short term. Group II corresponds to the high-impact, high-difficulty measures that appear in the northeast quadrant of the two-dimensional mapping. These are high-priority measures and hence should be initiated straightaway; however, given their greater level of difficulty it is unrealistic to expect results until the longer term. Group III corresponds to the medium-impact, low- difficulty measures that appear in the southwest quadrant of the two-dimensional space. Although of medium priority, they could be undertaken rapidly if so desired. Finally, Group IV corresponds to the medium- impact, high-difficulty measures that should be implemented, but are not as urgent as other recommendations. The results of performing this exercise on the primary list of recommendations (divided into four primary areas of policy reform: Governance, Finance, Infrastructure, and Innovation) arising from this report are listed in Table 1. As the table indicates, there are several opportunities for near-term high-impact reforms, the so-called “low-hanging fruit� where the political and regional coordination costs are relatively low, but gains to the investment climate are potentially high. In the area of governance, for example, this would include replicating regional best practices in simplifying government regulations and greater use of performance indicators. In infrastructure, one of the high-impact near-term policies would be to establish procedures for strengthening regional planning through a common coordinated information database. 2 This framework also used in the World Bank report, Colombia: Recent Developments in Infrastructure (2004) and the Colombia Country Economic Memorandum (2005). xiv Figure 1: Conceptual Framework for Prioritization of Measures I. II. High Impact High Priority High Priority Short Term Long Term Medium Impact III. IV. Medium Medium Priority Priority Short Term Long Term Low Difficulty High Difficulty Table 1 also shows the high-impact long-term policies, options that would be more difficult to implement, either because of political considerations or fiscal costs, but potentially have large pay-offs. It is interesting to note that many of these policies include regional initiatives that may require coordination and regional investment efforts. In the area of improving access to finance, this includes investing in appropriate payments and securities-settlement technologies as well as greater regional sharing of information between regulatory and supervisory financial agencies, which would require countries to change existing legal frameworks. In terms of improving infrastructure for a better investment climate, actions would include, among others, developing a framework for intra-regional coastal shipping and completing the intra-regional electricity interconnection framework agenda. xv Table 1: Factors for Enhancing the Investment Climate in Central America Governance Finance Infrastructure Innovation • Replicate regional best practices • Grant adequate supervisory and • Re-launch the Plan Puebla Panama I. High-impact in the area of regulatory regulatory powers on payments (PPP) with a focus on Short-term Policy simplification and use of systems to central banks. strengthening regional institutional Objectives performance indicators to • Take steps to improve the quality frameworks for infrastructure improve public sector systems and of the accounting profession in investment and operations (Objectives that could be services. the region. • Set up regional country working achieved within 1 year with • Increase transparency by setting groups by subsector that could little or no immediate fiscal up regional monitoring help to inform governments and cost. Some political mechanisms that combine user to promote policy continuity constraints, but has large satisfaction surveys and other during administrative transitions gains for competitiveness.) indicators to create greater • Provide improved security along accountability among private and principal road transit corridors. public providers of public • Strengthen the coordination group services. for power sector development • Focus resources on crime • Establish procedures for prevention for countries with strengthening regional planning currently low levels of crime, through a common coordinated such as Nicaragua, Panama, and information database. Costa Rica. • Increase transparency through • Provide supervisory authorities • Update Public Private Partnerships • Develop standards and related III. and IV. implementation of standardized with the legal powers and the (PPP) frameworks for enhancing quality infrastructure based on Medium-impact legal frameworks to regulate material resources necessary for private sector participation in the needs of key value chains. Short- to Long-term freedom of information implementing risk-based, infrastructure • Strengthen vocational and Policy Objectives mechanisms across the region. consolidated supervision of the • Identify mechanisms to develop technical training programs in financial groups operating in their regional power plants and sectors with high demand by (Objectives that could be respective jurisdictions. diversify the energy matrix providing incentives to private achieved after 2 to 3 years • Make consistent payments system • Strengthen regulation of individual providers and encouraging with modest fiscal cost and rules with corporate insolvency power sector entities by sharing modular, competency-based some political constraints but procedures. best practice experiences. worker certification programs have long-term • Improve incentives for competitiveness gains.) researchers in universities and public research institutions to collaborate with the private sector. xvi Governance Finance Infrastructure Innovation II. • Enhance and/or establish region- • Invest in appropriate payments • Develop framework for intra- • Build a functioning, High-impact wide institutions to facilitate the and securities-settlement regional coastal shipping and road internationally-recognized Long-term enforcement of common rules and technologies. networks. quality infrastructure, including Policy Objectives regulations in sensitive areas such • Move toward greater regional • Set standards and criteria for intra- standards and technical as: (a) regional prosecutors’ sharing of information between regional road transport operators regulations, metrology, (Objectives that could be networks; (b) oversight of courts’ regulatory and supervisory and develop regional road calibration and traceability, achieved after 3 to 6 years performance; and (c) overall state financial agencies. maintenance and investment conformity assessment, testing with fiscal costs and/or more management, transparency and standards. and inspection, certification, difficult political issues but accountability by establishing a • Develop procedures for selecting and accreditation. Regional have significant long-term regional control system and a and building regional power plants cooperation should be utilized competitiveness gains.) transparency regulator. and complete the intra-regional where appropriate. • Implement the customs electricity interconnection agenda • Improve the quality of integration process and improve education by institutionalizing customs at the country level. assessment systems, enhancing • Institutional strengthening and teacher performance, development of integrated and diversifying teaching multi-agency strategies to combat methodologies, and increasing crime in more affected countries instructional time. like Guatemala, El Salvador and • Foster links between supply Honduras. and demand for innovation by improving incentives for universities, research institutes, and the private sector to collaborate, strengthening public research institutions, and facilitating regional knowledge networks. • xvii Chapter 1. CENTRAL AMERICA’S MACROECONOMIC ENVIRONMENT AND THE ROLE OF THE INVESTMENT CLIMATE UNDER FREE TRADE 1 INTRODUCTION 1.1 Central America’s economic performance in recent years has benefited from improved macroeconomic management, a favorable external environment, as well as rising investor confidence since the region has pursued greater access to global markets, particularly with the signing DR- CAFTA in 2004. Nonetheless, while important reforms have been made, much remains to be done, and the context of a less favorable global environment underlines the urgent need to improve competitiveness and enhance productivity. An improving investment climate would contribute to both. 1.2 This chapter discusses the broad economic factors that shape the investment climate in Central America. Section 1 describes the overall strategic context, highlighted by the opportunities and challenges emerging from the new trade deal. Section 2 discusses the region’s macroeconomic environment—economic growth, sectoral changes, the business cycles with the US, and overall stability. Section 3 covers trade, financial and remittances flows, which are particularly important to the region. Section 4 summarizes and takes a look ahead. 1. STRATEGIC CONTEXT 1.3 In recent years, the Central American countries—individually or as a region—have negotiated, or are in the process of negotiating, free trade agreements with the United States, the European Union, Chile, Taiwan, as well as other countries. Dominating this agenda has been the free trade agreement signed with its largest trading partner, the United States. DR-CAFTA has been in the forefront because it has constituted a significant turning point for the region and its integration with the global economy. A comprehensive agreement, DR-CAFTA solidifies the framework of trade relations between the US and Central America and within Central America itself. It provides a context for regional integration and greater access to the Central America’s largest export market and has shown promise in increasing the region’s foreign direct investment (FDI) inflows. DR-CAFTA has been approved and implemented in El Salvador, Guatemala, Honduras and Nicaragua. Costa Rica has passed the accord and anticipates legislative approval of several laws necessary for implementation in 2008. Panama, which was not part of DR- CAFTA, is seeking its own bilateral FTA with the US. 1.4 Free trade implies significant advantages for economic performance. Growth will likely speed up for participating countries in the medium to long run, mainly because of expected gains from greater trade and domestic and foreign direct investment. 2 How much growth freer trade delivers will largely depend on Central America’s ability to pursue complementary policies, particularly those expanding economic opportunities. 1 This chapter was prepared by David Gould and Josef Loening. The authors thank María Ivanova Reyes for research assistance and comments. 2 Gould and Gruben (2005). 1 Box 1.1: What does DR-CAFTA do and has there been an impact The Dominican Republic–Central America Free Trade Agreement (DR-CAFTA) reduces barriers to imports, exports and financial flows, giving the region the same kind of benefits the US, Canada, and Mexico enjoy under the North American Free Trade Agreement (NAFTA). After the US Senate approved the DR-CAFTA in 2005, El Salvador became the first country to approve and implement the agreement on March 2006. Honduras and Nicaragua implemented it a month later, followed by Guatemala in July. The Dominican Republic came on board in March 2007, while Costa Rica is awaiting passage of several laws necessary for implementation after approving the agreement in a national referendum in October 2007. Under DR-CAFTA, tariffs on about 80 percent of US exports to the participating countries will be eliminated immediately, and the rest will be phased out over the next decade. US trade barriers also fall, but because the vast majority of goods produced in the participating countries already entered the US duty-free under the Caribbean Basin Initiative (CBI) of 1983, US barriers had less to decline. The CBI is a unilateral trade preference granted by the US and, as such, its continuation was subject to periodic review and approval by the U.S. Congress. The DR-CAFTA solidifies trade liberalization efforts and should help facilitate long-term investments in the region. DR-CAFTA should not be regarded as a dramatic shift in trade policy, but rather as a step in a gradual process of trade liberalization that began in the region over a decade ago. The agreement addresses the following areas: Cross-border trade in services, financial services, investment, government procurement, market access, agriculture, intellectual property rights, antidumping, and dispute resolution. It also includes environmental protections, labor standards, and transparency rules. Figure 1.1: Trade between the US and Central America before and after DR-CAFTA implementation (three-month moving average) 20 18 16 Pre DR-CAFTA Average Pos t DR-CAFTA Average 14 12 % y-o-y 10 8 6 4 2 0 t-17 t-16 t-15 t-14 t-13 t-12 t-11 t-10 t10 t11 t12 t13 t14 t15 t16 t17 t-9 t-8 t-7 t-6 t-5 t-4 t-3 t-2 t-1 t0 t1 t2 t3 t4 t5 t6 t7 t8 t9 *Central America growth average refers to El Salvador, Guatemala, Honduras and Nicaragua Note: Central America growth average refers to El Salvador, Guatemala, Honduras and NicaraguaThe figure shows monthly data that has not been seasonally adjusted. Source: Moller 2008 Because DR-CAFTA became effective in Central America only in the last 1-2 years, statistically significant evidence that it has increased trade flows is not yet available (taking into consideration other factors that would influence trade, such as income growth, exchange rates, and other variables). Nonetheless, as the figure above indicates, average trade growth in the year and a half after implementing DR-CAFTA is higher than it was over the same period before its implementation. 2 While it may be too early to precisely measure DR-CAFTA’s economic impact to date, early signs have been encouraging, although not all of them are necessarily due to the FTA. They include: o Increased inflows of foreign direct investment; o Increased investment plans from multinational companies in Central America—for example, Nicaragua ICT-Cone Demin and HBSC acquiring the Banisto Group, which has operations in the entire region; o Job creation in some sectors, such as in the textile exporting sector; o Improvement in Country Risk Classification—for example, Standard and Poor’s has reclassified Guatemala’s country risk from BB- to BB; o Improvements and harmonization in customs and trade regulations. 1.5 Many of these policies affect the investment climate, which shapes international competitiveness and the ability to diversify into non-traditional exports. A stable macroeconomic environment and sectoral factors are also key factors in domestic growth. Such topics as governance challenges, infrastructure needs, and technological and skill deficiencies will be discussed in later chapters. 1.6 Better economic fundamentals not only improve the business climate but also allow the poor to take better advantage of freer trade’s opportunities. About half of Central America’s population lives in poverty; some 20 percent endures extreme poverty, and inequality remains high, particularly in the poorest countries Figure 1.2: Poverty (%) and Inequality (Gini %) (Figure 1.2). 3 In part, poverty and inequality reflect low-wage employment in the informal sector, and it is Panama concentrated in rural areas where access to public services remains limited. A lesson Costa Rica from Mexico's experience with the North El Salvador American Free Trade Agreement Guatemala (NAFTA) is that lower commercial Honduras barriers, while important, are not a Nicaragua sufficient condition for growth, development, and poverty reduction. The 0 20 40 60 80 states in Mexico that had better infrastructure and public services prior to Inequality (of equivalized household income) Poverty ($2/day Purchasing Power Parity) NAFTA enjoyed higher growth after NAFTA. This suggests the importance of complementary polices to boost the Notes: Poverty: average 1999-2003, except Honduras (1999 only); impact of market openness. Inequality: latest available year (1998-2000). Source: World Bank,Inequality in Latin America & Poverty at a 1.7 Trading opportunities are likely Glance Tables make a significant contribution to growth, but so far the impact has not been sufficient to raise aggregate growth rates, transform economies or significantly reduce poverty rates. Evidence suggests that Central America’s competitiveness has remained flat over the past year. Agricultural products still dominate exports, leaving countries vulnerable to terms-of-trade 3 The numbers reported in Figure 1.2 are based on a poverty line of 1 dollar a day at 2000 purchasing power parity (PPP). Recently, Ravaillion et al. (2008) proposed to use 1.25,dollars a day as the new international poverty line. 3 shocks that can derail economic growth and fledging attempts toward policy reforms. Regional and national analyses typically reveal that increases in output per worker almost exclusively reflect capital deepening, rather than the productivity increases usually associated with greater efficiency and technological advances (Loayza et al. 2002; Desruelle and Schipke, 2007). 2. REGIONAL MACROECONOMIC ENVIRONMENT Economic growth returns 1.8 Central American countries have relatively small domestic markets, but together they form a significant regional market. With about 40 million people, the region accounts for about 7 percent of Latin America’s population and 5 percent of its total output. Dependence on traditional exports and close economic ties to the US still characterize the region’s economies. Central America faces challenges from increased global competition in some key export products, such as agricultural commodities and textiles, while its economies are exposed to common shocks, such as natural disasters and terms-of-trade changes. Table 1.1: Per capita GDP Growth in Central America, 2000-2007 (in percent) 2000 2001 2002 2003 2004 2005 2006 2007 Costa Rica -0.5 -1.1 0.8 4.4 2.4 4.1 6.4 5.3 El Salvador 0.4 0.1 0.8 0.8 0.4 1.7 2.7 2.8 Guatemala 1.2 -0.1 -0.2 -0.4 0.1 0.7 1.9 2.2 Honduras 3.6 0.6 0.7 1.5 3.0 2.1 4.0 4.2 Nicaragua 2.5 1.5 -0.6 1.2 4.0 3.0 2.4 2.5 Panama 0.7 -1.3 0.3 2.3 5.6 5.1 6.3 7.8 Note: per capita GDP growth based on per capita GDP in constant dollars of 2000. Source: 2000-2006 from WDI, 2007 estimated by CEPAL based on WDI data. 1.9 Economic growth accelerated in Central America since 2003 (Table 1.1), benefiting from the global economic expansion and, in particular, the US economy’s dynamism, which spurred demand for exports and increased remittances. Driven initially by a pickup in exports and rising commodity prices, the recovery spilled over to domestic demand. Figure 1.3 shows the difference between actual and predicted GDP growth for the six CA countries and the LAC average during 2002 and 2006. The predicted GDP growth is estimated as a function of internal factors, assuming that external factors had remained constant at 2002 levels. In all cases, the actual growth is higher than the predicted growth, revealing that a considerable proportion of recent growth is due to favorable external factors. 1.10 Overall, the region’s per capita GDP growth increased to about 3 percent (Figure 1.4, Panel A). Between 2004 and 2007, Panama achieved per capita growth rates of around 6 percent, while Costa Rica 5 percent, and Honduras and Nicaragua 3 percent. El Salvador and Guatemala had growth rates of 1 percent from 2004 to 2006, but both improved to more than 2 percent in 2007. 4 Figure 1.3: Per capita growth in Central America 30% Actual 25% Long-run cumulative growth, Predicted variation 2002/06 20% 15% 10% 5% 0% PAN GTM SLV HND NIC CRI LAC Note: Panel Regression: dlog(gdp)=f(dlog(ToT), dlog(Ind.Prod.G7), dlog(US5Y), dlog(HighYield)); 1991-2002. Predicted: setting external factors to zero. Source: World Bank LCRCE staff calculations 1.11 Nonetheless, Central America still lags behind the rest of Latin America, East Asia, and middle-income countries in growth (Figure 1.4, Panels B-E). The underperformance may be explained by the rise in global raw materials prices, including oil. In addition, Central American light manufacturing exports may have lost some dynamism in the face of competition from China and other Asian countries (SIECA, 2007). Growth also remains volatile, continuing the pattern of previous decades and reflecting the region’s vulnerability to external economic shocks (commodity price increases), natural disasters (El Niño, hurricanes, and earthquakes) and domestic policy reversals (pre-election spending). Over the near-term, regional growth remains vulnerable to the expected slowdown in US growth. 1.12 Structural change is important to growth in the region. Advancements in technology and export performance imply shifts from one sector to another. Central America’s economy is highly diverse and shows some changes over time, suggesting that technological and/or terms of trade changes have influenced the structures of the economies (Montobbio and Rampa, 2005; Yuki, 2007). The manufacturing industry has grown in all countries except Panama (Figure 1.5). In Guatemala, Honduras, and Nicaragua, economic activity is highest in agricultural sectors, but has declined in most countries with the exception of Nicaragua. Panama derives a large fraction of its GDPs from financial services, with the share increasing slightly over the past decade. The trade, restaurants, and hotels sector has been volatile in all countries. The sector with the strongest growth performance has been transport and communication services. 5 Figure 1.4: Per capita growth in Central America Deviations with respect to selected country groups, 1990-2008 A. Central America growth, B. Rest of LAC C. Middle Income Countries 1990-2008 4.0 4.0 3.0 3.0 2.0 2.0 1.0 1.0 0.0 0.0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 1990 1993 1996 1999 2002 2005 2008 -1.0 -1.0 4.0 -2.0 -2.0 -3.0 -3.0 3.0 2.0 1.0 D. OECD E. East Asia 0.0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 4.0 4.0 3.0 3.0 2.0 2.0 1.0 1.0 0.0 0.0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 -1.0 -1.0 -2.0 -2.0 -3.0 -3.0 Note: Deviations with respect to selected country groups are computed as Central America’s annual median growth rate minus the reference group’s annual median growth rate. All series have been smoothed with a backward looking three-year moving average. Source: Authors’ calculations based on data from WDI. 6 5 10 15 20 10 15 20 25 5 10 15 20 25 30 1990 1990 1990 1991 1991 1991 1992 CRI 1992 1992 CRI CRI 1993 1993 1993 1994 1994 1994 SLV SLV SLV 1995 1995 1995 1996 1996 1996 Source: National authorities 1997 1997 1997 GTM GTM GTM 1998 1998 1998 1999 1999 1999 Personal Services Manufacturing Industry 2000 2000 2000 HND HND HND Trade, Restaurants and Hotels 2001 2001 2001 2002 2002 2002 2003 2003 NIC NIC NIC 2003 2004 2004 2004 2005 2005 2005 PAN PAN PAN 2006 2006 2006 5 10 15 20 25 0 5 10 15 20 0 5 10 15 1990 1990 1990 1991 1991 1991 1992 1992 1992 CRI CRI CRI 1993 1993 1993 1994 1994 1994 SLV SLV SLV 1995 1995 1995 1996 1996 1996 1997 1997 1997 GTM GTM GTM 1998 1998 1998 Agriculture 1999 1999 1999 Financial Services Figure 1.5: Sectoral GDP Shares in Central America, 1990-2006 2000 2000 2000 HND HND HND Transport and Communications 2001 2001 2001 2002 2002 2002 NIC NIC NIC 2003 2003 2003 2004 2004 2004 2005 2005 2005 PAN PAN PAN 2006 2006 2006 7 Box 1.2: Country Specific Challenges Central America faces many common challenges. At the same time, a number of issues characterize each country’s domestic business environment and macroeconomic context. Among the most important are: Costa Rica is a middle-income country of about 4.4 million people with a per capita income of about US$4,900. It boasts one of Latin America’s most stable democracies, with uninterrupted elected governments since 1949. Since the mid-l980s, Costa Rica has followed a successful strategy of export-led growth, openness to foreign investment, and gradual trade liberalization. As a result, the structure of exports and the economy itself have been transformed from dependence on agriculture and agro-industry to greater diversity, featuring computer and electronic industries, services, non-traditional agriculture, and eco-tourism. Costa Rica attracts significant FDI, including high-tech investors such as Intel. El Salvador has been a regional leader in economic reforms. Starting in the early 1990s, successive governments have tackled trade liberalization, tax reform, strengthening of the financial sector, and promotion of private participation in telecoms, energy, and pensions. A decision to dollarize the economy in 2001 has resulted in lower inflation and interest rates and reduced business uncertainty. Despite the country’s impressive reforms and prudent macroeconomic policies, growth has been volatile. Instability largely reflects external shocks—earthquakes, terms of trade deterioration, and swings in coffee exports and the large maquila sector. Poverty has declined, but about one-third of the 7.7 million Salvadoreños remain poor. Per capita income is US$2,500. Guatemala has a record of sound macroeconomic management, with small fiscal deficits. Tax collections are low, which creates a challenge in meeting infrastructure needs. As with the other Central American countries, Guatemala is vulnerable to natural disasters and commodity price shocks. After 35 years of civil war, ending with the 1996 peace accords, Guatemala faces a post-conflict situation with a very large share of the population remaining poor, particular among the indigenous population suffering from social exclusion. With 13.4 million people, Guatemala has the largest population and biggest economy in Central America, a regional share of 33 percent. Per capita income is US$ 2,600. Honduras, with a population of 7.2 million, is a lower middle-income country, with an open economy and sluggish growth. Honduras transitioned from an authoritarian military regime to a pluralistic democracy. State institutions are considered fragile, and the governance framework is weak. The economy remains highly vulnerable to shocks from natural disasters or commodity price shifts. Recent years have seen a gradual shift away from traditional exports, which have been superseded by the rapid growth in maquila sales. Little progress has been made in reducing poverty in recent years, with volatile growth putting achievements at risk. About half of the population is considered poor. The per capita income is US$1,200. Nicaragua remains the second poorest country in Latin America after Haiti, with a per capita income of only US$1,000. Although economic gains have reduced poverty, 46 percent of the population still lives below the poverty line. Growth has been modest, averaging around 3 percent per year, even though exports have doubled. Nicaragua remains highly aid dependent. More than 30 percent of its budget comes from official development assistance. The country’s 5.4 million people are also vulnerable to natural shocks, evidenced by the devastation of Hurricane Mitch in October 1998. Maintaining fiscal discipline and improving governance remain key policy challenges. Panama, a country with a population of about 3.5 million, offers an international transport corridor, a modern financial center, a dollarized economy, and a per capita GDP of about US$ 4,900 that ranks among the upper-middle income nations. Despite its small population, Panama produces 15 percent of the region’s GDP, making it the second largest economy in Central America. The country has experienced spectacular economic growth in recent years. The government’s decisions to expand the Panama Canal over the next six to seven years and an expected bilateral free trade agreement with the US are likely to boost the economy for some time. Even so, Panama is still a country of stark contrasts. Perpetuated by educational disparities, more than one-third of Panama’s 3.3 million people were living in poverty in 2003, and 16 percent faced extreme poverty Source: Country Assistance Strategies; various documents and unofficial reports 8 Volatility and Business Cycles 1.13 GDP volatility has declined in most Central American economies during the past decade. 4 In particular, significant decreases occurred Table 1.2: Volatility of per Capita GDP in El Salvador, Honduras, and Nicaragua (Table 1.2). Growth, 2000-2006 (standard deviations) Trade and financial flows may have helped to reduce volatility by accelerating diversification of the export 1991-1999 2000-2006 base (Kose and Rebucci 2005). Another reason for Costa Rica 2.8 2.3 decreasing volatility may have been the US El Salvador 1.9 0.6 economy’s relative stability and increased integration Guatemala 0.7 0.6 with U.S. markets. Honduras 2.7 1.4 1.14 Shocks originating in the US tend to play a Nicaragua 3.0 1.5 prominent role in the region. Increasing trade, Panama 2.6 2.8 investment, and remittances suggest that the Central Source: WDI America has become more attuned to US economic developments. With an economic slowdown expected in the US, additional domestic policies may be needed to address regional growth, including policies that stimulate the investment climate and diversify business for exporting firms. 1.15 How dependent are Central America’s economies on the US? Roache (2008) finds strong GDP growth co-movement between the US and most countries in the region. Distinguishing between the contributions of cyclical and trend GDP growth, Roache finds that the cycle contributes most to GDP changes for the majority of Central American countries. Growth elasticities are generally high, ranging from about 1.07 to 0.41, implying that a 1 percent increase (or decrease) in US growth would increase (or reduce) regional growth by about 1.07 to 0.41 percent, depending on the country. The study does not identify the direct sources of cyclical fluctuations, but the most likely transmission mechanisms are trade, financial flows, and remittances. 1.16 Likewise, Fiess (2007) measures business cycle synchronization for Central America and finds evidence of close relationships between Costa Rica, El Salvador, Guatemala, and Honduras and between these countries as a group and the U.S., suggesting that a significant portion of Central American variability is driven by external factors. The US is by far the region’s most important trading partner; the financial linkages via remittances are weak. Overall, Central America has become more sensitive to developments in the US economy. Macroeconomic stability 1.17 Central America has made substantial progress in regaining macroeconomic stability. It has also continued to further integrate at both the global and regional levels. Inflation peaked in 2005, but recent increases in oil and food prices, along with stronger demand, have caused inflation to rise throughout the region (Figure 1.6). Nevertheless, inflation still remains modest in all countries. Although still relatively modest by historical standards, in the region inflation is 4 Growth volatility is measured as the standard deviation of GDP growth. 9 highest in Costa Rica and Nicaragua. As might be expected, it is lowest in the two dollarized economies—El Salvador and Panama. Figure 1.6: Inflation and Fiscal Deficits Inflation in Central America, 1998-2007 Fiscal Deficits, 2003 and 2006 17 15 13 Central CPI Change in percent (y/y) 11 2003 America 9 2006 7 Panama 5 Costa Rica 3 El S alvador 1 -1 Guatemala Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Honduras Nicaragua CRI SLV GTM HND NIC PAN -9 -6 -3 0 3 Non-Financial Public Sector Balance (% of GDP) Source: WDI and national authorities 1.18 The region made progress in reducing fiscal deficits. While most countries saw debt-to- GDP ratios rise over the past decade, the current cyclical upturn has allowed them to strengthen policies and improve fiscal accounts. Deficits declined from an average of nearly 3 percent of GDP in 2000 to about 1 percent in 2006, helped by buoyant revenue collections that reflected a growth dividend and ongoing reforms in tax policy and administration. Nevertheless, public debt remains a concern in most countries, with little room to expand public borrowing. 1.19 In terms of current account balances, Central American countries external positions have continued to improve. Current account deficits remained largely unchanged at about 6 percent of GDP in 2007, reflecting in part the higher oil import bill. Capital inflows have boosted international reserves. Integration and coordination 1.20 Movements toward regional integration have been modest, although momentum toward greater coordination is improving. The past decade has witnessed growing trade and financial linkages among the region’s economies and with their largest trading partner, the United States. Central America has a relatively open trade regime, with a tariff structure largely determined by the common external duties of the Central American Common Market. Although the region had already preferential access to the US market under the Caribbean Basin Initiative, the DR- CAFTA should make access more permanent and extend it further in some areas, leading to continued deepening of trade integration. Countries have seen a significant increase of intra- regional financial sector ties. 1.21 Economic policy coordination is at an early stage. Countries continue to pursue their own fiscal policies; while central banks consult regularly, monetary and exchange rate policies evolve largely independently. The region is working toward better coordination and has a set of regional institutions—for example, the Central American Monetary Council, Central American 10 Bank for Economic Integration, and the Secretariat for Central American Economic Integration. They are increasingly involved in information sharing, regulatory harmonization, and policy coordination, particularly in areas such as banking supervision, central banking, and trade. 1.22 El Salvador, Guatemala, and Honduras have made some progress toward a customs union. Other areas that could be strengthened include the standardization of norms and regulations, the flow of information among government agencies (such as financial sector supervisors), and policy coordination in certain areas (such as tax policies). At the center of this coordination effort is the Central American System of Integration, which brings together heads of state as well as regional councils and committees of ministers, central bank presidents, and superintendents. 3. TRADE AND FINANCIAL FLOWS Trade and tariffs Figure 1.7: Central American Trade Flows with the US and EU, 1996-2007 6000 5000 4000 US$Millions 3000 2000 1000 0 1996Q1 1996Q3 1997Q1 1997Q3 1998Q1 1998Q3 1999Q1 1999Q3 2000Q1 2000Q3 2001Q1 2001Q3 2002Q1 2002Q3 2003Q1 2003Q3 2004Q1 2004Q3 2005Q1 2005Q3 2006Q1 2006Q3 2007Q1 2007Q3 EX to USA EX to EU IM from USA IM from EU Source: IMF Direction of Trade Statistics 1.23 Since the 1990s, Central American economies have been very open, leading to a surge in international trade. The region’s average openness ratio (exports plus imports of goods and services divided by GDP) was roughly 76 percent during 1980–2007, but it showed some variation across countries. From 1980 to 2007, for example, the average openness ratio was less than 60 percent in El Salvador and Guatemala, about 80 percent in Costa Rica and Honduras, and above 150 percent in Panama. 11 1.24 Trade linkages between the US and Central America have grown rapidly over the past decade, in absolute terms as well as relative terms compared to other large trading blocks, such as the European Union (Figure 1.7). The region’s trade with the US increased fivefold in dollar terms from 1996 to 2006. While it is too early to attribute substantial gains to the impact of DR- CAFTA, non-traditional exports started to pick up at the end of 2006 in most countries. Exports from Central America to the US grew over 13 percent in dollar terms in 2006 and continued to show significant gains in 2007. The main increases in DR-CAFTA’s exports are in textiles, clothing, and processed crops. 1.25 Central American countries are similar in the general categories of exports and imports they send to and receive from the US (Appendix Table A.1.1.). Figure 1.8: Share of Central American Exports in World Exports, 1994-2006 Differences are greater in dependence on the US market 2.5%Share of Central American Exports in World Exports for exports. In 2007, Honduras sent more than 70 percent of its exports to the US, while 2.0% Guatemala shipped about 44 percent. Costa Rica’s export 1.5% share was the smallest at less than 30 percent. As a whole, Central America’s share of the 1.0% world’s export market has declined as other countries have increased their share, 0.5% particularly China and other 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 East Asian countries (Figure Food and animals All other items 1.8). Source: COMTRADE 1.26 Over the past decade, the Central American countries substantially diversified their trade (Table 1.3). For the region as a whole, manufacturing’s share of total exports rose from 24 percent to 34 percent from 1996 to 2006. Costa Rica, El Salvador, and Guatemala posted the largest gains in manufacturing exports. Agricultural and food exports’ dominance has declined. However, these goods still account for the majority of total exports. 1.27 The extent of export diversification is somewhat lower than expected. Countries continue to focus on agricultural commodities and textiles. In the long run, however, DR-CAFTA can lead to a substantial increase in trade flows through its impact on productivity and specialization. These gains could be substantial since DR-CAFTA includes various provisions about the flows of investment, financial services, and intellectual property. Countries have already begun expanding the scope trade with the US (Kose and Rebucci, 2005). For example, most US electrical machinery and apparel imports are being used as intermediate inputs for other goods that are then exported back to the US. 12 Table 1.3: Structure of Merchandise Exports, 1996-2007 (Percentage) Costa Rica El Salvador Guatemala Honduras Nicaragua Panama Manufacturing 1996-2000 50.94 45.00 31.97 22.84 16.55 17.52 2001-2005 63.95 57.62 41.53 30.36 13.23 12.92 Agriculture and food 1996-2000 47.51 48.44 63.32 70.81 81.65 75.02 2001-2005 34.53 34.01 50.65 64.46 84.03 81.60 Fuel and metals 1996-2000 1.40 6.26 4.71 6.29 1.57 7.46 2001-2005 1.51 8.22 7.82 5.14 2.53 5.47 Source: WDI Box 1.3: Export Diversification and Growth Export diversification may contribute to growth through a number of channels. First, it may reduce the reliance on a limited number of export commodities that may be subject to extreme price or volume fluctuations—such as coffee, natural resources, or textiles. Swings in foreign exchange revenues may hamper efforts at policy planning, reduce import capacity, and contribute to an undersupply of investment by risk-adverse producers. Therefore, decreasing economic vulnerabilities through export diversification may provide significant benefits. Second, export diversification may affect long-run growth through its role in increasing returns to scale and dynamic spillovers. Some analysts have argued that the quality of the export basket matters for growth--that is, specializing in some products may bring higher growth than specializing in others. This implies an important role for policy in encouraging a broad-based production structure (Hausmann et al. 2005). Relatively few detailed empirical investigations have looked into the links between export diversification and growth. Based on panel data, de Ferranti et al. (2003) find positive linkages for Latin America. To spur growth, the authors point to the importance of combining diversification with complementary factors associated with improving the investment climate, such as technology and skills. Further research is needed to distinguish between the role of a diversified export base and the overall investment climate in determining growth 1.28 Central American tariffs have decreased (Appendix Table A.1.2.). 5 Countries began to reduce import duties in the early 1990s, making rates among the lowest in Latin America. Average tariffs fell from about 10 percent in 1995 to 7 percent in 2000 and 5 percent in 2006. Costa Rica has the lowest rates; Nicaragua the highest. Evidence suggests that Central America has also reduced non-tariff barriers, largely sanitary and phytosanitary standards that had been used as trade barriers prior to reforms (Jaramillo and Lederman, 2006). Most of the remaining restrictions are limited to sanitary and technical standards. Foreign direct investment and remittances 1.29 Central American countries were able to increase FDI flows significantly in recent years. Panama, for example, has the highest gross FDI flows as percentage of GDP at about 15 percent 5 Tariffs reported in Table A.1.2 refer to Trade-weighted MFN averages. 13 in 2006. For the region, these flows were sizeable, averaging about 14 percent of domestic investment or 5 percent of GDP. The US is the largest source of FDI flows. Provided a favorable investment climate, DR-CAFTA may therefore serve as a “commitment device� to prevent further increases in trade barriers and encourage greater FDI inflows, including from outside the United States, while inducing a change in the nature of trade flows in favor of vertically integrated trade. DR-CAFTA could also help attract foreign multinational corporations to the Central American countries. 1.30 Remittances sent from the US are an important source of revenue in Central America, with some analysts suggesting they might surpass FDI flows. Estimates of these transfers range from about 2 percent of GDP in Costa Rica to 25 percent in Honduras (Table 1.4). Remittances may have grown rapidly over in recent years, reflecting the positive economic environment in the US. Unlike FDI and trade, remittances seem to be less cyclical, suggesting that migrant workers smooth their remittance flows by sending relatively steady US dollar Table 1.4: Estimated Size of Remittances, 2005-2006 amounts each month or quarter (Roache, US$ in Shares in percent 2008). Nonetheless, the recent US billions GDP FDI Inflows Exports slowdown has reduced remittance flows Costa Rica 0.5 2% 74% 4% to Central America. Migrant workers El Salvador 3.3 18% 667% 69% typically transfer monthly amounts of Guatemala 3.6 10% 1111% 66% about US$250-300 on a frequent basis Honduras 2.2 25% 774% 60% (Cheikhrouhou et al., 2006; World Bank Nicaragua 0.7 12% 235% 28% 2006 and 2007b). Source: Cheikhrouhou et al. (2006); World Bank (2006); World Bank (2007b); Roache (2008) 4. LOOKING AHEAD 1.31 The international economic environment over the past few years has been favorable to Central American countries. The region was able to make important gains in improving fiscal balances, reducing external debt exposure, and accelerating economic growth. Despite these gains, the region did not grow as fast it did in prior decades, or as fast as the rest of Latin America or other regions of the world. This suggests the need to improve the investment climate and accelerate growth. 1.32 The region’s governments continue to embrace macroeconomic and political stability. A series of presidential and congressional elections have demonstrated that democracies are functioning well, transferring power smoothly. Moving economic reforms forward will depend critically on the ability to build alliances because many governments lack majority support in their legislatures. 1.33 The outlook for the near future suggests that the world economy will be less favorable for Central America. Slowing global growth—most importantly, in the United States—could stall the momentum of Central American exports, FDI inflows, and remittances. With oil prices at record highs and Central America facing increased competition from Asia, the region needs to improve on policies that foster productive development and innovation. Economic integration within the region and with the US can be beneficial, but it needs to be backed with complementary policies improving the investment climate. 14 1.34 Achieving higher growth rates and improving the investment climate will rest largely on implementing institutional and governance reforms. The Doing Business Reports show the region improved over the past year, albeit from a low base. Central American countries have been particularly focusing on trade facilitation, registration of property procedures, and credit access. A continued effort to improve the investment climate is of central importance to ensure that the DR-CAFTA will lead to the expected productivity improvements and more tangible benefits for Central America. 1.35 By looking in detail at the microeconomic and sectoral determinants of domestic growth, and the determinants of growth of exporting firms, this report can help to identify important policy priorities to achieve this goal. 15 Chapter 2. BENCHMARKING THE INVESTMENT CLIMATE: WHERE DOES CENTRAL AMERICA STAND? 1 INTRODUCTION 2.1 Recent empirical studies on the determinants of long-run growth focus on the importance of the investment climate. To a large extent, this work was spurred by the endogenous growth theories of Romer (1986) and Lucas (1988). It is now well accepted that macroeconomic stability and avoiding financial and balance of payments crises have the same kind of impact on economic performance as investing in human and physical capital. Similarly, a number of studies, including Frankel and Romer (1999) and Dollar and Kraay (2002), have found that openness to trade and foreign direct investment accelerate growth, confirming the role of increased competition and market size in creating and absorbing technological innovations and exploiting economies of scale. 2.2 In addition, an ongoing debate centers on identifying the most important channels through which international trade affects economic growth. Cross-country research has been inconclusive, but Lopez-Cordova (2007) points out that several micro-level studies find that openness to trade improves firms’ efficiency. Furthermore, evidence suggests that reallocating resources from less to more efficient producers is a major channel through which trade liberalization leads to increases in average productivity. 2 However, international trade’s potential for raising productivity may be affected by a business climate that hinders the ability to efficiently allocate and reallocate resources. 2.3 To reap local gains from global opportunities, it is necessary to complement the trade agenda with reforms that reduce burdensome regulations, lower the costs of opening and closing businesses, improve access to finance, and provide the necessary infrastructure. Lopez-Cordova (2007) uses firm-level data from 90 countries to show, among other things, that labor regulations have a negative effect on export firms’ ability of to grow. Lederman (2007) finds evidence that trade policies and the investment climate affect entrepreneurs’ propensity for product innovation. 2.4 What are the implications for Central America? The recent findings on growth and trade suggest that the region will need to address key policy challenges and economic constraints to take full advantage of opportunities offered by greater market openness. Mexico's experience in joining the North American Free Trade Agreement (NAFTA) offers an important lesson: While freer trade is certainly important, it is not a sufficient condition for growth, development, and poverty reduction. 3 Freer trade’s benefits in terms of higher economic growth, technology transfer, and capital inflows derive from a country's ability to shift labor and other resources 1 This chapter was prepared by Veronica Alaimo. 2 The role of reallocation is highlighted in new theories of international trade that allow for heterogeneity in productivity levels (Lopez-Cordova, 2007; Melitz, 2003; Bernard et al., 2003). 3 See, for example, Lopez-Cordova (2007), Lederman, Maloney, and Servén (2003), Jaramillo and Lederman (2006), De Ferranti et al. (2002). 16 toward sectors that reflect its comparative advantage and out of sectors with a comparative disadvantage. For this to occur markets must be unfettered and resources must be agile. 2.5 How much additional growth and poverty reduction Central America ultimately extracts from freer trade will largely depend on the investment climate of its six countries—Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and Panama. Investment climate can be broadly defined as “key factors that shape the opportunities and incentives for firms to invest productively, create jobs, and grow� (World Development Report, 2005). It includes such factors a stable macroeconomic environment, the ease of opening or closing businesses, access to credit, the labor force’s educational level and flexibility, the ease of transport from ports and on roads, and the legal and judicial system’s ability to enforce contracts and facilitate business activity. 2.6 This chapter reviews Central America’s investment climate and discusses the constraints on gains from trade and regional integration. It provides a benchmark of regional investment climate indicators, including governance and regulation, finance, infrastructure (energy, transport, and telecommunications) and innovations and human capital. Central America’s indicators are compared to other relevant regions. Data come from Enterprise Surveys, Doing Business, World Economic Forum, World Development Indicators, and Logistic Performance Index. Two waves of Enterprise Surveys for El Salvador, Guatemala, Honduras, and Nicaragua allow comparisons between 2003 and 2007. Within countries, the numbers show some variations according to firm size, exporter status, and foreign ownership. 1. BENCHMARKING CENTRAL AMERICA 2.7 This section puts Central American investment climate into perspective by summarizing international rankings and micro-data indicators and benchmarking the data with respect to other income groups and regions. The regions include the rest of Latin America and the Caribbean (LAC), high income OECD countries (Hi-OECD), Eastern Europe and Central Asia (ECA), East Asia and Pacific (EAP), Middle East and North Africa (MENA), South Asia (SAR), and Sub- Saharan Africa (SSA). 4 2.8 It is important to note the strengths and weaknesses of aggregate indicators. On one hand, rankings from Doing Business, the Global Competitiveness Report and other sources provide a simple way to compare countries. On the other hand, the rankings may not be completely objective and may hide important variations between and within countries. These caveats are particularly relevant when indicators are aggregated into regions or income groups. Given these problems, it is important to use the indicators with caution. 4 The list of countries included in each region can be found in the Appendix. Note that the exact countries included in each indicator vary by source. Not all countries in a region are evaluated on every category of information. 17 Table 2.1: Central America’s Position in International Rankings Average Rankings Global Ease of Trading Logistic Region Competitiveness Doing Across Performance Index 2007-2008 Business Borders Index 1-131 1-178 1-178 1-150 REGION/GROUP Hi-OECD 18 22 25 15 Middle East & North Africa 49 96 80 80 East Asia & Pacific 54 77 75 66 Eastern Europe & Central Asia 70 76 91 80 Central America 78 96 73 78 Rest of Latin America & Caribbean 84 85 90 81 South Asia 86 107 116 99 Sub-Saharan Africa 107 136 131 104 INCOME LEVEL High income countries 20 27 30 20 Upper middle income countries 54 67 77 59 Lower middle income countries 83 99 97 92 Low income countries 108 137 133 108 Note: Within Central America, Costa Rica and Panama are upper middle income countries, while El Salvador, Guatemala, Honduras and Nicaragua are lower middle income countries. We exclude Central American countries in their income level groups. Source: Doing Business (2008) and Global Competitiveness Report 2007-2008. 2.9 Central America ranks relatively low in international assessments, especially when compared to other countries in the same income group. The first column of Table 2.1 reports the average ranking of regions and income groups on the comprehensive Global Competitiveness Index (GCI). It summarizes 12 pillars of competitiveness—Institutions, Infrastructure, Macroeconomic Stability, Health and Primary Education, Higher Education and Training, Goods Market Efficiency, Labor Market Efficiency, Financial Market Sophistication, Technological Readiness, Market Size, Business Sophistication, and Innovation. The GCI rankings are calculated from both publicly available data and the Executive Opinion Survey, an annual poll the World Economic Forum conducts in partnership with leading research institutes and business organizations in countries the report covers. This year, the surveys polled more than 11,000 business leaders in a record 131 countries. 5 2.10 Lower GCI scores indicate greater competitiveness. Central America’s average GCI ranking is 78, better than the rest of LAC (84), SAR (86), and SSA (107). However, simple averages may not be telling the right story about relative performance, particularly for countries within a region. The maps in Figure 2.1 divide the GCI rankings into deciles and show significant differences within Central America. Panama comes in at 59 and Costa Rica 63; Nicaragua is at 111—below the Sub-Saharan African mean. 5 For more information, visit http://www.weforum.org/en/initiatives/gcp/index.htm 18 Figure 2.1: Global Competitive Index, 2007-2008 Source: Own calculations using ArcGIS software to map CGI ranking 2.11 According to the Ease of Doing Business indicator (Table 2.1, second column), the Central American average ranking is 96, compared with 85 in the rest of LAC and 77 in EAP. The measure summarizes a wide range of characteristics, such as starting a business, dealing with licenses, and closing a business. As with the GCI, economies are ranked from best to worst. A low number on the ease of doing business index connotes a regulatory environment conducive to operating a business. This index averages the country's percentile rankings on 10 topics made up of a variety of indicators, giving equal weight to each topic. 6 6 For more information, visit http://www.doingbusiness.org/ 19 2.12 The Trading across Borders Index combines information on the required documents, time, and cost of exporting and importing (Table 2.1, third column). Central America ranks better than the rest of LAC and EAP. Once again the average obscures significant differences among countries. Panama ranks 9, while Honduras is 103 and Guatemala 116. 2.13 The Logistics Performance Index (LPI) sheds light on countries’ ability to move goods in a rapid, reliable, and cheap manner, a factor increasingly important to regional integration and global markets. A measure of trade logistics in 150 countries, LPI is based on a survey of operators on the ground—such as, global freight forwarders and express carriers. 7 It provides feedback on the logistics “friendliness� of countries in which they operate and those in which they trade. Data on the performance of key components of the home country logistics chain supplements the feedback from the operators 2.14 Central America has an LPI average ranking of 78 (Table 2.1, fourth column).The region is close to ECA (80), MENA (80), and the rest of LAC (81), but it trails EAP (66) and Hi-OECD (15). Within the region, Panama and El Salvador are below only Hi-OECD countries, while Nicaragua is below the SAR average of 99. Microdata from Enterprise Surveys 2.15 The World Bank’s Enterprise Surveys cover entrepreneurs’ subjective perceptions and objective assessments for more than 90 countries on the importance of various investment climate constraints. To maximize comparability across countries, most data measure a cross- section of firms in a country in a given year. The analysis focuses on surveys carried out between 2003 and 2006. 2.16 Enterprise Surveys are complementary to the Doing Business project—but the two differ in important ways. Briefly, the Enterprise Surveys ask entrepreneurs the degree of concern on key areas of the investment climate. The Doing Business project measures the annual costs for a hypothetical firm. 8 Its frequent observations, based on the same business concept description in 175 economies, can be extremely useful in monitoring progress in covered areas, which include the costs of starting and closing a business, employing workers, trading across borders, registering property, getting credit, dealing with licenses, and paying taxes as well as investor protection issues and contract enforcement. Enterprise Survey data allow analysis of the various firm-level attributes within a country and their productivity effects on the investment climate. 9 Subjective Investment Climate Indicators 2.17 The Enterprise Survey asks firms whether aspects of the investment climate represent no obstacle, minor, moderate, major or a severe obstacle to business operations. We compute the 7 For more information, visit http://go.worldbank.org/88X6PU5GV0 8 For example, Doing Business—a collaboration between the World Bank and IFC– collects the procedures, time, and cost involved in launching a commercial or industrial firm with up to 50 employees and start-up capital of 10 times the economy's per-capita gross national income. Similarly, it reports the administrative burden of paying taxes resulting from a case study that assume a firm that has 60 employees—four managers, eight assistants and 48 workers, and a start-up capital of 102 times income per capita at the end of 2004. 9 For more details, see http://www.enterprisesurveys.org 20 percentage of firms in each country or region that report certain areas to be major or severe constraints for business. 2.18 Figure 2.2 shows the five areas that Central American firms find to be the worst obstacles for business. 10 When compared to other regions, Central America has the highest proportion of firms reporting corruption and crime as a major obstacles, followed by the rest of LAC and South Asia. 11 Only the African region reports a higher percentage of firms perceiving electricity as a major concern. Practices from informal competitors and tax rates also appear as major concerns for Central American entrepreneurs. Figure 2.2: Obstacles for business according to entrepreneurs’ subjective perceptions 70 60.0 60 50 41.9 40.4 33.9 31.2 % firms 40 30 20 10 0 Corruption Electricity Crime Practices from Tax Rates informal competitors CA LAC AFR EAP ECA HiOECD MENA SAR Note: The bars show the percentage of firms that consider each area a major or severe obstacle for business. Central America (CA), rest of Latin America and the Caribbean (LAC), High income OECD countries (Hi- OECD), Eastern Europe and Central Asia (ECA), East Asia and Pacific (EAP), Middle East and North Africa (MENA), South Asia (SAR), and Sub-Saharan Africa (SSA). See Table A.1 in the Appendix for a complete list of obstacles in each region. Source: Enterprise Surveys 2003-2007 2.19 Some caveats are called for. These averages are based on entrepreneurs’ subjective perceptions. These perceptions may be influenced by recent events reported in the media, by recent macroeconomic performance, or by other cultural factors. They may also reflect specific cultural and socioeconomic backgrounds. For instance, managers of firms that concentrate on 10 We exclude “macroeconomic instability� in this analysis. This question is the most ambiguous of all those asked because it does not specify what “macroeconomic instability� means, leaving each respondent to interpret it in a different way. Table A.1 in the Appendix contains a list of all obstacles in the survey. 11 It is important to distinguish these percentages from the ones provided by the Worldwide Governance Indicators (WGI) produced by Kauffman and Kraay (2007). The percentages in Figure 2.2 represent the responses of formally registered firms that were sampled by the latest Enterprise Surveys, while the index Control of Corruption from WGI reflects the perceptions of a very diverse group of respondents, including several surveys of individuals or domestic firms with first-hand knowledge of the governance situation in the country; the perceptions of country analysts at the major multilateral development agencies; data sources provided by various nongovernmental organizations; and commercial business information providers. For more details see, http://info.worldbank.org/governance/wgi2007/ 21 local as opposed to national or international markets may lack the necessary information to judge problems in their cities or provinces and make comparisons to national or international best practices. Managers’ perceptions provide some insight on obstacles to business activity, but they certainly are not the whole story. 2.20 The Enterprise Surveys also identify the top three areas of concern for business operations. These measures improve upon the previous “obstacle variables� because respondents select only three areas they regard as main constraints for their business, thus reducing the bias introduced by optimists or pessimists. For example, downbeat entrepreneurs could respond that all areas were obstacles to business; now, they have to decide on the three most important constraints. 2.21 These top-three questions are only available in the latest surveys gathered in Latin America. Figure 2.3 compares the responses for Central America and the rest of LAC. The areas are grouped into five categories: rule of law (corruption, crime, and practices of informal competitors), regulations (license and permits, custom and trade rules, labor regulations, tax rates and administration, and functioning of courts), infrastructure (electricity, transportation, and access to land), access to finance, and human capital (inadequate education of the workforce). 12 Figure 2.3: Investment Climate constraints identified among the three most serious by entrepreneurs Central America (CA5) Latin America (LAC10) Finance Finance Rule of Law Human Cap ital Rule of Law Human Cap ital 9% 9% 38% 8% 45% 7% Infrastructure 10% Infrastructure 16% Regulations 22% Regulations 36% Note: The rankings are subjective perceptions of entrepreneurs, but in this case they are asked to choose the top three constraints for business. CA5 includes El Salvador, Guatemala, Honduras, Nicaragua, and Panama. LAC10 includes Argentina, Bolivia, Chile, Colombia, Ecuador, Mexico, Paraguay, Peru, Uruguay, and Venezuela. Source: Enterprise Surveys 2.22 In Central America, 45 percent of firms consider the (lack of) rule of law a top constraint for their businesses, a figure significantly higher than the 38 percent reported by the rest of LAC firms. Despite this, or perhaps consistent with a weak rule of law, regulations appear less problematic in Central America than in the rest of LAC. This is consistent with Doing Business reports in terms of the number of reforms Central American has implemented in the past few years (see Chapter 3 for more detail). In addition, the proportion of firms ranking infrastructure 12 As in the previous case, we exclude instability variables (macroeconomic and political instability) because they are not defined in the questionnaires and are subject to interpretation by each respondent. 22 as a top-three constraint is higher in Central America than the rest of LAC, while human capital and finance constraints are of similar importance in both regions. Objective Investment Climate Indicators 2.23 This section summarizes selected data that describe how entrepreneurs responding to the Enterprise Surveys view Central America’s investment climate. Their perspectives show where the region has room for improving its investment climate. Table 2.2: Governance indicators Percentage of annual sales Paid to "get things done" Region Lost due to Declared for Firms that say How much represents Spent in security crime tax purposes "yes" (%) of annual sales? (%) CA 1.86 2.14 73.47 17.67 1.53 LAC 1.27 1.16 77.70 15.73 0.85 SSA 1.49 0.88 66.67 41.64 2.49 EAP 2.33 0.56 75.74 47.96 1.72 ECA 1.10 0.60 85.18 33.25 0.98 Hi-OECD 1.38 0.53 93.13 21.09 0.34 MENA 1.07 0.64 83.30 31.02 2.24 SAR 1.07 0.48 92.44 64.37 1.82 Note: Central America (CA), rest of Latin America and the Caribbean (LAC), High income OECD countries (Hi-OECD), Eastern Europe and Central Asia (ECA), East Asia and Pacific (EAP), Middle East and North Africa (MENA), South Asia (SAR), and Sub-Saharan Africa (SSA). Source: Enterprise Surveys 2003-2007 2.24 Table 2.2 provides the regional averages for four governance indicators. 13 The first two columns capture crime at the firm level. Crime can deter firms from investment and divert money from productive activities to prevention. Crime can also affect workers by creating fears that reduce productivity or increase absenteeism. 14 At the macroeconomic level, previous studies suggest that law and order and corruption are key factors for growth and development (see Jaramillo and Lederman 2006, and research cited therein). These findings also hold from the microeconomic perspective of individual firms (The World Bank, 2007, Chapter 2). 2.25 Spending on security should help reduce the chances of being crime victims. For example, East Asia and Pacific firms report the highest percentage spent on crime prevention, but they also report one of the lowest percentages of sales lost due to crime. This is not the case for Central America. The region reports the highest percentage of sales lost due to crime (2.1 percent), well above the second highest region, the rest of LAC. The percentage of sales spent on security in Central America is also among the highest, only surpassed by East Asia and Pacific firms’ average. 13 Note that these averages are computed using the sampling weights of the firms that responded the Enterprise Survey in each country. 14 See Chapter 3 for a detailed analysis of the cost of crime in Central America. 23 2.26 Following The World Bank (2007), we approximate the level of regulatory compliance as the share of sales declared for tax purposes, a reflection of excessive and/or arbitrarily enforced rules and regulations. 15 Central American entrepreneurs report, on average, only 73 percent of their annual sales are reported for tax purposes. This percentage is significantly smaller than the LAC average, and only Sub-Saharan African firms report a lower percentage. Finally, firms are asked whether establishments like theirs make informal payments to “get things done� with regard to customs, taxes, licenses, regulations, etc. The percentage of Central American firms that say “yes� is among the lowest—18 percent versus 16 percent in the rest of LAC. In addition, informal payments as a percent of annual sales is higher in Central America than in the rest of LAC -1.53 percent versus 0.85 percent-, but it is lower than in Africa (2.49), South Asia (1.82) and East Asia and Pacific (1.72). 16 2.27 Other areas of the investment climate that promote productivity enhancement and firm growth are infrastructure, technology, and innovation. Access to finance, measured by the percentage of firms facing constrained credit, does not appear as a major problem in any region except Sub-Saharan Africa (Table 2.3, first column). 17 This may reflect a selection problem in Enterprise Surveys: only firms that do not need access to credit survive, and only surviving firms are included. However, Chapter 5 uses household survey data for Guatemala and Nicaragua and finds that firms do not seem extremely credit constrained. These findings do not suggest Central America has no finance issues. In fact, Chapter 5 discusses many limits on the region’s financial development, such as regulations, legal framework/protection of creditor rights, financial information, bank supervision, and payments system. 2.28 Data limitations make it difficult to compare regions on such infrastructure factors as transport or communications. Table 2.3 reports an analysis for electricity. It shows that Central American firms lost an average of 2.6 percent of annual sales due to power outages, lower than in Sub-Saharan Africa, the Middle East and North Africa, and South Asia. These responses reflect recent power problems, while the proportion of firms that own or share a generator may relate to past energy crises. On this aspect of electricity, Central America ranks low. Energy security is a top priority for the world, due to rising oil prices and climate changes (droughts) that may affect hydropower generation in some regions. Chapter 4 provides a detailed analysis of this issue for Central America. 15 There is evidence in the empirical literature that supports the view that regulatory compliance is directly related to the quantity, quality, and enforcement of regulations (see, for instance, Olson, 1999), hence we believe our proxy to correctly reflect regulatory differences within a country. 16 It is interesting to note, however, that among firms responding affirmatively, informal payments as a percent of annual sales is highest in Central American—10 percent, followed by 7 percent in the Middle East and North Africa, 6 percent in the rest of LAC and in Sub-Saharan Africa, less than 4 percent in East Asia and Pacific, less than 3 percent in Europe and Central Asia and South Asia, and less than 2 percent in high income OECD countries. 17 Firms that need a loan and don’t apply are considered credit constrained. Two conditions define constrained: (1) the firm has applied for a loan but has been rejected or (2) the firm has not applied for a loan for reasons other than “don’t need a loan.� 24 Table 2.3: Other investment climate indicators Finance Infrastructure Technology, Innovation Region Not credit Suffered Sales lost due Owns/shares Use foreign Webpage Provides constrained power to power a generator technology (%) training (%) (%) outages (%) outages (%) (%) (%) CA 90.32 30.83 2.56 12.79 16.81 30.67 34.57 LAC 84.41 17.34 0.87 13.64 9.94 39.50 42.92 SSA 49.28 79.01 7.37 32.78 12.59 16.69 34.73 EAP 95.84 43.84 1.90 24.35 16.35 28.42 62.60 ECA 93.90 29.60 1.47 52.54 14.79 63.53 43.05 Hi-OECD 97.08 2.41 0.07 NA NA 80.55 42.24 MENA 96.04 59.19 4.55 27.70 8.62 25.09 19.65 SAR 97.47 70.14 6.15 61.70 NA 29.92 19.97 Note: Central America (CA), rest of Latin America and the Caribbean (LAC), High income OECD countries (Hi- OECD), Eastern Europe and Central Asia (ECA), East Asia and Pacific (EAP), Middle East and North Africa (MENA), South Asia (SAR), and Sub-Saharan Africa (SSA). Source: Enterprise Surveys 2003-2007 2.29 Many studies consider innovation, technology and human capital to be key ingredients for development and growth. As Jaramillo and Lederman (2006) point out, one of the main channels through which international trade can raise long-term productivity growth is the importation of foreign technology (capital goods). However, foreign technology per se is not enough. It needs to be complemented by training and, in many cases, by adapting machinery to the receiving country’s production function, which may require developing new processes. According to Enterprise Survey data, Central America fares well when compared with other regions: 16.8 percent of firms use foreign technology, the highest regional average. 2.30 However, Central America does not do as well on other technology and innovation indicators in Table 2.3. The likelihood of firms’ using Web pages to communicate with clients and suppliers and providing training to workers is significantly lower than in the rest of LAC, but they are above Middle East and North Africa and South Asia averages. Chapter 6 provides an analysis of these issues for Central American countries. 2. MICRO DATA LESSONS 2.31 One of the Enterprise Surveys’ advantages, compared to such alternatives as Doing Business, is the possibility of exploring differences within countries. The micro data allows the study of differences by firm size, export status, and ownership. 18 In addition, there are two surveys available for El Salvador, Guatemala, Honduras, and Nicaragua. The first survey was conducted in 2003 for manufacturing firms only. Therefore, the panel analysis is only limited to manufacturing firms. 2.32 As we discussed on the previous section, one advantage of ES data is that it provides subjective and objective measures of the investment climate as reported by entrepreneurs. This allows us to evaluate the consistency between the subjective opinions of businessmen and the 18 Table A.2 in the appendix describes the sample size for each country and category. 25 objective measures reported in the surveys. Table A.2.3 in the Appendix shows the correlation between the variables that measure the percentage of firms reporting each area as a major or severe obstacle for business operations, and selected objective variables such as the percentage of firms paying to ‘get things done’, the average percentage of sales lost due to power outages, the use of email to communicate with suppliers and customers, among other variables. Interestingly, we find a positive and significant correlation (at 5 percent significance level) in most of the correlations studied. 19 2.33 For example, the correlations show a positive and significant relationship between electricity being an obstacle and the losses due to power outages, the use of own or shared generator, and the proportion of energy from it; there is also a positive association between tax rates obstacles and the manager’s time spent dealing with regulation (and the same for tax administration as an obstacle for business). In addition, those firms that consider transport to be an obstacle for firms operations also report relatively higher levels of cargo lost during domestic transportation due to breakage or spoilage. Although these are simple correlations, they show a consistency in the information that the private sector reports in the Enterprise Surveys. Figure 2.4: Differences in the investment climate by firm characteristics, Central America average, 2007 Panel B. By exporter and foreign-owned Panel A. By firm size classification Small M edium Large Exp orter Non-exp orter Foreign owned Locally owned Sales not lost due to Sales not lost due to crime crime 100 100 Has ISO certification 50 Regulatory compliance Has ISO certification 50 Regulatory comp liance 0 0 Sales not lost due to Provides training Provides training Sales not lost due to outges outges Uses foreign technology Uses foreign technology Note: all variables are defined such as the higher value, the better. The average values are reported in the Appendix. 2.34 Firms that report inadequate educated workforce as a major or severe obstacle are also those who provide training to their workers. Regional averages by firm characteristics show significant differences by firms’ size, export status, and presence of foreign ownership (Figure 2.4; see Table A.2.4 in the Appendix for the test of differences in means). However, the differences are noteworthy for innovation, human capital, and technology variables. The percentage of firms that have ISO certifications, provide training, and use foreign technology increases with firm size; it is also higher for firms selling at least part of their output in international markets and those with foreign participation. Interestingly, as Figure 2.4 (panel B) shows, the pattern for non-exporters is very similar to the one for domestically owned firms. 19 In a few cases, the correlation between objective and subjective variables was negative, but not statistically significant. 26 2.35 These descriptive findings cannot shed light on causation in one direction or the other. They merely show a positive and significant association between firm size, being an exporter, and having some degree of foreign ownership. 2.36 It is interesting to consider the relationship between IC variables and firm’s characteristics jointly considered. Table 2.4 present a multivariate analysis between the probability of facing different IC constraints and firm’s characteristics such as age, size, exporter status, foreign ownership, economic sector, region, taking into account country differences. The first column shows that the probability of making informal payments to ‘get things done’ is not statistically related with the firm characteristics considered here. The second column shows that exporter firms are less likely to report losses due to crime; the same is observed for small and medium enterprises when compared to large firms (the omitted category in the regressions). However, as Box 3.2 in chapter 3 points out, the average losses as a percentage of sales are much higher for small firms. 2.37 The third column of Table 2.4 shows that exporter firms are more likely to suffer losses due to power outages, and the same is true for large firms. As in the case of crime, when the proportion of sales lost due to power outages is considered, small and medium firms suffer more than large firms. The next column measures the probability of not being credit constrained. The multivariate analysis indicates that small and medium firms are more likely to be credit constrained. This result does not strike as surprising, however, it contradicts new evidence from household survey analysis for Guatemala and Nicaragua presented in Chapter 5. The problem may be related to the sample considered in each study and the way in which “credit constrained� is defined. See Box 5.1 in Chapter 5 for a discussion of this issue. Table 2.4: Multivariate analysis of IC variables and firm’s characteristics Pay to 'get Lost sales Experienced Not credit Offer Use webpage things done' due to crime power outages constrained training Exporter -0.007 -0.064*** 0.040* 0.016 0.100*** 0.136*** [0.024] [0.022] [0.023] [0.013] [0.029] [0.024] Foreign owned -0.003 -0.038 0.012 0.010 0.071 0.092*** [0.032] [0.029] [0.030] [0.019] [0.045] [0.032] Log of firm's age -0.012 -0.002 0.017 0.009 0.013 0.032** [0.013] [0.011] [0.012] [0.006] [0.016] [0.013] Small 0.002 -0.202*** -0.118*** -0.121*** -0.497*** -0.497*** [0.032] [0.027] [0.029] [0.020] [0.042] [0.030] Medium 0.042 -0.077*** -0.048* -0.064*** -0.254*** -0.250*** [0.031] [0.026] [0.028] [0.020] [0.040] [0.029] Services 0.000 0.062*** 0.017 0.02 -0.097 0.121*** [0.025] [0.021] [0.022] [0.013] [0.343] [0.023] Other sector 0.012 -0.027 -0.055* 0.031 -0.171 0.197*** [0.036] [0.034] [0.032] [0.022] [0.202] [0.034] Observations 1850 2929 2690 3028 1932 2842 Standard errors in brackets. * significant at 10%; ** 5%; *** 1%. The dependent variables are binary variables that take value 1 if the firm reports the IC characteristics described above and zero otherwise. The regression is estimated using probit model. The omitted categories are: Large firms, Costa Rica, and manufacturing sector. All regressions include country and regional dummies and a constant term. Source: Own calculations using Enterprise Survey 2003-2007 27 2.38 Finally, the last two columns of Table 2.4 refer to technology and human capital. Exporters and large firms are more likely to provide training for their workers and to use a webpage to communicate with clients and suppliers. The latter is also true for firms that have any level of foreign ownership, for older firms, and for non-manufacturing firms. 3. INVESTMENT CLIMATE, PRODUCTIVITY AND TRADE 2.39 In this section, we address the role of the investment climate on firm productivity and growth. The basic idea is that productivity is a function of firms’ characteristics (age, size, industry, and location, for example) and investment climate attributes (regulations, access to finance, infrastructure, technology, etc.). As an econometric proposition, however, it is difficult to establish causality from investment climate attributes to firm productivity and growth. 20 Are firms in better environments more productive, or do productive firms generate a better investment climate? 2.40 Escribano and Guasch (2005) addressed this issue by using a statistical method that takes into account the endogeneity between firm performance and investment climate attributes at the firm level. 21 They applied this methodology using Enterprise Survey data for Guatemala, Honduras, and Nicaragua. 22 This method has been used in many Central American country ICAs (Guatemala ICA 2004, 2008, Honduras ICA 2004, El Salvador ICA 2005, Costa Rica ICA 2007, Panama ICA 2007) and in a regional study for Latin America (The World Bank, 2007). 23 All studies find a significant relationship between the investment climate and firms’ productivity (see Box 2.1 for summaries of previous studies). 2.41 Enterprise Surveys contain more than 20 variables that can be used to characterize the business environment. In several studies, Escribano and Guasch use a “general to particular� approach where they include a large set of characteristics as explanatory variables and then reduce the number based on the significance level of each variable. Alternatively, some studies construct indexes that summarize investment climate attributes into a few variables. For example, the Panama ICA (2007) uses the first factor of principal component analysis to construct indexes for the four main areas—governance, infrastructure, finance, and human capital and technology. 24 We follow this approach for our sample of six Central American 20 The argument is similar to the one for exporter premium and productivity. Are productive firms more likely to export? Or do firms become more productive by being exporters? 21 The statistical method is known as an instrumental variables approach where the potentially endogenous investment climate variables—at the firm level—are instrumented using the same variables aggregated by industry and region. Following the regional study for LAC, we impose a minimum size of 10 observations to compute the region-industry-size average. If every group had at least 10 observations, the estimates corresponding to the group sample average and weighted average should be the same because the weights are identical for firms within a region-industry-size group. Where we do not have a minimum of 10 observations, we follow the same procedure for calculating group means for the purpose of constructing instrumental variables. We use the average of the corresponding region-industry regardless of firm size; if we are still we short of 10 observations, we use industry or region averages. In the worst case scenario, we treat the observation as missing. In the cases where weighted averages have to be computed at higher levels of aggregation than region-industry-size, we redefine the weights to take into account the different selection probability selection associated to the different groups. 22 Escribano, Guasch, and de Orte (2006) perform a similar analysis for Chile. 23 There are other individual ICA studies using the same methodology, see for example Ecuador ICA 2005. 24 Similar approaches have been used in El Salvador ICA 2005, and Ecuador ICA 2005. 28 countries. The surveys for El Salvador, Honduras, and Nicaragua have not been used yet in any of previous studies. Box 2.1: Investment Climate and Productivity in Previous Studies The World Bank has conducted several studies focusing on the relationship between the investment climate and firm productivity. These studies, known as Investment Climate Assessment (ICA), combine the analysis from Enterprise Surveys with relevant information to other sources and provide a practical basis for identifying areas of reform aimed at improving the investment climate (IC). Their findings are discussed with the private sector and other stakeholders in the countries. From a statistically rigorous point of view, it is difficult to establish a causal relationship between the IC and firm productivity. There are several factors that are not captured (unobserved) by the surveys that may be affecting both productivity and the business environment. Nonetheless, the econometric analysis conducted in these studies can be used to establish a significant association between the IC and different measures of competitiveness. Overall, the studies grouped quantifiable IC-variables into four broad categories: Governance (red tape, corruption, and crime); Infrastructure (transport, energy, telecommunications, water); technology and human capital (quality, innovation, and labor skills); and Finance (access to credit, corporate governance). In what follows we present a brief summary of the main studies. All estimations control for firm characteristics such as firm size, age, industry, region, among other variables. Region (Guatemala, Honduras, and Nicaragua): Escribano and Guasch (EG 2005) was a background study that served as an input in individual ICAs for those countries. Their estimates show consistently a high association between the IC and productivity. Overall, they consider that the business environment accounts for over 30 percent of productivity. The two most impacting areas are red tape, corruption and crime, and infrastructure, accounting respectively for about 12 and 9 percent of productivity. El Salvador: instead of evaluating the impact of each IC variable on firm productivity, El Salvador ICA (2005) generates an index for each of the four areas described above. Considering the sample of Salvadorian firms and a pooled sample that includes Guatemala, Honduras, and Nicaragua, it finds the technology index to have the largest quantitative effect on firms’ productivity. The index that measures governance appears as the second largest factor. In addition, small and medium-sized firms stand to benefit the most from IC-best practices, particularly in infrastructure and technology. Costa Rica: the Costa Rica ICA (2007) follows the approach of EG (2005) and concludes that the econometric analysis provides evidence in favor of a significant relationship between the IC and total factor productivity. In particular, it suggests that and increased emphasis on governance and business regulation, infrastructure, and innovation is appropriate in this country. The results are consistent with firms’ perceptions that put regulatory issues and infrastructure as severe constraints for firms operations. Panama: the Panama ICA (2007) follows the approach of El Salvador ICA by grouping the IC variables in four indexes. It finds that four areas have a significant impact on labor productivity and wages. For the average firm, the largest improvements on firm performance can be achieved though improvements in innovation and governance areas. Guatemala: Using the 2007 survey of Guatemalan firms, the Guatemala ICA (2008) performs an econometric analysis based on EG (2005) and finds that governance and regulation, infrastructure and access to finance have a significant effect on firm productivity. By solving a system of simultaneous equations, it is possible to estimate the contribution of each area of IC to economic performance measures such as productivity, employment, real wages, exports, and foreign direct investments (FDI). The biggest contribution to firm productivity is given by improvements in red tape, corruption and crime. This area is also important to improve real wages. 2.42 The governance index is based on the percentage of annual sales spent in security expenses, lost due to crime, and not reported for tax purposes. We expect all these variables to be negatively related to firm performance because enterprises are diverting resources from 29 productive uses to, for example, fight crime and violence. 25 The infrastructure index measures the proportion of sales lost due to power outages and the proportion of shipment value lost in domestic transport. It is also expected to be negatively associated with firms’ performance. The finance index captures whether firms have overdraft facilities or lines of credit and whether they are constrained in their access to bank credit. We expect a positive relationship to firm performance. Finally, the technology index includes indicators for training offered to workers, for Internet use to communicate with clients and suppliers, and for R&D activities. These factors should enhance firm performance. The models control for firm characteristics such as age, size, exporter status, foreign ownership, establishment uniqueness, and industry and region. Table: 2.5: Firm performance and the investment climate in Central America Sales per worker Total Factor Sales per worker Wages (*) (TFP sample) Productivity Weak Governance -0.339*** -0.345** -0.161 -0.214 [0.091] [0.149] [0.184] [0.162] Access to Finance 0.081 0.101 0.139 0.071 [0.103] [0.120] [0.141] [0.121] Poor Infrastructure -0.256** -0.264* -0.23 -0.159 [0.112] [0.145] [0.164] [0.141] Better Technology 0.195* 0.546*** 0.482** 0.542** [0.110] [0.208] [0.241] [0.225] Single location -0.254*** 0.079 -0.186 0.224* [0.079] [0.102] [0.118] [0.130] Log of firm's age 0.023 0.024 0.122*** -0.016 [0.032] [0.043] [0.046] [0.043] Foreign owned 0.488*** 0.405*** 0.707*** 0.067 [0.090] [0.118] [0.138] [0.114] Exporter 0.205*** 0.231** 0.470*** 0.058 [0.070] [0.102] [0.118] [0.104] 20-99 employees 0.306*** -0.109 1.356*** -0.29 [0.099] [0.196] [0.235] [0.222] 100+ employees 0.237* -0.315 2.977*** -0.25 [0.142] [0.306] [0.358] [0.339] Observations 2715 1538 1538 1226 Note: Dependent variables are expressed in logs, so that the reported coefficients can be interpreted as relative changes in respectively sales per worker, total factor productivity and wages. The four explanatory variables in the top panel have been normalized so that the coefficients represent changes in the dependent variable resulting from a one standard deviation improvement in the corresponding investment climate index (e.g. a 33.9% increase in sales per worker for each decrease of one standard deviation in the governance index). All regressions include industry and region dummies and a constant term. (*) in the wage equation we also control for the education level of the average worker. Robust standard errors in brackets. * significant at 10%; ** 5%; *** 1%. Source: Own calculations using Enterprise Survey 2003-2007. 25 By not reporting total annual sales for tax purposes, firms also hurt the firm performance observed for example by financial institutions, therefore reducing investment opportunities, credit access, etc. 30 2.43 Table 2.5 presents the relationship between the four investment climate indexes and firm productivity as measured by sales per worker, total factor productivity (TFP), and wages. 26 The first two columns show that a better business environment is associated with higher firm productivity in Central America. Three out of four investment climate areas appear significantly related to sales per worker. When it comes to TFP and wages, technology appears as the only significant investment climate area associated with firm performance. 2.44 An alternative statistical approach to address the endogeneity problem between firm performance and the investment climate at the firm level is to measure the investment climate using objective measures from other sources, such as Doing Business and Kauffman indicators.27 Lederman (2007) uses this method to address the relationship between innovation, R&D and the investment climate. Lopez-Cordova (2007) follows a similar approach to address the role of trade liberalization and regulations in firm growth. 28 Melitz (2003) demonstrates that greater trade openness raises industry productivity through selection effects and production re-allocation effects. Lopez-Cordova extends Melitz’s model to show that labor market regulations affect economies’ ability to allocate resources to more efficient firms, thus limiting the impact of trade liberalization on average productivity. 2.45 We explore the links between firm productivity, trade liberalization, and other investment climate characteristics. Following Lopez-Cordova (2007) and Lederman (2007), we combine Enterprise Surveys’ firm-level data with trade variables at the country-industry level and country-level investment climate variables from other sources (Doing Business, World Development Indicators, Kauffman, etc.). 29 Tariff data come from the World Bank’s World Integrated Trade System (WITS). 30 We aggregate tariff information at the industry level to match the Enterprise Surveys classifications—food, garments, textiles, machinery and equipment, chemicals, electronics, and non-metallic minerals. 2.46 Table 2.6 presents econometric evidence on the association between firm productivity, trade liberalization, and the investment climate. Firm productivity is measured with sales per 26 In the case of total factor productivity (TFP), the dependent variable is value added per worker controlled for capital per worker. Since many firms do not report their stock of capital, we also estimate the model for sales per worker restricting the sample to the TFP one. 27 Lederman (2007) points out that using aggregate firm-level data as exogenous variables (instruments) is based on the assumption that individual firms do not affect the average used as instrument. 28 A limitation of this approach is that it includes variables that are constant within countries, thus limiting the degrees of freedom to identify these variables. Therefore, it is not possible to perform this analysis for a single country or even a small number of countries –like CA region. It is necessary to consider a broader set of countries for which different sources of information are available. 29 The main criticism received by the standard productivity analysis of individual ICAs is the endogeneity problem between firm productivity and investment climate variables based on Enterprise Survey microdata. If endogeneity concerns are not properly addressed, it is not possible to make policy recommendations from that analysis. We tried to address this problem using an instrumental variable approach where the endogenous variables are the first component of principal component analysis (PCA) for each area of investment climate (governance, access to finance, infrastructure, etc.) and the instruments are the averages of the single variables used in the PCA. By doing this, we have more instruments than endogenous variables and we can test for the validity of the instruments. This approach improves upon previous studies (Economic performance, ICA Panama, etc.) that had a one-to-one relationship between endogenous variables and instruments. Unfortunately, our estimations show that the instruments used are not valid, i.e., they do not pass the standard exogeneity tests. 30 We thank Ernesto Lopez-Cordova for providing the tariff data. 31 worker. Trade liberalization is captured by the difference between the average import tariff at the time of the survey and three years earlier (negative values indicate tariff reductions over the period). Then we construct an indicator variable that equals one for the industries that benefited from trade liberalization, i.e. a reduction in the average import tariffs. This variable varies by country and industry. Table 2.6: Firm productivity, trade liberalization, and the investment climate Log of Sales per worker Dependent variable: (1) (2) (3) (4) (5) Country-level characteristics Quality of Regulation index 0.501* 0.988* 0.003 0.017 0.004 [0.295] [0.566] [0.046] [0.043] [0.045] Quality of Institutional index 0.172*** 0.210*** 0.568** 0.145*** 0.239*** [0.044] [0.045] [0.260] [0.041] [0.051] Quality of Infrastructure index 0.112* 0.098 0.080 1.403*** 0.061 [0.067] [0.100] [0.106] [0.496] [0.104] Access to Finance index 0.001 0.002 0.002 0.004* -0.007 [0.003] [0.003] [0.002] [0.002] [0.010] Industry-level characteristics Trade openness (TO) 1.330** 1.360** 0.481 1.550*** 4.327*** [0.623] [0.576] [0.533] [0.540] [1.165] TO * Regulation index 1.014* [0.572] TO * Institutional index 0.811*** [0.274] TO * Infrastructure index 0.616 [0.500] TO * Finance index 0.034*** [0.010] Firm-level characteristics Foreign owned 0.459*** 0.484*** 0.513*** 0.461*** 0.508*** [0.088] [0.091] [0.092] [0.085] [0.093] Has ISO 0.870*** 0.815*** 0.819*** 0.789*** 0.804*** [0.119] [0.093] [0.100] [0.096] [0.098] Exporter 0.512*** 0.517*** 0.508*** 0.540*** 0.500*** [0.088] [0.086] [0.082] [0.075] [0.079] Observations 12768 12768 12768 12768 12768 Regulation, Infrastructure, Institutional, and Finance indexes were defined so higher values connote better performance. Robust standard errors in brackets. * significant at 10%; ** 5%; *** 1% Source: Own calculations using Enterprise Survey 2003-2007 2.47 Each country’s investment climate is characterized using rankings and indexes constructed from sources other than Enterprise Surveys. All indexes were defined so higher values connote better performance. They are: 32 • Regulation index, which focuses on three Doing Business indicators—starting a business, difficulty in hiring, and difficult in firing. • Institutional index, which incorporates three Kauffman indicators—control of corruption, rule of law, and political stability. • Infrastructure index, using measures for roads and telephone data from World Development Indicators. • Finance index, which explores two sets of issues—credit information registries and the effectiveness of collateral and bankruptcy laws in facilitating lending. 2.48 Finally, firm level data comes from the Enterprise Surveys. We introduce the following characteristics: being an exporter (at least 5 percent of sales directly or 10 percent indirectly), having FDI (at least 1 percent is owned by foreign capital), having a quality certification, and firm size. 2.49 Table 2.6 shows a positive association between trade liberalization and higher labor productivity. The first column also indicates that the investment climate—captured by the regulation, institutional, and infrastructure indexes—is also positively related to firm performance. The next four columns incorporate the interaction between trade openness and the four investment climate indexes. For example, the second column shows that better regulations appear to enhance the positive effects of trade openness. This result is consistent with Lopez- Cordova (2007), who finds that burdensome labor regulations reduce the benefits of trade liberalization for firms’ growth. 4. SUMMARY AND CONCLUSIONS 2.50 This chapter discussed Central America’s investment climate and compared it to other regions. Central America appears better positioned than the rest of LAC on the Global Competitiveness Index, the Trading across Borders ranking, and the Logistic Performance Index. There is diversity within the region, with Panama and Costa Rica, and sometimes El Salvador performing generally better than Guatemala, Honduras, and Nicaragua. Over all, though, the region could still benefit from improvements in the business environment. 2.51 Subjective and objective data collected in Enterprise Surveys provide another way to examine Central America’s investment climate. The reports incorporate new data for El Salvador, Honduras, and Nicaragua. The subjective views of entrepreneurs responding to Enterprise Surveys indicate that corruption, energy, crime, the practices of informal competitors, and tax rates are areas of major concern. These opinions are supported by objective measures. The region has high percentages of sales lost due to crime, undeclared for tax purposes, and used to make informal payments “to get things done.� In addition, the region fares worse than the rest of LAC in power outages—both the proportion of firms reporting them and the proportion of sales lost due to them. It also trails in technology and human capital indicators—the use of Web pages, providing worker training, and the proportion of firms with ISO certifications. On a positive note, Central America performs better than the rest of LAC in terms of access to finance, measured by the proportion of firms that are not credit constrained. 33 2.52 Micro data are useful for studying investment climate constraints on firms of different size, exporter status, and type of ownership. Firms with more than 100 employees, significant exports or some level of foreign ownership report better investment climate indicators, according to the latest surveys. A productivity analysis shows a positive association between the business environment and firm performance. In addition, it shows that trade openness is positively associated with firm performance, and a better investment climate may enhance the benefits of trade openness. 2.53 This chapter concentrated on four broad areas of the investment climate—governance, infrastructure, finance, and technology and innovation. They are the subjects of this study’s remaining chapters. Each of them augments discussion of current trends with policy recommendations for the region and for individual countries. 34 Chapter 3. GOVERNANCE AND CORRUPTION: IMPLICATIONS FOR BUSINESS INVESTMENT 1 INTRODUCTION 3.1 Governance definitions tend to be rather broad and include components that are in themselves complex to the point that measurement difficulties could render them useless for applied research. Components can include parliamentary development, electoral system and practices, accountability, social capital, public administration, institutional quality, and decentralization. The Worldwide Governance Indicators (WGI) project proposes a definition that is broad, but has a fairly simple and relatively brief formulation: “... the traditions and institutions by which authority in a country is exercised. This includes the process by which governments are selected, monitored, and replaced; the capacity of the government to effectively formulate and implement sound policies; and the respect of citizens and the state for the institutions that govern economic and social interactions among them.� In addition, part of the UNDP definition provides a useful complement in line with this report’s focus: “It is the rules, institutions, and practices that set limits and provide incentives for individuals, organizations, and firms (2004).� 3.2 Most governance definitions center on the proposition that citizens’ support and respect for democratic institutions is fundamental. This is relevant to the business and investment environment because this dimension of governance directly affects the credibility and stability of the overall society and could constitute a source of uncertainty for investors if core democratic institutions were to be shaken. 3.3 On this level of measurement, governance in Central America, as well as in much of the rest of Latin America, has been undergoing a steady decline for some years now. Three different indicators of system support register a substantial weakening in citizens’ commitment to democracy. An additional element is citizens’ wavering support for the economic system. A review of determinants of citizen support for democracy points to the need for improving public service delivery, civic education, and public security, as well as stepping up efforts to combat corruption. 3.4 Government policies and behavior exert a strong influence on the investment climate through their impacts on costs, risks, and barriers to competition (Escribano, Guasch, and de Orte, 2006). 2 Among other governance-related factors, constraints on starting and closing businesses, mandatory compliance with time consuming norms and regulations, excessive red tape and bureaucracy, and weak rule of law increase significantly the cost of doing business, discourage investments, and harm overall competitiveness and growth (World Bank, 2007). 1 This chapter was prepared by Alberto Leyton and Amalia Prado, with contributions from Gabriel Demombynes, Enrique Fanta, Tanya Gupta, and Diana Hincapie. 2 See Guasch and Hahn, 1999, WDR 2005. Kaufmann, Kraay, and Zoido-Lobatón (1999); Knack and Keefer (1995); Mauro (1995). Fisman and Svensson (2000), Svensson (2003), and Murphy, Shleifer and Vishny (1991, 1993) Loayza, Oviedo and Serven (2005b) Djankov, McLiesh, and Ramalho (2006). 35 Many studies highlight how improvements in governance—as well as in areas such as infrastructure, innovation, and education—enhance prospects for aggregate growth. Trade liberalization pacts, such as the Dominican Republic-Central American Free Trade Agreement (DR-CAFTA), are likely to provide additional incentives for governments to consolidate recent policy reforms and push forward in the areas where weaknesses remain across Central America. 3.5 Overall institutional quality and business constraints and incentives both matter from a growth perspective. This chapter examines the regional governance situation along these dimensions and at different levels of analysis, focusing on regional patterns and trends, as well as differences between countries. The first section describes regional public opinion trends about democratic governance and institutions that show a decline in system support. The second section provides a more aggregate regional perspective of governance based on the views of a large number of enterprises, citizen, and expert survey respondents combined in the WGI’s six components. The third section examines business-specific measures of governance using information on rules and regulations from the Doing Business project and information on firms’ reports of actual practices from the Enterprise Surveys project. Also from a business-specific perspective, the determinants of crime and the relationship between corruption and other areas of the investment climate are addressed in greater detail, given their importance for the investment climate. The fourth section discusses customs regulations, a governance topic determinant for growth and a key for success in the context of DR-CAFTA. 3.6 In light of free-trade agreements and regional integration initiatives, the recommendations suggest focusing reform on topics with highly unfavorable average positions and large differences between countries (even if the average position is favorable). Such an approach would propel the region as a whole forward on topics where a majority of countries perform poorly, while bridging gaps and bringing laggards up to speed. Some interventions intended to reform and modernize country systems could be scaled up to a regional dimension, amplifying both their impacts and their chances to succeed. This report identifies the areas of opportunity and challenge best suited for a regional approach. 3.7 The region needs a gradual shift in attention toward more decisive interventions aimed at strengthening and modernizing the rule of law and enforcement-related systems and institutions. This should include actions in formal enforcement, investigation, prosecution, and judging mechanisms. However, it should also encompass setting the right incentives; implementing effective actions to reduce vulnerabilities to informality, crime, and corruption; and strengthening capacities to mitigate risks of state capture practices over these institutions. 3.8 A cross-cutting issue is the need for monitoring and evaluation mechanisms that focus on impact rather than activities and products; that is, recording improvements on specific indicators does not ensure the final objective of improving the investment climate is being met. 3.9 More concrete actions that could be taken at the regional level include: • As part of the regional approach, enhance and/or establish region-wide institutions to facilitate the enforcement of common rules and regulations in sensitive areas such as: (a) strengthened investigation and prosecution of crime through regional prosecutors’ networks; (b) oversight of courts’ performance by strengthening the existing regional court; and (c) 36 strengthening of the overall state management, transparency and accountability in each country by establishing a regional control system and a transparency regulator. • Increase transparency through implementation of standardized legal frameworks to regulate freedom of information mechanisms across the region. Thus far, only Honduras and Nicaragua have issued legislation to guarantee access to public information, but they have only started the implementation process. Common frameworks to access information in all countries will facilitate monitoring and benchmarking. Regional standards could emerge in the context of regional institutions like those proposed here. • Replicate regional best practices in the area of process simplification and use of actionable performance indicators to improve public sector systems and services. One of the best practices worth replicating is El Salvador Eficiente. It has taken a practical approach to tackle Salvador’s red tape constraints by using performance indicators as a reference and by unbundling determinants of poorly rated indicators to correct problems in a focused manner. • Hold privatized and non-privatized providers of public services accountable by setting up regional monitoring mechanisms that combine user satisfaction surveys and other indicators (frequency and duration of power outages, for example). Such mechanisms would identify bottlenecks, detect quality and coverage issues, and keep track of delivery gaps. By promoting user satisfaction, it would also indirectly attack the decline in support for the political and economic system. • Reform customs with an eye toward two types of changes. (1) Short-term actions, or “quick wins,� that could have a relatively rapid and significant impact. They include enhanced capacity to manage rules of origin compliance to avoid repeat customs inspections and increased quality of security, especially at key international ports. (2) Medium- to long-term actions related to institutional strengthening within customs administration. They include: o Undertaking civil service and human resource reforms in customs administration, including the private as well as the public sector; o Implementing fully the common customs manual; o Improving customs systems by integrating them with tax administration techniques, using process analysis to identify areas of simplification and integrating risk analysis techniques without compromising transparency; o Improving integration by creating an automated integrated clearance system that is linked with other systems, completing work on tariff elimination, limiting each border to one control post, and implementing the common customs manual; • Use information and communication technology to automate processes and mitigate risks of mismanagement and corruption in sensitive areas like tax collection and customs control. The Mexican experience in modernizing the tax administration agency by using process reengineering and automation serves as a model for the region. In this same area, border- level integration of customs information proved to be very effective in both reducing risks of contraband and corruption and facilitating legal trade by reducing hassles for legal importers and exporters. The US-Mexico customs agreement could serve as a model for the region 37 • Make crime a primary focus of public policy, particularly in the high-crime countries of Guatemala, Honduras, and El Salvador. Narrow criminal justice approaches to strengthen and improve police and court systems need to be complemented with crime-prevention efforts, which range from anti-violence education programs to crime-deterring environmental design. Focusing resources on prevention may be an advisable strategy for Nicaragua, Panama, and Costa Rica, which still have relatively low crime rates but have seen increases in recent years. Institutional and political economy analyses of incentives and better understanding of organized crime are also needed in all cases. 3.10 Finally, policies linking business environment improvements with delivery of services to citizens and overall welfare are needed to address potential trends that could threaten the democratic system and its institutions and spill over into the economic realm. Reconciliation of business- and citizen-oriented policies should be encouraged to counter a growing perception that businesses are growing at the expense of higher prices for the basic goods and services needed by the poor. More explicit government interventions to help the poor benefit from economic growth are strongly urged. 1. POLITICAL SYSTEM MEASURES 3.11 Stability of the political and economic system and its institutions is ensured by citizens’ support and respect for the “rules of the game� and confidence that the institutional framework responds to their needs. This is particularly relevant to the regional investment climate because Central America, along with most of Latin America, has for some years now been undergoing a steady decline in confidence in government institutions and the political process (system support). Moreover, in a context of resurgent populism, this trend increases the risk of mounting uncertainty about the long-term investment policies as disaffected citizens’ votes are cast for political options that offer to do away with institutions considered corrupt and/or inefficient. 3.12 Three different indicators register substantial deterioration in citizens’ commitment to the current political system and its institutions—and to the economic system as well. Citizens’ support for democracy, satisfaction with its performance, and confidence in its institutions have declined. A review of predictors of system support points to the need for improving public service delivery, civic education, and public security, as well as stepping up efforts to combat corruption. 3.13 At the start of this decade, analysts hypothesized that citizen support for democracy was consolidating worldwide, representing a move away from more authoritarian forms of government (Norris, 1999). This assumption, however, has not played out uniformly across Central America, where only two countries register higher levels of support for democracy than a decade ago; perhaps because initial promises of fast development and greater voice in the political process have progressed more slowly than expected. A second indicator, citizens’ satisfaction with the way democracy is working in their country, taps respondents’ perception of whether democracy is delivering the political, economic, and social goods they expect. On this measure, system support in Central America has deteriorated across the board. 3 3 See Chapter 3 appendix. 38 3.14 Trust in institutions is yet another measure that shows declines in recent years, although lack of data for all of Central America makes this information illustrative rather than representative. Average indictors show declines for Costa Rica, El Salvador and Panama from 2004 to 2006 (Table 3.1). Nonetheless, some indicators have improved. Panama, for example, registers an increase in confidence in government, Congress, and political parties. Interestingly, the institutions suffering the largest declines are the electoral board and justice system, both of which are a determinant for citizen trust in numerous other institutions. Table 3.1: Confidence in institutions in 2004 and 2006 (scale: 0 – 100) Costa Rica El Salvador Panama 2004 2006 2004 2006 2004 2006 Government 58 53 61 52 38 47 Congress 53 49 53 49 37 45 Political parties 35 36 40 35 33 40 Courts/Justice system 57 51 53 48 48 41 Electoral board 71 67 60 50 66 49 Ombudsman 73 71 65 65 54 47 Average difference -3.3 -5.5 -1.2 Sources: Latin America Public Opinion Project (LAPOP) 3.15 Predictors of system support in the political economy literature vary considerably by study and region. Individual-level socioeconomic variables are routinely included in analyses, but their effect disappears once political knowledge is included, pointing to the importance of civic education in the political socialization of new generations. This aspect gains relevance because youth tend to be less supportive of the political system. For the LAC region, Latin American Public Opinion Project (LAPOP) country studies point to two other factors that are negatively associated with system support—corruption and public insecurity. That is, the more pervasive the perception of corruption, the less system support; the same is true for insecurity. 3.16 The Latinobarometer finds that support for and satisfaction with the government are somewhat related to state of the economy, as measured by per capita GNP. The relation is not lineal, nor is it proportional. An interesting preliminary finding in Latinobarometer (2007) is a positive association between support for the government and satisfaction with public services’ delivery. The report hypothesizes that, over the past two years, a decline in the satisfaction with public services throughout the region would explain part of the decrease in support for new democracies. 3.17 An additional measure of system support is especially relevant to the investment climate: perceptions of the country’s economic model. In 2007, an average of 52 percent of Latin Americans agreed with the following statement: “A market economy is the most convenient for the country.� By contrast, only 34 percent of Guatemalans reported the same belief. Since 1998, the average drop for the six Central American countries is 23 percentage points, with Guatemala posting the largest decline and Nicaragua the smallest (Table 3.2). The Central American average decreased by 9 percentage points more than the LAC average. A similar measure asks about agreement with the statement: “A market economy is the only system whereby the country can achieve development.� Overall, decreases between 1998 and 2007 are smaller than the previous question, with Central American averaging 12 percentage points. 39 Table 3.2: Agreement with “A market economy is the most convenient for the country� over time (% that “agree� and “highly agree�) 1998 2000 2002 2007 1998 to 2007 change* Costa Rica 76 65 70 56 -20 El Salvador 78 57 55 52 -26 Guatemala 77 66 68 34 -43 Honduras 70 65 66 49 -21 Nicaragua 73 75 78 66 -7 Panama 63 47 56 41 -22 Latin America 66 57 59 52 -14 * Percentage points. Source: Informe Latinobarómetro 2007 3.18 Agreement with the statement “The private sector is indispensable for the country’s development� also shows declines, some of which are quite large considering the question has only been asked for four years. Central America’s average drop of 14 percentage points is larger than the LAC average, and Guatemala again registers the largest decrease. 3.19 Negative trends in citizen support for the political system, its institutions and rules could be countered by changes in its determinants. Statistical analyses of public opinion data for the Central American countries suggest improvements in the delivery of public services and in public security, as well as more effective initiatives to combat corruption, could counter these negative trends. Additionally, waning confidence in institutions could be mitigated by improvements in their efficiency. Process simplification and use of actionable performance indicators to improve public sector systems could follow a regional best practice, the El Salvador Eficiente program whereby red tape constraints were tackled using performance indicators as a reference and unbundling determinants of poor ratings to correct problems in a focused manner. 3.20 Declines in public support for the economic system require that business- and citizen- oriented policies be combined to counter a growing perception that businesses are growing at the expense of higher prices for basic goods and services, further disadvantaging the poor. More targeted policies to help the poor benefit from economic growth are also strongly urged. 3.21 In addition to public opinion surveys, other indicators show the challenges that Central American countries face in terms of the quality of institutions, regulations (including customs), corruption, crime and violence, and other governance issues. The rest of the chapter provides descriptive an analytical data on these issues in Central America. 2. OVERALL GOVERNANCE INDICATORS 3.22 Since 1996, the WGI project has provided aggregate quantitative information—scores and percentile rankings—on six dimensions of governance 4: 4 See Kaufmann and Kraay, 2007. The WGI’s frequency and coverage have increased over the years; its latest version covers 212 countries and territories, gathering information from 33 data sources. 40 • Voice and Accountability—covers the extent of citizen participation in selection of the government, freedom of expression, freedom of association, and freedom of the media. • Political Stability and Absence of Violence—refers to the likelihood that the government will be destabilized and overthrown. • Government Effectiveness—a measure of the quality of public services, the quality of the civil service and its degree of independence from political pressures; the quality of policy formulation and implementation; and the credibility of the government’s commitment to them. • Regulatory Quality—a measure of government’s ability to formulate and implement regulations that permit and promote private sector development. • Rule of Law—refers to the confidence in and abiding by the rules of society; the quality of contract enforcement; the police; the courts; and the likelihood of crime and violence. • Control of Corruption—a measure of the extent to which public power is exercised for private gain. 3.23 The aggregate indicators combine the views of a large number of enterprise, citizen and expert survey respondents in industrial and developing countries. The individual data sources underlying the aggregate indicators are drawn from a variety of survey institutes, think tanks, non-governmental organizations, and international organizations. 3.24 WGI data point to three main inferences: (1) with regards to both the current situation and longer-term trends, the region as a whole is not particularly different from the rest of Latin America, revealing a very slow or stagnant situation in the overall perception of governance quality; (2) country comparisons suggest the region could be divided into two groups—Costa Rica, Panama, and El Salvador, which fare as well or better than the LAC average, and Honduras, Nicaragua, and Guatemala, which rate lower; and (3) Rule of Law is the region’s most problematic dimension in comparison to both the LAC and worldwide averages. 3.25 Percentile rankings for the Central American countries are evenly distributed on both sides of the LAC average. Generally speaking, Costa Rica, Panama, and El Salvador are above and the rest are below. The two exceptions are El Salvador being below the LAC average on Voice and Accountability, and Guatemala ranking above the LAC on Regulatory Quality (Figure 3.1). This appears to be a fair depiction of the state of affairs in the region (Romero and Arndt, 2007). 3.26 Nonetheless, as the WGI team notes, all measurement is fraught with error, and precision estimates for these indicators in Central America are so wide they blur the differences between countries. Costa Rica presents the only consistent and statistically significant differences from the rest. Its scores are significantly above those of Guatemala, Honduras, and Nicaragua and overlap with those of Panama and El Salvador in only a couple of instances. Otherwise, the regional panorama is relatively homogeneous: scores are not too different, and their precision estimates overlap. Table A.3.3 in the Appendix describes the indicator scores and their confidence intervals. 41 Figure 3.1: Percentile ranking of Central American countries and LAC average on governance dimensions 100 CRI SLV GTM HND NIC PAN LAC 80 Percentile Rank 60 40 20 0 Voice & Political Government Regulatory Rule of Law Control of Accountability Stability & Effectiveness Quality Corruption Absence of Violence Source: Worldwide Governance Indicators 3.27 With regard to comparing the six countries with the 2006 LAC average score—for which no precision estimates are available—scores for Costa Rica and Panama are consistently higher across indicators, while those of El Salvador overlap on all. Scores for Honduras, Nicaragua, and Guatemala are lower than the LAC average on three indicators—Government Effectiveness, Rule of Law, and Control of Corruption. On the other three indicators, results are mixed. 3.28 Each year, the worldwide WGI average is centered on zero with a standard error of one, which implies that country scores are in themselves a ranking of sorts. Comparisons over time, then, are not very meaningful because country scores can vary as a result of true changes in governance or because other countries changed. With that caveat in mind, a comparison of Central American countries across years was nonetheless undertaken to determine any patterns. A review of the 2000 and 2006 scores and confidence intervals for each country on all indicators reveals no significant differences. In addition, the trends in that period do not show consistent patterns by country or by indicator. In sum, WGI indicators should be considered as a tool to describe the region’s governance, but the analysis cannot be taken any further. 5 3. BUSINESS-RELATED MEASURES 3.29 Business-specific measures portray a relatively optimistic picture of Central American governance, with regard to both the current situation and trends over time. Using information and data from Doing Business and Enterprise Surveys, this section examines reform efforts from de 5 There are, moreover, other quite serious methodological observations to the WGI, the more important ones being: the likelihood of correlation of errors, which violates an assumption of the statistical model used; the weighting procedure for different sources; and the lack of transparency on the substantive content of each dimension of governance (Arndt and Oman, 2006; Hermann, 2004). 42 jure and de facto perspectives, highlighting trends, regional commonalities, and intraregional gaps. 3.30 In terms of Doing Business, Central America’s standing varies by topic. The region does best on Getting Credit and Closing a Business and worst on Paying Taxes, Protecting Investors, and Enforcing Contracts. These challenges show small differences between countries, which suggest reform could take a regional approach. By contrast, other topics present marked intraregional differences, which would require gap-bridging interventions aimed at leveling the field. It is important to note that “best� and “worst� performances are quite evenly distributed among countries. 3.31 The region has been actively reforming regulations. Reform has sped up recently, but a considerable number of indicators show steady improvements over at least three years, some of which managed to bridge gaps between countries. However, the topics on which the regional ranking is poorest did not fare as well in reform efforts. Paying Taxes registers scarce improvement, while Protecting Investors and Enforcing Contracts show no change. 3.32 Enterprise Surveys of actual practice provide complementary information on some topics and new information on Informality and Corruption. With regard to changes over time, a comparison of four countries between 2003 and 2006 shows as many improvements as declines. Despite this region-wide semblance of stagnation, some topics show improvements that are not offset by declines. For example, the four countries with trend data register decreases in corruption on two indicators associated with public acquisitions. In addition, some countries made considerable headway on specific topics—for example, Nicaraguan firms’ reports of decreasing corruption in nine of 10 indicators. 3.33 Comparison of Doing Business and Enterprise Surveys qualifies some conclusions, as the firm data do not corroborate improvements made in regulations. Measures of laws and regulations - Doing Business 3.34 The Doing Business report is a joint World Bank–International Finance Corporation project that provides measures of the business environment across 178 countries, based surveys of experts, most of them lawyers and accountants. As confirmed by successive reports, greater ease of doing business is associated with faster economic growth, less informality, and lower unemployment, with greater benefits accruing to women and youth. 3.35 Chapter 2 briefly discussed the Ease of Doing Business rankings, showing that Central America was on average below the rest of LAC. It also had sharp intraregional differences. The Ease of Doing Business ranking is itself an average of the rankings across 10 topics or stages of a businesses’ life—a total of 39 indicators. Figure 3.2 summarizes the ranking by country, along with the six-country average, across all 10 topics. Lower rankings reflect greater ease, so the shorter the columns, the better for business. The uneven distribution of highs and lows suggests no patterns by country, a contrast with preceding governance indicators. 3.36 Splitting the Ease of Doing Business rankings in two helps untangle the data. Simply, positions in the upper half (1 to 89) qualify as better than positions on the lower half (90 to 178). On average, the region does best (upper half) on Getting Credit (43), Registering Property (62), 43 Trading across Borders (73), and Closing a Business (88). However, sharp intraregional differences on Registering Property and Trading Across Borders complicate the picture. Considering both average position and dispersion, only Getting Credit and Closing a Business qualify as regional best performances. Figure 3.2: Central American country rankings by Doing Business topic (2008) CRI SLV GTM 160 HND NIC PAN 144 CA 140 116 119 120 101 101 99 100 88 80 73 62 60 43 40 20 0 Starting a Dealing Employing Registering Getting Protecting Paying Trading Enforcing Closing a Business with W orkers Property Credit Investors Taxes across Contracts Business Licenses Borders Source: Doing Business 2008 3.37 On average, the region does worst (lower half) on Paying Taxes (144), Protecting Investors (119), Dealing with Licenses (116), Starting a Business (101), Employing Workers (101), and Enforcing Contracts (99). Of these, Paying Taxes, Protecting Investors, and Enforcing Contracts present smaller differences between countries, suggesting these topics require attention across the region. Dealing with Licenses, Starting a Business, and Employing Workers, on the other hand, show large gaps between countries. For example, on Employing Workers, Nicaragua ranks 59th while Panama ranks 170th; on Starting a Business, Panama ranks 31st and Honduras 135th. 3.38 Doing Business reports make another noteworthy comparison about reform performance by region. In 2006, the LAC region was ranked fifth out of seven regions in the average number of reforms undertaken by each country –Central America is not considered separately. In 2007, LAC also ranked fifth on a slightly different measure, the percentage of countries that undertook at least one positive reform. According to the 2008 report, however, LAC dropped to the bottom, becoming the region that reformed the least. Reform efforts appear to have slowed down, possibly as an effect of the large number of elections taking place. In Central America alone, between late 2005 and late 2007, there were three general elections, one balloting for regional and local representatives, and two referenda. Box 3.1 summarizes reforms made by Central American countries, as reported in Doing Business. 44 Box 3.1: Reforms in Central America Costa Rica allowed traders to transmit customs declarations electronically, speeding up cross-border trade by six days for imports and seven days for exports. El Salvador established a one-stop shop for importers, facilitating the documentation and approval process. Guatemala expedited a number of document requirements by making them electronic. The time to register property has been reduced from 37 to 30 days. The implementation of a new Electronic Data Interchange (EDI) system for electronic submission of customs declarations and the introduction of risk-based inspection regime decreased the time for exporting procedures to one day. With the full implementation of the one-stop shop, time for new company registrations was cut from 30 to 26 days, and 13 procedures were reduced to 11. Guatemala focused on increasing efficiency across the board, reducing the decision- processing time for construction from 286 to 235 days. Court reforms increased the number of cases decided by justices of peace, expanding the small-claims courts. Honduras simplified municipal licensing procedures, reducing the time to build a warehouse 32 percent and the time to start a business from 44 to 21 days. Registering property has also become easier; the creation of time limits on certain procedures has reduced the process from 36 to 24 days. Honduras has begun facilitating access to credit by allowing borrowers to get copies of their personal data and obtain a free credit report once a year. In addition, parties to a security agreement can agree to enforce a security out of court by using a notary to take care of the out-of-court enforcement of collateral agreements. Source: Retrieved from http://www.doingbusiness.org/Reformers/LAC.aspx on April 8, 2008 3.39 A comparison of Doing Business reports for 2007 and 2008 shows the ranking of four Central American countries improved, while two nations declined. A more detailed analysis of specific topics and their corresponding indicators sheds added light on these changes. Table A.3.4 in the Appendix shows the number of indicators that changed over the past two years and the direction of change. For example, four indicators make up the Starting a Business topic in the first row—the number of procedures, the time in days, the cost, and the minimum capital required. From the 2007 to 2008 reports, Guatemala improved on all four, hence the “4+.� 3.40 Overall, the number of improvements far exceeds the declines. Starting a Business and Dealing with Licenses, both topics with an unfavorable regional standing, were reformed the most in terms of number of changes and their distribution across the board; cost-reducing changes were the most popular reforms. Interestingly, the number of procedures to obtain a license registered no changes at all, despite the fact that there are considerable differences between countries, suggesting room for improvement. 3.41 Getting Credit and Trading across Borders also show a high number of positive reforms, but they are offset by deteriorations. Credit information coverage accounts for both positive and negative changes, while the number of documents to import and export increased slightly. Nicaraguan private credit bureau coverage went from 3 percent to 100 percent of adults from 2007 to 2008. Getting Credit represents the region’s best performance, and Honduras has the best ranking within the region. 3.42 Two topics register no recent changes: Protecting Investors and Enforcing Contracts. Central America ranks in the lower half on both, with relatively small intraregional differences. 45 The data suggest that these topics merit attention by all countries but have gotten none. A climate conducive to investment relies heavily on these two topics, and the types of indicators involved point to their complexity. Countries’ commitment to highly targeted and simpler changes, however effective these may be in improving their rankings, appear to be leaving fundamental reforms by the wayside. 3.43 Employing Workers, Paying Taxes, and Closing a Business were reformed the least. The first is probably more politically controversial, but the latter two include indicators that measure administrative burden, which differ markedly between countries. It takes Nicaraguan firms 240 hours a year to pay taxes; in Panama, it takes 482. Closing a Business takes 2.2 years in Nicaragua and four in El Salvador. 3.44 Summary information by country shows that Honduras and Guatemala have made positive changes on approximately one-fourth of the 39 indicators, followed by Costa Rica and El Salvador. Despite registering sizable changes, some countries’ final values remain highly unfavorable. For example, Guatemala reduced its cost of Dealing with Licenses by 117 percentage points, but the final value is still nearly ten times higher than Panama. Similarly, Nicaragua’s cost-related indicators still remain quite high. Firm-reported measures - Enterprise Surveys 3.45 Doing Business’ self-acknowledged limitations include focus on specific business forms and on each country’s most populated business center. In contrast, Enterprise Surveys deal with investment climate indicators on a larger number of issues, with samples that are representative of the whole economy because they are drawn from all sectors and business forms in each country. Entrepreneurs’ responses reflect their actual experiences, which may differ from what is prescribed by the relevant regulations. 3.46 Enterprise Surveys do not provide aggregate values for each topic, and country comparisons of the 21 specific governance-related indicators would be too cumbersome. 6 Instead, the following aggregation procedure has been devised: first, country values are compared to LAC averages; second, differences between countries are briefly reviewed; and, finally, trends are identified for the four Central American countries with surveys for 2003 and 2006. As in the preceding section, analysis highlights regional challenges in terms of commonalities and differences; only those differences that are substantively important and statistically significant are discussed. 3.47 Compared to the corresponding LAC averages, Central America in 2006 registers a number of statistically significant differences (see Table A.3.5.). On the two regulations and tax indicators, for example, Costa Rica has values below the LAC average on “senior management time spent dealing with requirements of government regulations� and “average number of meetings with tax officials.� 7 Overall, Central American countries present a more business- friendly environment on 55 of 109 comparisons, do worse on 34, and equally well on 20. Their 6 A regional perspective on entrepreneurs’ perception of major constraints for business is presented in the preceding chapter; those indicators are not repeated here. 7 The Costa Rica survey was run in 2005, but the corresponding LAC average is based on only three countries. Thus, the Costa Rica values are compared to the LAC 2006 average. 46 performance rates better on Regulations and Tax, with 10 of 12 comparisons showing favorable values. When it comes to Informality, Central America does worse on 3 of 15 comparisons. For the rest of the topics, Central America rates better than the LAC region on 42 of 82 comparisons. 3.48 The differences can be more easily visualized in Figure 3.3, which isolates six of the most illustrative indicators. They are: senior management time dealing with regulations, a Regulations and Tax (R&T) subtopic; the average of the number of days required to obtain an operating license, a construction-related permit, and an import license, a Permits and Licenses (P&L) subtopic; the percentage of firms reporting they are expected to give gifts to obtain operating and import licenses and construction permits and the percentage of firms reporting they are expected to give gifts to get connections to public services, both Corruption (C) subtopics; the percent of firms underreporting sales for tax purposes, an Informality (I) subtopic; and the average of the number of days required to clear imports and exports through customs, a combination of two Trade (T) subtopics. For all indicators lower values reflect a more business- friendly environment. Figure 3.3: Firm-reported country values and LAC average on selected and aggregated indicators 80 CRI SLV GTM HND NIC PAN LAC 70 Permits & Licenses Informality 60 48 50 46 Regulations 40 & tax Corruption 30 Trade 20 17 10 9 10 6 0 % of senior Average days to Average % Average % % that expres sed Average days to management time obtain permits expected to give expected to give a typical firm clear imports and dealing with and licens es gifts to obtain gifts to get underreports exports through government permits and public s ervices sales for tax cus toms regulation licens es connections purposes Source: Enterprise Surveys, various years 3.49 El Salvador does best in the region on the R&T subtopic, having lowest average for number of days to obtain permits and licenses at 36. The country also registers the lowest percentages both Corruption subtopics. Costa Rica has no data available on three of the six indicators, but it leads on the Trade subtopic, with the fewest average number of days to clear imports and exports. Honduras reports the lowest percentages on the Informality subtopic, firms’ perceptions of the prevalence of underreporting of sales for tax purposes. 47 3.50 Figure 3.3, of course, is not comprehensive. Broadening the field to include the two other Informality subtopics, Panama replaces El Salvador as the regional leader. Honduras ties with Costa Rica on R&T when we include the other subtopic, the average number of meetings with tax officials. 3.51 Some of the highs and lows mentioned above are a result of changes over the past few years. Summary data indicate the region saw as many improvements as deteriorations over the course of three years. Nicaragua’s improvements far outnumbered its negative changes. By contrast, negative changes outnumbered positive ones in El Salvador and Honduras. 3.52 Of the four countries, only El Salvador increased the time spent by senior management dealing with regulation and tax requirements—and then only slightly. Guatemala, Honduras, and Nicaragua decreased the time by an average of 4 percentage points. Honduras improved enough to have the region’s lowest (“best�) value on this topic. Nicaragua also improved the second R&T indicator, decreasing the number of meetings with tax officials by two-thirds. Over the same period, the LAC region on average nearly doubled the percentage of senior management time spent on regulations and tax. 3.53 On Permits and Licenses, there were nine declines and only one improvement. The average time needed to obtain a construction-related permit across the four countries increased by an average of 75 days. The LAC region registered a smaller increase—33 days. Nicaragua increased its number of days by 178, accounting for the tall column in Figure 3.3. Obtaining import licenses required an average of 19 more days in 2006 in three of the four countries; the average increase in delays for an operating license was 20. The LAC region also worsened its values, but by seven days for import licenses and 19 days for operating licenses. In sum, entrepreneurs suffered increasing delays for permits and licenses throughout the LAC area, but more so in Central America. Reports of corruption related to obtaining operating and import licenses more than tripled in Honduras, and expected gifts for construction-related permits rose as well. El Salvador and Nicaragua showed considerable increases in construction-permit corruption. It was, however, the only deterioration for Nicaragua, which otherwise achieved a substantial decrease in reports of corruption across nine of the 10 indicators. Most of them fell to less than a third of their original values. 3.54 Guatemala and Nicaragua substantially reduced their reports of “expected gifts� on some indicators, while Honduras doubled reports of “expected gifts� for electricity and phone connections. Even though it privatized telecommunications nearly a decade ago, El Salvador quadrupled reports of “expected gifts� for phone connections. There appears to be a need for monitoring mechanisms to hold privatized and non-privatized public service providers accountable. Monitoring could combine user satisfaction surveys and other indicators, identifying bottlenecks, quality and coverage issues, and keeping track of delivery gaps. 3.55 On tax-related indicators, Guatemala shows an interesting pattern. While the average number of meetings with tax officials has not changed significantly, firms’ perception of tax officials’ expecting “gifts� declined to a fourth its original value and reports of underreporting sales for tax purposes decreased 20 percentage points. It seems Guatemala’s administrative burden has not decreased—something corroborated by Doing Business indicators—but its 48 implementation has improved both in terms of officials’ ethical standards and firms’ tax-paying responsibility. 3.56 El Salvador and Honduras also registered better scores on Informality, measured by the underreporting of sales for tax purposes. All four countries with data for 2003 and 2006 had decreases averaging 50 percent in reports of corruption in government contracts. The progress may be noteworthy, but the LAC region reduced corresponding reports by 80 percent over the same period. Putting two and two together – regulatory changes and actual experiences 3.57 The two types of business-specific governance measures record different perspectives on the same investment climate; they are not, however, strictly convergent. For example, Doing Business’ measures of Trading across Borders show reductions of import and export times that are not corroborated by Enterprise Survey reports. The four countries with Enterprise Surveys in 2003 and 2006 report significant time increases for exporting and importing, the average number of days doubling in most cases. 3.58 In the Doing Business reports, Central America’s average ranking on Dealing with Licenses is poor, despite its being the second-highest target of reforms over the course of the past two years. As previously noted, reforms have focused on costs. The number of procedures did not change for any country, nor did the stipulated times. Enterprise Survey reports of the same issue, on the other hand, register a general increase in the delays for obtaining permits and licenses, as well as increases in reports of corruption. 3.59 Such crosscurrents increase the relevance of a Doing Business 2008 recommendation that rests on a 2007 study by David Hummels: Trade reformers still focus too much on cutting tariffs and not enough on cutting delays for exporters and importers. This attention is misplaced: a recent study finds that the cost of import delays exceeds tariff costs in every region, while the cost of export delays exceeds tariff costs in every region but East Asia and Western Europe. (p. 47) 3.60 In these words is an admonishment to judge reforms on results as well as intents. Even when reforms are well-meaning, their translation to actual practice may miss the mark. By itself, information on regulatory change is not enough to conclude whether there is greater or lesser ease of doing business. Appropriate monitoring and evaluation mechanisms should be developed to track the impact of reforms and ensure objectives are being met. Crime and Corruption and their influence on Investment Climate 3.61 Two areas are of great concern for the region—crime and violence and corruption. Crime and violence have long plagued Central America. Guatemala, El Salvador, and Nicaragua in particular suffered through long periods of violent conflict, which spilled over to their neighbors. In recent decades, the twin threats of organized gangs and narcotics trafficking have emerged as major drivers of crime and violence in the region, with myriad consequences for human welfare. 49 3.62 A number of theories link corruption to slower economic growth. In some cases, the problem is related to entrepreneurs choosing not to start or expand their business because of the risk of potential profits being usurped by corruption. In others, the problem is related to an inefficient allocation of investment, technology, and human capital as resources are pulled from their most productive uses. Finally, corruption may also affect entrepreneurial skills, with firms devoting more time on securing preferential treatments from government and dealing with regulatory barriers than on improving productivity (Svensson 2005, Murphy, Shleifer and Vishny, 1991, 1993). Following the approach of The World Bank (2007), we provide a brief analysis of the determinants of corruption for firms in Central American countries. Crime 8 3.63 Federico Colorado, head of El Salvador’s National Private Enterprise Association, has called crime “a stone in the shoe of development�(Latinnews.com (2005), cited in UNODC (2007)). Crime contaminates the investment climate, both deterring firms from investing and diverting resources from productive activities to crime prevention. Crime can also potentially affect workers by creating fears that may reduce productivity and increase absenteeism. 3.64 Figure 3.4 shows murder rates for 2006, collected from various sources by the Observatorio Centroamericano sobre Violencia (2007). The data illustrate the clear divide between crime levels in El Salvador, Guatemala, and Honduras and Nicaragua, Costa Rica, and Panama. The homicide levels in the three higher-crime countries are among the world’s highest, while levels in the other three countries are close to the world average. 3.65 Many businesses in Central America’s higher-crime countries are subjected to extortion. Gangs are believed to be responsible for much of it. According to Figure 3.4: Homicide Rates in Central America, 2006 press reports, the Chamber of Commerce Homicide Rates in Central America, 2006 and Industry in San Miguel, El Salvador, claimed in 2006 that gangs preyed on 80 SLV 55 percent of its membership, and that some GT M 45 HND 43 businesses paid as much as US$1,200 to NIC 12 $US1,400 a month in protection money. PAN 11 Transport unions in the same city CRI 8 reported that gangs demanded a “tax� of 0 10 20 30 40 50 60 $US3 to $US5 a day from minibus taxi Homicides p er 100,000 Pop ulation drivers (Latinnews.com (2006) and Inter Press Service, 2007). Note: The World Health Organization (2002) estimated that in 2000 there were 8.8 murders per 100,000 population 3.66 When people refuse to make worldwide. This figure does not include war-related deaths, extortion payments, they risk becoming which were 5.2 per 100,000 population worldwide). Source: Observatorio Centroamericano sobre Violencia victims of violence. Media reports describe Guatemalans closing small businesses after being subjected to extortion demands of US$394 per month (Agence France-Presse, 2007). In El Salvador, at least 70 drivers were killed for refusing to pay protection money. In January 2007, another six people were killed and 25 8 This section is adapted from Demombynes and Hincapie (2008) 50 buses were set on fire. According to unofficial police statistics reported in the press, incidents of extortion increased dramatically in 2006 in El Salvador. There were 2,485 complaints filed with the police that year, compared with 493 in 2005 and 290 in 2003. 3.67 Apart from the suffering crime against business causes in the form of violence and lost property, a high-crime environment reduces business productivity. A World Bank (2007) analysis, for example, shows that high crime rates greatly reduce both firm sales and total factor productivity. 3.68 The 2006 Enterprise Surveys collected data on security costs and direct losses due to crime. For five Central American countries (Costa Rica was not surveyed in 2006), the total of security costs and losses averaged 3.7 percent of firm sales, significantly higher than the 2.8 percent for the LAC region as a whole. The combined costs of crime were 4.5 percent of firm sales in El Salvador and Honduras and 3.9 percent in Guatemala (Figure 3.5). 3.69 More detailed analysis presented in Demombynes and Hincapie (2008) shows that crime is not confined to businesses in certain sectors or those with particular characteristics. The costs of crime as a percentage of sales are lowest in manufacturing firms, but Figure 3.5: Firms Security Costs and Losses Due to Crime as a Percentage of Firm Sales, 2006 they are still high, averaging 2.9 percent of sales for the five Central 5 American countries. Multivariate 4.5% 3.9% 4.5% Security Costs 3.7% analysis shows that larger firms are 4 Losses % of Firm Sales 3.1% 2.8% more likely to be victimized by crime, 3 2.5% but average losses as a percentage of 2 sales are much higher for small firms. 9 1 Small firms, which are often owned and operated by less well-off 0 SLV GTM HND NIC PAN 5 CA LAC individuals, suffer most from crime. 3.70 Controlling for regression covariates, firms that export are 29 Source: Authors’ analysis of 2006/7 Enterprise Surveys percent more likely to face some security costs. 10 They are, however, less likely to experience losses from crime, possibly due to their greater protective measures. Average losses due to crime are no higher for exporting firms than for companies overall. 3.71 The threat of crime can also diminish business productivity through its effects on workers. Enterprise Surveys asked whether firms had experienced increased worker absenteeism as a result of crime and insecurity in the previous year—for example, workers unwilling to report 9 Demombynes and Hincapie (2008) used multivariate regression analysis to examine the correlates of security costs and losses from crime. They used probit analysis to examine correlates of binary outcomes: having security costs or losses from crime. Ordinary least squares regressions were used to examine security costs and losses from crime as a percentage of sales. Dependent variables employed were a binary variable for whether the firm exports, the log of annual sales, sector of firm activity, dummy variables for firm size (small, medium or large), and dummy variables for each country. 10 The greater risk of security costs for exporting firms is even higher if firm size is controlled for. 51 for the night shift, or employees unable to work because they had been crime victims. For the average LAC country, only one in ten firms responded affirmatively, compared to 14 percent in Central America. Large increases in absenteeism were also reported in El Salvador (20 percent), Guatemala (16 percent), and Honduras (23 percent). 3.72 Addressing the very high crime levels of the region’s three high-crime countries has become a primary focus of public policy. A consensus is emerging in government and civil society that a narrow criminal justice approach is not effective. Policies should aim to improve the police and court system, but they should be supplemented by crime prevention efforts, ranging from anti-violence education programs to crime prevention through environmental design. Recent policy has focused on community-based approaches. Focusing resources on criminal prevention may be an advisable strategy for Nicaragua, Panama, and Costa Rica, which still have relatively low crime rates but have seen increases in recent years. Corruption 3.73 Corruption is a big Figure 3.6: Bribe payments in Central American countries concern for Central American firms, but it needs to be addressed 30 together with other problems, 24.6 such as regulatory compliance 25 and the quality of the judiciary 18.5 % firms 20 system. The proportion of firms 14.9 15.0 15.0 paying bribes to “get things done� 15 10.8 or to obtain public services ranges 10 from 10.8 percent in Panama to 5 24.6 percent in Honduras, according to Enterprise Surveys 0 (Figure 3.6). In particular, bribery s r a a a a a do gu m ic reports are significantly higher for al ur R na m va r d ica ta te on Pa al medium-sized firms (20-99 os ua N lS H C G E employees) than for small and large ones, and a similar gap Notes: Bribe payments are measured as a variable that equals one if (1) when the interviewer asks “When establishments like this one do business with the exists between manufacturing government, what percent of the contract value would be typically paid in firms and services. additional or informal payments or gifts to secure the contract?� the firm’s manager answers with a percentage greater than zero, or (2) if the firm answers 3.74 Bribe payments reflects affirmatively to any of the questions “Was an informal gift or payment expected or requested to obtain the [service]?� in reference to a telephone, a wide range of factors and water, or electricity connection; an import, or operating license; or a circumstances that contribute to construction permit. Bribe equals zero if the manager answers “zero� to the an environment of corruption, not first question and “no� to all other questions (public services). Source: Enterprise Surveys 2005/7 a policy variable that can be changed at will. If the policy objective is attacking corruption, it is extremely important to identify its roots, so that effective interventions can be designed 3.75 Some studies point to inadequate institutional capacity and quality and policy distortions as causes for corruption. For example, the 1997 World Development Report found that, after controlling for other factors, the predictability of the judiciary appears to lower the 52 level of corruption. A similar relationship between corruption and the independence of the judicial system is presented by Ades and Di Tella (1996). Table 3.3 shows that corruption is negatively associated both with regulatory compliance (measured as the percentage of sales that firms report for tax purposes) and courts’ enforcement powers. Table 3.3: Determinants of Bribes in Central American countries Bribes given for procurement or public services (0 if not, 1 if yes) Explanatory variables (1) (2) Regulatory compliance (proportion of sales reported -0.001* -0.001* for taxes) [1.81] [1.7] Court enforcement (perception of strength of -0.039* enforcement) [1.67] Observations 1,496 1,496 Regulatory compliance is measured as the proportion of sales reported for tax purposes. Court’s enforcement power is measured with a variable that equals one if the manager agrees (at least partially) with the statement “I am confident that the judicial system will enforce my contractual and property rights in business disputes.� Controls included: capital per worker, foreign owned, exporter, firm's age and country, sector, and firm size dummies. Robust z statistics in brackets. * significant at 10%; Source: Author’s calculations using Enterprise Surveys 2005/7. 3.76 These findings are in line with previous studies (World Bank 2007, Chapter 4) and provide evidence of two hypotheses on the mechanisms behind corruption. One, known as the “control rights hypothesis,� states that regulations give public officials extensive discretionary power to extract bribes from firms. The other, known as the “grease the wheels hypotheses,� states that firms have incentives to bribe officials to avoid regulations or secure contracts. 3.77 From a policy perspective, the results suggest that a good regulatory framework should provide minimal discretion to responsible public servants and use information and communication technologies to cut users’ costs, increase transparency, and reduce opportunities for corruption. In this context, administrative simplification, together with efforts to professionalize the civil service, could lower the cost and improve the efficiency of regulations, reducing the incidence of corruption. 3.78 As the World Bank (2007) points out, competition and openness to trade deserve some attention. To the extent that competition lowers the rents of economic activities, it might also reduce the motive of public officials to capture a share of these rents by means of extortion and corruption. In this regard, Ades and Di Tella (1995, 1997) use openness to trade (measured by imports to GDP) as an indicator of competition, finding that it is negatively related to corruption in a cross-section of 55 countries. Similar results are found by Brunetti and Weder (1998) in a 122-country study. Treisman (1999) and Leite and Weidmann (1999) use the Sachs and Warner index to measure the number of years a country has been open to trade, finding it negatively correlated with the level of corruption. 53 3.79 In sum, this section highlights the need for a combined approach to reform. It identifies specific topics which are of concern to the whole region and some which show wide gaps between countries. On the first, it behooves all Central American countries to move forward, on the latter, improvement of standards in those countries that are behind would improve the overall regional status. However, if countries do not also look to longer-term institution-building needs, exclusive focus on business-related governance measures might fail to generate recommendations for sustainable growth. In other words, while relaxation of specific constraints on private economic activity may trigger a growth spurt, sustaining such growth requires solid institutions. Coming out of this analysis, the main regional challenge is the strengthening and modernization of rule of law and enforcement-related systems and institutions and it would best be met by region-wide interventions. This would not only include the necessary strengthening of formal enforcement, investigation, prosecution and judging mechanisms; but also the need to intervene in setting the right incentives as well as implementing effective preventive actions to reduce vulnerabilities to informality, crime and corruption, as well as to strengthen capacities to mitigate risks of state-capture practices over these institutions. A key step is the enhancement of transparency through public information access legislation and mechanisms. 4. CUSTOMS REGULATIONS -- A KEY ELEMENT OF THE DR-CAFTA AGENDA 3.80 Chapter 4 will discuss the importance of infrastructure, especially transportation, for greater trade in the region. While better roads and ports are necessary for moving goods from one place to another, firms will find it difficult to benefit from new trade opportunities without a well functioning customs system. 3.81 Tariff revenue has accounted for about 3 percent of the region’s GDP in recent years, but customs regimes’ importance and impact goes far beyond tax collections. Reform can contribute to improved governance in Central America, where excessive red tape, rampant fraud, and corruption within customs agencies keep trade costs high. Customs reform will also help improve competitiveness and encourage public-private sector partnerships. Availability of customs information to the public will enhance transparency and improve perceptions of the Central American customs departments. Governance measures should also benefit from sustained reform, both through increased transparency of the customs processes and strengthening of associated institutions. 3.82 Improving import processes and installing efficient export regimes have become increasingly important in light of expanding trade and the demands of DR-CAFTA, the WTO, international rules imposed by the EU and other entities, and security compliance. Expertise is required in the technologies and practices that help improve the efficiency of processes by moving away from forms and paperwork to the paperless processing associated with automated risk management. 11 Success relies on comprehensive reform that stretches from issue diagnosis to impact assessment. Multiple stakeholders need to be involved—cargo handlers, warehouse operators, banks, the private sector, and the police. 3.83 The Logistics Performance Index (LPI) provides a sense of how Central American countries stack up against each other. The LPI is an interactive benchmarking tool that consists 11 IFC, 2007 http://rru.worldbank.org/Documents/Toolkits/ReformBookFinal.pdf 54 of both perceptions and objective measures and helps build profiles of countries’ logistics friendliness. It measures performance along a country’s logistics supply chain in three parts: perceptions of trading partners, information from the home country, and real time-cost performance data. The LPI Index is highest for Panama and El Salvador, both of which are above the LAC average. They are followed by Costa Rica, Guatemala, Honduras, and Nicaragua (Table 3.4). Table 3.4: Logistic Performance in Central America Panama El Salvador Costa Rica Guatemala Honduras Nicargua LPI (overall index) 2.89 2.66 2.55 2.53 2.50 2.21 Customs 2.68 2.38 2.49 2.27 2.48 2.14 Infrastructure 2.79 2.42 2.43 2.13 2.32 1.86 International shipments 2.80 2.78 2.53 2.62 2.48 2.18 Logistics competence 2.73 2.53 2.43 2.50 2.41 2.41 Tracking & tracing 2.93 2.82 2.57 2.43 2.41 2.19 Domestic logistics costs 3.21 2.94 3.08 3.00 2.88 3.04 Timeliness 3.43 3.06 2.89 3.23 2.88 2.50 Note: Based on a 5-point scale (the higher the better). Source: Logistic Performance Index, 2007 3.84 According to Doing Business, some countries introduced reforms in the trade area between 2006 and 2008. Guatemala leads with two; Honduras is the only country without any. The four other countries made one reform each between 2006 and 2008. Although Guatemala made the largest improvements during the period, Doing Business reports that it increased the number of documents needed for export and import. Panama and Honduras did not make changes in any of the six indicators. (Box 3.2 provides a brief summary of custom reforms in the region.) 3.85 Looked at broadly, customs reform has a mixed record. The legal groundwork for customs and trade-related reforms is in the process of being established. Commitment has been received from all countries, signifying full support for this legal framework, but it is not yet fully operational. It is most important that the legal and regulatory framework provide (a) an adequate and coherent authority structure for the essential trade-related institutions and (b) clearly stated regulations and procedures for implementing these authorities that strike a balance between facilitation and necessary control. 3.86 The latest version of the Central American Uniform Customs Code (Código Aduanero Uniforme de Centroamérica, or CAUCA III) and its implementing regulations (RECAUCA III) are in force in Central America. DR-CAFTA members have adopted CAUCA and its implementing regulations, and this provides a good legal framework for customs authorities. CAUCA and RECAUCA provide a legal foundation for efficient processing of goods. CAUCA requires that all declarations be electronically transmitted. The electronic transmission can be the legal declaration. Procedures are standardized, written and available on the customs Web site. 55 The procedures are in general conformity with Kyoto Convention guidelines. Although CAUCA III, as an international agreement, should take precedence over domestic laws, it appears that some countries are still in the process of fully implementing many provisions of the code. Box 3.2: Status of customs reforms by country Costa Rica: Customs modernization has included improvements in the areas of in-works clearance, advance ruling, single window (SIVUCE, which handles 69 percent of the exported goods), customs information systems (CIS), customs transit control, and risk assessment. Customs clearance time went from six days in 1994 to 112 minutes in 2004, with physical inspections. The average time to obtain custom clearance for imports was only an hour. All customs units had computerized systems; the proportion of professionals in total custom staff was 21 percent. The collection rate per custom employee was $US968,000. Work is still needed in the area of human resource reform and integration with other areas, such as tax authorities. Guatemala: The country has made the region’s most rapid progress on customs reforms in recent years. Since 2001, some customs units have been equipped with computers, including an on-line system allowing a close monitoring of custom operations and “workflow-type� management of procedures. A training program has also being implemented. Local banks collect customs fees and taxes directly. The system still has a ways to go. According to Romero, E. and J. Barahona, only about a quarter of the customs units had computerized systems, and the proportion of professionals in total custom staff was 5 percent. The average time to obtain custom clearance (for imports) was 22 hours. The collection rate per custom employee was also low at $US600,000. Honduras: In 2000, Honduras implemented the WTO Customs Valuation Agreement (CVA), which stipulates the use invoice value (the price actually paid for the goods). The same year, Honduras and Nicaragua joined the Customs Union formed in 1996 by Guatemala and El Salvador. To facilitate customs' processing, El Salvador and Guatemala established satellite offices at the Honduran port of Puerto Cortés and the El Amatillo border crossing between Honduras and El Salvador. Honduras still has work to do in the areas of personnel recruitment, training, and equipment. Most facilities need refurbishing, and additional management and technical training would benefit the customs function. Nicaragua: Customs reform has been limited. Nicaragua is behind schedule on implementing the WTO Customs Valuation Agreement Foreign investors have complained about arbitrary customs procedures and valuations. Tariffs and import taxes for most used goods are not assessed on a CIF/bill of lading basis but rather on "reference prices" determined by customs at the time of inspection. This reference price can be significantly higher than the actual amount importers pay. The presentation of a bill of sale (or other evidence of purchase price) that is certified by a Nicaraguan consular official is often, but not always, accepted by inspectors as proof of used goods’ value. The door is still open for a wider, integrated reform effort, especially in the area of human resources/civil service reform, modernization of customs systems, and process improvements. El Salvador: The country has taken steps to simplify customs procedures and has been implementing the WTO Customs Valuation Agreement since March 2002. It has put in place the Central American Uniform Customs Code (CAUCA III) procedures, which modernize customs by implementing uniform documents, allowing electronic transmission of customs information, and permitting electronic prepayment of charges, tariffs, and taxes. The Central American countries have agreed to apply a single manual for customs procedures. Its application began during the first quarter of 2003 as a pilot project involving El Salvador and Guatemala. El Salvador has also implemented the "Teledespacho" system that allows goods to be presented without a customs broker. The importer/exporter is electronically linked to the Central Customs Service and can present and process all his documents from his office. Panama: The U.S.-Panama Free Trade Agreement, if implemented, will result in improved customs procedures and rules of origin, potentially resulting in enhanced transparency, heightened predictability, greater accountability, improved procedures for express delivery shipments, greater customs efficiency through information technology, and customs cooperation and information sharing (at least between the U.S. and Panama). However there are policy gaps remaining in the area of customs reform in Panama, mainly in the area of further integration and human resources By Enrique Fanta and Tanya Gupta 56 3.87 RECAUCA regulations adequately cover the brokerage community, requiring either a written test or a university degree in customs matters to qualify for a license. The law includes a process for administrative and criminal penalties. CAUCA provides that the importer has a right to both administrative and judicial review of customs decisions. CAUCA and RECAUCA have adequate authority for expedited processing of express shipments, and a process allows importers to request tariff classification rulings that are binding. On December 12, 2007, Central American governments subscribed to the Framework Agreement Establishing the Central American Customs Union, which sets forth the legal provisions necessary to achieve the customs union. Currently, it is in the process of ratification and approval by the respective legislatures. The Single Manual on Customs Procedures, approved in 2004, is being implemented at the integrated and peripheral customs with the objective of standardized procedures. It will be reviewed to make sure it meets countries’ needs. 12 3.88 On the critical issue of customs staffing and professionalism, Central America still has significant work to do. A good civil service career is difficult, if not impossible, in most Central American countries’ customs offices. Most Customs agencies do not have a professionalized workforce and operate in a politicized environment. Officers are often assigned for the duration of an administration, suggesting no apparent career path. They often do not have sufficient training or the customs expertise to work in today’s complex environment. Customs personnel practices need to be reviewed to provide a career path that creates incentives for personnel development. The perception that the customs function is highly corrupt serves as a disincentive for qualified personnel. Guatemala has undertaken some efforts in civil service reform and human resource development, but most other countries lack qualified personnel management systems. Both the World Bank and Inter-American Development Bank have projects that cover this area, and this work needs to be further developed and coordinated across institutions. We suggest including the private sector while working on customs-related civil service reform; the private sector is an active stakeholder in the customs modernization process. As an example, private sector personnel in airports and port authorities are often quite weak. 3.89 The flow of Figure 3.7: Tons of cocaine seized in Central American countries (rounded) illicit goods into and through Central America remains a problem. Stringent physical checks in Central America are meant to control contraband and smuggling, and this is very time-consuming. It is important to determine the extent of the problem and incorporate risk- analysis techniques to determine optimal times for the checks. However, Source: UNODC 12 http://www.sieca.org.gt/site/VisorDocs.aspx?IDDOC=CacheING/17990000002611/17990000002611.swf 57 not enough information is available. The perception is that contraband and smuggling is a significant problem. 13 A United Nations Study on Crime in Central America found that South America produces an estimated 900 tons of cocaine, shipping most of it to 10 million users in the US and Europe, a market worth some US$60 billion in 2003. In the past, volumes and values were much greater. Most of these drugs transit either Central America or the Caribbean, or both, by land, sea, or air, and a vigilant customs regime can play a strong deterrent role. Drug smuggling was an important source of income for both insurgent and state groups during the civil wars, and enduring corruption combined with limited law-enforcement capacity have ensured that well-worn smuggling channels remain viable. 3.90 Better coordination within the nation and region can improve the performance of customs regimes. Integration needs to take place at the national level between customs agencies and other arms of government, such as tax administration. It also needs to take place between cross-border agencies and between countries—so an import declaration in one country automatically becomes an export declaration in another country. More work is needed on developing a set of integrated procedures at the border. Today, the region is very far away from a “one-stop shop� or a single window. There even are “overlapped customs� in some border crossings. The information of a common clearance of goods is not automatic, and no interchange occurs with agencies other than customs. 3.91 Like integration, technology can improve customs operations. Central America countries are at different levels in terms of customs information systems. Some have in-house customs systems, whereas others have large-scale customs information systems, such as ASYCUDA (developed by UNCTAD). The biggest challenge, though, is that these systems are, for the most part, not integrated with tax administration systems, even when customs and tax functions are housed in the same unit (such as in El Salvador). Guatemala has the most integrated systems, largely because of the Integrated Financial Management System, a World Bank project. It is vital that customs reforms in the area of information systems, simplification, and process improvement be linked in with the governance agenda. Most of the elements of such a reform can contribute to improved transparency and reduced corruption. 3.92 The private sector is a key stakeholder for the customs modernization process. Central America’s governments, for the most part, do not have a medium- or long-term strategy for working with enterprises, industry groups, and others. This is a new area and needs more efforts from governments because customs reform cannot succeed without the private sector’s active support and cooperation. Other regions may provide interesting ideas. The APEC Sub- Committee on Customs Procedures (SCCP), for example, is seeking SMEs for its continued customs-related trade facilitation initiatives, providing them the necessary leverage in their home countries to open a dialogue on SMEs needs, raise the priority of their work, and secure the political will to effect the desired changes. The private sector should also be used to improve the quality of human resources in customs administration. The DR-CAFTA report notes that public provision of training has been characteristically inefficient and unresponsive to private sector needs across Central America. 13 http://www.unodc.org/pdf/Central%20America%20Study.pdf 58 3.93 Many countries do not have adequate controls on imports and exports or special regimes, such as maquilas. Today, a maquila refers to a company operating under a special customs regime, which allows for temporary import on a duty-free basis of machinery, equipment, materials, parts and components, and other items needed for assembly or manufacture of finished goods for subsequent export. Free zones are also weak on controls. Special emphasis is need on inventory control. Internal and international transit control is weak in most Central American countries, creating a need for integrated transit control systems. 14 3.94 Central American countries have made advances in customs reform in recent years, most notably in the area of legal and regulatory framework. However, much work remains. Further customs reform can contribute to transparency and improved governance and combat fraud and corruption within customs agencies. Customs reform will also contribute to competitiveness, particularly in trade, and encourage public-private sector partnerships. Ongoing reform efforts should be deepened, with particular focus on the areas just mentioned—the staffing and professionalism, legal and regulatory framework, contraband/smuggling data, systems development, harmonization and integration, public/private partnerships, and improvements in the control structures. 14 http://www-personal.umich.edu/~deanyang/papers/yang_psi.pdf 59 Chapter 4. INFRASTRUCTURE 1 INTRODUCTION 4.1 Economic or physical infrastructure, defined mainly as transport, telecommunications, power and water, plays a fundamental role in supporting economic growth and facilitating a country’s ability to compete. Transport and telecommunications are necessary to link producers with input and product markets. Power and water are essential inputs into the production process for most activities, whether industrial, agricultural, or service industries. 4.2 The adequate supply and efficient operation of infrastructure promotes the production and distribution of goods by removing bottlenecks to economic activities and lowering firms’ production costs. By affecting both production possibilities and costs, the availability of efficient infrastructure exerts an important effect on firms’ and on countries’ competitiveness in international markets. Many of these infrastructure services also directly contribute to employment and improving living standards. For these reasons, productive infrastructure has been positively linked with economic growth and competitiveness in the findings of several World Bank analytical reports (Calderón and Servén, 2003; Guasch, 2005; Loayza, Fajnzylber, and Calderón, 2005; among others). 4.3 Poor infrastructure can have debilitating effects. Without additional infrastructure development in Central America, serious bottlenecks would persist, hindering the region’s prospects for growth with macroeconomic stability and making it difficult or nearly impossible to compete in a globalized world. In most Central American countries, direct trade barriers, such as duties and quotas, appear to be less constraining than physical bottlenecks in the production and movement of goods. 4.4 Across Latin America, the importance of infrastructure on firm’s competitiveness has been confirmed in survey after survey. More than half the firms questioned in World Bank Enterprise Surveys (ES) considered infrastructure to be a major obstacle to their businesses’ operations and growth—a level that is among the highest in the world. ESs probably tend to understate the importance of infrastructure because they only capture the opinions of firms that chose to locate in a country, despite its infrastructure constraints. Left out are the probably larger number of firms that simply decided not to locate in the Central American countries in the first place due to infrastructure inadequacies and the difficulties of establishing reliable connections with key markets. 4.5 This chapter summarizes some of the key findings on physical infrastructure in Central America and why it is important for competitiveness, looking in particular at transportation and energy. For Central America, these infrastructure sectors are of prime importance: In order to unlock the benefits of trade, infrastructure must be available, reliable, and cost effective and a wide range of firms must be provided with reliable access and efficient connectivity as well as adequate and reliable energy. The chapter begins with an overview of infrastructure’s importance for growth and competitiveness, emphasizing the Latin and Central American context. The 1 This chapter was prepared by Emmanuel James (Transport), Fernando Lecaros and Patricia Rodriguez (Energy). 60 second section deals with transportation, particularly roads and ports. The third section shifts the focus to energy, with an emphasis on electricity. The final section looks at the challenges Central America faces in improving its infrastructure and some strategies to make improvements. 4.6 The chapter’s key policy recommendations include: • Promote the efficient use of infrastructure through (1) appropriate policies to manage demand for services, (2) effective institutional organization and regulatory policies, and (3) sufficient attention to operations and maintenance. • Institute efficient pricing and subsidy policies that (1) require public utilities to cover operating expenses, interest, and depreciation and self-finance a part of their investment program; (2) target the needs of an area on a holistic basis, especially for poor urban and rural sub-areas, and (3) streamline mechanisms for the periodic adjustment of infrastructure prices. • Improve operations and maintenance by fully establishing road-maintenance funds and using private contractors that are awarded contracts via a public bid process. • Set priorities for infrastructure investment, for example in the road sector by looking at (1) overall investment requirements for meeting growth and competitiveness goals; (2) implementation capacity constraints; (3) potential role of the private sector, and (4) availability of financing. • Foster private sector participation in ways that (1) reduce the financing requirement for the public sector, (2) reduce investment and implementation constraints; (3) improve competition and contestability of markets, and (4) contribute to overall efficiency. • Improve administrative efficiency of utilities, for example by implementing enhanced customer and load management procedures, such as reducing commercial losses through improved metering, and increasing collection rates, to reduce crippling financial losses • Promote regional interconnections that expand the role of regional institutions in developing and putting in place a framework that would bolster private investor confidence and eases constraints on each Central American country’s public infrastructure investment budget. 1. INFRASTRUCTURE OVERVIEW 4.7 In covering the launch of the Dominican Republic-Central America Free Trade Agreement (DR-CAFTA), the mainstream media focus mostly on its provisions for lowering tariffs and fostering trade and economic competition. However, a less noticed but more substantive benefit to producers may be the long-term reduction of uncertainties regarding the import and export regimes within the region and the US. Predictable trade rules could translate directly into a firm’s objective function for determining location and production decisions, such as factories and logistical hubs. This aspect of free trade will be important to Central America because the presence of supporting infrastructure—or its absence—will be a key factor in determining a country’s competitiveness and ability to attract and retain productive investments. 4.8 With trade liberalization spreading, transport and logistics have become more relevant for Central American countries in a strategic sense. Firms need to move inputs from suppliers to their factories, and they also need to move their outputs to distributors and customers. 61 Throughout the production and distribution process, firms need to rely on physical infrastructure for their transport (e.g., roads and ports), their transport-related services (e.g., logistic operators, trucking industry), their trade facilitation needs (e.g., customs, border agencies, tariffs), and their inventory processes. 4.9 DR-CAFTA may well foster increased trade, but trade alone does not account for all growth. More targeted analysis shows that the ease of shifting capital and labor from low- productivity firms to higher-productivity ones and the underlying conditions of firm productivity have at least as much impact on growth. Available and efficient infrastructure is a pre-condition to firm productivity. An analysis of the factors that were important for Central America’s growth from the 1980s to the 1990s suggests that infrastructure has had about as much impact on growth as increasing trade openness (Table 4.1). Trade and infrastructure also complement each other in their joint impact on growth. Efficient infrastructure helps trade to boost growth while increasing trade leads to greater demand for infrastructure, resulting in greater investment, lower per unit costs, and higher firm productivity. Table 4.1: Analysis of Changes in Growth Costa Rica El Salvador Guatemala Honduras Nicaragua Structural determinants Education 0.15 0.42 0.46 -0.10 0.48 Financial depth -0.10 0.13 -0.06 -0.16 0.41 Trade openness 0.41 0.37 0.21 -0.07 0.30 Government burden 0.26 0.65 0.43 0.44 1.00 Infrastructure 0.35 0.63 0.41 0.60 0.37 Other 0.05 -0.11 -1.45 0.10 -0.72 Change in growth 1990s vs. 1980s Predicted 1.13 2.09 2.44 0.82 1.84 Actual 3.80 4.14 3.05 0.84 4.40 Source: Jaramillo and Lederman (2006) as adapted from Loayza, Fajnzylber and Calderón (2002) 4.10 A wide range of empirical studies supports the conclusion that infrastructure is, in its own right, a major contributor to economic growth, particularly for developing countries. The approach to understanding the linkage has varied, with different analyses considering: • The impact of infrastructure on aggregate total factor productivity (Krugman, 2004); • The impact of individual sectors on growth (Roller and Waverman (2001) for telecommunications or Fernald (1999) for roads); • The region-specific impact of infrastructure stocks across sectors (Calderón and Servén, 2003); and • The separate impact of service quality on growth as opposed to just infrastructure endowment (Esfahani and Ramírez, 2003; and Calderón and Servén, 2004). 4.11 The cumulative result of this research is a robust demonstration of infrastructure’s role as a driver of growth. Underlying the “direct linkage literature� (infrastructure and growth) is the 62 recognition that infrastructure is an important determinant of firm productivity. That is, the supply, quality, and price of infrastructure are defining elements of firms’ cost structure and competitiveness. In an era of greater trade openness, logistics and transport costs alone tend to be higher than duties imposed on imports as well as the cost of quotas and other non-tariff barriers. 4.12 Sound trade policies can contribute to growth; poor infrastructure can have the opposite effect. In most Central America countries, direct barriers to trade, such as duties and quotas, appear to be less constraining than physical bottlenecks in the production and movement of goods. In Costa Rica, for example, 52 percent of firms surveyed for the ES regarded infrastructure as a major or very severe constraint to their investment climate, while only 7 percent said trade regulations were a major or very severe constraint. When asked in more detail about the infrastructure constraints, 42 percent found transportation to be a major constraint, while 28 percent identified electricity and telecommunications services. Such responses are common in the Central American countries. 4.13 Across Latin America, infrastructure’s importance to firm competitiveness has been confirmed in survey after survey. Beyond perceptions of managers, detailed cost results from the ES suggest that competitiveness is eroded by poor infrastructure. Indeed, an in-depth ES cross- country analysis confirms that infrastructure is a major determinant of total factor productivity (TFP). The cumulated effect of infrastructure-related variables on TFP adds up to about 55 percent for the countries in the study. 2 Infrastructure variables with the highest impact on average productivity include poor electricity and transport services. 4.14 Perhaps more important, the analysis also found that infrastructure in Latin America has affected firms’ integration into global markets. Poor infrastructure affects the capacity of firms to export, as well as the ability of countries to attract foreign investments. It thus reduces opportunities for greater international integration, higher competitiveness, and enhanced technology and innovation. Because of the importance of infrastructure in the physical movement of goods, bottlenecks contribute to high logistics costs which, in turn, lead to rising inventory levels in Latin America and the Caribbean. Average logistics costs through the region range from a low of 15 percent of product value in countries such as Chile to a high of 34 percent in Peru. Unreliable infrastructure will result in higher losses in transit, the need to hold larger inventories rather than ordering just-in-time, and generally higher transport costs. 4.15 The Logistics Survey conducted for Costa Rica focused on three high value-added goods and found costs ranged from 13 to 15 percent of value. This suggests the national average—which would include the movement of such low-value goods as cement, the primary import into Puerto Caldera—is considerably higher. 3 It is worthwhile to note that the average share of logistics costs to product value in OECD countries is around 10 percent. 4 This average includes the full range of high-value and low-value shipments, suggesting that Costa Rican firms are faced with a significant cost disadvantage when competing against firms in industrialized countries. Much of this extra logistics cost burden on Costa Rican products can be attributed to deficiencies in infrastructure quality and reliability—particular in transport, given the poor 2 Escribano and Guasch (2005) 3 Transport costs alone often represent more than 50 percent of the delivered cost of low-value goods such as cement, coal, coke, grains and other products shipped in bulk. 4 See Guasch, The World Bank (2002) 63 condition of, and the many bottlenecks along, the main road between San Jose and Limon, the country’s largest port. 4.16 Poor quality and reliability result in damaged goods, demurrage charges, lost sales, and higher inventory levels. While US businesses typically hold inventories of around 15 percent of GDP, inventories in Latin America and other developing regions are often twice that (Guasch 2004). Such levels are costly to maintain, principally because they tie up capital, a high cost asset in most of LAC. This significantly increases unit costs, diminishing competitiveness and productivity. Guasch estimates that, assuming an interest rate for financing holdings of 15 percent to 20 percent, the cost to an economy of additional inventory holdings is more than 2 percent of GDP. 4.17 For Central America as a whole, the issue of infrastructure is critically important: To unlock the benefits of trade, infrastructure must be available, reliable and cost effective, and a wide range of firms should have efficient connectivity. 2. TRANSPORT SECTOR 4.18 In the context of DR-CAFTA implementation, Central American firms are conscious of the relevance of transport issues for growth. In 2007, the ES found that about 20 percent of Guatemala’s surveyed manufacturing firms perceive transport as a major or severe constraint to growth (Figure 4.1). In the LAC region, Guatemalan firms rank among the top five countries where transport is considered a major or severe constraint. 4.19 Large firms are more likely to perceive transport as an obstacle to growth than small and micro ones. Compared to 2003, Guatemalan firms, as well as their counterparts in Central America, perceived that transport is becoming a bigger barrier to growth.. It is important to note that firms’ perception may not coincide always with actual changes—for good or bad— in the transport sector. 4.20 In addition to the impact on firm competitiveness, Central America’s infrastructure may also affect its ability to attract foreign investment. Guatemala, Honduras, and Nicaragua ranked below average in overall infrastructure quality in the survey of business executives contained in the World Economic Forum’s Global Competitiveness Report (Table 4.2). In regards to ports, Guatemala shows some improvements, but it ranks below El Salvador and Honduras; however, Costa Rica declined. In the area of air transport, Guatemala, Costa Rica, and Honduras scores were below average. However, Guatemala showed marked improvements in recent years, partly attributed to its upgraded La Aurora airport and improved safety standards. In air transport, Costa Rica again seems to show a marginal decline. 4.21 Central America’s governments have long recognized the importance of infrastructure development for promoting and sustaining economic growth and competitiveness. They have included projects in successive national development plans. After the “lost decade of the 1980s,� significant progress was made in all sectors during the 1990s, but the low starting levels mean Central America’s infrastructure endowment still sits at the low end of the world scale. Since the 1990s, the most heavily trafficked trunk roads have been substantially rehabilitated, especially in Costa Rica. 64 Figure 4.1: Perceptions of Transport as a Major or Severe Constraint to Growth Central America, 2003 & 2007 25 20 15 2003 2007 10 5 0 El Salvador Guatemala Honduras Nicaragua Latin America, 2006 (*) 30 25 Percent of firms 20 15 10 5 0 Colombia Bolivia Brazil El Salvador Guatemala Chile Venezuela Mexico Costa Rica Ecuador Panama Nicaragua Guyana Argentina Paraguay Peru Uruguay Honduras Notes: (*) The latest survey available was used- Costa Rica (2005), rest of CA countries, 2006/7. Source: Enterprise Surveys 4.22 Some very good projects have also been carried out to meet some of the needs for ports and railways. These included projects to modernize and expand infrastructure—for example, the La Union port in El Salvador as well as to meet post 9/11 safety requirements such as at Puerto 65 Cortez in Honduras. In Panama, the old trans-isthmus railroad was concessioned, refurbished and reopened and now provides a very good connection to Colon. In addition, a regional initiative called Plan Puebla Panama was launched in 2001, see Box 4.1 for a brief description of the PPP. Table 4.2: Comparative Survey on the Quality of Infrastructure Overall infrastructure Road infrastructure Port infrastructure Air transport quality quality quality infrastructure quality 2006 2004 2006 2006 2004 2006 2004 Guatemala 3.5 2.7 3.7 3.5 2.5 4.4 3.5 Nicaragua 2.5 2.2 2.4 2.3 2.0 3.8 3.7 Costa Rica 4.0 3.0 2.1 2.3 2.5 4.3 4.8 El Salvador 3.8 4.4 4.9 3.6 3.3 5.7 5.6 Honduras 3.0 3.0 3.0 4.3 3.8 4.0 3.3 Panama 3.8 4.0 4.0 5.7 5.7 5.2 5.2 Argentina 3.0 3.6 3.3 3.4 3.6 3.8 4.1 Brazil 3.3 3.5 2.4 2.7 3.1 4.6 5.1 Chile 4.9 4.9 5.3 4.9 4.8 5.6 5.7 Colombia 3.7 2.9 2.6 2.9 3.0 4.9 4.4 Mexico 3.7 3.4 3.8 3.4 3.3 4.7 5.0 Average 3.7 3.5 3.5 3.7 3.6 4.7 4.7 Notes: Survey based subjective evaluation on scale from 1-“poorly developed and inefficient� to 7-“among the best in the world.� Source: World Economic Forum, Global Competitiveness Report 2004-05 and 2005-2006. 4.23 The bulk of public infrastructure investment has focused on the transport sector, mostly roads. However, the slightly improved availability of transport infrastructure in Central America since the 1990s has not been enough to keep up with the region’s economic growth, and bottlenecks are developing, especially in land transport through cities and at border crossings. The investments were mostly well-placed but limited compared to needs. Central America’s annual spending on public infrastructure has remained low at typically below than 2 percent of GDP, compared with 4 percent to 5 percent in such Asian giants as India and China. Asia is essentially pulling away from Central America in this aspect of supporting competitiveness. Road transport 4.24 Most of Central American transportation takes place by land, indicating that investment in roads and related infrastructure services may be crucial to facilitating traffic within each country and across the region. Road transport dominates for both passengers and freight in Central America (Table 4.3). 66 Box 4.1: Plan Puebla Panama The Plan Puebla Panama (PPP) is a multi-billion dollar development plan that was formally initiated in 2001 under the leadership of the Mexican President. The PPP was intended to "promote regional integration and development" among eight countries (six Central American countries, Colombia, and Mexico), especially by fostering improvements in their infrastructure sectors. The initial goal in the transport subsector was to create a fully linked, 8,977km Central American Highway network -Red Internacional de Carreteras Mesoamericanas (RICAM)-, along two main corridors, (Pacific and Atlantic) that would be complemented by branches and connections in each country. The individual governments were expected to adhere to a uniform set of technical specifications for rehabilitation or expansion of the network and they were also responsible for its subsequent maintenance in their territories. The PPP did not have any special funding arrangements of its own but rather served as a mechanism to be used by governments when prioritizing investments in terms of their regional attributes and for choosing among various infrastructure projects such as highways, air and sea ports, or electric and telecommunications grids. The IDB promoted the concept of the PPP and also essentially provided its Secretariat. By 2004, the first stages of regional interconnection, especially for the transmission of electricity, were generally considered to have been achieved successfully. For the transport sector, during the first five years of the PPP, approximately US$4,000 million were allocated by member governments for rehabilitation and improvement of 4,300 km of the RICAM; most of those projects should be concluded by end 2008. Emphasis was then placed on completing the Atlantic corridor, and also for the first time other transport modes and private sector participation were accorded more prominent roles in the PPP agenda. The scope of the RICAM was extended to 10,627 km by adding mostly existing roads, the majority of which (56 percent) required reconstruction and rehabilitation. It is estimated that an additional US$1,350 million is needed to rehabilitate and to improve the additional segments of the RICAM; however this time the Guatemala, Costa Rica and Mexico governments are trying to encourage the private sector to finance most of the investments. Apart from the physical investments, in the transport sector important advances were made in terms of the implementation of programs for the Harmonization of Regulations and Technical Norms, Air Transport and Airport Security, Fostering regional competitiveness, Modernization of border crossings, and Harmonization of Customs Schedules. Source: IADB-SIECA, Informe de Cierre - Primera Etapa, Plan Puebla Panama, Iniciativa Mesoamericana de Transporte," December 2007. 4.25 However, the region’s road networks are less than adequate. For Table 4.3: Logistics, transportation model paved roads, the typical Central % firms using each transportation mode American country’s endowment started Land Air & other Port & other Total at low levels in the 1980s and has not improved much. All six countries are El Salvador 90.0 2.5 7.6 100.0 Guatemala 94.8 2.6 2.6 100.0 below average for their level of Honduras 82.7 4.5 12.9 100.0 development in the percentage of roads Nicaragua 93.9 0.8 5.4 100.0 that are paved (Figure 4.2). Close to Panama 57.8 3.6 38.6 100.0 two thirds of the traffic flow in Central Total 90.8 2.6 6.6 100.0 America is concentrated in urban areas. Source: Enterprise Surveys 2006, 2007 4.26 Nicaragua, Honduras and Guatemala face particularly acute road shortages—measured by either the total roads or paved roads per 1,000 people (Figure 4.2). Panamá also has a limited road network, but it is less problematic due to the concentration of its population in Panamá City and the availability of other modes. 67 Figure 4.2: Overall Network and Paved Road coverage characteristics Total road density Paved road density 0.8 10 0.20 2.5 km paved/1,000 people km roads/1,000 people 8 2.0 km paved/km2 km roads/km2 0.6 0.15 6 1.5 0.4 0.10 4 1.0 0.2 2 0.05 0.5 0.0 0 0.00 0.0 NIC HND GTM SLV PAN CRI NIC HND GTM SLV PAN CRI territorial density population density territorial density population density Note: Data for 2000 (latest available). Source: WDI. 4.27 Central America’s road transport industry is characterized by Figure 4.3: Vehicle fleets in Central America atomized firms that provide poor quality vehicles per 1,000 people and costly services. Mobility is further 160 140 compromised by a relatively small fleet 120 100 and the age and often poor condition of 80 vehicles. The number of vehicles per 60 40 capita is low but compares reasonably 20 0 well with countries at similar development levels. Some countries, a a a a r as am do al ic gu r R m du a ra n te lv ta such as Nicaragua, have very small Pa a on Sa ua os ic H N G C vehicle fleets (Figure 4.3). The country passenger cars buses trucks had an estimated 250,000 vehicles in 2007, including about 21,000 heavy trucks and 5,500 buses. About 50 Source: World Development Indicators percent of Nicaragua’s freight transport fleet is more than 14 years old and a significant share is over 20 years old. Table 4.4: Percentage of firms using each type of transportation No 100% own Own & 3rd Party All 3rd Party Total transportation El Salvador 46.4 30.5 13.2 9.9 100 Guatemala 42.2 36.8 14.2 6.8 100 Honduras 30.9 17.1 23.9 28.1 100 Nicaragua 43.9 23.0 15.8 17.3 100 Panama 52.9 28.2 6.9 12.0 100 CA5 42.3 30.3 15.0 12.4 100 Source: Enterprise Surveys, 2006-2007 68 4.28 An unreliable and inefficient transport industry creates uncertainties and higher costs for firms that need to move inputs and goods. Many producers decide to bypass the commercial system by purchasing their own equipment and hiring their own personnel. Use of owned transport is unexpectedly high in Central America (Table 4.4). Table 4.5: Theft or breakage during transportation % firms reporting theft or breakage of % of cargo value lost to theft or breakage Domestic Cargo Export Cargo Domestic Cargo Export Cargo Costa Rica 19.4 26.4 6.6 6.2 El Salvador 36.9 19.3 4.8 7.1 Guatemala 25.0 40.1 10.8 2.2 Honduras 30.2 53.4 7.5 4.8 Nicaragua 28.7 17.3 8.4 10.6 Panama 27.0 21.4 4.8 2.0 CA6 28.0 32.0 8.1 3.6 LAC9 30.3 13.8 6.5 4.3 Note: LAC9 includes Argentina, Bolivia, Chile, Colombia, Ecuador, Mexico, Paraguay, Peru & Uruguay Source: Enterprise Surveys 4.29 The poor quality of transport infrastructure affects firms’ competitiveness beyond the direct and indirect cost of delays. In an underdeveloped transport system, goods are damaged relatively often in transit, and congestion leads to greater opportunities for theft. ES data show this is particularly true for El Salvador, where almost 37 percent of firms reported theft or damage in their domestic shipments (Table 4.5). The figure for export cargo exceeds 53 percent in Honduras. Nicaraguan firms estimated that they lost almost 11 percent of the value of their exports due to theft or breakage. These theft and damage problems lead to high insurance costs. Figure 4.4: Impact of Breakage and Theft in Transit on Costa Rican Firms a. Percent of Firms Reporting Breakage and Theft b. Percent of Sales Lost to Breakage and Theft Breakage Theft Breakage Theft 25 12 20 10 Percent 8 Percent 15 6 10 4 2 5 - - % of cargo value lost to % of cargo value lost to % firms that ship % firms that export and breakage or theft (average breakage or theft (average domestically and report report theft or breakage of for all exporting firms) for exporting firms that theft or break age of cargo cargo experienced breakage or theft) Source: Enterprise Survey 2005 69 4.30 ES data for Costa Rica suggest subtle differences between breakage and theft. More domestic shippers and exporters report problems with breakage. For example, 23 percent of exporters report breakage and only 11 percent report theft (Figure 4.4). Among exporters encountering problems, however, losses as a percent of value are 10 percent for theft, compared with only 2 percent for breakage. These findings help shift the spotlight to bottlenecks at port facilities. Port facilities 4.31 Much of Central America’s Table 4.6: Cargo volumes and market shares commerce transits the region’s seaports on the way to international markets, Unloaded Loaded Total Share increasingly through the Atlantic-side (%) facilities for containerized cargo. As might Guatemala 10,572 5,509 16,081 19 El Salvador 4,910 1,055 5,965 7 be expected, Panamanian ports move the Honduras 6,228 3,165 9,393 11 largest share of the region’s shipping— Nicaragua 2,354 353 2,707 3 about 46 percent of the total of 86.2 million Costa Rica 7,336 5,488 12,824 15 tons (Table 4.6 and Figure 4.5). Guatemala Panama 20,606 18,639 39,245 46 ranks second with 16.1 million tons. Total 52,006 34,208 86,214 100 Source: COCATRAM (2006, Table 4) 4.32 Table 4.7 shows a doubling of the container volumes (TEU) handled by Figure 4.5: Market share by Country Central American ports since 1999. The Panama figures should be interpreted cautiously in 46% the case of Panama, where they include transshipment due to Canal activities and Costa Rica do not really represent solely the national 15% economy. With the growth of most commerce being on the Atlantic side, it is Nicaragua not surprising that Nicaraguan and 3% Salvadoran ports handled the smallest Honduras volumes. Figure 4.6 shows that 11% Guatemala 19% Panamanian ports had the highest growth El Salvador 7% rate during 1997-2006—almost 20 percent a year. Source: COCATRAM, 2006 Table 4.7: Containerized Cargo volumes (TEU) 1999 2006 % Change 4.33 A lack of port efficiency Guatemala 507,676 835,253 65 and capacity hinder Central El Salvador 11,132 124,331 1017 American exporters. This is Honduras 280,197 593,800 112 particularly true for El Salvador and Nicaragua 9,211 47,948 421 Nicaragua, which do not have their Costa Rica 640,890 880,436 37 own facilities on the Atlantic and Panama 1,263,619 3,027,562 140 rely primarily on Puerto Cortez in Total 2,712,725 5,509,330 103 Honduras for their shipments. Source: COCATRAM, 2006 Puerto Caldera in Costa Rica was 70 designed to move a maximum of 600,000 metric tons a year. The limit was surpassed in 1988; since 2004, Puerto Caldera has operated at more than 300 percent of capacity. Strained ports and red tape mean bottlenecks for businesses depending on imported materials. The Costa Rica ICA (World Bank, 2007) reports estimations made by Bontempo, Chávez, and Rivera. These authors estimated that a company importing grain through Puerto Caldera in 1996 could see a CIF price increase of at least 6 percent just because of warehousing expenses. Although there are no updated studies, the situation probably has not changed significantly. Figure 4.6: Total Cost per container, including Operating and Inventory Expenses caused by Inefficiency Market S hare of TEUs, 2006* Annual Growth, 1997-2006 60 60 50 50 40 40 Percent Percent 30 30 20 20 10 10 - - Costa Rica El S alvador Guatemala Honduras Nicaragua Panama * TEUs refers to Transport Equivalent Units (20-feet container) Source: American Association of Port Authority 4.34 Limon, an Atlantic port in Costa Rica, appears to enjoy a cost advantage over other facilities in the region. According to a 2004 study by Kent and Fox, charges per vessel and per container are lower at Limon than at Cartagena Figure 4.7: Total Cost per Container, Including Colombia (Figure 4.7). This comparison seems Operating and Inventory Expenses Caused by impressive because Cartagena is widely Inefficiency considered an efficient port, which has been under a concession contract since 1993 and 450 393,9 400 operates in a highly competitive environment. 350 However, any cost saving from Limon’s lower 300 US dollars port charges is more than offset by higher 250 222,0 operating and inventory expenses. For 200 example, crane productivity is significantly 150 100 lower—38 moves per ship-hour at Limon, 50 compared with 52 at Cartagena. Moreover, 0 waiting times to dock are 12 hours longer at Limon Cartagena Limon than at Cartagena, and waiting times at the berth are 1.5 hours longer. These waiting Source: Kent and Fox (2004) times result in additional expenses associated with idling ships as well as higher inventory expenses. After taking these expenses into account, 71 Kent and Fox estimate that the total cost per container for a call at Limon exceeds the cost at Port of Cartagena by $171.89. 4.35 In considering Central America’s exports, volume and value offer a different perspective on the relative importance of each transport mode. Maritime corridors represent 81 percent of exports on a volume basis and about half of exports by value. This difference reflects the value of time and the use of air shipments for high-value, low-weight products, such as medical equipment, cut flowers, and computer chips. In fact, air shipments in Costa Rica represent 1 percent of export volume but 37 percent of value. Another characteristic is the large number of firms reporting a dependence on surface transport (roads) for exports. The data suggest smaller firms are more dependent on regional markets, and thus the road network, for their exports. 4.36 Because trade outside of Central America represents more than 70 percent of Costa Rica’s total trade, port access and efficiency are major concerns for the country’s firms. Limon is the primary port for Costa Rica in terms of both imports and exports, handling nearly 76 percent of the country’s freight. Limon’s exports are concentrated (82 percent) in primary sector-related products in which logistics play a central role in competitiveness. However, poor road connections to San Jose and capacity and performance constraints at Limon and the nearby port of Moín combine to make this trade channel costly and inefficient. Port infrastructure was heavily criticized by firms interviewed in the Logistics Survey. The deficiencies include road access, cranes, container berths, and pier depth. Firms also pointed to outdated technology, which hinder loading and unloading and embarkation and debarkation. Figure 4.8: Liner Connectivity Index of CA and Comparator Countries, 2006 Liner Shipping Connectivity Index GDP per capita GDP per capita, PPP (constant 2005 Liner Connectivity Index (1=max) 0,35 14.000 0,30 12.000 international $) 0,25 10.000 0,20 8.000 0,15 6.000 0,10 4.000 0,05 2.000 0,00 - El Salvador Guatemala Colombia Thailand Jamaica Panama Honduras Nicaragua Mexico Costa Rica Philipines Chile Note: The Liner Shipping Connectivity Index is a composite index that includes fleet assignment, liner services, and vessel and fleet sizes. Source: UNCTAD (2006), WDI 72 4.37 The liner connectivity index sums up shipping services in a single score based upon the number of containers handled at a country’s ports, the frequency of regular container shipping services, and the size of vessels that call at the country’s ports. Based on this measure, Central America as a region is not underserved, primarily due to Panama’s performance (Figure 4.8). However, as a result of capacity constraints and inefficiencies, Costa Rica ranks well below the liner-connectivity standard set by other countries at its level of income, such as Chile and Mexico. 4.38 Inadequate port capacity, inefficient port services, and poor liner connectivity have important implications for the future of Central America’s export competitiveness. Maritime- trade forecasts for Central America’s ports show a significant and continuous growth, with some changes in relative market share by country (Figure 4.9). Indeed, Guatemala is expected to nearly double Costa Rica’s East Coast shipping movements by the year 2020, while Honduras is expected to surpass Costa Rica over the same period. At current performance levels, Puerto Caldera will be an even less significant player in Pacific cargo movements within 10 years. 5 Figure 4.9: Projected Port Throughout by Country and Coast (Volumes in Millions of Metric Tons) a. Caribbean Ports b. Pacific Ports Guatemala Guatemala 30 24 Honduras El Salvador Honduras Millions of Metric Tons Millions of Metric Tons Limon-Moin, CR 25 18 Nicaragua Caldera, CR 20 12 15 6 10 5 0 2003 2010 2020 2003 2010 2020 Source: Authors’ calculations based on forecasts of UNCTAD (2006) Note: Pacific ports’ forecasts do not include development of the Port of Cutuco at La Unión, El Salvador. Air transport 4.39 All Central American countries have modernized their airports since the 1990s and are connected by relatively frequent service to the US and other main markets. In general, the region’s airports are busy. For example, passenger traffic through Costa Rica’s primary airport, the Aeropuerto Internacional de Juan Santamaría, has grown rapidly—about 10 percent per year—over the past decade. At the same time, the increasing share of high-value, time-sensitive shipments in total exports have added to demand for the airport’s cargo facilities. 5 Imported volumes in Puerto Caldera were seven times export volumes in 2003. This suggests that all Asia and most West Coast-bound exports must travel to Panama to find a port of exit. 73 4.40 Most Central American airports are now operated by concessions and use a share of the passenger exit tax for their operation and maintenance. To expand its airport’s passenger and cargo capacity, Costa Rica signed a concession with a private company in 1997. About half of the planned construction was completed, with the cargo area significantly improved, before a dispute with the concessionaire halted the work. Business survey respondents give the airport the highest marks among Costa Rica’s transport infrastructure, but current capacity during tourism season is insufficient to meet demand. 4.41 For Central America as whole, garments and agricultural products dominate exports, and less than 2 percent of goods by tonnage are shipped by air. Given the small share of freight and the good connectivity, this sector is not considered a bottleneck to business competitiveness. 3. ENERGY SECTOR 4.42 Energy consumption and intensity are low for Central America. For example, the region’s per capita consumption in kilowatt hours/month is significantly below the LAC countries’ average (Figure 4.10). However, Costa Rica is above the LAC average, and Panama is similar to it. Figure 4.10: Per capita consumption per month 2000 160 148 2006 140 140 125 126 122 120 107 100 80 57 62 60 48 50 53 42 34 39 31 40 25 20 0 Costa El GuatemalaHonduras Nicaragua Panamá CA LAC Rica Salvador Source: Author’s calculations based on CEPAL 4.43 Energy intensity measures consumption per unit of GDP—a higher level suggests less efficient usage. Central America’s overall measure is higher than the LAC average, although the values in Costa Rica and Panama are lower (Figure 4.11), and El Salvador’s is almost the same as the LAC average. Honduras and Nicaragua are the most inefficient energy users in Central America. 4.44 Between 2000 and 2006, Nicaragua and Panama increased their energy intensity slightly—that is, they become a little more inefficient in their use of energy. Honduras and El 74 Salvador became more efficient, helping pull down the region’s overall energy intensity. Differences arise between electricity and oil. All countries are consuming more electricity per unit of GDP. Outside Honduras and Panama, however, countries have maintained or reduced their energy intensity with respect to oil, reflecting conservation associated with higher prices (Figure 4.12). Figure 4.11: Energy intensity (BEP/K US$) 4.5 4 2000 2006 3.5 3 2.5 2 1.5 1 0.5 0 Costa El GuatemalaHonduras Nicaragua Panamá Istmo LAC Rica Salvador Source: Author’s calculations based on CEPAL Figure 4.12: Electric and Energy intensity Electric intensity Energy intensity (GWh/106 USD constant prices 2000) (Bbl/103 dollar of PIB 2000) 0.6 Energy Intensity 2000 2006 0.5 2.50 0.4 2.00 0.3 3 1.50 0.2 BbL/ 10 dollar of PIB 2000 2000 2005 1.00 0.1 0.50 0 Costa Rica El Guatemala Honduras Panamá Nicaragua 0.00 Salvador Costa Rica El Salvador Guatemala Honduras Nicaragua Panamá Source: Author’s calculations based on CEPAL and OLADE 4.45 With the exception of Guatemala, which produces some oil (output has declined since 2000), Central American countries import all their needs for oil and oil products. Initiatives for 75 oil exploration are being supported in several countries. With the exception of Costa Rica, all Central American countries produce a large proportion of their electricity with oil-based products. Most nations support the development of renewable energy sources, and thermal generation has increased in recent years. 4.46 Like all other regions of the world, Central America faces higher prices for energy, particularly oil. Despite the rising prices, the region’s diesel and gasoline consumption have increased, except in Nicaragua, where consumption of diesel fuel in 2006 remains practically at 2000 levels. Consumption increases would have been even greater if not for increased mileage efficiency in the transport fleet. Figure 4.13: Hydroelectric potential and net generation a. Hydroelectric Potential and its utilization 7,000 Potential Installed Cap acity 6,000 700,000 600,000 5,000 500,000 4,000 400,000 MW MW 3,000 300,000 2,000 200,000 100,000 1,000 0 0 CA LAC Cost a Rica El Salvador Guat emala Honduras Nicaragua P anamá Potential Installed Cap acity b. Net Generation by type, Central America Isthmus Renewable Charcoal 2000 2006 Oil 0% 20% 40% 60% 80% USD/gallon Source: Author’s calculations based on CEPAL and OLADE 4.47 In procuring oil, Central American countries have benefited from the availability of oil products under preferential conditions, based upon the cooperation of other governments 76 (Venezuela and Mexico). The two principal mechanisms in place are the San José accord and the Petrocaribe accord. Both accords provide preferential terms on payment, but not necessarily on price. In the San José accord, all Central American countries are beneficiaries of oil originating in Mexico and Venezuela. In Petrocaribe, the beneficiaries are limited to Nicaragua and Honduras. A large proportion of the fuel purchased under the Petrocaribe accord is financed at subsidized rates over 20 or more years. The two Petrocaribe treaties differ somewhat, but their essentials are similar: the accord is managed by the Government or by a state corporation that sells fuel at market prices to the Exxon refinery (in the case of Nicaragua) or to large users and fuel distributors in Honduras. Consequently, the benefits from the treaties do not translate into lower fuel prices for consumers, or lower electricity rates, thereby preserving incentives for reduced consumption. According to the treaty’s terms, profits derived from the accord are expected to be channeled towards social and economic development projects as part of the poverty reduction strategy. The treaties also reduce rationing risks. 4.48 The main option for addressing high oil prices lies in changing the energy matrix and making it less dependent on oil, mainly because the price of many of its substitutes have not increased proportionally. For example, although there is a spot market for coal, it is mostly traded on the basis of long term contracts which provide greater price stability than bunker or other oil product contracts, and the same holds true for LNG. In the case of geothermal exploration, which uses similar technologies to oil exploration, drilling costs have increased but—as in the case of oil—the rewards have become proportionally greater. A similar case can be made for wind farms, whose equipment cost has increased with booming demand in the last years. Diversification is particularly feasible in the power subsector, where plants based on diesel oil could be replaced by those that use bunker fuel (about half the price of diesel), by coal plants or, even better, by hydroelectric or geothermal plants. 4.49 While the region lacks significant oil production, it exhibits an important potential in terms of hydroelectricity and geothermal resources. Costa Rica, Guatemala, and Honduras have the highest hydroelectricity potential. However, resources have been harnessed unevenly (Figure 4.13). Usage of the hydroelectric potential tops out at 25 percent in Panama. Nicaragua, by contrast, has only harnessed 6 percent. Overall, the region’s use of energy from renewable sources has declined since 2000. However, the scenario of high oil prices has sparked a renewed interest in developing hydro plants of all sizes, geothermal fields, and wind resources. Figure 4.14: Electrification Index 100% 2000 2006 2000 2006 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% P anamá El Salvador Guat emala Honduras Nicaragua CA LAC 77 Source: Author’s calculations based on CEPAL 4.50 Electrification has improved in all countries in Central America. Although the regional average is still below the LAC average, the difference is much smaller than it was in 2000 (Figure 4.14). Wide differences remain among countries, with Costa Rica having an electrification ratio of around 99 percent, while Nicaragua and Honduras are on the order of 60 percent to 70 percent. Figure 4.15: Electricity tariffs and prices Electricity prices US$/kWh Source: LAC Benchmarking Database, The World Bank 2007 4.51 Electricity prices vary significantly among Central American countries, with residential prices lowest in Costa Rica and Honduras, and highest in El Salvador, Nicaragua, and Panama. Commercial customers generally pay higher prices than residential or industrial users. Countries with strong sources of renewable energy, such as Costa Rica, Panama, and Honduras, are among those with lower prices (Figure 4.15). 4.52 Effective electricity capacity and demand are very close in all countries in 2007, leaving little margin for reserves (Figure 4.16). Markets are already tight in El Salvador, Honduras, and Nicaragua. According to the expansion plan developed by the regional planning group “Grupo de Trabajo Planificación Indicativa Regional� (GTPIR), El Salvador would have zero (or negative) reserves by 2009. Potential capacity restraints create severe constraints on the development of new industry by imposing additional costs associated with power disruptions, scheduling breakdowns and the need for investing in backup equipment. 4.53 Importing electricity may help alleviate capacity restraints. Central America is strategically located with feasible interconnections to energy-rich neighbors Mexico and Colombia. At present, a 400-kilovolt line is being built between Mexico and Guatemala, with an initial transfer capacity of 200 Megawatts from Mexico to Guatemala and 70 Megawatts in the opposite direction (to be increased by 30 percent upon completion of transmission expansion projects in Guatemala). A 220-kilovolt interconnection line is being studied between Panama and 78 Colombia, with 300 Megawatts of transfer capacity. Ultimately, the Central American nations hope to link the interconnection lines with grids in North and South America. 4.54 The Sistema de Interconexion Electrica para America Central (SIEPAC) grew out of discussions that began in the late 1980s. Its two main objectives are: (1) to support development of the regional electric market (Mercado Eléctrico Regional, or MER) by establishing the legal, institutional, and technical mechanisms to facilitate private sector participation in the development of additional electricity generation capacity, Figure 4.16: Peak Load/Effective Capacity and (2) to establish the 1,4 necessary transmission 1,2 infrastructure for the 1,0 interconnections to permit the 0,8 electricity exchanges between 0,6 MER members, with their 37 0,4 million consumers. The 0,2 framework agreement (Tratado 0,0 Marco) was signed in 1996. The do r ra s ua al a ica am á l va ndu ra g em aR P an Inter-American Development E l Sa Ho N ica G ua t C o st Bank (IADB) is financing SIEPAC’s institutional and Source: Expansion Plan, Grupo de Trabajo Planificación Indicativa regulatory components. The Regional (GTPIR), 2007 infrastructure is under the responsibility of the Empresa Propietaria de la Red (EPR), the company that owns the line. EPR’s shareholders include Endesa (Spain), ISA (Colombia), and the six Central American countries. 4.55 This infrastructure, along with the support of national transmission systems, will allow regional countries to have a reliable interconnection capacity of 300 megawatts. Electricity trade is developing and, despite a reduction in 2006, it is expected to increase with the new line in 2010. The reduction in trade during 2006 reflected the tight constraints on supply within all countries in the region (Figure 4.17). This has forced countries such as El Salvador to put into service very high cost production units, which will impact rates. Figure 4.17: Imports and Exports of Electricity By country Regional Market CA 79 1000 800 2006 Exp orts Imp orts 600 GWh 400 200 ` 2000 0 Costa Rica El Guatemala Honduras Nicaragua Panamá Salvador 0 300 600 900 1,200 1,500 Exports 2000 Imports 2000 Exports 2006 Imports 2006 GWh Source: Author’s calculations based on CEPAL and OLADE 4.56 The reliability of power supplies is important for assessing competitive conditions in the region’s energy sector. Reliability refers to grid-related performance, the failures in the transmission or distribution systems rather than interruptions related to the supply and demand imbalances at the generation level. The two usual indicators used to track reliability are known as SAIFI (System Average Interruption Frequency Index) and SAIDI (System Average Duration Index). The calculation of these indices follows rules that vary from one system to another; for example, major disturbances originating principally in weather conditions (e.g. hurricanes) are often ignored to limit tracked outages to technical factors, such as line overloads, that may be overcome through appropriate investments. 4.57 Not all countries track reliability. In general, the indicators are used by regulators to establish service quality conditions for regulated (usually private) distribution companies. Quality indicators are tracked in Panama, El Salvador, and Guatemala, where private distribution predominates and where relatively strong regulation prevails. Across the region, it appears that the most serious reliability problems occurred before 2000 (Figure 4.18). El Salvador includes rebates to consumers affected by service interruptions. 4.58 In 2007, the World Bank undertook a benchmarking exercise for energy regulatory agencies in the LAC region, with a view to categorizing them according to the quality of their regulatory governance. 6 It was defined in terms of three attributes: autonomy, accountability, and transparency. Some characteristics of the Central American regulators are summarized in Table 4.8. The similarity between the setup of the different agencies is not random: they were designed around the same time and in many cases the same consultants advised the governments. Figure 4.18: Electricity interruptions 6 Andres, Luis et.al. (2007) 80 Source: LAC Electricity Benchmarking Database, The World Bank, 2007 4.59 The benchmarking resulted in an overall Electricity Regulatory Governance index, which could be used to rank the different agencies. Five Central American agencies fell near the LAC’s median—in order, El Salvador, Guatemala, Nicaragua, Costa Rica, and Panama. One agency—in Honduras—is the LAC’s worst performer. Over all, the study suggests that, with the exception of Honduras, the investment climate in the power subsector is enhanced by functional regulating agencies. Table 4.8: Characteristics of Central American Energy Regulatory Authorities Name Year Legal Status Budget Appeals Accountability Staff SIGET 1997 Separate entity with Regulation Judicial Government 106 (El Salvador) autonomy from the line Tax Review and Congress ministry CNEE 1996 Separate entity with no Regulation Executive and Government 51-100 (Guatemala) autonomy from the line Tax Judicial and Congress ministry Reviews INE 1994 Separate entity with Regulation Judicial Government 200 (Nicaragua) autonomy from the line Tax Review and Congress ministry ARESEP 1996 Separate entity with Regulation Judicial Congress 167 (Costa Rica) autonomy from the line Tax Review ministry ASEP 1996 Separate entity with Regulation Judicial Government 182 (Panama) autonomy from the line Tax Review ministry CNE 1995 Separate entity with Governmen Judicial Government <20 (Honduras) autonomy from the line t Budget Review and Congress ministry Source: Andres, Luis et.al. (2007) 4.60 This result should be interpreted cautiously because it does not take into account sector performance. For example, Panama is ranked low in terms of formal organization, but the country boasts one of the better functioning and market-oriented sectors in the region. On the 81 other hand, Nicaragua’s agency achieved a relatively high ranking, despite having had constant struggles with the distribution company it regulates, including a claim against MIGA. 4.61 In addition to the national regulators, the SIEPAC interconnection has given rise to a regional agency, currently based in Guatemala, that is oriented towards the regulation of interchanges within the Central American market (MER). Its mandate does not include regulation of prices within individual countries, and its scope is somewhat limited. 4.62 Electricity market performance depends on private investors as well as public regulations (see Tables A.4.1 and A.4.2 in the Appendix). For generation, Panama stands out for its high participation by the private sector at 85 percent. El Salvador, Honduras, Guatemala, and Nicaragua all rely on the private sector for at least 60 percent of their supplies. Costa Rica’s 17 percent in private generation is primarily renewable, mainly hydros of 50 megawatts or less. Outside of Panama and Costa Rica, the private sector has invested in thermal generation. Despite its vertical integration, Honduras has the highest dependence on private thermal generation. Only Guatemala has coal-based private generation at present. 4.63 A number of generation projects in excess of 150 megawatts are under consideration (Table 4.9). They correspond to the latest expansion plan of the Electrification Council for Central America (CEAC). In El Salvador, two additional projects could be added—a 250 megawatt, coal-based plant and a 500 megawatt liquid natural gas plant. The former has been officially announced by the private investors and has the greatest chance of being developed. In Nicaragua, the government advised the private company promoting Copalar to desist due to environmental and social problems associated with the project. Table 4.9: Possible regional hydro projects Country Project Capacity MW Mill $ $/KW Period Xalalá 181 312 1722 2013-2020 Guatemala Chulac 446 666 1491 2016-2020 El Tigre 704 1517 2155 2014-2020 El Salvador Cimarrón 261 674 2583 2014-2020 Jicatuyo 173 529 3060 2014-2020 Honduras Patuca 2 270 685 2536 2013-2020 Tumarin 160 321 2005 2014-2020 Nicaragua Copalar 350 808 2309 2012-2014 Savegre 200 503 2515 2014-2020 Costa Rica Reventazón 300 670 2232 2015-2020 Diquís 622 1347 2166 2016-2020 Barú 150 685 4564 2012-2020 Panamá Chan75 153 387 2530 2010-2020 Source: National Companies’and Regulators’websites 4.64 All of these projects involve investments in excess of $500 million, which should be mobilized in partnership with the private sector. PPP arrangements are convenient because risks 82 are shared between the Government and the private developer, with benefits to the host country in the form of royalties, income taxes, and electricity supply. 4.65 Private companies operate most of the distribution systems in El Salvador, Guatemala, Nicaragua, and Panama. Performance varies considerably. The El Salvador and Panama distribution companies are profitable and sustainable, with healthy indicators in terms of losses. In Guatemala, there is a mixed picture, with private companies that appear solvent overall but have serious problems associated with the regulator’s decisions. In Nicaragua, the sole private distributor appears to be on the verge of insolvency and has entered into frequent confrontation with the regulator. The two other countries are extremes in terms of performance: the Costa Rican distribution has the lowest losses of the region, whereas Union Fenosa, the Nicaraguan distribution company, has the highest. From 2000 to 2006, loss levels have increased in El Salvador, Guatemala and Honduras; they fell in Costa Rica, Panama and Nicaragua (Figure 4.19). For 2006, the region’s losses were similar to the LAC average. 83 Figure 4.19: Electrical Losses (%) 35 25 2000 2006 30 2000 2006 25 20 20 15 % % 15 10 10 5 5 0 0 Cost a Rica El Salvador P anamá Guat emala Honduras Nicaragua CA LAC Source: Author’s calculations based on CEPAL and OLADE 4. CHALLENGES AND STRATEGIES 4.66 Pricing and subsidy policies: The role of pricing policies and cost recovery cannot be overstated in managing demand and financing of infrastructure. These two concepts are key to the efficient use of facilities, mobilizing resources to pay for maintenance and operations, and providing incentives for private sector participation. Most Central American countries have only partly adopted policies that (1) require public utilities to cover operating expenses, interest, and depreciation and self-finance a part of their investment program; (2) target the needs of an area on a holistic basis, especially true for poor urban and rural sub-areas, and (3) streamline mechanisms for the periodic adjustment of infrastructure prices. If pricing is based on full cost recovery, affordability may remain a major obstacle to expanding services to low-income beneficiaries for some types of projects—for example, modernizing urban transport or connecting maquila zones and poor neighborhood. In such cases, explicitly targeted subsidies are better than the existing complex and non-transparent pricing policies, with their hidden cross- subsidies. This applies, for example, to improving the abysmal quality of the strike-plagued urban bus services in virtually all Central American cities. In electricity, there remains a great deal to be done in terms of cost recovery levels, particularly with the increase of oil prices, and in targeting subsidies to the poor. 4.67 Setting appropriate priorities for infrastructure investment: Investors tend to locate near transportation nodes and labor sources, especially in the maquila industries. Central America’s governments have at times focused their efforts on such productive zones, with an eye toward concentrating infrastructure investments and realizing considerable synergies. This strategy may be a necessity. Countries may not be able to address broad infrastructure needs due to public spending ceilings included in agreements with the International Monetary Fund. Limited funding makes it more important to reflect the priorities of this productive zones strategy in public expenditure programs. In terms of productivity in a particular zone, for example, improving the network of secondary and rural roads that connect to existing trunk roads may be 84 more important than building more trunk national highways. For Costa Rica, it may be more important to improve the trunk road connecting the port of Limon to San Jose city. In the power sector, the shorter horizon of private investors and prevailing cost considerations led to an imbalance in the thermal/renewable proportion of generation in the 90s in most Central American countries. Maintaining a more balanced approach by taking an active role in promoting the development of hydro and geothermal sites would have left the region better equipped to addressing the challenges of higher oil costs. Reverting the imbalance is a short and medium term priority in the region, but it is not likely to take place fast enough to moderate consumer prices. 4.68 Promote regional interconnections: The soon to be re-launched Plan Puebla Panama (PPP) for economic integration among Central America, Colombia, and southern Mexico may provide an important boost to attaining regional goals. So will completion of the Atlantic road axis, which follows the direction of most Central American commerce. In the maritime sector, there is a need to develop cabotage and short-haul sea shipping, especially linking the Atlantic ports where most Central American containers are moved. With World Bank assistance, COCATRAM, the regional maritime advisory group completed studies on the potential for these shipping services and identified the lack of an enabling framework in all Central American countries as the main obstacle. 4.69 Improved regional maritime services should be a component of the re-launched PPP. The eight-nation forum should also target the strengthening of regional institutions so they could be capable of managing region-wide infrastructure operations and services. This would help provide the confidence needed to spur private companies to invest in the major projects that serve markets that extend beyond borders. The role of regional institutions in developing and putting in place a framework that would bolster private investor confidence takes on greater importance in light of the constraints on each Central American country’s public infrastructure investment budget. The region may need to place greater reliance on the private sector to help meet PPP goals. This would be especially true for regional transport fleet modernization, the development of cabotage services, or for major port and logistics terminal investments. Some specific steps for addressing infrastructure bottlenecks and gaps: 4.70 In Central America, the most extensive and crippling bottlenecks to enhancing competitiveness lie in the road sector, and significant investment will be required to address them. This has long been recognized in country priorities, and the strategy has been to remove bottlenecks by focusing first on the core trunk network. Strategies in the transport area need to focus on improving coverage, efficiencies, self-financing ratios, and on deepening private-sector participation. 4.71 The program for addressing gaps and bottlenecks could include: • Improve and expand the paved road network. The region needs trunk road rehabilitation programs as well as improvements in the grid of secondary and major rural roads (the distributor network) that link to the trunk roads and serve the productive zones. The needs are vast, averaging about 400 kilometers per country per year. If managed well, addressing these needs could provide an opportunity for 85 strengthening Central America’s construction and finance sectors as well as generating high and consistent employment levels. Improving the distributor networks that serve countries’ more productive zones would in addition address the local infrastructure needs and help to boost productivity in the agriculture sector. • Emphasizing maintenance. As roads are improved, it will be essential that they are not left to deteriorate and once again become bottlenecks. A comprehensive maintenance program would include guaranteed funding arrangements via dedicated road or airport funds and the use of least-cost public bid contractual approaches for implementation. From a macroeconomic perspective, extra-budgetary mechanisms, such as dedicated road and airport funds, are conceptually problematic because they could hamper the normal budget process’ allocation of scarce public resources among competing goals. Nevertheless, a number of microeconomic arguments can sometimes justify their use. First, they provide a means of helping to ensure that users receive the services they pay for. Second, they help to stabilize the flow of resources for maintenance expenditure, which is typically highly vulnerable to cutbacks during fiscal downturns. Third, with public-private co-financing of infrastructure investments, the use of a trust-fund mechanism can provide additional assurance to private investors that public funds will be forthcoming, thereby reducing the perceived risk of participation. Fourth, the creation of an autonomous source of funding can improve managerial incentives by reducing the potential for political interference. In addition, the microeconomic case for extra-budgetary mechanisms is strengthened whenever there are suitable governance arrangements to ensure that funds are used effectively. In short, there appears to be a conceptual case for the use of trust funds, particularly where they are funding public goods components of the infrastructure network (such as road maintenance or airport operations and maintenance), or socially motivated subsidies to privately operated infrastructure services (such as for coastal shipping or air transport services). In the case of the roads sector, in particular, trust funds for road maintenance have been quite widely advocated and applied under World Bank projects in many countries. • Strengthen planning and management. The Central American countries need to develop the basic capacity to carry out road-condition inventories, fully install their Pavement Management systems, seek least-cost solutions to technical and engineering problems associated with road and bridge maintenance, and develop their domestic consulting and contracting industries. In addition, all Central American entities should receive appropriate technical assistance, particularly in contracting and procurement procedures to provide them with the capacity to manage the maintenance of the entire core road network. For overall sustainability and efficient management, further analytical work is needed to identify the appropriate roles and responsibilities of the key players (both at central and local government levels) and to promote synergies while avoiding the duplication of efforts. • Improve urban transport. To improve traffic circulation and remove restrictions to reliable mobility, priority should be accorded to building busways for where justifiable in the major cities. The World Bank is about to support the development of such a busway for Panama City. Apart from relieving congestion and boosting city 86 competitiveness, the busways could link, for example, workers’ population centers to their employment in the maquiladora industries. City public transport services, although already largely operating under concessions, need to be restructured as part of this integrated busway development. Furthermore, improved traffic management is sorely needed if costly urban road space is to be freed up and used more efficiently for the movement of both people and freight. • Finally, key steps are needed to maximize the impact of investment priorities, pricing policies, and the institutional framework on development, which could encompass both rural and urban zones. For example, granting highest priority to strategically enlarging and upgrading domestic transport and utilities infrastructure could help to promote the modernization of agribusiness and the expansion of tourism. The City Port project in Limon, Costa Rica, could serve as a model for this territorial approach. The current market imperfections and the implicit and explicit transaction costs can be greatly reduced by such targeted policies and investments to improve transport infrastructure and service provision. 4.72 In the power sector a number of recommendations emerge, related to integration issues, such as: • Strengthen the institutional framework in the power sector. Several of the failings in electricity sectors where private sector participation became significant can be traced to weak and unpredictable regulation. Lessons learned from countries which have been successful in establishing effective regulating agencies should be disseminated, and proven approaches should be adopted to increase supply, such as energy and capacity auctions for new capacity; planning capabilities were neglected during the last fifteen years when private sector investment flourished. In order to address current challenges and to establish effective PPPs the public sector must be strengthened and effective regional cooperation should be put in place. The existing electrification committee for Central America (CEAC) should be strengthened with participating Governments and multilateral support (IDB has taken a leading role in this sense). • Strengthen utilities financially through appropriate regulations and investments. There is a disparity of performance among utilities in the region which could be reduced by adopting best practice regulations, based upon experience and proven performance of selected systems, such as the Panamanian and Salvadoran setups; cooperation in this sense could take place through the Plan Puebla Panama committees; • Establish common standards of service for power utilities. At present there is a wide divergence of regulations controlling quality of service, which could be reduced by agreeing on common standards and mechanisms for improving service, thereby raising the level of competitiveness in the region; • Establish regional cooperation mechanisms. The natural follow-up to the SIEPAC line should include coordination of operations and large investments among the countries with a view to capturing economies of scale in new investments and taking advantage of the opportunities offered by an enlarged market; 87 • Identify methods to capture economies of scale in power generation. The development of regional power plants is a priority in the region, particularly for hydro-based resources, which requires the adoption of a common approach, such as regional auctions as indicated above; although oriented towards native resources, such an approach could encompass regional plants to diversify the energy matrix by introducing alternative fuel sources such as coal; • Coordinate approaches to the economic provision of oil products. A key decision for supplying oil products to the region lies in the possible construction of a refinery; such an enterprise should be closely coordinated with private sector investors, and particular care should be taken to ensure its economic viability within the competitive market for oil and oil products in the Caribbean. 88 Chapter 5. FINANCIAL SECTOR 1 INTRODUCTION 5.1 Expanding access to credit and other financial services in Central America requires progressively overcoming the constraints inherent in the existing legal and operational infrastructure. Through years of work on the issue, the World Bank has accumulated a wealth of in-depth knowledge on the legal and operational environment surrounding financial sector activities in nearly all Central American countries. 2 An examination of the investment climate provides an ideal platform to look at the region as a whole and to identify common issues and possible policy options. 5.2 This chapter focuses on four key areas of financial sector infrastructure: (1) Financial Regulation and Supervision; (2) Protection of Creditor Rights and the Legal Framework for Corporate Insolvency; (3) Payment Systems; and (4) Accounting and Auditing Standards. The single source of information is the series of Reports on the Observance of Standards and Codes (ROSC) produced for individual Central American countries, either within the Financial Sector Assessment Program (FSAP) or independently by staff and consultants of the International Monetary Fund (IMF) and World Bank. 5.3 The ROSCs behind the four sections are: (1) Basel Core Principles for Effective Bank Supervision (BCP); (2) Insolvency and Creditor Rights (ICR); (3) Core Principles for Systemically Important Payment Systems (CPSIPS); and (4) Accounting and Auditing (A&A). Each report was based on an expert, qualitative assessment of the degree of compliance with a series of well-known and generally accepted international standards or principles of good practice. 5.4 Most country-specific ROSCs are confidential and remain unpublished, and this chapter will include few references to individual countries. Instead, the analysis will look at the common challenges found across the region, as well as recommended prescriptions. 5.5 Assessments of compliance with international standards can help countries pinpoint problem areas within legal and/or operational systems that may constrain growth. In this way, the assessments can serve as a guide for setting reform and development priorities, and they can provide an impetus for improved transparency and disclosure. 5.6 ROSCs provide summary findings of compliance with the corresponding standards and recommend specific actions to improve the functioning of the financial sector. Hence, this 1 The production of this chapter was coordinated by Aquiles A. Almansi (LCSPF), with advice from Adolfo Rouillon (LEGPS), Mario Guadamillas (LCSPF), Henri Fortin (LCSPF), and the collaboration of Cecilia Brady (consultant), Rashmi Shankar (LCSPE), and Daniel Alonso (LEGPS consultant). Susana Sanchez prepared the section on Central America’s financial sector. 2 The only important exception being Panama. 88 chapter will include a large number of specific recommendations. However, the key findings and associated recommendations can be summarized as follows: • Financial sector observers often fault bank regulations and supervisory zeal, for the high costs and limited outreach of credit and other financial services. Reading the BCP assessment of supervisory challenges in Central America should, however, help lead to a more balanced perspective. Despite the rapid regionalization of financial services over the past few years, Central American supervisors still cannot, for a variety of legal and logistical reasons, adequately implement the consolidated supervision of all local entities of each financial group operating in their jurisdictions. This is true for both local and foreign (regional and extra regional) entities. As a result, it is extremely hard for supervisors to enforce standard limits on minimum-provisioning and related-party lending. Their attempts to mitigate this shortcoming lead to complementary, and quite frequently distortionary, regulations. Similarly, most supervisory practices are still backward-looking and compliance-based. The capacity to perform forward-looking, risk-based supervision is still being developed in the region. Until this capacity is developed, authorities deal with the trade-off between stability and efficiency as best they can, which frequently implies higher costs and limited outreach for providers of financial services. Not surprisingly, the main recommendation coming from BCP assessments, repeated in multiple instances, is simply to provide supervisory authorities with the legal powers and the material resources necessary for implementing risk-based, consolidated supervision of the financial groups operating in their respective jurisdictions. • When considering granting a new loan, lenders take into account all possible scenarios they may face if the prospective debtor defaults. The type of contracts that can be enforced when acting individually or collectively against a delinquent debtor, and the quality and costs of this enforcement, directly determine the cost and outreach of credit in a particular jurisdiction. The ICR ROSCs reveal that it is still difficult to pledge assets other than real estate property in Central America. Several countries lack functional institutions that can deal with corporate insolvency without destroying potentially viable corporations. The ICR assessments recommend the creation of new enforceable contracts, such as pledges on moveable assets, and the enactment of modern legal frameworks for dealing with corporate insolvency. • While considerable regional progress has been made in modernizing national payments systems, the CPSIS ROSCs identified several persistent shortcomings. The finality of payments by means other than cash is not yet assured in most Central American jurisdictions, and not all central banks have been granted authority to adequately regulate and systemically supervise important payment and securities-settlement systems. The ROSCs found that the interconnection among different payments systems within each jurisdiction is usually far from fully functional. In addition, the interconnection of payments and settlement systems among different jurisdictions has lagged even further. The CPSIS assessments generally recommend: (1) granting adequate supervisory and regulatory powers to central banks; (2) creating a consistent legal and regulatory framework that incorporates payments system rules and bank and corporate insolvency procedures; and (3) and investing in appropriate payments and securities-settlement technologies. • Finally, a primary tenet of financial systems is that lending cannot take place without adequate financial information on prospective borrowers. Such information depends on 89 standard and trustworthy financial reports. The A&A ROSCs describe an uneven approach toward compliance with international financial reporting standards and a lack of adequate training and policing of practitioners. The A&A assessments propose concrete steps to improve the quality of the accounting profession in the region. 1. CENTRAL AMERICA’S FINANCIAL SECTOR 5.7 This section presents a succinct overview of the banking sector in Central America, using aggregate information on financial services for the period 2000-2006. It does not intend to provide an analysis of performance but rather to highlight the main features that may affect banking sector outreach. It then examines traditional indicators of domestic financial development like credit or deposits to GDP, comparing the Central America countries and identifying trends. 5.8 Figure 5.1 presents some indicators for Central America’s banking sector in 2000 and 2006. The region had 151 institutions with 4,631 branches in 2006. Assets added up to $95.2 billion in 2006, with about 47 percent in Panama, followed by 15 percent in Guatemala and 14 percent in Costa Rica. Since 2000, banking sector assets in Guatemala, Honduras, and Costa Rica have grown the fastest, doubling in size (in current US dollars). The banking system has gone through a process of consolidation (Morales and Shipke, 2005; Fitch Ratings, 2008). The number of banks contracted from 172 to 151 between 2000 and 2006. Costa Rica and Guatemala have seen the largest reduction in the number of banks. This consolidation process has resulted in increased concentration of bank assets in fewer banks in all countries. The share of the top five banks in total assets, an indicator of banking concentration, rose in all countries. The largest increase was in Guatemala, which went through a series of mergers as a result of bank restructuring, followed by Nicaragua. 5.9 Bank consolidation has been accompanied by an increased presence of foreign banks. Regional financial conglomerates, other financial groups, and foreign banks hold half of Central America’s total bank assets. According to a recent Fitch Rating report, foreign banks control more than 80 percent of El Salvador’s assets, and Panama is second with 40 percent. Guatemala, Nicaragua, Honduras, and Costa Rica have much smaller shares of foreign bank ownership. Factors contributing to increased participation of foreign players likely include prospects of the Dominican Republic-Central American Free Trade Agreement (DR-CAFTA), a growing remittances market, and sound macroeconomic conditions. 5.10 An increased presence of foreign banks may raise questions regarding the availability of lending for small and medium-size enterprises and banking sector outreach. However, evidence from other Latin American countries suggests that foreign bank entry has been associated with increased lending for all firm sizes—although large firms benefited the most (Clarke, Cull, and Martinez Peria). In terms of the geographical distribution of branches and loans and deposits, Beck and Martinez Peria (2007) looked at Mexico, finding that the number of municipalities with bank presence increased after foreign bank entry, but only very rich and urban areas experienced an increase in branches. This latter finding may raise concern for the Central American countries with a growing presence of foreign banks. 90 Figure.5.1: Indicators of the Banking Sector in Central America, 2000 & 2006 A. Number of Banks in Central America 2000 2006 90 75 77 Number of Banks 60 32 30 22 24 21 16 16 14 11 8 7 - Costa Rica El Salvador Guatemala Honduras Nicaragua Panama B. Regional Share of Banking Sector Assets Costa Rica Costa Rica 10% 14% El Salvador 13% El Salvador 12% Panama Guatemala 47% Panama 10% 58% Honduras 6% Guatemala 15% Nicaragua 3% Nicaragua Honduras 3% 9% C. Assets of Five Largest Banks / Total Assets 2000 2006 120 91 95 90 76 77 81 64 65 67 Percent 56 60 41 39 45 30 - Costa Rica El Salvador Guatemala Honduras Nicaragua Panama Source: SECMCA (Secretaría Ejecutiva del Consejo Monetario Centroamericano) 91 Table 5.1: Financial depth indicators in Latin America, 2006 Bank Deposits Bank Lending to Stock Market Country / GDP Private Sector / GDP Capitalization / GDP /a Costa Rica 0.44 0.35 0.07 El Salvador 0.37 0.42 0.19 Guatemala 0.32 0.25 n.a. Honduras 0.48 0.42 n.a. Nicaragua /b 0.33 0.24 n.a. Panama /b 0.71 0.74 0.27 Central America average 0.44 0.40 0.18 Argentina 0.21 0.11 0.33 Bolivia 0.34 0.35 0.22 Brazil 0.52 0.32 0.56 Chile 0.47 0.62 1.07 Colombia 0.26 0.22 0.38 Dominican Republic 0.28 0.19 n.a. Ecuador 0.22 0.22 0.09 Guyana 0.85 0.42 0.20 Haiti 0.30 0.13 n.a. Jamaica 0.44 0.24 1.21 Mexico 0.23 0.17 0.35 Paraguay 0.17 0.16 0.03 Peru 0.21 0.17 0.52 Uruguay 0.43 0.25 0.02 Rest of LAC average 0.35 0.26 0.42 a/ Data as of 2005 for Central American countries, Bolivia, Guyana, Paraguay, and Uruguay b/ Data as of 2005. Source: Financial Structure Dataset Figure.5.2: Trends in Financial Sector Depth, 2000 - 2006 Bank Lending to Private Sector / GDP Bank Deposit / GDP 2000 2006 2000 2006 - 0.3 0.5 0.8 1.0 - 0.3 0.5 0.8 Costa Rica 0.44 Costa Rica 0.35 El Salvador 0.37 El Salvador 0.42 Guatemala 0.32 Guatemala 0.25 Honduras 0.48 Honduras 0.42 Nicaragua 0.33 Nicaragua 0.24 Panama 0.71 Panama 0.74 CA region 0.40 CA region 0.40 LAC region 0.36 LAC region 0.28 Source: WDI 5.11 A number of aggregate indicators are commonly used to measure financial depth. They include bank credit as a share of GDP, the ratio of bank deposits to GDP, and stock market capitalization as a percent of GDP. One caveat in measuring financial depth with domestic 92 banking indicators is that the integration of financial markets allows residents to borrow from foreign sources. Another qualification involves the off-shore banking sector, which according to some estimates amounts to 50 percent of deposits and 62 percent of lending (Rivera and Rodriguez, 2006). We excluded non-banks because they account for a small share of financial sector assets. 5.12 According to data from International Financial Statistics (IFS), Central American countries seem to have better indicators of banking sector depth than the rest of Latin America (Table 5.1). Bank deposits amounted to 44 percent of GDP in Central America, compared with 40 percent in Latin American. Central America’s bank credit exceeded 35 percent of GDP, well above Latin America’s 26 percent. Nevertheless, Central American countries exhibit wide variation in banking depth indicators. Due to its status as a banking center, Panama ranks at the top of the list, while Guatemala and Nicaragua rank at the bottom, with figures similar to the rest of Latin American. Capital market development in Central America, measured by the ratio of stock market capitalization to GDP, is lower than the rest of Latin American. Among Central American countries, Panama with 27 percent and El Salvador with 17 percent have the best performance; Costa Rica has the worst with 7 percent. Table .5.2: Business environment indicators that can influence financial depth Getting Credit Enforcing Contracts (1) (2) (3) (4) (5) (6) (7) Public Private Region or Economy Legal Credit registry bureau Procedure Duration Cost (% Rights Informatio coverage (% coverage s (number) (days) of claim) Index n Index adults) (% adults) Costa Rica 4.0 5.0 6.1 52.7 40.0 877.0 24.3 El Salvador 3.0 6.0 17.2 74.6 30.0 786.0 19.2 Guatemala 3.0 5.0 20.7 13.1 28.0 1,459.0 26.5 Honduras 6.0 6.0 12.7 58.0 45.0 480.0 30.4 Nicaragua 3.0 5.0 14.8 100.0 35.0 540.0 26.8 Panama 6.0 6.0 - 41.6 31.0 686.0 50.0 Average - Central America 4.2 5.5 11.9 56.7 34.8 804.7 29.5 Latin America & Caribbean 4.1 3.4 8.1 32.1 37.7 754.0 30.3 East Asia & Pacific 4.5 1.9 5.5 10.8 37.3 549.8 47.8 Eastern Europe & Central Asia 5.6 3.4 2.4 15.4 35.9 443.0 22.7 Middle East & North Africa 3.7 2.6 3.6 8.1 42.5 659.3 23.7 OECD 6.4 4.8 8.6 59.3 31.3 443.3 17.7 South Asia 3.9 1.9 0.7 1.9 43.5 1,047.1 27.2 Sub-Saharan Africa 4.0 1.3 2.1 4.5 39.4 643.0 48.7 (1) Legal Rights Index, which measures the degree to which collateral and bankruptcy laws facilitate lending. The index ranges from 0 to 10, with higher scores indicating that collateral and bankruptcy laws are better designed to expand access to credit. (2) Credit Information Index measures rules affecting the scope, accessibility and quality of credit information available through either public or private credit registries. The index ranges from 0 to 6, with higher values indicating that more credit information is available from either a public registry or a private bureau to facilitate lending decisions. (5) Number of procedures from the moment the plaintiff files a lawsuit in court until the moment of payment. (6) Time in calendar days to resolve the dispute. (7) Cost in court fees and attorney fees, where the use of attorneys is mandatory or common, expressed as a percentage of the debt value. Source: Doing Business 93 5.13 Since 2000, financial sector depth has remained fairly stable in Central America, with the Central America average for bank lending hovering around 40 percent and bank deposits showing a slight improvement (Figure 5.2). However, differences appear in country-level performance. From 2000 to 2006, financial depth performance improved in Costa Rica, Guatemala, and Honduras, while deterioration took place in El Salvador, Panama, and Nicaragua. 5.14 An assessment of the determinants of banking depth is beyond the scope of this Table .5.3: Indicators of branch intensity across section, it is interesting and relevant to countries, 2003 consider the likely impact of institutions and Country Branches per Branches per 1,000 100,000 people sq km the business environment. Table 5.2 Costa Rica 9.6 7.5 compares a number of business environment 4.6 14.6 El Salvador variables known to influence credit market 10.1 11.5 Guatemala development. In particular, it shows Doing Honduras 7.4 4.6 Business report data on Getting Credit and Nicaragua 2.8 1.3 Enforcing Contracts. Central America stands Panama 12.9 5.2 out with better credit information index and Central America 7.9 7.4 average coverage of their public and private credit Argentina 10.0 1.4 bureaus. Nevertheless, the region still needs 14.7 1.7 Belize improvement in the area of legal rights and Bolivia 1.5 0.1 enforcing contracts. Brazil 14.6 3.0 Chile 9.4 2.0 5.15 Financial sector depth indicators do Colombia 8.7 3.7 not shed light on the accessibility, outreach, Dominican 6.0 10.8 Republic or breadth of financial services in an Ecuador 9.3 4.4 economy. Access to financial services is Guyana 3.1 0.1 considered important for economic Mexico 7.6 4.1 development and income inequalities. In a Peru 4.2 0.9 World Bank study, Demirguc-Kunt, Beck Venezuela 4.4 1.3 Uruguay 6.4 1.2 and Honohan (2008) summarized recent Rest of LAC findings on access to finance and its links to average 7.7 2.7 economic growth, poverty, and income Source: Beck, Demirguc-Kunt, and Martinez Peria (2006) inequality. Without access to financial services, enterprises have to rely on internal resources for their investment needs and individuals have to use their own wealth to invest in education, pay for health-care needs, or finance promising business opportunities. Access to finance also matters for business creation. To start their operations, new businesses often need to tap external sources of finance. Another argument for the importance of access to finance is its ability to reduce income inequality. A recent case study on access to finance for SMEs in Guatemala and Nicaragua based on household surveys suggests that access problem may be modest (see Box 5.1). 3 5.16 Using a recently constructed database on countries’ access and use of financial services (Beck et al. 2006), Table 5.3 looks at two indicators of branch intensity—branches per people and geographical area. These indicators show that Central America outperforms the rest of LAC 3 Box 5.1 summarizes a background paper prepared for this report. The full paper is available upon request. 94 in terms of the number of branches per 1,000 square kilometers—7.4 versus 2.7, respectively. Among Central American countries, Guatemala and Costa Rica have the best and Nicaragua the worst performance on bank accessibility indicators. While Guatemala has relatively good indicators for financial access, this is mostly indicative of urban areas. Branch network shows modest development in rural sectors. Figure.5.3: Trend in branch intensity in Central America, 2003-2007 Number of Branches 2003 2007 5,000 4,631 4,000 3,000 2,245 2,000 1,000 571 716 363 323 413 - Costa Rica El Salvador Guatemala Honduras Nicaragua Panama CA region Branches per 100,000 people 18 17.0 2003 2007 12.8 12.4 12 10.5 9.5 6.0 6 5.1 - Costa Rica El Salvador Guatemala Honduras Nicaragua Panama CA region Branches per 1,000 sq km 24 2003 2007 20.7 17.5 18 11.2 10.7 12 6.4 5.5 6 2.7 - Costa Rica El Salvador Guatemala Honduras Nicaragua Panama CA region Source : Beck, et al (2005) for 2003 data, Bank Superintendencies' website for 2007, and authors' calculations. 95 5.17 Central American banks have expanded their branch networks in the past four years, improving their bank accessibility indicators. The number of bank branches increased by 54 percent between 2003 and 2007, mainly due to a rapid expansion that added almost 1,000 branches in Guatemala (Figure 5.3). Guatemala accounts for 48 percent of bank branches in Central America, but only for 15 percent of bank assets (see Figure 5.1, panel B). Box 5.2: SME Access to Credit in Guatemala and Nicaragua The study summarized here develops a conceptual framework and offers new statistical evidence on the access to credit by micro, small, and medium enterprises (MSMEs) in Guatemala and Nicaragua. To this end, and after reviewing the existing literature on the topic, it produces new empirical evidence drawn from official household surveys conducted in both countries in 2006. The core contribution of the paper lies in the critical revision of three pieces of common knowledge, namely: (1) A large fraction of MSMEs has an excess demand for credit; (2) In the presence of credit market failures, governments must and actually do assist MSMEs in gaining access to loan facilities; and (3) Alternative credit instruments, such as leasing, factoring, microcredit, and third-party guarantee schemes, can be a suitable and massive solution for the lack of financing. Our analysis refutes to a large extent these assertions and advances some basic policy prescriptions that should help improve the resource allocation and impact of specific MSME financial programs. A major statement of the study summarized here is that the conventional wisdom has been excessively concerned about supply over demand factors. The literature (see, for example, Levine, 2005; IDB, 2005; and Bebczuk, 2007) enumerates reasons why banks and other lenders would be indisposed to extend credit to this segment of firms, implicitly suggesting that a large number of MSMEs firms have good investment opportunities and would be willing to borrow in order to put such projects in motion, but they are unable to do so because credit is too costly or unavailable. This belief overlooks the demand side of the credit market. From the perspective of the typical MSME owner, a battery of arguments backs up our position that in many circumstances financial debt may be profit- and welfare- reducing vis-à-vis internal funds, inducing in practice a scarce use of debt (see Bebczuk and Garegnani (2006) and the references therein). This invites the question as to what we understand by a financially constrained MSME. Strictly speaking, that would be a firm willing to use internal funds to undertake profitable projects but, lacking those funds, is unable to obtain capital at a similar cost (or at any cost, for that matter). In the case of Guatemala, the econometric analysis shows that the probability of applying for a loan increases with education, income and remittances. The probability of obtaining a loan depends significantly of gender (women have more chance of being granted one) and house ownership, but income and education, which came up as potential explanations earlier, have no bearing in the lender’s decision after controlling for other factors. This might be a consequence of self-selection (all applicants share homogeneous features) combined with the presence of unobservable characteristics driving the lender’s rejection –for example, lack of a well documented business plan or credit history. In turn, further analysis suggests that the probability of not having demand for a loan (preference for internal funding plus no need for credit) as the main reason for not applying goes up with education, marriage, urban residence, income, house ownership, and non-agricultural occupation. For Nicaragua, the probability of having a loan appears positively and significantly (at 5% or less) associated to education, income, and remittances. Additionally, married firm owners are more likely to have a loan. Note: The full paper is available upon request. Source: Bebczuk (2008) 2. FINANCIAL SECTOR REGULATION AND SUPERVISON 5.18 The growing sophistication and globalization of finance is creating both new opportunities and new risks. Central America has not been isolated from recent changes, and countries are at different stages of financial sector reforms to cope with the complexity of today’s financial practices. Financial groups have become increasingly dominant within Central America, leading to rapid growth in cross-border operations. These groups encompass a variety 96 of corporate structures and have been able to expand for several reasons—growing depositor confidence, improved efficiency, Panama’s role as a global financial center, and the incentives to diversify operations across the region provided by political instability. 5.19 This section looks at Central America’s compliance with the Basel Core Principles (BCP) as a way to assess the status of regional financial sector reforms. The analysis is based on various BCP compliance assessments conducted by the World Bank and IMF between 2002 and 2007. Findings are not reported for individual countries for reasons of confidentiality. 5.20 The analysis reveals substantial progress and recommends further regulatory and supervisory reforms. It finds persistent areas of weaknesses in individual countries. Progress has been hindered by the slow pace of legal and regulatory change and weaknesses in supervisory authorities’ technical capacity and resources. Challenges are especially acute in the effective implementation of risk-based supervision, consolidated supervision, and harmonized standards on anti-money laundering (AML) and countering the financing of terrorism (CFT). Supervisory Powers 5.21 The region does well with regard to the definition of responsibilities in banking supervision, although some weaknesses remain. Prudential regulations specify the minimum standards that banks are required to meet, though legal amendments to these regulations in Central America have not been keeping pace with financial market developments. In particular, the growth of cross–border operations and the diversification of financial products require laws that authorize regulators to supervise foreign subsidiaries and subsidiaries engaged in leasing, factoring, and credit-card lending. The current system of strong confidentiality requirements restricts cooperation and information sharing among supervisors on a regional basis. These deficiencies can prevent a supervisory authority from effective consolidated supervision, even where the legal framework for domestic supervision of credit institutions is comprehensive. While regional authorities have made some progress, supervisory arrangements have lagged behind financial groups’ development of cross-border operations under parallel structures. Supervisory arrangements should be strengthened regarding consolidation of accounts and consolidated monitoring of compliance with prudential standards, and cooperation between home- and host-country regulators. 5.22 In Central America, regional financial conglomerates, other financial groups, and foreign banks hold half of total bank assets. Lack of effective consolidated supervision exposes the system to multiple risks. Uneven treatment of credit risk, especially large credit exposures, in different countries creates opportunities for regulatory arbitrage. Cross-border deposits can amplify contagion effects through the transmission of liquidity risk. Faulty reporting of intra- group transactions and failure to maintain capital adequacy at the group and entity levels is likely given cross-country differences in actual and required ratios, measurement practices, accounting standards, and auditing consolidation. Outsourcing of contracts to related parties may generate non-transparencies in pricing. Asset dumping within the group (and off-shore) may lead to capital inflation as well. 5.23 Current legal frameworks are not adequate to respond to these risks and should be strengthened. By and large, they rely on quantitative assessments such as CAMEL (sometimes 97 CAMELS) ratings. The acronym stand for an international bank-rating system that supervisory authorities use to rate institutions according to the following factors: C - capital adequacy, A - asset quality, M - management quality, E – earnings, L – liquidity, and S - sensitivity to market risk. Even where qualitative opinions are permitted, these are not a basis for intervention, leading to insufficient early warning signals and reliance on punitive or corrective rather than preventative action. Supervisors in the region are also hindered by lack of legal protection and by the fact that interventions are often reversible on appeal. Judicial proceedings, even imprisonment of supervisors, remains a disincentive in several countries in the region, with no provision for reimbursement of legal expenses and no legal distinction between inadvertent policy errors and criminal behavior. 5.24 Regionally, compliance with recommended licensing criteria is improving but still weak. In particular, the following aspects of the licensing process need to be formalized into regulations: • Formal identification and evaluation of the ultimate beneficial owners of the bank; • Assessment of the transparency of the bank’s ownership structure and its wider group. Authorities should collect information on all shareholders’ related parties prior to authorizing a license, as well as on exact sources of capital; • Assessment of policies regarding risk management and internal controls; • Assessment to ensure that a bank’s board collectively understand the risks of the activities the institution intends to conduct. 5.25 Additionally, the authorities should have the power to prevent licensing of group configurations that may hamper consolidated supervision. 5.26 In general, ownership of a bank in Central America can change without prior approval of the authorities. The seller merely transfers the shares of the company controlling the bank, leaving the purchaser to inform the authorities of the transaction. Supervisory authorities are unable to object to or reverse such transfers, if deemed undesirable, because the transfers would have been made according to law. To prevent this, laws should be amended to clearly define the concepts of “significant ownership� and “controlling interest.� Prior approval and “fit and proper� assessment should be made applicable to any transfer that would result in a significant change in ownership, beneficial ownership, or the exercise of voting rights of a bank. The authorities should be capable of rejecting any potential ownership transfers if buyers do not meet the fit and proper criteria. 5.27 The region does not perform well on assessments of capital adequacy regulations. Accurate measurement of capital would require the region to progress to fully consolidated supervision, covering all types of institutions in a financial group. In addition, most countries give a zero risk-weighting to government securities and central bank debt in the portfolio. This practice contributes to already insufficient accounting of market risk, especially in countries that issue US dollar-denominated debt. 98 Risk Management 5.28 While some progress has been made, progress on risk management processes is still inadequate and further reform is needed. In general, banking supervision systems in Central America are in the early stages of modernization. In recent years, the authorities have expended a great deal of effort on moving towards risk-based supervision. Countries need to implement a strategy that includes the following: • Streamlining and upgrading the supervisory processes to focus on banking risks and the adequacy of policies, procedures, organization, and information systems to manage these risks; • Implementation of a human resource plan with training, recruitment, and incentives to produce the necessary cultural change and ensure that the new processes are properly internalized by supervisors and that the budget is appropriate to enhanced technical capacity; • Implementation of a system of communication, incentives, and enforcement to ensure that banks also internalize the need to focus on risk; • Issuing risk-specific regulation for market, interest rate, and operational risks, and upgrading regulation on credit risk; • Further developing the supervisory information and reporting systems. 5.29 Supervisors have considerable experience in loan evaluation, but they do not yet conduct a comprehensive assessment of the adequacy and effectiveness of bank strategies, policies, and procedures for the management of credit risk. In the current context of fast credit growth in consumer loans and mortgages, supervisors must establish processes for an appropriately controlled credit-risk environment. In addition, credit-risk evaluation should be specified by law for entities that are part of larger financial groups. Specific regulations should be in place for investments and on-site evaluation of controls. 5.30 Central American regulators have made substantial progress regarding problem assets, provisions, and reserves. However, the following areas need further attention: • Off-balance sheet risk should be subject to classification and reserve requirements; • Guidelines are necessary for restructured loans; • Provisioning should be based not only on minimum requirements, but also on a comprehensive assessment of loan quality, including the borrower’s repayment capacity; • The use of eligible collateral should be well-specified by law. Transfer and liquidation of collateral should be timely, especially if provisioning is based on the loan amount net of the collateral’s value; • Debtor rating should be independent of collateral to ensure that a single borrower does not generate several ratings for multiple operations in an institution or across institutions. 5.31 Lack of consolidated supervision weakens management of large exposure limits even where regulators have made progress with the relevant prudential regulations. In general, Central America lacks limits and guidelines to manage concentrations of credit—by industry sectors, by 99 geographic area, by currency, by collateral, or by any other common characteristic that links otherwise unrelated loans. If it exists at all, monitoring of financial groups’ large exposures is at the incipient stage. 5.32 The absence of consolidated supervision and the dominance of financial groups impede implementation of recommendations relating to connected lending and exposure to related parties. Authorities should issue specific definitions of what constitutes “connected lending� and strengthen the regulatory framework in this context. Due attention should be given to the following provisos: • Bank board members with conflicts of interest should be excluded from the lending approval process; • Persons benefiting from a potential exposure, or persons related to such a person, should be excluded from the process of granting and managing the exposure; • Prior approval of all transactions with related parties should be enforced. 5.33 Some progress has been made on establishing suitable policies and processes that clearly articulate roles and responsibilities related to the identification, measuring, monitoring, and control of market risk. However, authorities should introduce stress testing and periodic testing of systems used to measure this risk. Even where banks are required by corporate governance regulations to establish risk management policies and processes for market risk, supervisors typically do not assess these issues, largely because the regulations governing supervisor responsibilities have not kept pace with developments in market complexity. 5.34 The region was assessed poorly on liquidity risk, with all four countries analyzed receiving an MNC rating. Liquidity guidelines, as included in the overall CAMEL rating method, in general do not effectively analyze liquidity risk. The focus of liquidity guidelines is usually on short-term monitoring, without due consideration of undrawn commitments and other off- balance sheet and on-balance sheet liabilities. These risks should be regulated. Limits on cash- flow mismatches for foreign currency and stress testing procedures covering liquidity risk should be established. Regulators should require that banks inform authorities in the event of liquidity problems. Given the rapid growth of cross-border transactions in the region, authorities should accelerate the process of implementing consolidated supervision, so they can monitor and control for liquidity risk. 5.35 Regional practice still relies on reviews of internal controls rather than operational risk management activities. The region does not have a requirement for a comprehensive assessment of losses due to operational risk or for adjusting reserves to account for these losses. Existing regulations often primarily require supervisors to undertake an overall assessment of the CAMEL components of supervised banks and not an analysis of specific operational area risks. Internal controls must adapt to changing circumstances, especially in the capability to identify risks. Specific guidelines should be issued to assist the banks in identification, monitoring, and control of a full range of business risks. 5.36 More thorough operational reviews and interim goals are needed. As opposed to comprehensive regulation on processes to identify, assess, monitor, and mitigate operational risk, current supervision focuses primarily on IT systems, which are reviewed on-site. Inspections 100 review the quality of a bank’s IT contingency plan, but other elements of business resumption and contingency planning are not typically evaluated. Delays in implementation of reforms should be addressed through interim goals and deadlines, which would provide concrete steps for both banks and supervisors. Supervisors should have right of inspection of companies that provide “outsourced� IT support to banks. 5.37 Existing procedures covering interest rate risk are not comprehensive enough. The current practice of on-site inspections generally consists of a review of a bank’s operations, including the interest rates offered on domestic and foreign exchange accounts, core deposits, correspondent bank accounts, and inter-bank activities. In most cases, supervisors do not have qualified staff, nor established procedures, to address interest rate risk. Stress tests should be required to measure banks’ vulnerability to loss under adverse interest rate movements. Regulation should formalize the requirement that bank boards approve, implement, and periodically review the interest-rate risk strategy and policies and processes for identifying, measuring, monitoring, and controlling interest rate risk. 5.38 The region has made substantial progress in the area of internal control and audit, although efforts should be intensified to implement a risk-based approach. In all cases, there should be an appropriate balance in skills and resources of a bank’s back office and control function relative to its front office and business origination activities. Emphasis must be on segregation of functions, crosschecking, and organizational structures. IT supervision and resource analysis are critical to effective compliance. In addition, the internal auditor must have unfettered access to the bank’s business lines and support departments, reporting lines to the board, and separation from the operating and administrative functions. In addition, increased staff capacity, resources, and risk-assessment methodologies are necessary. Supervisory approach 5.39 In general, the current supervisory approach analyzes past performance and may miss incipient risks. Typically, regional supervisors rely heavily on a rigid weighting of CAMEL-type ratios, which should only be used for establishing the rating of individual banks and not to assess potential risk. Supervisors should require banks to notify them of substantive changes in activities or overall conditions, or material adverse developments once management becomes aware of them. On-site and off-site supervisory functions are compliance-based and generally ‘backward-looking’ in their analysis. For example, offsite supervisors monitor and identify trends and developments in the banking system as a whole and prepare summary reports on a monthly basis. Effective risk assessments are not performed. Due diligence should be exercised in reporting on the state of the corporate governance function, including risk management. 5.40 Coordination among regulatory authorities should be strengthened. The increasing complexity of market-based services and the dominance of financial groups suggest the need for formalization of coordination and information sharing between bank supervisors and other regulatory authorities (for example, pensions and insurance). 5.41 Training programs for risk-based supervision are needed. There is lack of qualified staff to assess the risk profiles of individual banks and banking groups. Training programs on risk- based supervision, with a focus on the challenges of consolidated supervision, should be 101 implemented and methodologies developed to facilitate the assessment of the nature, importance, and scope of risks to which individual banks are exposed. Reporting, Accounting, and Disclosure 5.42 Substantial progress has been made on supervisory reporting, although work remains to be done. Supervisors must have access to—and the right to demand—detailed information on banking groups’ subsidiaries. The legal framework should be amended to ensure the quality of external auditing of banks. Adequate investment in IT support and specialists will be necessary to ensure full compliance with the goal of validating information. Regular meetings between the supervisors and internal and external auditors are recommended. 5.43 Current practices require increased coordination and standardization. In the event that a bank’s Chart of Accounts does not cover a particular situation, authorities should rely first on the International Accounting Practices (NIC), followed by GAAP. Accounting practices should be well-defined regarding (a) loan loss provisioning, (b) the valuation of securities, (c) the valuation of property, (d) principles relating to consolidation, and (e) disclosure on sensitive aspects, such as related-party disclosure and business segment performance. Valuation rules should be consistent and prudent and show profits net of appropriate provisions for both credit and investments. 5.44 Supervisors should have increased authority regarding certain accounting and disclosure practices. They should have the legal power to hold bank management and a bank’s board responsible for ensuring the accuracy of annual financial statements. A country’s superintendent of banks should have the authority to reject or rescind the appointment of an external auditor deemed to have inadequate expertise or independence to carry out an external audit. Consolidated Supervision 5.45 The region ranked very poorly on the principle of consolidated supervision. Much remains to be done to mitigate risk. As previously discussed, supervisory authorities have not yet been able to implement an effective regulatory and legal framework, nor system processes and methodologies, for consolidated supervision. Even where consolidated supervision is required by law, the process has not yet begun in practice. There is a general lack of legal authority to supervise the subsidiaries of domestic groups or banks operating abroad (e.g. off-shore banks) as well as domestic financial activities that are not subject to functional supervision (leasing, factoring, credit cards, etc.). The latter is true even when domestic subsidiaries conduct such activities. This situation precludes an adequate appraisal process of risks incurred by financial institutions. Building supervisory skills in this area would require implementation of training programs to bolster existing capacity and possible support from international consultants. 5.46 Supervisory authorities have put in practice regulatory responses at both individual country and regional levels. These include attempts to combine supervision of domestic conglomerates and cross-border financial intermediation; regulatory ring fences that apply stricter prudential regulation to entities belonging to groups not submitting consolidated financial statements; and memoranda of understanding (MOU) within the Central American Council of Superintendents of Banks, Insurance, and Other Financial Institutions. The MOUs are especially 102 important in the Central American context, where insufficient political integration hinders support for the establishment of a common supervisory authority. 5.47 Substantial progress has been made regarding home-host relationships. However, information exchanges with foreign financial supervisors need to be formalized, which will also assist with implementation of consolidated supervision. This is especially important given the rapid growth in cross-border financial transactions within the region. There is need for harmonization of supervisory systems and practices relating to anti-money laundering (AML) and countering the financing of terrorism (CFT). 3. CREDITOR RIGHTS AND INSOLVENCY 5.48 This section offers a brief summary of findings and recommendations regarding creditor rights and insolvency systems, resulting from Insolvency and Creditor Rights (ICR) ROSC 4 and FSAP assessments conducted in Costa Rica (2007), El Salvador (2004), Guatemala (2006), Honduras (2003) and Nicaragua (2002). 5 The present summary of the insolvency and creditor rights legal and institutional framework and practice in the cited countries, at the time of the respective studies, has been pursued over five years. Since the respective ICR ROSC and/or FSAP reports have been delivered, some countries have addressed specific issues, enacting legislative reforms generally in line with the documents’ recommendations. Examples include: the new secured transactions law (movable assets) in Guatemala (2007), real estate registry improvement in Honduras (2004), and the Codes of Civil Procedures in El Salvador (2006) and Honduras (2007). 6 Other significant reforms are currently being considered, including new insolvency and secured transactions laws in El Salvador and Honduras. Legal Framework for Creditor Rights 5.49 The UN Commission on Legal Empowerment of the Poor has very recently (June 2008) argued that “four billion people around the world are robbed of the chance to better their lives and climb out of poverty, because they are excluded from the rule of law�. 7 This Commission does not see the rule of law as a “mere adornment� of development, but as “a vital source of progress�. While the reasons are generally well understood, the Commission very effectively reminds us of the basic economics behind such observation: “Effective and inclusive laws, enforced through well-functioning institutions, bring a host of economic benefits that are so 4 Creditor rights and insolvency systems have been initially benchmarked using the World Bank Principles and Guidelines for Effective Insolvency and Creditor Rights Systems (April 2001): see, www.worldbank.org/gild. Currently the ICR ROSC is conducted based on Principles for Effective Insolvency and Creditor Rights Systems (“Principles�), World Bank, revised version 2005, www.worldbank.org/gild and Legislative Recommendations from the Legislative Guide on Insolvency Law, United Nations Commission for International Trade Law, www.uncitral.org. 5 In El Salvador, Guatemala, Honduras and Nicaragua the ICR ROSC was conducted in parallel with the FSAP. In Costa Rica the main features of the creditor rights and insolvency systems were evaluated in the FSAP report. 6 Guatemala: Ley de Garantías Mobiliarias (Law on Secured Transactions over Movable Assets) - Decreto 51-2007, effective as of January 1, 2008. Honduras: Ley de la Propiedad (Property Law) - Decreto 82-2004, effective as of June 29, 2004. Código Procesal Civil (Civil Procedure Code) - it was passed in 2007. El Salvador: Código Procesal Civil y Mercantil (Civil and Commercial Procedural Code) - The latest and final version of the Project is November 2006 (t has not been passed yet). 7 UN Commission on Legal Empowerment of the Poor, Making the Law Work for Everyone, UNDP, June 2008. 103 fundamental that they are often forgotten. They make transactions easier and cheaper. They foster predictability, security, and trust. They make enforceable long-term contracts between strangers possible. That, in turn, permits a greater specialisation and division of labour, economies of scale, long-distance trade, and essential financial functions such as credit and insurance. Such features mark the difference between a rudimentary economy with a simple pattern of production and exchange and a vastly more complex and productive developed economy.� 8 5.50 Given the critical importance of the legal framework in making economic and social progress possible, the Commission declares that “the legal empowerment agenda must be integrated as a core concern of global multilateral agencies such as the World Bank, UNDP, ILO, FAO and UN-HABITAT.� The World Bank has done so since many years ago, and it has developed key instrument to evaluate the legal framework within which transactions take place: the Insolvency and Creditors Rights (ICR) ROSC. Under the ICR ROSC exercise and after an exhaustive information gathering process, two final deliverables are produced for the client country: (1) A principle by principle analysis of the insolvency and creditors’ rights system, based on the unified standard encompassing the Principles for Effective Insolvency and Creditor Rights Systems (World Bank) and the Legislative Guide on Insolvency Law (United Nations Commission on International Trade Law, or UNCITRAL). This involves, inter alia, an analysis outlining where the local insolvency regime is consistent with best practices, where it is not so and recommendations, by principle, of what might be done to improve the system; and (2) A Summary of the overall findings and recommendations. This section summarizes the findings and recommendations of the ICR ROSC for Central American countries. 5.51 Central American legal frameworks for creditor rights are generally inconsistent with a modern system. Nor are they in accordance with international best practices. The existing legal frameworks do not meet the current needs and expectations of market participants. Some frameworks appear obsolete, and require urgent updating, the exception being Costa Rica’s. Though some areas of security legislation are broadly consistent with best practices, particularly security interests over real estate, the systems as a whole are perceived as limited, rather rigid, and unreliable. This lack of clearly defined rights limits access to credit by debtors who may only offer movable assets as collateral. A broader use of security interests over movables should be welcomed. Limited use of insolvency proceedings makes it difficult to assess the balance between commercial enforcement and insolvency. 5.52 The secured lending frameworks could be further improved to enhance legal certainty and foster access to credit. Financial credit is mostly secured credit. Traditional security interests such as mortgages (hipotecas) are broadly used. Collateral systems other than mortgages are insufficiently developed, not widely accepted, and underutilized. The exception is Costa Rica, where leasing contracts and “guarantee trusts� have grown significantly in banking practices. In some countries, non-possessory pledges are utilized, but with different degrees of frequency; in others, they have limited use due to registration flaws or registration unavailability. 5.53 Affordability, reliability, and predictability of the creation and registration processes of secured transactions vary from country to country. Legal frameworks for creating, recognizing, 8 ibid, page 48. 104 and enforcing security interests on immovable assets are relatively well developed. Some real estate and movable assets registries are perceived as efficient and reliable (Costa Rica, Guatemala, and El Salvador), 9 while others still require significant improvement because shortcomings in the registration systems weaken property and creditor rights. The relatively recent upgrading of registration in El Salvador and Guatemala has increased transparency and access to information. Furthermore, some countries lack an updated real estate survey (catastro) coordinated with the real estate registry. Guatemala has made efforts in this regard, though coordination difficulties still cause uncertainty about immovable property. The legal frameworks lack sufficient range and/or flexibility to broaden the spectrum of movables that could be offered as pledges, generally requiring an “identifiable� asset for registration. 5.54 Commercial enforcement systems rely heavily on the judiciary, with infrequent extrajudicial execution. The sluggishness of judicial procedures is generally perceived as a drawback for effective enforcement and provides for limited recovery by creditors. The extremely slow pace is usually explained by: (1) the antiquated or cumbersome procedural rules in force 10, and (2) overburdened and backlogged courts due to the excessive number of cases and lack of adequate training and facilities, with specific exceptions for some large cities in such countries as Nicaragua and Guatemala. Alternative dispute resolution (ADR) mechanisms are rarely used, even after the enactment of recent legislation allowing ADR in some countries. 5.55 Unpredictability in debt recovery is the practical result of some current procedural rules and/or the procedural attitude of the parties. Executive proceedings are available for secured and some unsecured credit in such countries as El Salvador and Nicaragua. In practice, however, these proceedings can be protracted and the outcomes challenged, hindering debt recovery. Ordinary proceedings are lengthy and inefficient. In some countries, all proceedings suffer from abuses of defenses and excessive appeals, or the unfair and frequent use of the constitutional injunction (amparo). Legal Framework for Corporate Insolvency 5.56 Limited use of out-of-court corporate arrangements (workouts) may be the result of the legal frameworks´ lack of adequate incentives and their failure to create an environment that would encourage participants to negotiate informal agreements. Formal insolvency proceedings are scarcely known, little used, and do not represent a compelling incentive to engage in alternative, less formal negotiations. Absence of legal incentives includes inefficient insolvency frameworks, lack of neutral tax treatment for restructurings, or inadequate norms related to the efficacy of these workouts in case of the debtor’s subsequent bankruptcy. Furthermore, there are no institutions or methods for supervising an informal process that would guarantee a fair negotiation environment. Finally, an adversarial culture prevails in most countries. Corporate credit is almost always secured, with creditors preferring foreclosure to facilitating a distressed debtor’s reorganization. 9 In Honduras, improvements to the registry system have been introduced after the ICR ROSC. 10 In October 2007 the Parliament approved a new Law on Execution (effective after six months) that would address the weaknesses of the current procedural rules in Costa Rica. In El Salvador a new Code of Civil Procedure was enacted after the ICR ROSC. 105 5.57 The insolvency system receives little practical use and must be updated to meet modern business needs. Although existing insolvency legislation contains a number of sound features, the most remarkable finding of the assessments is its lack of practical utilization across the region. The general perception is that insolvency proceedings are complicated, unpredictable, ineffective, antiquated, and obsolete, or even incompatible with the country’s broader legal and commercial systems. Nevertheless, the legal features discussed below deserve attention. 5.58 Insolvency legislation is fragmented, causing legal inconsistencies, making interpretation unpredictable, and failing to provide an efficient, orderly, and fair mechanism to achieve a collective resolution of insolvencies. Substantive and procedural legislation often empowers multiple legal bodies, which then simultaneously regulate different aspects of commercial insolvency proceedings. As a whole, existing insolvency laws fail to provide an equitable mechanism to resolve a debtor’s financial difficulties or maximize assets’ value. Furthermore, current laws do not strike a careful balance between liquidation and reorganization. In some cases, the creditors’ rights system and the insolvency system are not designed to work in harmony. 5.59 Bankruptcy implies significant stigma, discouraging the early use of either reorganization or liquidation mechanisms. The repressive nature of insolvency proceedings conspires against their use as a mechanism for liquidation and distribution of assets, or as a rehabilitation method for viable businesses. In Guatemala, a bankruptcy declaration includes a detention order for the debtor. 5.60 Treatment of stakeholders’ rights and priorities may hinder the possibilities of predictable liquidations or successful reorganizations. Mortgages and pledges usually are not stayed by commencement of insolvency proceedings. Absence of such moratoria may hinder reorganization prospects when certain encumbered assets are indispensable for corporate activity, contributing to the scarce use of the system. Although legal fragmentation of priority regulation may reduce certainty, pre-insolvency security rights are usually respected. In Honduras, however, mortgages and pledges are substantially eroded, since they rank lower than labor claims. In Nicaragua, the confusing scheme that gives preferential status to a large number of claims changes the legitimate expectations of creditors and may impede the efficient rehabilitation or liquidation of troubled enterprises. Reorganization proceedings 5.61 Reorganization proceedings are rarely used, and they are perceived as inadequate schemes for the real-time recovery of viable distressed or insolvent companies. Although some legal aspects of current reorganization proceedings are in theory consistent with best practices, the extremely scarce actual experience makes an implementation assessment difficult and underlines the urgent need for legislative reform. The current environment fails to strike a balance between liquidation and reorganization, permitting dismemberment of enterprises. The following country-specific shortcomings deserve particular attention: • Access to reorganization proceedings is excessively difficult (Honduras); • The lack of a stay in all legal actions means insufficient protection for the creditors willing to participate (Nicaragua); 106 • No sound form of priority is contemplated for meeting a debtor’s ongoing business needs (El Salvador, Guatemala, Honduras, and Nicaragua); • Requirements for debtor-filed information or its reliability is insufficient (El Salvador, Nicaragua); • The content and nature of the reorganization plan and classification of creditors for proposal or voting purposes are legally constraining and lack flexibility (Guatemala, Honduras); • There is an absence of rules on implementation (Guatemala, Nicaragua) and amendment of judicially-approved reorganization plans (Guatemala, Honduras, Nicaragua). International considerations 5.62 Central American countries are ill-equipped to deal with cross-border insolvency cases. Existing legal frameworks do not establish a comprehensive regime for cross-border insolvency. Present regulations are inefficient. They are inadequate for dealing with the complex and increasing number of problems currently caused by most cross-border insolvency cases. The countries are part of the Private International Law Code (Código de Derecho Internacional Privado), known as Havana Bustamante Code (or Código Bustamante), signed in Havana in 1928. The Code’s rules on insolvency proceedings are considered outdated, and its scope is limited to adopting countries. Rules of jurisdiction, recognition of foreign judgments, cooperation among courts in different countries, and choice of law are not materially consistent with international best practices. Institutional considerations 5.63 Creditor rights and insolvency law implementation are often hindered by shortcomings in institutional frameworks. Although the role of the insolvency judge is generally consistent with international best practice in some countries; overall, it is difficult to assess the judge’s non- intrusive and supervisory role because insolvency proceedings have rarely been used. The general perception is that parties avoid the judicial treatment of insolvencies, due in part to the perception of procedural sluggishness. 5.64 Judicial organizational and performance standards may be inadequate to deal effectively with corporate insolvency cases and meet commercial needs. Except for the few, specialized commercial and/or bankruptcy courts in such cities as San Salvador, San Jose de Costa Rica, commercial enforcement and insolvency cases are handled by courts of general (civil and commercial) jurisdiction. The system’ users perceive that: (1 judicial selection and qualification does not take into account a candidate’s knowledge and experience in commercial and business law; and (2) the court’s staff is inadequately trained and bureaucratic. Furthermore, there is limited or no continuing education for judges or staff on insolvency, commercial law, and business issues. Most training efforts focus on procedural and criminal law issues. Lack of formal court evaluations and performance improvement programs is another problem that should be addressed. In general, no clear and objectively established standards are in place for competency assessments of court services and judges. 107 5.65 Widespread opinion in the region holds that existing procedural legal frameworks are inadequate to meet current business needs. Operational systems for processing commercial enforcement and insolvency cases are generally perceived as non-functional, sluggish, excessively formal, and lacking options for real-time solutions. Abuses of defenses or improper use of procedural mechanisms—particularly constitutional injunctions—tend to be tolerated by courts. In addition, courts are backlogged and overburdened, and court facilities and equipment are deficient, with specific exceptions for some large cities. Court administration rules are deemed unsuitable, with no publicly available rules in place for court operation or case management. Court decisions on commercial law and creditor rights issues are insufficiently disseminated to the public. 5.66 Due to the limited use of insolvency proceedings, it is not possible to ascertain the integrity of the system. Despite the perception of recent improvements, users express concerns about insufficient objectivity, political influence, and/or cronyism in judicial appointments and promotions in some countries, or judicial vulnerability to undue political influence in others. Undue favoritism and corruption issues affecting court officials and some judges have also been mentioned. In some countries, the continuing inadequacy of physical security for judges is a concern. Finally, judicial budgets’ insufficiency has been an issue, even in countries with constitutional norms granting a minimum budget. Regulatory considerations 5.67 Although legal frameworks contain provisions regulating the conduct of insolvency administrators, the system is often insufficiently developed and vulnerable due to lack of oversight or absence of adequate eligibility requirements. Regulatory bodies have not been created or are not involved in the process. While some countries have no particular standard to supplement legislation, others have extensive legal provisions but their practical observance remains dubious because of extremely limited experience. Eligibility requirements to serve as insolvency administrators or representatives do not include any specific proof of competence or expertise. 11 4. PAYMENT SYSTEMS 12 5.68 In recent years, Central America’s central banks have played an active role in ongoing reforms of national payment systems. These programs have resulted in better integration between central banks and individual banks, resulting in reduced settlement lags on accounts individual banks hold at the central bank. In some countries, these positive results have been achieved despite instability in the financial sector. While it is important to acknowledge these significant achievements, central banks still need to consolidate reforms and consider broadening reform efforts in additional areas. 11 Guatemala requires a merchant from the Commercial Registry and has no specific roster of qualified persons. El Salvador, Honduras, and Nicaragua require lawyers. In Nicaragua, the statute is silent as the required experience or training. 12 This chapter is a summary of a work done by Mario Guadamillas and Massimo Cirasino (World Bank) for the publication titled: “Central America: Structural Foundations for Regional Financial Integration� 108 5.69 To make the system sound and efficient, a payment system must be supported by a clear and robust legal framework. Except in El Salvador and Panama, Central American laws grant the central bank some authority over the payments system. However, the legal foundation for oversight of clearance and settlement systems is not always solid. For example, laws are often unclear about the scope of the function and the relative roles of the central bank and other authorities. 5.70 The region’s central banks should have the ability to carry out their oversight roles effectively. To this end, central banks should: • Establish appropriate organizational arrangements and staffing; 13 • Ensure that an adequate degree of participant cooperation exists, sufficient to promote and realize the desired organizational and operational arrangements; • Verify that individual payment systems satisfy users’ needs as well as risk and efficiency requirements through appropriate interventions, both at the development stage and during the ongoing system implementation and operational phases; • Define and implement appropriate actions should participants fail to comply with published rules and regulations (e.g., the application of predetermined penalties and sanctions for compliance failures); and • Collect and distribute relevant statistical information to demonstrate how each system is being used and the extent to which they are satisfying end-user and other market needs. Information on material payment system matters should be disclosed in a manner that assures wide dissemination among stakeholders and the general public. 5.71 Central American central banks do not fully observe several of the responsibilities outlined in the Core Principles for Systemically Important Payment Systems regarding payment system oversight. In addition, securities settlement oversight should be strengthened by devoting adequate resources to regulators and establishing an effective cooperative framework with other agencies, self-regulatory organizations, and the private sector. In performing the oversight function and as system operators, central banks and securities regulators should ensure transparency in their policies and conditions for payment services offered. For resolution of conflicts related to payment services, the public should be able to resort to a bank’s ombudsperson, to the central bank or another appropriate supervisor, and to consumer protection agencies 5.72 Provisions are lacking regarding acceptance, irrevocability, and/or settlement finality of payment orders processed by the system. These concepts are especially important in the event of insolvency, where precise legal definitions could limit potential problems. In netting systems, having legal definitions of these concepts would reduce uncertainties and limit systemic liquidity risk from unwinding procedures. Similarly, although in general there are no explicit zero-hour rules, it is not clear if the countries’ courts could revoke pending or already executed operations 13 This includes forming a small unit in charge of payment system oversight to be separated to the extent possible from the units in charge of operating the systems offered by the Central bank. Skills of the staff involved in the function should be as wide as possible and include operational, technical, and policy expertise as well as proficiency in the areas of law and economics. 109 made by the defaulting institution. There is also a general lack of clarity regarding penalties and the conditions and procedures for removing participants from the system. 5.73 Explicit legal recognition of multilateral netting arrangements is not in place, creating legal uncertainty in the event of insolvency. Since netting is used on a broad scale for the settlement of stock exchange transactions, checks, and retail payments, the situation constitutes serious legal risk. This gap might also hamper further development of financial instruments such as derivatives. It is important to protect netting schemes from potential disruption, so that even if a system participant fails during the day, a liquidator cannot unwind settlements occurring on a net basis at a later time in the day. 5.74 Most Central American countries have recently approved laws for electronic documents and signatures. Such laws should be complemented by revisions of existing rules and regulations to ensure that their legal basis is effective. , The law should also apply to the electronic exchange of messages within payment systems operated by the central bank. 5.75 In general, no public or private body is responsible for the resolution of conflicts arising from payment systems’ operations. No provision exists regarding the responsibility of operators in case the system malfunction and, therefore, no rule dealing with possible compensation for those cases. 5.76 The judiciary generally lacks familiarity with the financial sector’s specific legal needs and the implications of the application of certain laws. Focused training programs should be put in place as soon as an overall assessment of the legal framework for the payments system is completed. 5.77 Protection of customer assets under custodian arrangements is not often clearly established. Therefore, assets pledged as collateral by clearing members are not adequately covered by the law, compromising the capacity to execute collateral. Sometimes, the protection of custody arrangements is included for the depository and brokers/dealers but not for other custodians, or vice versa. This limitation could bar potential settlement arrangements from taking place. 14 In the case of public securities, this situation could limit retail market development because beneficial owners are usually not identified in the depository’s accounts. 5.78 Custodian arrangements for government securities (which are often issued in physical form) do not have secure legal support because primary dealers keep the ownership in the registry of the finance ministry or central bank. Thus, it is often uncertain whether the certificate endorsements in subsequent repurchase (repo) operations can be legally considered a proof of ownership. 5.79 Although repo operations are legally defined, their use as guarantees for transactions in the payments system is uncertain. In the region, the legal basis for the pledge is typically included in the civil code but the pledge, as a tool for collateralized operations, is not regulated. Furthermore, there are no rules on the execution of those guarantees in case of a default. This 14 For example, markets with a high volume of transactions sometimes do not maintain sub-accounts with beneficial owner information but only accounts at the participant level. Such settlement systems should be accompanied by a strong legal protection of custody arrangements and supervisory framework. 110 lack of clarity creates an important impediment in the granting of collateralized intraday credit by the central bank for the purpose of settling payments. 5.80 In international practice, additional collateral is normally requested through a margin call if collateral no longer covers in full the obligation to pay back a loan or the value of the securities borrowed. In Central America, the definition of repo leaves no room for this market practice and makes it difficult to use international standard contracts. Margin calls during the contract period might give rise to legal risk if a regional court re-characterizes a repo agreement with interim margin calls as an improper pledge. 5.81 The legal framework covering securities and capital markets is inadequate. Legal basis for securities lending does not exist in many cases; where it does exist, detailed regulations have not been developed. 5.82 National legal frameworks should be strengthened in several specific areas, including the irrevocability of final settlement, protection of the systems in bankruptcy procedures, the legal basis for custody arrangements, the legal definition of repo operations, the legal recognition of multilateral netting arrangements, the legal definition of immobilization and dematerialization of securities (especially public securities), and the legal definition and regulation of central bank oversight powers. From a development viewpoint, stronger frameworks are also needed on the legal basis for collateral pledge and securities lending in all countries except El Salvador and Costa Rica, where the laws contain specific provisions for the creation, regulation, and enforcement of pledges. Due to the variety and importance of these legal issues, passage of separate payments system laws might be advisable in some countries. Interbank Exchange and Settlement Circuits 5.83 A large-value system is the most significant component of the national payment system because it can generate and transmit systemic disturbances to the financial sector. The development of Real Time Gross Settlement (RTGS) systems is one response to prevent these disturbances. The RTGS system can effect final settlement of individual fund transfers on a continuous basis by limiting settlements during the processing day. It can also reduce settlement risks in securities and foreign-exchange transactions by facilitating delivery versus payment (DvP) and payment versus payment (PvP) mechanisms. Large-value and systemically important payment systems in Central America do not fully observe several of the Core Principles for Systemically Important Payment Systems (CPSIPS). Central banks in El Salvador, Guatemala, and Honduras have initiated payment system reforms to improve the safety and efficiency of large-value systems, launch RTGS systems, and reduce the use of checks for large-value settlement. 5.84 Interbank payment methods are in transition to more sophisticated systems. The high value of checks settled in the region’s clearinghouses and their use in interbank payments confirm that these systems should be viewed as SIPS. Central banks should evaluate ways to provide intermediaries with incentives to use the RTGS system instead of checks for interbank transfers. On a positive note, progress has been made in launching RTGS systems in the region. Costa Rica already has a safe and efficient RTGS system, and new systems in line with the 111 CPSIPS are being launched in El Salvador, Honduras, and Guatemala. In Nicaragua, an overhaul of the gross settlement system (Transferencia telefónica segura de fondos, TTS) is under way. Retail Settlement Systems 5.85 New systems to process retail electronic credit and debit instruments have played a major part in spurring efforts to modernize national payment systems. Automated clearinghouses (ACHs) have been launched in some countries—Costa Rica, El Salvador, Guatemala, and Honduras. In January 2007, Honduras launched ACH Pronto, the electronic payment transactions clearing house run by CEPROBAN, a company owned by private banks. In most Central American countries, however, ACH projects are either too slow to keep pace with customer needs or too limited in scope (e.g., the project only focuses on improving check clearing procedures). Central banks should actively support the full deployment of efficient systems to process electronic retail payment instruments trough ACHs. 5.86 The region relies heavily on checks, which is far from optimal from the point of view of efficiency and risk control. Central banks and all stakeholders in the retail arena must work together with a clear strategy to promote the intensive use of retail electronic payment instruments and reduce the importance of checks. Furthermore, efficiency gains could be implemented in the check clearing system, such as a full or partial truncation (for checks under a given value). In Honduras, a new clearing and settlement service for dollar-dominated checks drawn on foreign banks was added to the clearinghouse in August 2007. 5.87 Retail circuits (e.g., ATMs and EFTPOS) are characterized by very low interoperability, resulting in the inefficient use of the current infrastructure. Many of the positive effects of a payment cards system for increased efficiency are not being captured because of the lack of electronic payment instruments for retail transactions. 5.88 Some banks are starting to offer retail payment instruments and services in multiple countries in the region. These efforts should be monitored and supported by central banks and banking supervisors. However, no project is planned or underway to develop a common regional infrastructure in the retail sector. Consideration should be given to fostering the standardization and harmonization in this area to allow for the creation of some form of regional ACH. 5.89 Government payments are a major source of liquidity for the banking system. They provide an opportunity to channel high transaction volumes (i.e. tax collection, salaries, purchase of goods and services, etc.). Government payments should make the ACH project for electronic payment instruments more attractive for potential investors. If coordinated effectively, government payments can facilitate the smooth functioning of the RTGS system being implemented in the region and increase its appeal to participants. Costa Rica, Guatemala, and Panama have implemented or are in the process of implementing projects to integrate the public sector into the national payments system. In Costa Rica, the integration has been particularly successful. In other countries, these projects are stand-alone and not fully consistent with a long- term strategic vision of the payments system. 5.90 The region’s foreign-exchange transactions are not settled on a payment-versus-payment (PvP) basis. Central banks should investigate the possibility of introducing measures to mitigate 112 the risks associated with these operations when PvP is not possible. Foreign-exchange transactions are of particular concern given the large amount of remittances channeled through the region. Currently, a large share of remittances are still channeled through unregulated specialized institutions, with no standards for such aspects as transparency of fees and other charges or the timing of accreditation of funds to end beneficiaries. Commercial banks’ proprietary mechanisms for cross-border payments can become a less costly and convenient alternative for customers. Another possible benefit is that remittances from other Central American countries can now be fully channeled through banks instead of the unregulated companies, so that, among other things, remittances need not be paid in cash. Clearing and Settlement Processes 5.91 Some systems (e.g., in Costa Rica and Panama) have close links between trading and settlement, such as blocking of transactions prior to matching. This procedure ensures the availability of securities but hampers back-to-back transactions and the effective arbitrage between trading and settlement platforms. The blockage also makes the rollover of repos difficult. The development of a new system that allows for the separation of trading and settlement is crucial for the efficient operation of stock exchanges as well as observance of international standards for settlement risk. 5.92 Some stock exchanges (e.g., Costa Rica) play a crucial role in money markets and have difficulties in accommodating different settlement, risk management, and communication needs. Should these stock exchanges eventually evolve into a more traditional role of handling secondary market transactions, the introduction of a standardized settlement cycle will be needed. With the exception of Costa Rica, communication networks do not follow international standards. Although local communication networks work correctly, the networks must be improved to be relevant for cross-border transactions and consistent with international communication procedures. 5.93 Central banks and finance ministries are making efforts to achieve complete dematerialization of government securities. The private sector is undertaking similar efforts. The physical handling of securities is still common in national security settlement systems. The clearing and settlement of securities transactions with physical certificates instead of a transfer via a book-entry system is not only risky but also cumbersome. In addition, such practices are costly and hamper the development of capital markets. 5.94 Systems in El Salvador, Guatemala, Honduras, and Nicaragua still do not settle on a delivery-versus-payment (DvP) basis. Consequently, payments are not necessarily linked to securities transfers and vice versa. An important risk exists and no measures have yet been taken towards its elimination, reduction, and/or mitigation. 5.95 Risk-management tools for covering settlement failures are largely absent. Some systems used in the region do not offer any tools, while others offer tools that are clearly insufficient. Authorities should determine whether risk management tools are robust enough to cover potential failures, especially when taking into account that existing guarantee funds can be used for failures other than those associated with settlement. 113 5.96 Linking settlement of securities and funds would allow stock-exchange transactions to be settled on a DvP basis, eliminating principal risk. The following aspects of current settlement risks need to be improved: achieving full dematerialization and immobilization of securities; establishing DvP procedures; upgrading risk management tools; mitigating credit and liquidity risk in the cash leg settlement (including eliminating the use of checks as a cash asset); providing better access to liquidity for securities settlement system (SSS) participants; and developing comprehensive strategic approaches to the reform of SSSs, as opposed to technology-driven and purely operational reform projects. 5.97 The various plans currently in place for backup sites and disaster recovery facilities should be accelerated. External audits should be undertaken, especially when the systems have been developed domestically and/or the oversight framework is weak. 5.98 Some legal and governance arrangements in the region introduce monopolistic situations that impede the adequate development of some markets (e.g., the money market). Such gaps include high entrance fees, inadequate facilities, and lack of facilities to process intraday repos used by the central bank to provide intraday liquidity. Finally, unresolved conflicts of interest are the main reason for the underdevelopment of basic SSS infrastructures such as depositories. In these cases, under the leadership of securities regulators and central banks, in coordination with finance ministries, a legally sound solution should be agreed with stakeholders to establish the depository function as soon as possible. 5.99 Since the transactions of most regional securities depositories include cross-border risk, the authorities should carefully analyze the inherent risks. Particular attention should be devoted to the multiple-jurisdiction profile of these transactions, especially from a legal and operational perspective. 5. ACCOUNTING AND AUDITING STANDARDS 5.100 This section is a summary of the common findings and recommendations of ROSCs for El Salvador, Honduras, and Guatemala. 15 Analysis of these shared observations may suggest a way forward for regional cooperation on the critical issues that face the accounting and auditing professions in these countries. 5.101 A transition is underway from traditional practices to new international norms. The financial systems in the countries evaluated are still developing and share common characteristics: • Financial groups (conglomerados financieros) are dominant; if the securities market has played a role at all, it has been very limited; • Liquidity is high and banks are seeking opportunities to lend to small and medium- sized enterprises (SMEs); • Banks are highly regulated but non-financial enterprises less so; • Domestic demand for financial information is low; and 15 Diagnostic reviews were carried out from 2005 to 2007. The reports were written and/or published in 2006 and 2007. Some reports remain unpublished and contain information that is considered confidential. 114 • International standards and norms are being integrated into domestic systems, sometimes causing conflict with established practices. 5.102 Change is being accelerated by outside forces. All countries assessed are signatories to DR-CAFTA, and some are pursuing additional trade agreements with the EU and others. International banking groups have been actively pursuing acquisitions in Central America, often introducing their own management methods and practices. Maximizing the opportunities of trade pacts is a key element of the development strategy of these countries. For Central America to realize the full benefits of DR-CAFTA, companies and their banks need to bring their practices and operations up to acceptable international standards. Another significant factor is the wave of international banks’ acquisitions of local financial groups, which should facilitate the region’s transition toward international standards. Common Findings 5.103 Statutory frameworks were generally observed to be inadequate, outdated, or incomplete, causing multiple problems for accounting and auditing functions, and sometimes encouraging non-compliance. Traditionally, vague or ambiguous standards on preparation and filing of financial data have limited the information available to potential investors and other third parties and often increased the burden on small businesses. In one country, for example, regulations require companies to prepare their financial reports according to Generally Accepted Accounting Practices (GAAP) without defining the standard. As a result, a majority of companies prepare their financial statements in accordance with tax rules, which are both fairly detailed and in conflict with International Financial Accounting Standards (IFRS). In El Salvador, companies are required to file financial information with multiple agencies, increasing costs to businesses, duplicating the efforts of the authorities, and encouraging non-compliance. The situation in Honduras is slightly better since the 2004 law defining accounting and auditing standards was promulgated, but small businesses are still held to the same reporting requirements as large companies. 5.104 There is wide variance on which types of companies must be audited, and on where and how much financial information is made public. El Salvador requires that a majority of companies (all joint-stock, or S.A. de C.V.), including SMEs, have their financial statements audited, and large companies 16 are required to have an independent audit. Listed companies, including by law all banks, are required to publish financial information in a newspaper. Information on non-listed companies is available to the public, but only the balance sheet, which does not provide robust information. Requirements in the other countries are minimal, especially for non-listed companies, and very little public information is available. 5.105 Banks usually have distinct reporting and auditing requirements, justified by the higher level of risk they entail. These distinct requirements, which involve additional layers of complexity, have a prudential focus and vary to greater and lesser degrees with IFRS. In all countries analyzed, banks are required to have independent audits, and in El Salvador the 16 Defined by law to be companies with assets in excess of 10 million colones (approx. US$ 1.1 million) or revenues in excess of 5 million colones (approx US$0.6 million) 115 requirements for independent bank auditors are quite stringent. Banks sometimes use accounting standards that are substantially different from IFRS. 5.106 Professional requirements for accountants are low. In El Salvador, they must have a bachelor’s degree in accounting, while in the other countries there are two levels of professional accountants differentiated by academic preparation. One level requires a university degree, while the other only a high school education. In no country is there a requirement for practical experience to register as a public accountant, nor is there a test of professional competence. In Honduras, an accountant with only a high school education can perform independent audits, raising serious concerns regarding quality. Educational preparation is uneven. Although it is possible in all countries to obtain a university degree in accounting, there are no standard curricula and usually little preparation to work with international standards. Up until recently, only one Honduran university—the Universidad Autonoma de Honduras (UNAH)—offered a degree in accounting. There is no entrance examination for UNAH, but the university is making an effort to update its curriculum to meet International Federation of Accountants (IFAC) standards. Private schools have flourished, offering courses and degrees in accounting, but the quality of education in such schools is often below that of public universities. 5.107 Continuing professional development (CPD) requirements are weakly enforced or non- existent. In El Salvador, there is a CPD requirement for accountants; however, it is not adequately monitored and sanctions are weak. Continuing education for professional accountants is not required at all in other countries. Professional associations are usually self-regulating and lack resources, limiting the services they offer members. Professional accounting bodies organize seminars for their membership, covering a variety of professional issues, but these are not substitutes for a CPD program. 5.108 IFRS has been adopted, officially or semi-officially, in the region, but uneven progress has resulted in limited application of the standards. In 2000, El Salvador passed a law mandating both IFRS and ISA, and in December 2004 the country adopted the 2003 version of IFRS. However, progress on the ground has been slow. Currently, the vast majority of licensed auditors do not know the standards well enough to apply them. Companies have not made wide efforts to prepare for applying IFRS, a fact corroborated by a sample review of audited financial statements. The situation is similar in other countries. In one, IFRS was formally adopted by the accounting profession, but companies were not bound by the decision and IFRS is rarely applied. That country’s legal framework also fails to designate an institution to set accounting and auditing standards, further hampering efforts to apply IFRS. In another country, an institution was established to set such standards, but it has neither offices nor full-time staff. 5.109 Preparation for implementation of IFRS has been uniformly inadequate, due in part to the lack of CPD. In general, adoption processes have not included a review of each standard to identify which areas would pose the most significant challenges for domestic enterprises. As previously discussed, accountants and auditors have not received proper training, nor has the private sector made adequate plans for transition to IFRS. 5.110 In general, monitoring of banks and other Central American financial institutions is much stronger than other corporate obligations. In two countries, enforcement is weak for the few listed companies’ reporting requirements, although bank monitoring is stronger and uses a two- 116 tiered system. In El Salvador, three agencies enforce reporting requirements to varying degrees, but cooperation among them could be enhanced. The country’s Superintendence of Corporate Obligations lacks capacity to monitor reporting requirements, and sanctions are weak. 5.111 Official codes of ethics for accountants are converging with international norms, but could be strengthened. In Honduras, both accounting professional organizations (colegios) have codes of ethics, and one has adopted IFAC’s code, although it is unclear how effectively it has been implemented. Another country promulgated an ethics code in 2005 based on the 1998 version of the IFAC code. Still another country's accounting profession has developed its own auditing and ethical standards, which are broadly in line with international norms, although not as comprehensive and thorough. 5.112 Professional associations do not actively monitor the quality of practice among their members. In two countries, the professional associations for accountants are self-regulating and have played a limited role in technical and professional development issues. There is no enforcement of professional standards. In one country, the professional associations’ disciplinary panel has not taken any sanctions against a member in more than 10 years. In El Salvador, the situation is better, and recent steps have been taken to monitor auditing practices through the Accounting and Auditing Oversight Body. Recommendations 5.113 Legal changes should include measures to define, clarify, and/or codify accounting and auditing standards, to clarify the accounting and financial reporting requirements of various types of enterprises, to update and harmonize audit regulation, to strengthen enforcement, and to delineate the roles of and guarantee resources for various agencies involved in the practice or oversight of accounting and auditing. Other recommended changes included increasing public access to financial information and making that information more robust. New legal measures should recognize the need for a simpler yet internationally acceptable accounting and reporting framework for SMEs. Many of these changes support the goal of making the transition toward IFRS possible. 5.114 Although IFRS has been adopted in all countries analyzed, the transition is still ongoing. This suggests the need for evaluation of the differences between current accounting practice and IFRS, subsequent review of each new standard, and adoption of the most current version of IFRS—or, in one case, a simplified IFRS. The recommendations call for detailed transition plans for further implementation, including the extension of deadlines where necessary. The need for training on IFRS standards is acknowledged. In the Honduran financial sector, where a primarily prudential approach to accounting seems justified for the near future, gradual steps are needed toward adoption of IFRS after an appropriate period of transition. 5.115 Update academic curricula in the field of accounting and auditing, and align them with IFAC’s education principles and guidelines. Honduras should update and strengthen the accounting curriculum at UNAH as well as the secondary education for accounting technicians because most practitioners do not have a university education. Similar recommendations were made for other countries. 117 5.116 Establish a licensing certification system and a quality assurance system for public accountants in all countries analyzed. The Honduras ROSC noted that the certification should be of such quality that it could be recognized in the US, pursuant to DR-CAFTA. A quality assurance system, perhaps within the professional associations, for all registered public accountants, was also recommended. 5.117 Fortify the professional associations and establish or strengthen continuing professional development programs. In one country, where no CPD program exists, it was recommended to establish a compulsory CPD program. Another country has such a program, but sanctions for non-compliance are not an effective deterrent. Honduran accountants have no continuing education program, but one is under development. 5.118 The business community should be made aware of the potential benefits of enhanced corporate financial transparency and accountability. The current regional emphasis on competitiveness and foreign investment requires that the private sector’s practices must become more aligned with the expectations of prospective international partners. The ROSCs of two countries recognized the need to engage the private sector in improving accounting and auditing. 5.119 All the ROSCs, to a greater or lesser extent, recommended collaboration among DR- CAFTA signatories on a range of issues. These include IFRS adoption, certification processes, and improving education. Harmonizing the regulatory environment within the DR-CAFTA area would contribute to reducing the cost of doing business for companies operating across that economic zone. Cooperation could also provide capacity-building for accounting technicians, quality assurance for auditors, and harmonizing of audit requirements so that public interest entities would be legally required to present audited financial statements and SMEs would be exempted from audits and unnecessarily burdensome accounting requirements. 5.120 Substantial technical assistance from international donors will be needed to support the ROSC recommendations. Three World Bank supported technical assistance projects were underway at the end of the first quarter of 2008, one dealing with the accounting curriculum improvements in five Central American countries, another in Honduras, and a third in El Salvador. 118 Chapter 6. INNOVATION, SKILLS, AND QUALITY 1 INTRODUCTION 6.1 Innovation is a key driver of economic growth. Studies have shown that differences in technological progress, rather than capital investment or labor growth, account for much of the widening gap between rich and poor countries. For example, Hall and Jones (1999) and Dollar and Wolf (1997) found that differences in total factor productivity (TFP), which are generally associated with technological progress, account for roughly half of the cross-country variations in per capita income and growth. Easterly and Levine (2003) argue that productivity differences largely explain global income gaps. To boost productivity, governments need to do all they can to foster an innovation-friendly business environment. 6.2 This chapter uses the term “innovation� in the broadest sense, going beyond such traditional sources of innovation as research and development (R&D). Innovation also encompasses technology absorption and the adoption and upgrading of existing products and processes. According to this definition, changes that a firm makes to improve a production process to meet a quality standard would be considered innovation. 6.3 An environment that fosters innovation will help countries to take advantage of the Dominican Republic-Central American Free Trade Agreement (DR-CAFTA), the pact among Costa Rica, the Dominican Republic, El Salvador, Guatemala, Honduras, Nicaragua, and the US. The trade deal offers the potential for significant economic gains from cross-border investment, access to export markets, and opportunities to adapt foreign technologies for domestics firms. To fully benefit from DR-CAFTA, well-functioning national innovation systems 2 are needed, as well as complementary actions on the educational front. 3 6.4 Firms need skilled workers to absorb and adapt new technologies. Due to poor education quality, low secondary enrollment rates, and weak worker training systems, Central American countries face varying skill deficits. With the possible exception of Costa Rica, the region has a shortage of appropriately skilled workers with the capacity to innovate and increase productivity by adopting new technologies and processes. 6.5 National systems for measuring and certifying product quality are critical for trade, innovation, and competitiveness. Such systems form a country’s quality infrastructure (QI)—the public and private entities required to establish and implement standardization, inspection, 1 This chapter was prepared by Thomas Haven, with assistance from Jane Hwang. 2 The term “national innovation system� generally refers to an interactive set of national institutions, enterprises and individuals that contributes to the country’s technological and economic development (Agapitova, 2005). Put differently, it is the network of organizations within a country that are directly involved in the creation, diffusion, and use of science and technology, as well as the organizations responsible for the coordination and support of those processes. 3 See De Ferranti et al (2003) and Jaramillo and Lederman (2006). 119 testing, product and system certification, and accreditation. These QI services are necessary to provide evidence that products and services meet requirements imposed by authorities or the marketplace (see Annex 1 for a full description of QI components).4 The services provided by QI institutions help boost the private sector’s competitiveness and facilitate exports. For example, auto-parts manufacturers that meet recognized and verifiable standards can sell their products to car makers abroad. Quality systems are gaining prominence in Central America because they are critical to overcoming technical barriers to trade. These impediments are increasingly likely to be binding constraints as DR-CAFTA eliminates formal barriers. Moreover, quality certification can be a stepping stone to new technologies. Standards embody technology, and they can act as a channel for technology transfer for firms that adopt them. For more information on the importance of QI, see Guasch et al (2007). 6.6 This chapter draws on a wide variety of indicators to benchmark Central America’s performance on innovation input and output indicators, technology adoption, human capital, and the environment for innovation. The data show that Central America underperforms the rest of the Latin American-Caribbean (LAC) region on most measures. Large intra-regional differences exist; namely, Costa Rica and Panama outperform their neighbors on almost all indicators. After benchmarking, the chapter analyzes possible reasons for disappointing innovation levels. Low worker skills and deficient infrastructure to measure and certify product quality emerge as the most probable causes, particularly within the less-developed countries in the region. 6.7 The chapter concludes with key policy recommendations: • Improve the quality of education by institutionalizing assessment systems, enhancing teacher performance, diversifying teaching methodologies, and increasing instructional time. • Increase the effectiveness of technical and vocational education and training by incentivizing nongovernment providers, setting standards and encouraging program accreditation, and providing skills training in priority areas where nongovernment providers are reluctant to invest. • Strengthen quality infrastructure by increasing awareness of governments’ coordination responsibilities, focusing on all interrelated components of the system, and encouraging regional cooperation. • Foster links between supply and demand for innovation by creating incentives for universities and the private sector to collaborate, strengthening public research institutions, and facilitating regional knowledge networks. 1. HOW CENTRAL AMERICAN COUNTRIES PERFORM 6.8 This section analyzes data from a wide variety of sources to benchmark Central America’s performance. Output indictors include patents, scientific publications, and introduction of new products and processes. Among the input indicators are expenditures on R&D, licensing of foreign technology, and use of information and communications technologies 4 This terminology could potentially be confused with high-quality infrastructure, which could refer to good electricity and water supply, roads, etc. However, quality infrastructure in the context of this chapter refers to a completely different aspect of competitiveness. 120 (ICT). Other measures include the availability of human resources, education levels and quality, worker training, and some indicators of the environment for innovation. 6.9 The performance indicators are clearly split between relatively high-performing Costa Rica and Panama and the rest of Central America. The dichotomy holds true for nearly every measure. What varies is whether Costa Rica and Panama outperform the LAC average. For instance, Costa Rican and Panamanian export discoveries and technology adoption rates are high by LAC standards, while R&D and patent levels are below average. Skilled workers as a percentage of the labor force are high in Costa Rica and Panama, but the environment for innovation seems exceptionally good only in Costa Rica. The number of researchers as a percentage of the workforce is low in all Central American countries. Innovation Output Indicators 6.10 Patents and scientific publications. Numbers of patents granted and scientific publications, as well as the introduction of new products and processes, can serve as proxies for innovation levels. A look at patents per million workers in the labor force over the last five years illustrates Central America’s severe underperformance (Figure 6.1). Panama, the region’s best performer, managed only about one-fourth the LAC average. In terms of scientific publications, the performance is mixed, with Costa Rica scoring above the LAC average and Panama publishing at about 75 percent of the LAC average (Figure 6.2). The four other countries are significantly lower. Figure 6.1: Patents Granted to Residents Per Figure 6.2: Scientific Publications as a Percent of Million Labor Force (2000-2005 Average) GDP, 2005 LAC Cos ta Rica 18.9 18.6 Panam a 4.9 LAC 15.1 El Salvador 3.7 Panam a 11.6 Honduras 1.9 Nicaragua 8.2 Guate m ala 1.8 Guate m ala 3.0 Nicaragua 1.1 Honduras 2.3 Cos ta Rica 1.0 El Salvador 1.5 0 5 10 15 20 0 5 10 15 20 Num be r of Re s ide nt Pate nts Grante d pe r Publications in SCI Se arch / GDP (US$ billions ) M illion Labor Force Source: RICYT (2008) Source: RICYT (2008) 6.11 New products and processes. Enterprise Survey data show that Costa Rican firms are the most likely to introduce new products; elsewhere, Central American firms are less likely than the rest of LAC to introduce new products (Figure 6.3). 5 When it comes to new processes, Central American countries cluster fairly close to the LAC average (Figure 6.4). The big exception is Costa Rica, where only 29 percent of firms report introducing new processes. There is no uniform definition of what constitutes a new product or process, so this data should be taken with a grain of salt. 5 In this context, new products and processes are those that are new to the firm, not the country. 121 Figure 6.3: Percent of Firms that Introduced Figure 6.4: Percent of Firms that Introduced New Products New Processes Percentage of Firm s that Introduced New Percentage of Firm s that Introduce New Products Processes Costa Rica 87.8 Panama 61.8 Chile 71.0 El Salvador 61.8 LAC10 63.1 Chile 61.1 El Salvador 62.8 LAC10 59.4 Honduras 61.5 Honduras 58.8 CA6 58.1 CA6 55.6 Guatemala 56.8 Nicaragua 55.5 Panama 51.2 Guatemala 53.3 Nicaragua 45.3 Costa Rica 29.2 0 20 40 60 80 100 0 20 40 60 80 Note: LAC10 includes Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Mexico, Paraguay, Peru & Uruguay. These questions represent the average for Manufacturing and Services firms (i.e. does not include firms in Construction & Transport sector). Source: Enterprise Surveys, 2006 6.12 Export discovery. When a country exports a product for the first time, it can be viewed as an indicator of innovation. Klinger and Lederman (2004) compared the predicted number of export discoveries based on GDP per capita with the observed number in the 1990s. In Honduras, Nicaragua, and El Salvador, the actual number was in line with the predicted value. Costa Rica and Panama, on the other hand, far exceeded expectations, while Guatemala fell short (see also Jaramillo and Lederman, 2006, for more details). Innovation Input Indicators 6.13 R&D expenditures. Based on data from 60 countries, Lederman (2007) shows that R&D expenditures tend to be significantly associated with product innovation—i.e. non-patentable innovation—in developing countries. On average, the LAC region spends 0.54 percent of GDP on R&D. With the exception of Costa Rica and Panama, Central American R&D expenditures are quite low by LAC standards. Honduras, Nicaragua, and Guatemala do not exceed 0.06 percent of GDP (Figure 6.5). 6.14 In the Global Competitiveness Report 2007-2008, the World Economic Forum (WEF) reached similar conclusions when ranking company spending on R&D. Out of 131 countries, the WEF ranked Costa Rica 30th and Panama 79th (lower rankings mean higher R&D spending). El Salvador, Nicaragua, and Honduras all ranked over 100. Surprisingly, Guatemala ranked 64th, which somewhat contradicts the low values reported in other data. There is some evidence that Central American countries’ low levels of R&D are consistent with their low income levels (Rodriguez-Clare, 2005). This finding suggests that focusing on less sophisticated measures of innovation—such as technology adoption and QI—might make more sense in the Central American context. 6.15 Foreign technology adoption and ICT use. Firms can boost productivity and enhance the sophistication of their operations by importing technology, particularly in the areas of 122 information and communications (ICT). Costa Rica and, somewhat surprisingly, Guatemala exceed the Latin American average in licensing foreign technology. Honduras and Nicaragua bring up the rear, with less than 10 percent of firms in each country using technology licensed from abroad (Figure 6.6). Use of personal computers also reflects basic technology adoption, and Costa Rica and Guatemala again come out ahead, with rates that are more than twice the LAC average. Internet use is by far the highest in Costa Rica, with Nicaragua scoring lowest (Table 6.1). Figure 6.5: Expenditure on R&D as a % of GDP, latest Figure 6.6: Percentage of Firms That Use Technology year available Licensed from a Foreign Company LAC 0.54% Cos ta Rica 32 Cos ta Rica 0.41% Guate m al 21 Panam a 0.25% CA6 17 El 17 Honduras 0.06% Panam a 15 Nicaragua 0.05% Chile 12 LAC10 10 Guate m ala 0.03% Honduras 9 Nicaragua 6 0.0% 0.2% 0.4% 0.6% 0 10 20 30 40 Expe nditure on R&D as a % of GDP % of Firm s Note: Data for LAC is from 2005; Costa Rica 2004; Panama Source: Enterprise Surveys 2005; Honduras 2003; Nicaragua 2002; and Guatemala 2005. Source: RICYT (2008 Table 6.1: Use of ICT, 2005 Costa Rica* El Salvador Guatemala* Honduras* Nicaragua Panama LAC* Personal computers (per 1,000 219 51 191 16 43 46 88 people) Mobile phone subscribers (per 207 271 258 100 144 397 318 1,000 people) Price basket for mobile (US$ per 4 14 4 7 16 18 10 month) Internet users (per 1,000 people) 254 93 79 36 27 64 156 Price basket for Internet (US$ per 28 23 54 33 28 38 26 month) Secure Internet servers (per 1 62 5 6 4 2 56 10 million people) * Mobile phone subscribers, price basket for mobile, and personal computers data from 2004 Source: World Bank (2008), Development Data Platform Human Capital, Skills, and Educational Quality 6.16 Skill levels. Inadequately trained workers are a greater handicap in Central America’s less developed countries. World Bank Enterprise Surveys show that about three of 10 firms in El 123 Salvador and Guatemala see an insufficiently educated workforce as a major or severe obstacle. Panama and Costa Rica report less than half that (Figure 6.7). When it comes to the researchers as a percentage of the labor force, Costa Rica leads the rest of Central America, with Panama a distant second. Nevertheless, Costa Rica’s research potential is only about half the LAC average (Figure 6.8). As noted by De Ferranti et. al. (2003), the number of scientists, engineers, and researchers found in Central American countries is roughly in line with levels expected at their respective per capita GDP levels. Figure 6.7: Percentage of Firms That See Inadequately Educated Workforce as a Major Figure 6.8: Number of Researchers, 2005 or Severe Obstacle El Salvador 32% Guatemala 29% LAC 1.42 Cos ta Rica 0.76 Honduras 23% Panam a 0.36 Nicaragua 23% Honduras 0.21 Panama 14% Nicaragua 0.12 Costa Rica 13% Guate m ala 0.12 El Salvador 0.09 0% 10% 20% 30% 40% % of Firms Re s e arche rs pe r Thous and Labor Force Source: Enterprise Surveys Note: Honduras data is from 2003, Nicaragua from 2004. Source: RICYT (2008 6.17 Educational performance. Costa Rica and Panama are the exceptions to Central America’s low educational levels. Gross primary enrollment is well over 100 percent in all Central American countries. 6 However, secondary-school enrollment falls steadily from the most- to the least- developed nations. Gross secondary-school enrollment is 79 percent in Costa Rica—9 percentage points below the LAC average—compared to the 60s in Nicaragua, Honduras, and El Salvador (Figure 6.9). Mean years of schooling show similar trends, although Panama bests Costa Rica on this measure (Figure 6.10). Guatemala scores well below its neighbors both in terms of secondary enrollment and mean years of schooling. 6.18 Highly skilled workers. According to Rodriguez-Clare (2005), an important determinant of R&D investment rates is the share of the labor force with at least some post-secondary schooling. For Central America overall, the percentage of adults aged 25-65 educated for more than 13 years is below the LAC average. However, Panama has the same proportion of highly skilled workers as Chile, and Costa Rica nearly matches the LAC average. Only 5 percent of the Guatemalan and Honduran labor forces have some post-secondary education (Table 6.2). 6.19 Educational quality and learning outcomes. According to a recent World Bank study, most Central American children complete their educations without gaining sufficient skills to earn a good living. 7 Except in Costa Rica, average test scores are low or low-intermediate 6 The gross primary enrollment ratio is the ratio of total primary enrollment, regardless of age, to the population of the age group that officially corresponds to primary education. Gross primary enrollment can be higher than 100 percent due to students that are older than the official primary age, but still enrolled in primary school. 7 See Di Gropello (2005). 124 relative to each country’s standards. 8 Where assessments are comparable over time, results showed no clear improvement in the second half of the 1990s or early 2000s. Disappointing achievement is associated with low levels of reading comprehension, an indicator often regarded as the most relevant measure of educational quality in the third grade. In Honduras, it appears that a majority of third-grade students cannot identify characters or ideas in what they read, nor can they identify parts of speech. Results are substantially lower in disadvantaged schools. In El Salvador in 2001, around 25 percent of students were not able to identify the main message of a simple written test. The proportion rose to 50 percent after correcting for the fraction of correct answers due to mere chance. 9 Figure 6.10: Mean Years of Schooling, Workers Age 25- Figure 6.9: Secondary School Enrollment, 2005 65 (Latest Year Available) Chile 91 Chile 10.6 LAC 88 Panama 9.5 Costa Rica 79 Costa Rica 8.3 Panama 70 El Salvador 6.7 Nicaragua 66 Honduras 66 Honduras 6.1 El Salvador 63 Nicaragua 5.8 Guatemala 51 Guatemala 4.3 0 20 40 60 80 100 0 5 10 15 Secondary School Enrollment (% Gross) Mean Years of Schooling Source: World Bank, Development Data Platform (2008) Note: Chile data is from 2003; Panama 2004; Costa Rica 2006; El Salvador 2004; Honduras 2006; Nicaragua 2005; and Guatemala 2004. Source: SEDLAC (2008) 6.20 Regional exams. Central American learning assessment systems cannot be compared, and countries do not participate in cross-regional international exams. Nevertheless, limited data from regional exams suggests deficiencies in student learning. A 1997 Latin American assessment shows that Honduras underperforms vis-à-vis other LAC countries, even considering its low GDP. The LAC countries, in turn, perform poorly vis-à-vis the OECD, a fact illustrated by Chile, one of Latin America’s best performers, ranking poorly on the internationally comparable Program for International Student Assessment (PISA) exam. 10 Vegas and Paltrow (2007) also present considerable evidence of exceedingly low educational performance in Latin America. 8 It is important to keep in mind that being “intermediate� in El Salvador is not the same as being “intermediate� in Nicaragua. El Salvador puts students in the intermediate achievement range if they can identify and understand material but cannot elaborate on it or apply their knowledge. In Nicaragua, this description corresponds more closely to a “basic� level of achievement. 9 World Bank (2005a). 10 World Bank (2005a). 125 Table 6.2: Share of adult population in each skill group, 2004* All adults 25-65 Low Medium High Central America 65.4 23.5 11.1 Costa Rica 57.6 26.4 16 El Salvador 61 27.8 11.3 Guatemala 80.1 15.4 4.5 Honduras 78.1 16.5 5.4 Nicaragua 71.5 19.4 9.1 Panama 43.9 35.7 20.4 Rest of Latin America # 50.9 32.2 16.9 Chile 32.6 47.1 20.2 Notes: Three groups are formed according to years of formal education. Low=0 to 8 years Medium=9 to 13 years, and High=more than 13 years. * 2003 for Chile, 2003-04 for Bolivia and 2005 for Nicaragua. # LAC average includes Chile Sources: SEDLAC (based on household surveys) (2008) 6.21 Worker training programs. According to enterprise survey data, on-the-job training programs are more prevalent in El Salvador, Costa Rica, and Panama, with nearly half of firms offering them. In Honduras, Nicaragua, and Guatemala, one-third or less firms offer training programs for workers (Figure 6.11). Environment for Innovation 6.22 Outside of Costa Rica, Central America’s Figure 6.11: Percentage of Firms Offering environment for innovation is relatively weak Training Programs for Employees (Table 6.3). The World Economic Forum finds that the quality of scientific research institutions El Salvador 50% is low across the board—with the major exception Costa Rica 46% of Costa Rica, which ranks 34th out of 131 countries. University/industry research Panama 44% collaboration is weak in most countries, Honduras 33% particularly El Salvador, Honduras, and Nicaragua 29% Nicaragua. Government procurement of advanced Guatemala 28% technology products seems unlikely to benefit firms in any of the countries. Rankings for 0% 20% 40% 60% intellectual property right (IPR) protection are decent in a few countries. Innovation can occur in Source: Enterprise Surveys the absence of sound research institutions, university/industry collaboration, government procurement, and adequate IPR laws, but these factors make it much more likely. 6.23 In most countries, labor regulations do not appear to be a significant constraint to hiring. According to Enterprise Surveys, less than 10 percent of firms in El Salvador, Guatemala, Honduras, Nicaragua, and Panama consider labor regulations a major or severe obstacle. In contrast, it is 24 percent in Costa Rica. This probably reflects the prevalence of unions. About 67 percent of Costa Rica’s workforce is unionized, compared with 7 percent in Panama and 1 126 percent or less in the region’s other countries. For firms regarding labor regulations as an impediment to hiring, severance pay was considered the biggest obstacle. Table 6.3: World Economic Forum Innovation Indicators 2007-2008 Rankings Out of 131 Costa Rica El Salvador Guatemala Honduras Nicaragua Panama Countries Quality of Scientific Research Institutions 34 121 103 120 125 95 University/Industry Research Collaboration 35 115 57 94 114 83 Government Procurement of Advanced Technology Products 61 90 85 88 112 79 Intellectual Property Protection* 48 60 72 94 91 49 *WEF 2006-2007 (Out of 125 Countries) Source: World Economic Forum, Global Competitiveness Report 2007-2008 6.24 With the exception of Costa Rica and Panama, quality certification appears low. The most common indicator focuses on an internationally recognized standard for quality—ISO 9001:2000 certification. After controlling for industrial value-added, such certifications in Panama and Costa Rica are on par with the LAC average. The other Central American countries fall well behind (Figure 6.12). 2. REASONS FOR LOW INNOVATION 6.25 The previous section detailed the Figure 6.12: ISO 9001:2000 Certifications per US$ Billion underperformance of Central American Industrial Value-Added, 2006 countries—particularly Honduras, Nicaragua, and Guatemala—on indictors Panam a 38 LAC 36 that shaped the environment for Cos ta Rica 35 innovation. This section explores possible Nicaragua 20 reasons for the poor showing. They are El Salvador 19 grouped into three main subjects: market Honduras 14 Guate m ala 9 failures, worker skills, and education, and quality infrastructure. While these $0 $10 $20 $30 $40 constraints play a role across the region, ISO 9001:2000 Ce rtifications pe r US$ they are likely to be more binding in the Billion Indus trial V alue -Adde d less developed countries, particularly Note: Industrial value-added data for Costa Rica is from 2005. Honduras, Guatemala, Nicaragua, and El Source: ISO (2007) for ISO data and World Bank (2008), DDP for Salvador. Lack of access to finance is industrial value-added. another possible factor, but it is not included because of the extensive discussion in Chapter 5. The external environment has also played a role—exposure to foreign technology through the Panama Canal, for example—but it is beyond the scope of this chapter. 6.26 The degree of intellectual property rights protection is another factor that could influence innovation. However, IPRs are not discussed in detail in this chapter, and the reader is referred to 127 the extensive existing literature on the subject.11 Even in small countries, strong IPR systems may spur innovation in areas of significant comparative advantage, especially if the economy has the human capital, infrastructure, incentives, and funding necessary to nurture new technologies. This has likely been the case in Singapore, Taiwan, and Ireland. Central American countries (with the possible exception of Costa Rica) are at a stage of development where technology adoption is likely to be much more important than new-to-the-world innovation. Therefore, mechanisms such as petty patents that provide protection for small adaptations of existing products or technologies could play a role. 12 FDI is more likely to flow to countries where investors feel that their rights are protected, meaning that strong IPRs can also help attract FDI and hence technology from abroad. Market Failures 6.27 Market failures in the creation and diffusion of knowledge lead to underinvestment in innovation. Measurement difficulties inhibit quantitative analysis on market failures’ impacts in Central America. Nevertheless, the presence of the following types of market failures is well accepted: 13 • Knowledge cannot be fully appropriated. Firms that invest in generating knowledge cannot reap all the benefits from their activities. As a result, their investments tend to be less than socially optimal. This reasoning applies not only to the generation of knowledge that is new from a global perspective but also to the first introduction of foreign ideas in countries where they have not been tested commercially. Hausman and Rodrik (2003) argue that patents may help realize private benefits in the case of new products, but formal protection mechanisms do not exist for the first production of existing products in a country. Thus, entrepreneurs assume the full cost of failure, but they are forced to share the benefits with imitators if the products succeed. This reduces the incentive to invest in developing new products and processes and search for technologies in which a country might have an untested comparative advantage. • Knowledge spillovers generate positive externalities and social returns that exceed private returns. As suggested by Griliches (1992) and others, the social rate of return on R&D is often three times larger than the private rate of return. 14 In other words, gains increase several fold when taking into account the impact of R&D on the economy through knowledge spillovers that improve other companies’ products and production techniques. Lederman (2007) also finds evidence of knowledge spillovers. Looking at patent density, he finds that firms with access to more commercial ideas are more likely to innovate. • R&D investments can be large, long-term, and risky. Less developed financial markets seldom provide adequate instruments and term structures for financing R&D 11 For example, for information on IPRs and R&D, see Lederman and Maloney (2003); for IPRs and DR-CAFTA, see Jaramillo and Lederman (2006); for IPRs and growth, see Gould and Gruben (1996). 12 For more details, see Rodriguez-Clare (2005). 13 For discussions of market failures that shape private sector investments in innovation, see De Ferranti et al (2003); Jones and Williams (1999); and Hausmann and Rodrik (2003), among others. 14 See also Lederman and Maloney (2003) and Jones and Williams (1998) for more on social returns to R&D. 128 expenditures. Moreover, many R&D operations have increasing returns to scale and large sunk costs, creating obstacles for funding R&D departments at smaller firms. Worker Skills and Education 6.28 A country’s skill level is an important factor in its capacity to adopt foreign technologies. De Ferranti et. al. (2003) find that upgrading skills and technology transfer reinforce each other. Innovation raises demand for skilled workers, and the improved job prospects increase incentives to stay in school. In turn, a better educated workforce can further stimulate firms’ demand for new technologies. Some countries may enter a virtuous cycle of upgrading both technology and skills, resulting in higher productivity and growth rates. 6.29 By contrast, countries with low skill levels may not be able to absorb much technology through trade and may not attract much FDI. These countries may also be unable to generate technological upgrading through domestic R&D, another skill-intensive process. Therefore, countries with very low skill levels may be trapped in a vicious cycle of little technology transfer, no domestic innovation, low productivity, and stagnant growth prospects. 15 Adequate levels of secondary schooling and above are crucial to breaking out of this cycle and facilitating technological upgrading for local manufacturers, attracting FDI with high technological content, and benefiting from potential spillovers from those investments. 16 Consistent with this argument, Rodriguez-Clare (2005) finds a strong positive relationship between R&D spending relative to GDP and education levels. 6.30 Skill levels are highest in Costa Rica and Panama. As noted in the previous section, Costa Rica has the lowest percentage of firms that see an inadequately an educated workforce as a major obstacle. It also leads the region in secondary-school enrollment, university graduates and master’s degree holders. Panama scores a close second on most measures, and its workforce average of 9.5 years of schooling surpasses Costa Rica. The four other Central America countries fall consistently behind the leaders on human capital indicators. Guatemala, with a secondary- school enrollment rate of only 51 percent, is generally at or near the bottom in most measures. 6.31 Skill levels are driven by schooling, not simply by attendance but learning and education quality as well. 17 Better education will make workers more productive, increase rates of return to education, and provide incentives for further private investments in education. Attending school does not automatically provide the skills and knowledge needed in the job market. Factors inside and outside the educational system affect outcomes: students, their families and communities, teachers, pedagogy and curriculum, educational materials and infrastructure, and school management. Poor learning outcomes lead to absenteeism, low grades, repeating courses and dropping out. These issues snowball as students attempt to move to higher grades without necessary skills and knowledge, resulting in continuing problems regardless of the quality of later teachers. Better learning can directly improve cognitive skills and abilities and reduce dropout and repetition rates. 15 De Ferranti et al (2003) 16 Jaramillo and Lederman (2006) 17 This section is adapted from Di Gropello (2005). 129 6.32 When trying to understand low skill levels, it is useful to look at returns to education. Evidence from Latin America during the 1990s suggests that integration into the global economy—through increasing trade liberalization, rising FDI, and more flexible licensing arrangements—facilitated technology transfer. These technologies required higher skills, pushing up demand for educated workers. 18 This skill-biased technological change suggests increasing returns to education in Central America over the past several years, a time of substantial trade liberalization. On the other hand, as Pritchett (1996) notes, a negative relationship exists between returns and the stock of human capital. If so, returns to education should be lower where education rates are higher. However, Maloney (2008) points out that returns to education in Chile and South Korea remain similar over the past decades, despite the latter’s more dramatic educational gains. This may be reflecting a rise in the quality of education in Asia relative to Chile. 19 6.33 We conducted an econometric analysis of changes in returns to education over time in Central America, focusing on the median urban salaried worker aged 25 to 55 in each country.20 Multiple survey years were used, covering periods between five and 15 years, depending on the country. The analysis found rates of return vary with the survey year, but estimated returns for one year are, with a few exceptions, not statistically significantly different than other years. This simple analysis suggests that further research is necessary to untangle the determinants of changes in labor supply and demand, as well as how these changes impact wages and returns to education. Possible influencing factors include skill-biased technological change and changes in trade patterns and policies, unionization, labor market reforms, and quality of schooling. 21,22 6.34 A separate econometric analysis pooled survey results from multiple years (using year dummies) to highlight differences in returns to education across countries and between educational levels. In Table 6.4, each coefficient represents the percentage gain from completing an additional level of education. In Costa Rica, for example, the median salaried, urban worker who completed secondary education would earn 34 percent more per hour than a similar worker with only a primary education. The returns to primary education are lowest at 8.5 percent in Panama and 9.7 percent in Costa Rica. They are highest at 20.9 percent in Honduras and 17.0 percent in Guatemala. The results likely reflect higher average educational levels of the Panamanian and Costa Rican labor forces (see Figure 6.10 and Table 6.2). Returns to secondary 18 For more on skill-biased technological change, see De Ferranti et al (2003). 19 Maloney (2008): “…Over these four decades, almost all the Asian countries close to doubled their human capital without, apparently, massive falls in the return. Korea started below Chile in 1960 -5.0 years compared to 5.2, but by the 1990s was at 10.4 years compared with 7.2 in Chile...� 20 We estimate the model by Heckman maximum likelihood methods. We estimated the model with quantile regression for the median (q=0.5). As regressors, we include educational dummies, age, age squared, gender, an urban dummy, regional dummies and year dummies. We also include firm size dummies and economic sector dummies. The selection equation also includes school enrollment and the number of children. In the tables we report the marginal “returns� to completing each educational level. 21 For a discussion of some drivers of wage changes in Latin America, see Sanchez and Schady (2003). 22 The results should be interpreted with caution. The idea behind Heckman selection methods is to address the potential presence of unobserved characteristics of individual that may be affecting both education and wage outcomes. The variables used for the exclusion restriction may not be capturing the unobserved characteristics of the subjects. However, due to data limitations (lack of panel data set or a larger set of variables to measure unobserved ability, among other things) the performed analysis is the best possible given data availability. In addition, we estimated the returns for the median person instead of the mean person, to reduce the incidence of outilers. 130 education vary between 34 percent for Costa Rica and about 56 percent for both Guatemala and Honduras. If high returns are at least partially driven by high demand for the corresponding types of workers, then there could be bottlenecks to secondary education in Guatemala and Honduras. Table 6.4: Returns to education, pooled years Median urban salaried worker aged 25-55 Costa Rica El Salvador Guatemala Honduras Nicaragua Panama Primary 9.7% 12.6% 17.0% 20.9% 14.3% 8.5% Secondary 34.0% 26.4% 56.7% 55.6% 31.8% 34.1% College 94.2% 65.5% 70.5% 78.9% 65.0% 76.2% Note: The percentages are the increase in hourly wage that the median urban salaried worker would enjoy compared to a similar worker with one level less education. Source: Authors’ calculations using household surveys for several years in each country 6.35 Costa Rica has the highest returns to college education at 94 percent, most likely due to the high quality of the country’s universities. Because the share of the Costa Rican workforce with a college education is also relatively high, the evidence is consistent with skill-biased technological change. Panama presents a slightly different picture: the share of its workforce with a university education is the region’s highest; yet, at 73 percent, returns to university education are much lower than they are in Costa Rica. Deficient Quality Infrastructure 23 6.36 Central American countries could benefit from improving their QI. If successful, their efforts should lead to a reduction in technical barriers to trade, better products and processes, and more competitive and innovative firms. QI is particularly important for small and medium enterprises (SMEs) because they typically do not have the resources to test and verify product quality. 6.37 Recent trade agreements, including DR-CAFTA, give Central American countries a chance to sell their products around the world. These exports must meet consumer expectations, of course. Just as important, they must fulfill target markets’ increasingly stringent legal, health, safety, and environmental requirements. The industrial nations’ high standards are a significant challenge for Central America, given its current low level of quality consciousness. 6.38 SMEs can only meet international quality standards if they have local access to a competent, integrated, and recognized QI. Important features include: access to national, regional, and international standards and technical regulations; participation in related international working groups; internationally recognized calibration of measuring instruments; internationally recognized testing; and the certification of products and quality management systems. Additionally, quality management systems must be introduced to SMEs, especially those not directly linked to multinational activities or companies. As a rule, these interlocking services are provided by a nationally organized, yet internationally integrated, QI. 23 This section is based on inputs from the technical cooperation arm of the German national metrology institute (Physikalisch-Technische Bundesanstalt, PTB). 131 6.39 Sound QI facilitates innovation and technology adoption. New products and processes often require exact measurements and analyses during development and production stages. Foreign measuring and testing facilities usually are not feasible because the necessary oversight has to be performed locally. Hence, reliable, local QI is necessary. Furthermore, standards that contain technical specifications for methods or products can smooth technology transfer. In doing so, international standards may be adjusted to fit local or regional needs, a process best undertaken by a competent standards institute. 6.40 QI is a public good that contributes to safety and environmental protection, and it is thus a key to sustainable economic, environmental, and social development in Central America. Exact and reliable lab analyses, for example, cut health systems’ costs by reducing false diagnoses. Consumers can have access to products that do not endanger their health, and they can rely on the manufacturer's product specifications. Implementing environmental protection regulations relies on measuring and analysis that can stand up in court. 6.41 Non-tariff barriers to trade are gaining prominence as customs duties fall. Technical regulations and conformity assessment procedures are aimed mainly at protecting consumers, health, and the environment and are thus justified and essential. They can, however, obstruct or even prevent trade flows, especially if captured by special interests hoping to thwart competition. Voluntary standards can have the same impact as technical regulations. They reflect customers’ expectations in terms of quality, making them increasingly important in international trade. Both technical regulations and voluntary standards can act as a barrier to trade, when—because of a weak or non-existent quality infrastructure—exports fail to meet the required standards and no proof of conformity assessment can be provided. See Annex 2 for more details on technical barriers to trade under DR-CAFTA. 6.42 Development of metrology and standardization services varies by country, as does participation in international networks. In metrology, only Panama and Costa Rica are associate (not full) members of the International Bureau of Weights and Measures (Bureau International de Poids et Mésures, BIPM). For several units in the official BIPM database, the National Metrology Institute Panama (CENAMEP) has already proved its Calibration and Measurement Capacity (CMC), a prerequisite BIPM established for the international recognition of metrology capacity and competence. Costa Rica has successfully presented its Quality Management System, but not yet declared a specific CMC. No countries in the region are members of the Organisation Internationale de Metrologie Legale (OIML), the world body that deals with legislative, administrative, and technical procedures. 24 6.43 International Standards Organization (ISO) membership is relevant for standardization, and Central American countries fall into three categories. Standards Institute Costa Rica (ENN, formerly INTECO) and Panama’s Comisión Panameña de Normas Industriales y Técnicas (COPANIT) are full ISO members, giving them access to information and participation in technical committees. Correspondent members are El Salvador’s Consejo Nacional de Ciencia y Tecnología (CONACYT), Guatemala’s Comisión Guatemalteca de Normas (COGUANOR), and Nicaragua’s Dirección Técnica de Normalización y Metrología (DTNM). They have full access to information but do not participate in technical committees. Finally, Honduras is a subscriber 24 See www.bipm.org and www.oiml.org. 132 member, providing a developing economy the chance to integrate itself gradually into the processes of international standardization. 25 6.44 Certification activities are hard to quantify, but Costa Rica appears to be a leader. The huge variety of private actors in certification activities make it difficult to get actual data. However, it is possible to list the most relevant certified standards (or families of standards) according to their priority or diffusion in the region (Table 6.5). The US and EU standards have to be applied in the certification of organic products, a fast growing market in recent years. Costa Rica is the only Central American country with status as third-country producer with general access of its organic products to the EU. In other cases, every product has to be certified separately. The certification bodies are European as well as Latin America based. Table 6.5: Most Relevant Standards for Central America Standard or Guideline Subject Area ISO 9001 Quality Management ISO 14001 Environmental Management SA 8000 Social Accountability OHSAS 18001 Occupational Health and Safety Assessment Series ISO 27000 Information Security Management System ISO 22000 Food Sector Management, new standard for general implementation in production, processing and distribution HACCP Hazard Critical Control Point, management scheme that is fundamental for the legal framework of the food sector FSC Forest Stewardship Council, guideline for sustainable management of forests GLOBALGAP Formerly EUREPGAP, private Retailer Standard with global extension GMP Good Manufacture Practices, that find applications in different production value chains 6.45 Progress in accreditation has been made through the Central American Accreditation Forum (FOCA) and the establishment of national accreditation bodies. Costa Rica, El Salvador, Honduras, Guatemala, Nicaragua, and Panama are FOCA members. Table 6.6 shows the number and type of organizations in each country that had received accreditations or had entered the accreditation process at the end of 2007. The testing laboratories category account for by far the highest number. They are important for agro-industry and environmental monitoring. 6.46 The range of QI implementation is wide in Central America, varying from government institutes, to highly sector specific entities, to pragmatic and self-made private sector solutions. Where QI services are not fully developed or available only with severe restrictions, individual activities predominate. Larger enterprises create internal laboratories and assure traceability to foreign external institutes. At the same time, there are small and informal service providers that are not connected to the required recognition structures. These mechanisms can function at a small scale, but the lack of national coordination is evident when it comes to integration into global structures and market access. SMEs tend to suffer most from underdeveloped QI because they do not have the resources to achieve the certifications for international markets on their own. 25 See www.iso.org. 133 Table 6.6: Organizations With or In the Process of Accreditation, 2007 Testing Calibration Certification Inspection Medical Country/Accreditation Laboratories Laboratories Bodies Bodies Laboratories Body With IP With IP With IP With IP With IP Guatemala/OGA 5 7 2 0 0 1 1 0 3 1 El Salvador/CONACYT 12 3 0 0 0 0 0 0 0 0 Honduras/OHA 0 2 0 1 0 0 0 0 0 0 Nicaragua/ONA-MIFIC 6 2 1 0 0 0 1 1 0 0 Costa Rica/ECA 28 9 6 6 1 0 4 0 3 0 Panamá/CAN 8 4 0 1 0 0 5 0 0 0 Total 59 27 9 8 1 1 11 1 6 1 Note: IP denotes In Process of receiving accreditation. Source: Authors’ elaboration. 6.47 The following country-by-country survey describes Central America’s QI development. This overview is complemented by case studies showing how the value chains of selected Central American products are supported or hindered by home-country QI. Box 1 contains the case studies for balsam in El Salvador, honey in Honduras, the Panama Canal, and shrimp in Costa Rica. 6.48 Costa Rica. The individual components and cooperation within Costa Rica’s QI system are the most developed in the region. The 5-year-old national Quality Act defines each of the QI components: the private-sector managed Standards Institute Costa Rica (ENN/INTECO) for standardization; the state-run National Metrology Institute Costa Rica (LACOMET) for metrology; the Ente Costarricense de Acreditación (ECA), jointly managed state, private sector and independent bodies, for accreditation; and state-run Oficina de Reglamentación Técnica, (ORT) for technical regulation. The law determines and coordinates policies via the National Quality Council, which consists of a range of stakeholders. 6.49 The strength of the system is shown by the metrology institute’s membership in BIPM, the relevant international body. The Mutual Recognition Arrangement for testing laboratories has been accepted by the international accreditation bodies—the Inter American Accreditation Cooperation (IAAC), the International Laboratory Accreditation Cooperation (ILAC), and the International Accreditation Forum (IAF). 6.50 ECA, Costa Rica’s accreditation agency, has the region’s largest number of accreditations (42), including several private calibration laboratories. ENN/ INTECO has performed the most work on standardization in the region, with over 500 national standards. It is also the only nationally accredited certification body in Central America. Other countries have only branches of international certification companies. Despite this acceptance in the economy and the extension of QI services in the region, the shrimp case study in Box 6.1 shows the system still has deficiencies. 6.51 Honduras. The draft of a modern Quality Act exists, but it still awaits legislative approval. Hence, the country has neither a legal framework nor legally functioning QI components. The Ministry of Economy’s rudimentary standardization activities do not fulfill the 134 criteria of an internationally recognized institute. Less than 10 national standards have been issued. QI services have to be bought from external providers. This could be a solution in a strongly developed region with open borders, but it is not realistic for a country with limited resources. 6.52 Recently, Honduras’ QI has undergone some improvements with World Bank support. Under the auspices of the Consejo Hondureño de Ciencia y Tecnología (COHCIT), a virtual QI was created in the form of the Organismo Hondureño de Normalización (OHN), the Centro Hondureño de Metrología (CEHM) and the Oficina Hondureña de Acreditación (OHA). These organizations have started to function, even though they do not yet exist from a legal view point. Once the Quality Act passes, the structure will already be in place with some capacity. 6.53 While Honduras develops particular components, some services are being purchased from neighboring countries. In metrology, for instance, more than 500 calibrations have been carried out in cooperation with laboratories in Costa Rica and Mexico. An agreement with Costa Rica’s ECA has enabled the OHA to offer internationally recognized accreditations to three Honduran laboratories. Standardization has also been taken up in the OHN according to international criteria and has been adjusted to meet national demand. 6.54 Guatemala. Guatemala has a national Quality Act that defines its QI components—the Comisión Guatemalteca de Normas (COGUANOR) for standardization, the National Metrology Institute (CENAME) for metrology, and the Oficina Guatemalteca de Acreditación (OGA) for accreditation. However, Guatemala’s QI is relatively underdeveloped. The accreditation body was set up with strong international support and guided to international recognition. There are 13 accredited entities in the country and demand appears to be increasing. OGA is currently in the final proceedings of the IAAC Peer Evaluation that could lead to Mutual Recognition Arrangement. In metrology, a modern building houses more than a half-dozen equipped laboratories, but human resources are not sufficient to use the infrastructure. As a consequence, the industrial sectors mostly look for service providers from neighboring countries or try to set up sector-specific calibration laboratories with traceability abroad. 6.55 El Salvador. As in most countries, the beginnings of QI for standardization, legal metrology, certification, and later accreditation have developed in the national Science and Technology Council (CONACYT). They have not yet managed to break off and establish themselves as independent institutions. This leads not only to conflicts of interest, but also hinders responding to demand for services (see balsam case study in Box 6.1). 6.56 The situation is especially drastic in metrology, which requires expensive physical infrastructure. A small and limited national metrology laboratory belongs to CONACYT, but it is run in cooperation with the national university. Sporadic participation in the regional metrology organization takes place without official recognition. More comprehensive and better metrological facilities for calibration services and investigation are found in a private university, but it lacks linkages to the international metrology network and offers only poor access to the national market. A private calibration laboratory has been established with branches in neighboring countries. It carries out more than 10 times the calibration services of the national laboratories. The demand is so significant and the customer relationships so well-established that quality aspects are no longer questioned, although accreditation or other forms of proof of 135 competence are lacking. El Salvador’s QI is currently in transition under the responsibility of the Ministry of Economics, which intends to separate the relevant departments from CONACYT and restructure the national metrology, standardization, and accreditation bodies. Box 6.1: Case Studies of Select Value Chains Balsam – El Salvador Balsam, a natural extract from the sap of balsam trees, is a raw material for medicinal and pharmaceutical products. The trees are scored like rubber plants, and the sap caught in containers. In El Salvador, balsam is produced mainly by family businesses. The extracted quantities are too small for direct marketing; furthermore, the country lacks export structures, and small businesses cannot set up their own. Thus, the raw material is sold to brokers, who provide interim storage, compile sufficient quantities, and handle the exporting. National QI could solve a crucial local problem in the value chain. Balsam quality at the transfer point between producers and brokers is not sufficiently defined and cannot be verified objectively. Thus, contamination occurs due to improper handling, the use of unsuitable collection containers (plastic or metal), or manipulations (dilution). On the other hand, brokers can reduce the price they pay if the extract’s claimed characteristics cannot be proven objectively. Weight and volume cannot be measured with calibrated instruments. This is a classic example of local requirements that do not require large resources. However, neither the value chain partners nor El Salvador’s QI had taken on the task. Both groups were brought together only through external intervention. Salvadoran criteria were developed with the national standards body, which specifies balsam quality based on local consensus and international requirements. Currently, the producers are implementing an adapted form of good manufacturing practice (GMP) to fulfill the new criteria. Then, competent laboratories will have to technically specify the characteristics. This will then lead to demand for the calibration of measuring instruments and laboratory accreditation. Panama – the Canal One of Panama’s largest customers for QI services is the Autoridad del Canal de Panamá, which has been certified by an internationally operating body according to ISO 9001:2000. It thus fulfills not only high standards of quality, but also demands them of its suppliers. A corresponding multiplication effect can be seen in the demand for national calibration services. Although Panama’s QI components are still weak in the sense of an integrated and recognized system, the country’s metrology is strongly developed, expressed by the declaration of Calibration and Measurement Capacity (CMC) and international recognition. Indirectly, the driving force is the international market and its quality requirements. The Panama Canal authority sells services to international shipping lines, which place stringent demands on safety and reliability. The services can be considered as an export product. The added value creation takes place in Panama and is dependent on reliable metrological services, which led to an the sector’s accelerated development. At the same time, the consolidation of the other QI components is still lacking. Costa Rica – Shrimp Costa Rica has developed an important export industry in aquaculture, including shrimp to Europe. Quality infrastructure is critical. The European market requires the country of origin to not only test the final product but also maintain extensive controls over the entire production process. US Food and Drug Administration regulations to a large extent give importers responsibility for compliance, but European requirements fall on exporters as well as importers. In the exporting country, Europe requires a public agency be designated as the competent authority, responsible for compliance with EU directives. It is the counterpart of Europe’s consumer protection agency, the Directorate General for Health & Consumer Protection (DG SANCO). The competent authority must operate within a “farm to fork� approach. This entails being in a position to document all the links of the value chain in accordance with the EU consumer protection directives (traceability); prove documentation by means of competent inspections; and confirm the inspection results by means of recognized laboratory analyses. Laboratories, inspection centers and other external service providers can be officially integrated into the process if they can prove their competence, generally through accreditation. Every few years, DG SANCO’s Food and Veterinary Office carries out inspections to check the competent authorities’ operations. Costa Rica’s competent authority in is the Servicio Nacional de Salud Animal (SENASA), under the authority of the Ministry of Agriculture. Following a Food and Veterinary Office mission in 2000, further inspections took 136 place in January 2007. 26 A range of nonconformities were identified, somewhat to the surprise in a country with some of Central America’s most developed QI. Problems included: Lack of inspections in the case of the Residue Monitoring Plan; deficiencies in residue controls; limited resources for their own control laboratory; laboratories that do not use the approved standards; lack of accreditation; non-compliance with the requirements for fishery products’ certification; and insufficient coordination with other bodies. The report led to an import embargo on shrimp bred in Costa Rica, imposed by the EU Commission on March 22, 2007. It shocked an unprepared industry. The subsidiary of a European firm that had invested more than € 3 million in 2006 was also affected, even though EU inspectors confirmed its production was in line with EU regulations. However, the “farm to fork� system gives the competent authority responsibility for monitoring the entire value chain. Thus, good performance of an individual producer cannot overcome the total system’s deficiencies. Costa Rica developed an action plan and sent it to the DG SANCO, and it is currently being implemented. Due to the broadly developed QI in Costa Rica, duplications can be avoided by means of better coordination and by the integration of the QI components into the competent authority’s tasks. A national working group was set up to help stakeholders work together to implement the activity plan. Rather than creating an ad hoc solution, they are attempting to install a sustainable system integrating as many existing components as possible. The example of shrimp in Costa Rica shows how solutions can be carried out by means of better stakeholder coordination, based on a legally anchored and high-capacity QI system. The responsible ministry adopted a new technical regulation, based on Codex Alimentarius, US FDA Regulations, and EU Directives, that specifies maximum residue values in fishery products. An additional technical regulation recognized external testing laboratories for official analyses by the competent authority. It refers to the accreditation by the national body, the ECA. The agency’s competence has been proven via international recognition, adding external weight to the competent authority’s decision. Both decrees were reported to the WTO via the notification authority and are available on SENASA’s Internet page: (www.senasa.go.cr/documentosenconsulta.html). Honduras – Honey Honey is a common product in several Central American countries. Many producers sell to brokers for marketing purposes, similar to the balsam commercialization but with the key difference that honey is already a final product. Until recently, Honduras did not have a functioning standards institute. For honey production, various recommendations were made, but no nationally adopted consensus emerged. The output is marketed either as “no name� products or as low-priced “bulk honey,� which is later blended with cheap brands, mostly from Asia. To break out of this low-value equilibrium, Honduras had to identify requirements for classifying various qualities of honey. The Organismo Hondureño de Normalización (OHN), a standards body, was created in 2007, although it still lacks legal recognition. OHN developed the first three national standards for honey in consultation with key members of the value chain. In the next step, the demand for calibration, testing, certification, and accreditation services will be developed for this sector. For exports to markets such as the EU, a national Residue Monitoring Plan would have to be set up and implemented by the competent authority within the scope of the “farm to fork� system. 6.57 Panama. In Panama, the Ministry of Industry and Trade takes the lead in standardization and accreditation. This arrangement is reasonable because of their impact on industrial and economic development. Institutionally, however, it is not compatible with the criteria for international recognition. The National Metrology Institute Panama (CENAMEP) has steadily progressed—from measurement capabilities in the university sector, to the Secretariat of Science and Technology and, finally, to an independent institute based on a legal framework. CENAMEP was the first in Central America to attain Calibration and Measurement Capacity (CMC) for mass, electricity/AC, frequency, and time in the BIPM data base. An important reason lies in the strong, although selective, demand stimulated by the Panama Canal (see Box 6.1). 26 The report of the mission is accessible at http://ec.europa.eu/food/fvo/index_en.htm . 137 6.58 Nicaragua. The country’s QI conditions are between those of El Salvador and Honduras, but the National Metrology Institute (LANAMET), which is associated with the national university, has developed well. No formal international recognition has yet been achieved, but preparations are underway for accreditation of the quality management system and the mass laboratory under a Mutual Recognition Arrangement from the International Laboratory Accreditation Cooperation (ILAC). The good metrological basis is not reflected in the number of accredited laboratories or in the strength of the national economy. Maintaining central administration of standardization, technical regulations, and accreditation in the Ministry of Industry and Trade suggests staffing and financial restrictions. 3. POLICIES TO ADDRESS LACK OF INNOVATION 6.59 This section focuses on potential policies to address the key constraints previously identified—low worker education and skills and deficient quality infrastructure. The section also discusses policies to overcome market failures and encourage more innovation and technology adoption through university/private-sector collaboration, public research institutions, and regional cooperation. When considering potential policies, it is useful to keep in mind that human capital levels shape the types of effective interventions. This means that policies in countries with higher education levels, such as Costa Rica and Panama, will not necessarily be the same as policies for other countries in the region. 6.60 The market failures described in the previous section imply that an unregulated and purely privately funded market for knowledge may lead to sub-optimal outcomes. Governments counter this in various ways—for example by establishing intellectual property rights systems. When it comes to creating knowledge, however, regulations alone do not align private and social returns, which is why many governments look to supplement private innovation-related expenditures with public funds. Many developed nations invest close to 1 percent of GDP in science, technology, and innovation, financing universities, technology transfer centers, metrology institutions, venture capital funds, private sector R&D, and knowledge sharing activities. 6.61 Regardless of the policies pursued, clear leadership at the national level is important. Governments would be well-served to coordinate efforts to promote innovation and technology adoption. One useful idea involves channeling all related programs through a single agency with a strong mandate to think strategically. This ensures that programs are part of a coherent plan, rather than a disjointed mix that misses potential opportunities for complementarities between, for example, agriculture and industry. Improving Worker Skills and Education 6.62 All Central American countries would benefit from improving worker skills and education, but appropriate strategies differ according to development levels. Costa Rica and Panama are relatively strong in nearly all human capital measures. In fact, Costa Rica does not seem to have a shortage of advanced human capital. University graduates as a percentage of the labor force is more than twice the LAC average, and the World Economic Forum ranks the country 37th out of 125 on the availability of scientists and engineers. Policies to improve worker skills and education are more critical in El Salvador, Guatemala, Honduras, and 138 Nicaragua. This section concentrates on strategies to improve worker skills in Central America’s less developed countries. 6.63 Policy recommendations focus on improving educational quality. The reason: higher quality should automatically attract and retain more students because they and their parents will perceive more value from attending school. Better quality should also increase returns to education. Based on Di Gropello (2005), the following recommendations take aim at barriers to learning and educational quality: • Further institutionalize assessment systems. This includes starting evaluation in first grade and establishing effective dissemination mechanisms. Several innovative dissemination methods are currently used internationally. One example is school report cards published in local newspapers or laminated and posted at school entrances. • Improve teacher education and preparation. This includes developing selection criteria and such diagnostic tools as accreditation exams. It may require implementing full teacher certification systems, like the ones in El Salvador and Costa Rica, and reducing demand and supply gaps in teacher specializations. • Improve teacher effort. This could include such incentive mechanisms as: (a) carefully designed team-based merit pay; (b) salary scales that promote higher effective hours of work; (c) decentralized systems of teacher monitoring; and (d) fixed-term contracts or local authority to hire and fire teachers. An analysis of school-based management with community participation in El Salvador, Guatemala, Honduras, and Nicaragua suggests that empowering parents to hire, fire, and monitor teachers’ results in greater teacher effort as measured by hours. • Diversify teaching strategies and methods. This entails emphasizing classroom practices more in pre-service and in-service training and developing innovative ways of organizing in-service teacher support. It would also include creating spaces for teachers to share ideas and methodologies and developing effective systems of school supervision that ensure continuous support for teachers applying new methodologies. • Increase instructional time. This includes lengthening the school year to 200 days and developing better ways of recording and monitoring teacher absences and school closings, perhaps by involving community stakeholders. 6.64 Improved quality of teaching and learning through these policies could be strengthened by complementary strategies that address demand-side constraints, such as poverty and low parental education. Demand-side interventions include: (a) publicity campaigns focused on learning quality that give families important information about exams, timely enrollment, or tips for supporting children’s success in school; and (b) policies that directly impact children’s background characteristics, such as health, nutrition, and access to clean water and safe homes. 6.65 Technical and vocational education and training can also play an important role in upgrading worker skills. Central America offers a variety of training institutions and policies in this area. For instance, Panama’s recent efforts to modernize its worker training system incorporates international best practices, such as accreditation for providers and programs and certification for workers participating in competency-based standard training. The modernization 139 is also expected to allow the supply of training services to respond to the labor market’s current and emerging needs. 6.66 Some other Central American countries have public training institutions that do not respond well to the private sector’s requirements. This is often the case where public training institutions are financed by taxes, meaning their incentives are not entirely demand driven. When considering policies to facilitate the provision of training services, governments should foster dialogue between training providers and the private sector. Training providers, be they public or private, should respond to private sector demand by developing competency profiles and certification programs in the economy’s most dynamic sectors, including industries with high growth potential and those facing competitive pressures from abroad. Programs should be modular, allowing workers to acquire defined competencies in easily digestible units that accommodate their schedules. 6.67 Governments can be proactive in the following areas related to technical and vocational education and training: (1) developing policies, setting standards, investing in training materials and instructors, improving public information about the training system, and carrying out evaluations of training; (2) financing training to meet equity objectives and fill strategic skill gaps; and (3) providing skills training in priority areas where nongovernment providers are reluctant to invest (but exercising caution to avoid crowding out nongovernment providers). 6.68 Finding the right balance between government and nongovernment provision and financing of training is important to ensuring that public resources are available for other spending priorities, such as basic education. Economic analysis of training markets—including both supply and demand—is needed to inform decisions regarding the government’s optimal role. The most important part of the analysis is understanding who other than government provides training in an economy, how cost-effective this training is, and what barriers exist to enhancing and expanding this capacity to reduce pressure on public spending. Strengthening Quality Infrastructure 27 6.69 This section looks at ways to improve QI. Central American governments should be cognizant of the importance of QI and try to enhance its development. They should create the relevant legal and institutional conditions and consider making the necessary resources available to set up QI. Countries should also involve their national institutions in regional and international bodies, such as COPANT, ISO/IEC, SIM, BIPM, IAAC, ILAC/IAF. Central America‘s widespread lack of coordination between competent authorities and QI establishments frequently creates institutional conflicts that, in turn, lead to user uncertainty and less interest from the private sector. It also encourages the growth of informal services, which lack such features as traceability of measurements and harmonization of regulations, and inefficient duplication of QI activities. 6.70 Attention should to be paid to the basic development of all QI components. Individual QI components are highly interrelated and can be provided by a mix of public and private entities. Increasing competitiveness requires the availability and application of standards. Accreditation is 27 This section is based on inputs from the technical cooperation arm of the German national metrology institute (Physikalisch-Technische Bundesanstalt, PTB). 140 based on solid quality management and reliable measurements. Product certification demands reliable lab testing. Therefore, a key development goal should be a functioning, basic QI, which is complete and relevant to sectors with greatest demand. A complete basic system is preferable to highly developed, isolated solutions that are not integrated into a whole and thus have limited economic impact. Moreover, systems should be tailored to the needs of specific sectors, with an eye toward services decentralized to areas with high concentrations of producers. Table 6.7: Recommendations for an Integrated Regional Quality Infrastructure System Metrology • Design and implement the structure for SI (International System) units’ traceability by identifying the needs and measurement range for each of the main units per national metrology institute. Staff capacity-building, space requirements, and laboratory equipment specifications would also have to be identified based on market demands. • Relate the capacity building strictly to the development of needed measurement capacities (once they have been identified). • Establish the CAMET Forum as a permanent professional platform; introduce and maintain continuous inter-comparison exercises. • Promote the coordination of capacities between the national metrology institutes to establish a regional metrology network. • Institutions that are leaders in calibration and measurement capacity (CMC) should implement a regionally agreed concept of traceability to SI units, with the aim of being internationally recognized through the International Bureau of Weights and Measures database. • Establish structures and procedures for Legal Metrology, including the coordination of required technical and institutional responsibilities. Standardization, • Harmonize regional standardization and the technical regulations as required for the Technical “Union Aduanera�. regulations • Support of competent authorities for food safety, consumer protection, health, environmental and natural resources management, and security. • Develop inquiry and focal points to make information on standards and technical regulations available to the public. Testing • Support the provision of reference materials for proficiency testing on a regional level. • Eliminate logistics and customs barriers to facilitate shipping of testing and reference materials. • Promote competitive services for national and global markets. • Create incentives and enforce existing legislation by means of governmental coordination in the fields of environment, health, consumer protection, security and food safety. Quality • Use ISO, CODEX, HACCP, and others standards and guidelines as references and Management introduce quality concepts into the curriculums of general and technical education. Systems Certification/ • Support the accreditation process for conformity assessment bodies on a national level, Inspection both for voluntary certifications and mandatory inspections. Accreditation • Support the implementation of cross-border policy and joint accreditation activities. • Integrate all of the accreditation bodies in the region as full IAAC members. • Promote the use of FOCA as a regional instrument Source: Authors’ elaboration. 6.71 State procurement can encourage a “culture of quality.� As a key purchasing agent, the state can facilitate the development of the market through legislation, technical regulations and creating incentives that stimulate demand for QI services. Calls for tenders can be oriented 141 towards national standards or, as a prerequisite, stipulate certification of companies, products, and laboratories. An example comes from Costa Rica, where the Ministry of the Economy requires all laboratories that provide services to official entities to be accredited to the ISO 17025 standard and by the national accreditation body, ECA. 6.72 To sustain an integrated QI, a critical mass of demand for quality services is necessary. Yet, such a critical mass can be hard to obtain, especially in small, less-developed economies. Regional cooperation can reduce the need for services in each country. An approach that includes not only harmonization but also sharing resources and quality services providers could help reach a critical mass. A promising example is the Foro Centroamericano de Acreditación (FOCA). Another possibility could be in the field of metrology. Each national metrology institute could provide measurement services for things that require lower accuracy, meaning less expensive lab conditions. At the same time, one national metrology institute in the region could house more sophisticated equipment and apply regional standards with internationally recognized traceability. Table 6.7 provides detailed recommendations for an integrated regional quality infrastructure system. To move beyond each bullet point in the table, it would be necessary for stakeholders to jointly identify and analyze the relevant issues and develop an action plan. Linking Innovation Supply and Demand 6.73 Concentrating on education and QI improvements is a sound strategy for all Central American countries. However, the region’s more advanced countries, such as Costa Rica, Panama, and El Salvador, should also consider more sophisticated policy approaches. One important set of activities involves strengthening public-private partnerships and ties between universities, public research institutions, and the private sector. Such linkages can help substitute for many firms’ lack of highly skilled scientists and engineers. 6.74 Supply and demand for knowledge and innovation tend to be disconnected. Supply emerges from the universities and public research institutions that generate knowledge, and demand comes from private enterprises that use the knowledge to boost productivity and profits. In nearly all Latin American countries with available data, the majority of R&D spending is financed by governments and universities. In contrast, 65 percent of R&D is financed by private enterprises in the US. In Central America, data is only available for Guatemala, where the statistics capture no private sector R&D financing, and Panama, where domestic enterprises account for less than 1 percent of R&D spending. 28 6.75 Linkages facilitate the sharing of costs, risks, and human resources between the public and private sectors. They also allow for the transfer of tacit knowledge through personal interactions, research projects, networks, and clusters, or by means of mobility between the public and private sectors. Without such linkages, firms are much less likely to undertake innovation on their own. Successful national innovation systems are characterized by interaction among all players. Across the OECD, governments have recognized the importance of such linkages, leading to a wave of initiatives to encourage public-private collaborations. The movement started in the 1990s and has continued to expand. In some countries, programs to 28 RICYT (2008) 142 foster public-private partnerships constitute a significant share of public R&D funding (OECD, 2005). In Latin America, a few countries are beginning to test the benefits of such partnerships. In 2004, Chile launched its first consortia program to bridge the gap between public research centers and the productive sector. Mexico and Uruguay are launching similar initiatives. 6.76 Barriers within higher education discourage the transfer of knowledge to the private sector. Based on the OECD, Canton et al (2004) identifies three key barriers to university-private sector collaboration that are all highly relevant for Central America. 29 • Mismatch between knowledge supplied by universities and knowledge demanded by businesses. Inward-oriented and centralized universities may be the cause of the mismatch. In the U.S., the educational system is more decentralized and competitive. It is similar to the Finnish approach, where academic departments receive limited basic funding that is supplemented by competitive financing and incentives to undertake work useful to business. In Chile, nine cooperative research consortia have been established, and they could provide an equally successful means of providing links with the business community. • Ineffective reward structure for academics. Many incentives that promote creativity may, to some degree, deter researchers from addressing the needs of business. An emphasis on publishing, while essential for guaranteeing output quality, does not necessarily facilitate dissemination. Local industry would benefit more from direct contact than from publications. The opposing cultures of disclosure in the academic world and protection of trade secrets in the business world can work against each other. However, the importance of this effect is mixed. • Lack of an entrepreneurial culture. Even if an academic idea’s originator has the ability to take it to market, little incentive may exist to pursue it beyond “proof of concept.� To encourage commercialization, US policies have included science parks and technology transfer offices (TTOs), which develop networks of industrial partners, set up guidelines for commercialization of research, and manage university intellectual property. In conjunction with their royalty sharing systems, the TTOs appear more likely to increase commercialization in private universities, which are less constrained by regulations. 6.77 Central American universities’ incentive structures should be adjusted to encourage collaboration with the private sector. Previously presented World Economic Forum evidence suggests that university/industry collaboration is quite strong in Costa Rica. Guatemala also does relatively well on this indicator (Table 6.3). The data suggest that Panama and El Salvador, given their levels of development, would have the most to gain from improving incentives for university researchers and tackling barriers to private sector linkages. 6.78 Universities and public research institutions (to the extent that they exist in Central America) can play an important coordination role and help overcome market failures. According to Link and Scott (2004), these institutions’ activities in Latin America are most likely to succeed in correcting underinvestment when research has definable outputs and outcomes aimed at 29 Canton et al (2004) as summarized in World Bank (2006a). 143 addressing market failures. Based on case studies, Link and Scott provide the following examples of ways these institutions can provide or facilitate socially valuable research: • Standards. Universities and public research institutions can act as honest and knowledgeable brokers, bringing together industrial firms to ensure appropriate standards are developed, promulgated, and used. • Coordinating R&D efforts. Universities and public research institutions are often in the best position to facilitate cooperative R&D efforts that join industry and government in publicly subsidized projects. • Oversight of (partly) publicly funded and privately performed research. In many cases, industry has the expertise to perform socially desirable research most efficiently. However, the incentives to perform the research must be provided by the public research institutions’ award processes, their subsidies for the research, and their oversight. The aim should be to leverage public expenditures by requiring as much private co-financing as possible. Just enough public funding should be supplied to allow each project’s private rate of return to exceed its hurdle rate, despite the externalities and spillover effects that limit the private sector’s ability to appropriate returns for socially valuable projects. 6.79 Central America’s public research institutions need to focus on clear, measurable, and transparent objectives. It is essential to demonstrate public accountability. That requires public research institutions to identify market failures before undertaking projects internally or funding them externally. Just as important is an evaluation program. In fact, establishing an ongoing evaluation program is a hallmark of successful public research institutions. For more details on appropriate evaluation methodologies, see Link and Scott (2004). 6.80 Central America countries could benefit from regional cooperation to collect knowledge from abroad and disseminate it to relevant industries. Given the countries’ small size and overlapping challenges, duplicating such efforts at the national level might not make sense (due to economies of scale). One possibility would be to establish a regional organization with a clear mandate. It could begin by focusing on sectors with high cross-country relevance, such as agro- industry, maquilas, and software. The organization would serve as a base of technical knowledge, collecting it from abroad and disseminating it to firms in the region. Other functions could include collaborating with existing knowledge centers in the region and elsewhere, including the INCAE Business School in Costa Rica and the Escuela Agrícola Panamericana (better known as Zamorano) in Honduras. 144 APPENDICES Chapter 1 Table A.1. 1: Central America's Trade with the United States: Top 7 Import and Export Items, 2006 Top 7 Import items from the United States Costa Rica Guatemala Honduras Nicaragua Panama El Salvador Central America Machinery/transp equipmt Machinery/transp equipmt Mineral fuel/lubricants Chemicals. Machinery/transp equipmt Machinery/transp equipmt Machinery/transp equipmt Manufactured goods Mineral fuel/lubricants Machinery/transp equipmt Machinery/transp equipmt Food & live animals Food & live animals Mineral fuel/lubricants Chemicals. Chemicals Food & live animals Food & live animals Manufactured goods Mineral fuel/lubricants Chemicals Miscellaneous manuf arts Food & live animals Chemicals Mineral fuel/lubricants Miscellaneous manuf arts Chemicals Manufactured goods Food & live animals Manufactured goods Manufactured goods Manufactured goods Chemicals Manufactured goods Food & live animals Mineral fuel/lubricants Miscellaneous manuf arts Miscellaneous manuf arts Miscellaneous manuf arts Mineral fuel/lubricants Miscellaneous manuf arts Miscellaneous manuf arts Crude mater.ex food/fuel Crude mater.ex food/fuel Crude mater.ex food/fuel Commodities n.e.c. Crude mater.ex food/fuel Crude mater.ex food/fuel Crude mater.ex food/fuel Top 7 Export items to the United States Costa Rica Guatemala Honduras Nicaragua Panama El Salvador Central America Food & live animals Food & live animals Food & live animals Food & live animals Food & live animals Chemicals Food & live animals Miscellaneous manuf arts Mineral fuel/lubricants Commodities n.e.c. Commodities n.e.c. Miscellaneous manuf arts Food & live animals Miscellaneous manuf arts Machinery/transp equipmt Miscellaneous manuf arts Crude mater.ex food/fuel Beverages and tobacco Manufactured goods Miscellaneous manuf arts Machinery/transp equipmt Manufactured goods Crude mater.ex food/fuel Miscellaneous manuf arts Miscellaneous manuf arts Commodities nes Manufactured goods Manufactured goods Crude mater.ex food/fuel Manufactured goods Manufactured goods Animal/veg oil/fat/wax Crude mater.ex food/fuel Mineral fuel/lubricants Mineral fuel/lubricants Chemicals Chemicals Beverages and tobacco Crude mater.ex food/fuel Chemicals Machinery/transp equipmt Crude mater.ex food/fuel Commodities n.e.c. Machinery/transp equipmt Machinery/transp equipmt Manufactured goods Machinery/transp equipmt Commodities n.e.c. Chemicals Note: n.e.c. not elsewhere classified Source: WITS database; SITC Classification, Revision 3. 145 Table A.1. 2: Tariff Structure in Central America, 1995-2006 Costa Rica El Salvador Guatemala Honduras Nicaragua Panama 1995-00 2001-06 1995-00 2001-06 1995-00 2001-06 1995-00 2001-06 1995-00 2001-06 1995-00 2001-06 Agricultural 10.9 10.3 11.8 12.4 12.0 10.3 12.0 10.4 13.0 10.9 13.4 13.4 materials Agricultural raw 4.3 2.0 3.5 3.1 5.9 5.1 5.4 6.8 10.7 7.7 6.2 4.0 materials Chemicals 4.4 3.1 5.8 6.1 4.6 3.6 3.9 2.8 5.4 3.8 5.8 3.5 Food 11.7 11.3 13.1 13.5 12.8 10.8 12.4 10.6 13.1 11.0 13.7 13.8 Machinery & transport 8.0 5.0 4.5 3.1 5.4 3.7 12.3 10.0 5.1 2.7 4.2 3.5 equipment Manufactures 4.0 2.1 5.2 4.6 5.8 5.6 5.6 5.5 5.2 4.2 8.9 7.6 Miscellaneous 5.0 5.0 5.0 5.7 6.0 5.5 6.0 5.0 6.7 5.0 na na goods Ores and metals 3.4 1.5 3.3 1.4 3.6 1.8 4.6 3.1 5.6 3.4 8.0 5.0 Other 8.1 6.6 9.1 8.5 8.7 7.7 10.1 7.7 8.8 6.6 10.6 7.9 manufactures Textiles 14.7 11.5 10.2 14.5 14.9 11.8 18.3 13.6 14.1 11.5 10.3 9.6 Total less 7.0 5.7 7.6 7.6 7.1 6.1 8.8 7.3 8.0 5.8 8.5 6.7 military Total non-oil 6.1 4.3 7.7 7.8 7.3 6.6 7.7 6.4 7.9 6.1 9.6 7.9 trade Note: Tariffs refer to Trade-weighted MFN average Source: Trade Analysis Information System Database prepared by the United Nations Conference on Trade and Development. 146 Chapter 2 The list of countries included in international indicators can be found in the corresponding source of each indicator Doing Business, WEF, LPI, etc. The descriptive statistics using Enterprise Survey cover a sample of more than 60 countries, including some from the MENA and SAR regions. However, due to data availability, the regressions of section 2.4 only use a sample of 47 countries. Those countries are: Table A.2. 1. Countries included in regression Central America Eastern Europe & Central Asia Costa Rica Albania El Salvador Armenia Guatemala Belarus Honduras Bosnia and Herzegovina Nicaragua Bulgaria Panama Croatia Estonia Rest of LAC Georgia Argentina Hungary Bolivia Kazakhstan Chile Kyrgyz Republic Colombia Latvia Ecuador Lithuania Mexico Macedonia, FYR Paraguay Moldova Peru Poland Uruguay Romania Russian Federation Africa Slovak Republic Madagascar Slovenia Mali Tajikistan Mauritius Turkey South Africa Ukraine Uzbekistan East Asia & Pacific Hi-OECD Philippines Czech Rep. Ireland Spain 147 Table A.2. 2: Obstacles for business according to entrepreneurs’ subjective perceptions Hi- CA LAC AFR EAP ECA MENA SAR Total OECD Access to Finance 22.4 22.4 44.4 22.4 33.8 22.1 57.1 38.9 24.5 Access to Land 20.3 8.9 17.3 7.2 10.3 8.5 35.0 16.8 11.5 Corruption 60.0 50.1 19.4 27.4 22.5 8.3 36.6 40.7 47.8 Courts 9.6 13.1 6.7 16.3 18.0 7.8 26.0 12.5 Crime 40.4 25.1 16.7 17.1 12.9 7.2 7.9 22.4 26.1 Customs & Trade regulations 15.2 9.8 13.4 17.3 14.1 7.3 23.3 23.6 11.6 Electricity 41.9 31.0 58.1 20.6 10.4 5.7 23.1 43.0 32.6 Labor Regulations 6.9 16.2 6.8 13.2 12.2 11.4 24.2 16.2 14.2 Licensing & Permits 14.8 17.6 9.4 10.5 12.9 9.1 25.0 14.6 16.4 Macro Instability 42.0 46.9 35.7 35.5 29.9 15.9 44.6 27.3 43.9 Practices from informal competitors 33.9 36.6 22.1 21.5 23.6 14.7 41.5 21.5 34.2 Tax Administration 23.9 27.3 18.7 18.9 26.4 22.4 44.7 35.1 26.4 Tax Rates 31.2 37.1 31.1 21.8 36.7 27.4 61.0 32.9 35.7 Telecommunications 13.0 14.2 8.9 8.8 4.3 5.2 8.0 11.7 11.4 Transportation 21.7 13.3 16.2 13.3 7.0 7.0 8.0 15.7 14.4 Workforce Education 26.6 24.8 13.7 16.8 13.7 10.7 28.4 15.2 23.7 Source: Enterprise Surveys 148 Table A.2. 3: Correlation between subjective and objective variables Practices of Tax Informal payments to… Corruption informal Tax rates administration competitors Manager's time dealing with "Get things done" 0.0516* regulations -0.0193 0.0398* 0.0365* Obtain public services 0.0983* Number of tax inspections 0.0051 0.0253 0.0199 Sales reported for tax purposes 0.0425* 0.0386* 0.0189 Crime Education Security cost 0.017 Training 0.0982* Crime losses 0.1266* Uses email to communicate 0.1053* Uses own website to communicate 0.0858* Electricity Has ISO certifications 0.0332 Power outages 0.1830* Foreign licenses 0.0398 Own/share generator 0.0523* New products 0.0703* Electricity from own/shared generator 0.1907* New processes 0.0629* Customs & trade Domestic shipment lost in transit due Wait to clear customs for Transport regulations to Exports (average) -0.008 Theft 0.0377 Exports (longest) 0.0656 Breakage 0.0614* Imports (average) 0.0209 Imports (longest) 0.0960* Note: the columns report the subjective variable, as measured by the proportion of firms that report each area to be a major or severe obstacle for business. The rows report the objective variables reported by firms. The star * means that the correlation between the subjective and objective variable is significant at 5 percent level. Source: Own calculations using Enterprise Surveys 2005/7 149 Table A.2. 4: Data description, 2007 Firm size Sector Small (<=19) Medium (20-99) Large (>=100) Manufacturing Services Other Costa Rica 63.0 25.4 11.7 100.0 0.0 0.0 El Salvador 39.5 37.2 23.2 67.4 27.8 4.8 Guatemala 39.8 38.1 22.0 62.8 28.4 8.8 Honduras 46.8 30.3 22.9 60.1 27.5 12.4 Nicaragua 57.1 33.1 9.6 76.4 19.7 4.0 Panama 57.1 29.8 12.1 40.2 40.4 19.4 Source: Enterprise Surveys Table A.2. 5: Data Description continued Exporter (*) Foreign owned (*) Number of observations Yes No Yes No Costa Rica 38.8 60.9 9.0 91.0 343 El Salvador 35.1 64.9 12.8 87.2 693 Guatemala 32.4 67.6 11.5 88.5 522 Honduras 20.6 79.4 14.4 85.6 436 Nicaragua 16.1 83.9 9.4 90.6 478 Panama 21.2 78.6 11.9 88.1 604 Note: (*) A firm is considered an exporter if at least 1% of firm’s revenue come from direct exports or at least 5% come from indirect exports. A firm is considered to be foreign owned if any percentage is owned by foreign individuals/companies/organizations Source: Enterprise Surveys 150 Table A.2. 6: Differences by firms’ characteristics, Central America average, 2007 Sales not lost Sales not lost Regulatory Uses foreign Provides Has ISO due to due to crime compliance technology training certification outages Small 97.8 71.5 96.9 13.5 26.1 6.9 Medium 97.3 75.0 98.3 19.0 46.8 12.8 Large 99.2 81.1 98.4 39.5 73.7 42.1 S vs M (p- value) 0.1760 0.0000 0.0477 0.0000 0.0000 0.0000 S vs L (p- value) 0.0059 0.0000 0.0000 0.0000 0.0000 0.0000 M vs L (p- value) 0.0293 0.2951 0.0007 0.0000 0.0000 0.0000 Sales not lost Sales not lost Regulatory Uses foreign Provides Has ISO due to due to crime compliance technology training certification outages Foreign owned 99.3 79.5 96.4 33.6 56.0 41.0 Locally owned 97.7 72.7 97.6 15.7 33.1 8.3 P-value 0.0190 0.0026 0.0035 0.0000 0.0000 0.0000 Sales not lost Sales not lost Regulatory Uses foreign Provides Has ISO due to due to crime compliance technology training certification outages Exporter 98.7 74.2 98.1 27.9 60.8 26.3 Non-exporter 97.7 73.3 97.3 12.8 25.2 9.4 P-value 0.0738 0.0028 0.0779 0.0000 0.0000 0.0000 Note: All variables are defined such as the higher value, the better. p-value for the test of difference in means. 151 Chapter 3 Political system measures Two regional public opinion surveys are examined in this Chapter, the Informe Latinobarómetro and the Latin American Public Opinion Project (LAPOP). They both cover the six Central American countries and contain a wealth of information on attitudes and beliefs about the democratic systems, as they has developed in several countries in recent years, in addition to related topics as perceptions of corruption and public insecurity. The first is older, but the latter has a more robust methodology and more elaborate country-specific reports. 1 The discussion that follows is based on a review of both projects’ published reports; for each topic, the specific source with more complete published data is used. Comparisons across countries and over time are reviewed on the following indicators: support for the current system (democracy) and trust in public institutions—judiciary, legislative, executive, electoral board, and political parties. Measures are simple because that is the way they have been framed in the region for over a decade. The construction of scales—sometimes drawing on new complementary measures—is more recent and limits comparability over time. Since 1996, support for the democratic system has waned El Salvador, Guatemala, Honduras and Panama, while it has increased Costa Rica and Nicaragua. The percentages shown in Table A.3.1 in bold are lows in support for democracy and highs in support for authoritarianism that were reached in 2007. They represent extreme figures not only in comparison with other Central American countries, but also with the rest of the Latin American and Caribbean (LAC) region. Only Paraguay registers similar levels of support: low for democracy (33 percent) and high for authoritarianism (36 percent). 2 Over the past two years, support for democracy increased in only five Latin American countries, among them Costa Rica, Panama, and Nicaragua. On the other hand, El Salvador, Guatemala, and Honduras show the greatest decreases, along with Chile and Argentina. For recent years, then, Central American countries lie at the two ends of the spectrum. Looking over a longer span, only Costa Rica and Panama have higher percentages expressing a preference for democracy in 2007 than in 1996. Central America‘s average changes, however, parallel those for the rest of the LAC region. 1 Overall, LAPOP results present a more optimistic picture of system support. A donor-funded review of the LAPOP surveys concludes that differences with Latinobarometer result on conceptually equivalent items may be the result of an urban bias in the latter’s samples (Mishler and Bratton, 2005). Both surveys, however, are consistent (highly correlated) in their country rankings. Despite LAPOP’s methodological strengths, its published data are less complete and cover a shorter period, which is why Latinobarometer results are used for two of the three indicators. 2 Although levels of statistical significance are not reported by the source, it is highly probable most of the differences are significant, given their size and the large samples (at least 1,000 per country). 152 Table A.3. 1:Support for political system over time (1996 – 2007) in Central America In certain circumstances an Democracy is preferable to any other authoritarian government can be type of government (%) preferable to a democratic one (%) 1996 2001 2006 2007 1996 2001 2006 2007 Costa Rica 80 71 75 83 7 8 9 5 El Salvador 56 25 51 38 12 10 15 20 Guatemala 50 33 41 32 21 21 35 33 Honduras 42 57 51 38 14 8 12 17 Nicaragua 59 43 56 61 14 22 14 10 Panama 75 34 55 62 10 23 19 13 Rest of LAC 61 49 58 55 18 21 17 18 Source: Informe Latinobarómetro 2007 A complementary measure of system Table A.3. 2: Evaluation of the performance of democracy support involves an evaluation of democracy found in asking: “Would you Would you say you are satisfied with the say you are satisfied with the way way democracy is working? Here, “Very democracy is working in [country]?� satisfied� and “Satisfied� (%) While preference for democracy 2004 2006 measures a more abstract type of Costa Rica 75 59 support, satisfaction taps respondents’ El Salvador 61 47 perceptions of whether democracy is Guatemala 58 41 delivering the political, economic, and Honduras 64 54 social goods they expect. As reported by Nicaragua 54 43 LAPOP, the percentage of people Panama 52 46 expressing satisfaction with democratic Sources: LAPOP performance decreased by an average of 12 percentage points across all six countries between 2004 and 2006 (Table A.3.2). 153 Other graphs and tables Table A.3.3: WGI scores and confidence intervals* for Central American countries Costa Rica El Salvador Guatemala Honduras Nicaragua Panama Indicator score score score score score score (CI) (CI) (CI) (CI) (CI) (CI) Voice & 0.84 -0.05 -0.29 -0.34 -0.22 0.49 accountability (0.58 to 1.1) (-0.31 to 0.21) (-0.52 to –0.06) (-0.59 to -0.09) (-0.45 to 0.01) (0.21 to 0.77) Political stability 0.93 -0.07 -0.82 -0.46 -0.44 0.09 (0.55 to 1.31) (-0.48 to 0.34) (-1.20 to -0.44) (-0.84 to -0.08) (-0.82 to -0.06) (-0.29 to 0.47) Government 0.29 -0.26 -0.67 -0.6 -0.97 0.12 effectiveness (0.01 to 0.57) (-0.56 to 0.04) (-0.95 to -0.39) (-0.88 to -0.32) (-1.25 to -0.69) (-0.16 to 0.40) Regulatory 0.44 0.13 -0.09 -0.44 -0.48 0.33 quality (0.13 to 0.75) (-0.18 to 0.44) (-0.40 to 0.22) (-0.75 to -0.13) (-0.79 to -0.17) (0.02 to 0.64) Rule of law 0.55 -0.52 -1.02 -0.88 -0.76 -0.13 (0.32 to 0.78) (-0.78 to -0.26) (-1.25 to -0.79) (-1.11 to -0.65) (-1.01 to -0.51) (-0.36 to 0.10) Control of 0.37 -0.18 -0.7 -0.78 -0.76 -0.28 corruption (0.11 to 0.63) (-0.49 to 0.13) (-0.96 to -0.44) (-1.04 to -0.52) (-1.06 to -0.46) (-0.54 to -0.02) Scores range from -2.5 to 2.5, intervals at a 90 percent confidence level are in parentheses. Source: Worldwide Governance Indicators Table A.3.4: Number of indicators by topic showing changes from 2007 to 2008 and direction of change* Topic Indicators CRI SLV GTM HND NIC PAN Summary Starting a Business 4 1+ 3+ 4+ 3+ 1+ 1+ 13+ Dealing with 3 1+ 1+ 2+ 2+ 1+ 2+ 9+ Licenses Employing Workers 6 -- -- -- 2+ -- -- 2+ Registering Property 3 -- 2+ 2+ 1+ 1+ -- 6+ Getting Credit 4 2+ 2- 2+ 4+ 2+ 1- 8+, 3- Protecting Investors 4 -- -- -- -- -- -- -- Paying Taxes 3 -- -- 1+ -- -- 1+ 2+ Trading across 6 3+ 2+, 1- 4+, 2- -- -- -- 8+, 3- Borders Enforcing Contracts 3 -- -- -- -- -- -- -- Closing a Business 3 1+ -- -- 1+, 1- -- -- 2+, 1- Summary 8+ 8+, 3- 12+, 2- 13+, 1- 5+ 4+, 1- 50+, 7- *Improvements are pluses, declines minuses, and no change is a double dash. Source: authors' calculations using Doing Business 2008 154 Table A.3.5: Firm-reported country indicator values as compared* to LAC average (2006) Indicators CRI ** SLV GTM HND NIC PAN Summary Regulations & Tax 2 2+ 1+, 1- 1+, 1- 2+ 2+ 2+ 10+, 2- Permits & Licenses 3 N/A 2+ 2+, 1- 1+, 2- 2+, 1- 1+, 1- 8+, 5- Corruption 11 N/A 7+, 4- 5+, 5- 3+, 6- 4+, 2- 9+, 1- 28+, 18- Informality 3 N/A 1+, 2- 3 1+, 2- 2- 1+, 2- 3+, 8- Trade 2 2+ 1+ 1+, 1- 1+ -- 1+ 6+, 1- Summary N/A 12+, 7- 9+, 8- 8+, 10- 8+, 5- 14+, 4- 55+, 34- * A “plus� means the country indicator value is more business-friendly than the LAC average, a “minus� the contrary. ** Costa Rica data are for 2005 and some indicators were not measured then (N/A). Source: authors' calculations using Enterprise Surveys 155 Chapter 4 Table A.4.1: Net Generation Private and Public Public 2000 Private 2000 Public 2006 Private 2006 Costa Rica 83.7% 16.3% 83.4% 16.6% El Salvador 56.3% 43.7% 35.4% 64.6% Guatemala 41.3% 58.7% 31.1% 68.9% Honduras 60.5% 39.5% 32.2% 67.8% Nicaragua 45.0% 55.0% 18.9% 81.1% Panamá na na 14.8% 85.2% Source: Author’s calculations based on CEPAL and OLADE Table A.4.2: Sector Organization and Private Participation % Private % Public # Gen. # Dist. # Large # Wholesale Generation Generation Cos Cos Consumers Suppliers Panama 85%, of which 37% hydro 19 3 64% hydro Costa Rica 17%, mostly 83%, of 37 1 mayor, 8 There is an hydro which 92% minor association renewables Nicaragua 62%, of which 61% 1 mayor, 4 65% thermal and Thermal and multiple geothermal Hydro small El Salvador 65%, of which Mostly 15 5 4 10 57% thermal Hydro Honduras 68%, mostly Mostly 29 1 1 0 thermal Hydro Guatemala 69%, of which 100% hydro 37 17 43 8 thermal 79%, Source: CEPAL 156 Chapter 6 This annex contains two sections. The first details the different components of a functioning, internationally-recognized quality infrastructure (QI). The second describes the linkages between technical barriers to trade, DR-CAFTA, and QI. 3 1. Quality Infrastructure Components The term Quality Infrastructure (QI) denotes the totality of the infrastructure (public and private) required to establish and implement standardization, metrology (scientific, industrial and legal), inspection, testing, certification (product and system) and accreditation services necessary to provide acceptable evidence that products and services meet defined requirements. Requirements can be demanded by authorities or by the market place. This term is slowly replacing limiting and confusing acronyms such as STQM, SQAM and MSTQ. Standards and technical regulations are related to each other. However, according to the WTO Technical Barriers to Trade (TBT) Agreement definitions, standards are voluntary in themselves, whereas technical regulations are mandatory. The concept of voluntary standards is not well established in Central America because traditionally government institutions were the standard setting bodies. The standards development process is fully described in the ISO/IEC directives and is mostly a process of adoption—except when there is a specific and locally defined need, product or market. Important elements of this process are the involvement of all the important stakeholders and their experience in the form of standardization committees in the respective fields and reaching a consensus among them. Furthermore, an adequate period for public comment is required. A National Standards Institute can be a private or public entity as long as it is a legal entity and officially recognized by the government. Standards can be classified as National Standards (e.g. ANSI, BS, DIN, etc.), Regional Standards (e.g. EN, COPANT etc.) and International Standards (e.g. ISO, IEC, ITU, Codex Alimentarius, IPPC, etc.). Furthermore, many standards used internationally are developed by industrial interest groups and national associations (e.g. GSM, ASTM, etc.). Standards can refer to high technology sectors, as well as to good practice methods of primary production, among others. The actual requirement for a standard in a national economy is mostly defined by the needs of a sector to develop its market. It is not the quantity of standards published, but the utility for the market which makes standardization efficient. The Central American states are all active members of COPANT (www.copant.org) and are stakeholders of the process of harmonization and cooperation that is a linkage of the regional interests with the global structures of standardization. The WTO TBT Agreement considers the development and the implementation of technical regulations in the case of national security requirements, the protection of human health and safety, the protection of animal and plant life, the protection of the environment and the prevention of deceptive practices. Technical regulations are in effect legislation and therefore 3 This annex is based on inputs from the technical cooperation arm of the German national metrology institute (Physikalisch-Technische Bundesanstalt, PTB). 157 authorities often develop them with the public interest in mind (i.e. without building consensus) before going to the legislature. If technical regulations are based on the consensus of certain stakeholders (for example an industrial lobby), they could be biased and would then violate WTO TBT Agreement requirements. Inquiry points as institutional units have to be installed, and the standard setting and technical regulations situation of each country made transparent to all parties. This permits clear monitoring and evaluation of the implementation of technical regulations. Implementation itself is based on inspection, market surveillance and the imposition of sanctions when products do not comply. Metrology is defined as the science and art of measurements or as a system of weights and measures. It encompasses experimental and theoretical determinations of a physical quantity at any level of uncertainty and in any field of science and technology. To specify the broad variety of those fields, metrology may be subdivided into scientific (or fundamental) metrology, industrial (or applied) metrology and legal metrology. 4 Calibration and traceability are core concepts in metrology. Calibration is used to establish the relationship between a standard (measurement reference standard defining the physical unit) input and the instrument’s output, achieved by means of a direct comparison against standards. Traceability, mostly obtained by calibration, refers to an unbroken chain of comparisons which all have stated uncertainties. Metrology Institutes provide essential metrology services to society and industry to guarantee the safety and the functions of all national infrastructure and to facilitate market access and trade. Industrial companies also define documentary standards and procedures to meet their particular needs for technically and economically competitive manufacturing. The measurement standards define what specific instrument technology will be used to measure each quantity, how often it will be measured and which definition of each quantity will be used as the basis for process control that their manufacturing and product specifications require. The regional metrology cooperation SIM (Sistema Interamericano de Metrología), where all the Central American countries are members, declares in its mission precisely “to promote and support an integrated measurement infrastructure in the Americas that ensures equity in the marketplace, improves the quality of life and facilitates international trade�. Conformity assessment is the demonstration that specified requirements have been fulfilled, including proceedings of and structures for inspection, testing and certification. Independent, third-party conformity assessment is required by both authorities and the market. Service providers therefore have an important role to fulfill, namely to bridge the confidence gap between the supplier and the purchaser/regulator. In advanced markets, the state is responsible for market surveillance, whereas conformity assessment is provided commercially. However, in many developing and transition economies, the state still requires testing and certification by its own laboratories, although the integrity of those institutes is frequently not sustained by the adequate means of conformity assessment. This can often lead to conflicts of interest, lack of 4 The Bureau International de Poids et Mésures (BIPM) is the global office for the concerns of all the members of the Comité International des Poids et Mésures (CIPM, that executes tasks in order of the GCPM, the general conference of the Meter Convention, founded in 1875), while the Organisation Internationale de Metrologie Legale (OIML, founded in 1955) defines the international key guidelines for the legal and regulatory subjects of metrology. 158 technical capability, “unofficial� payments that frequently lead to the duplication of efforts or an increase of costs. The two general kinds of certification are product and system certification. Product certification can be subdivided into six systems as defined in ISO/IEC Guide 67. Only a few internationally recognized schemes have evolved, mostly limited to the country of origin and maybe its neighbors, and much effort is spent on defining and establishing them. Management system certification operates at the international level, with the standard families of ISO 9000 and ISO 14000 having gained the highest profiles. The success of ISO 9000 has spawned many other, more focused certification schemes offered by industry or other groupings, such as FCCP, SA 8000, BRC, GLOBALGAP, EFQM etc. Obtaining such certification is a major barrier for SMEs in all countries because of the cost/utility relation of certification. The certification process starts with the supplier’s product or process and the quality control being assessed for compliance with the relevant standards. A certificate is issued, authorizing the supplier to utilize the certification mark on the product or promotional material. Furthermore, the supplier is subject to surveillance by the certification body on an ongoing basis which can lead to quite a large burden for smaller suppliers. Empirical data from Maskus, Otsuki and Wilson (2005) indicate that standard setting in developed countries can very significantly diminish exports of developing countries where standards cannot be met or increase production costs where certification is needed. Testing and inspection are performed in a variety of situations as support for the certification process to specify the chemical, microbiological or any physical characteristics of substances for the market requirements or to provide evidence of the compliance required by legal frameworks. The reliability of tests depends on the correct operations and methods and the accuracy of testing and measuring equipment. The complexity of chemical matrices and parameters makes it difficult for developing economies, especially in the field of primary agricultural production, medical health, and basic environmental monitoring, to create structures that can compare with the standard setting resources of the export markets. In the case of import inspection for instance, special criteria are often established rather than accepting conformity assessment evidence from the countries of origin, thereby creating technical barriers to trade. The elements of quality infrastructure described above provide the technical expertise and confirm the compliance of products, persons and processes with standards. Accreditation, the final quality assurance, however, takes place if there is any formal recognition that a body or person is competent to carry out specific tasks in a respective field of service. That can only be done in an independent and transparent manner by the recognized accreditation body, following the established criteria of the ISO/IEC 17011. In order to demonstrate their technical competence, conformity assessment service providers are accredited against ISO/IEC 17020, 17021, 17025, 17024, 15189 or ISO/IEC Guide 65, 62. The accreditation process is coordinated by two global cooperation platforms: International Laboratory Accreditation Cooperation (ILAC) for the laboratory accreditation and International Accreditation Forum (IAF) for the accreditation of system certification. Both are structured in American/IAAC, European/EA, and Asian-Pacific/APLAC subgroups, who hold Mutual 159 Recognition Arrangements signed with ILAC/IAF and the African Accreditation Cooperation that finds itself still in a consolidation process. Accreditation represents the top of the whole system of Conformity Assessment and thereby confirms the competences of an organization. This refers to certification bodies, inspection bodies, and calibration and testing laboratories. Furthermore, the accreditation body is, in many cases, entrusted by government to approve conformity assessment bodies in the field of technical regulation. Enterprises use the services of accredited institutes to ensure and to prove the high quality of products or services required by their clients. The increasing trust in their products is reflected by the increasing trust in the conformity assessment services that are growing as parts of the quality infrastructure. To increase the potential of national QI is particularly important for SMEs that do not have the capacities to create an internal quality system to control their own supply chain. They have to rely on a QI that is sustained through the accreditation by an internationally- recognized, nationally-located body. Market surveillance is a responsibility of the state and it is required where, beneath third party conformity assessment service, the suppliers are responsible for proving compliance with technical regulations by testing and certification of products. Market surveillance checks to see if technical regulations are met. If not, non-conforming products have to be removed from the market or sanctions have to be applied. 2. Technical Barriers to Trade, DR-CAFTA, and Quality Infrastructure The Technical Barriers to Trade Agreement lays the framework down for mutual interaction between all members of WTO. The member states of the WTO have pledged to eliminate unnecessary technical barriers to trade through the Agreement on Technical Barriers to Trade (TBT Agreement). The Agreement aims to reduce unnecessary technical barriers to trade through the use of international standards, more transparency, and bilateral negotiations. It follows the principle of non-discrimination to ensure that products imported from another WTO member are accorded the same treatment as products produced locally or by another member. It also adheres to the principles of the harmonization of technical regulations as widely as possible through the use of international standards where reasonable, and to those of transparency, by notifying the TBT Committee of draft technical regulations and conformity assessment procedures, as well as by the establishment of national enquiry points. These measures aim to ensure that products imported from another WTO member are accorded the same treatment as products produced locally. The TBT Agreement encourages the mutual recognition of conformity assessment procedures in a general manner. Furthermore it takes into consideration the large challenge this agreement signifies, especially for developing countries, through the possibilities of "special and differential treatment of developing country members" and the encouragement of "technical assistance to other members". Nevertheless, it is the responsibility and the commitment of every member state to implement the necessary quality infrastructure, to assure a meaningful development of technical regulations and standards, and to implement conformity assessment procedures to fulfill the TBT Agreement commitments. Conformity assessment is stated as an essential part of trade and therefore needs to be considered for both developed and developing economies. Chapter 7 of the DR-CAFTA 160 Agreement refers strongly to the TBT Agreement and encourages the parties to recognize each others' conformity assessment bodies and results, to accept each others' equivalent technical regulations, and to strengthen transparency. The principle of non-discrimination is echoed in the statement that "each Party shall accredit, approve, license or otherwise recognize conformity assessment bodies in the territories of the other Parties on terms no less favorable than those it accords to conformity assessment bodies in its own territory". Chapter 7 of the DR-CAFTA Agreement can be seen as a concrete embodiment of the TBT Agreement. But the possibility of "special and differential treatment of developing country members" expressed in the TBT Agreement is not explicitly mentioned in Chapter 7 of the DR-CAFTA Agreement. It is likely that during negotiation of the agreement, the implementation challenges were underestimated and certain possibilities for favoring the Central American countries were omitted. The DR-CAFTA Agreement refers to the importance of measurement for quality assurance. One interesting aspect is that the DR-CAFTA Agreement makes the following special reference: "standards, technical regulation, or conformity assessment procedure includes those related to metrology". For experts, metrology, as one of the pillars of quality infrastructure, has always been an essential component for reliable tests, product certificates, accreditation, and traceability in value chains. Yet it is often not awarded the necessary significance in the political decision-making structure. This special reference to ensuring its meaningful implementation is groundbreaking. The unilateral definition of quality by highly developed economies means many countries are unprepared to benefit from the TBT and DR CAFTA process. The asymmetry in the strength of the trading partners is reinforced by the non-discrimination principle specified in the trade agreement. The technical rules and conformity assessment procedures of the more developed trading partners like the U.S. and European Community can be a significant trade barrier for the less developed Central American countries. In contrast, the technical rules and conformity assessment procedures of the weaker trading partners in Central America generally present no barrier to trade for the more developed economies. Under-developed quality infrastructure and conformity assessment procedures in Central America mean that the Central American SMEs can barely attain the high standards required by the target markets. At the same time, the Central American countries cannot demand high standards themselves, as their own SMEs would not be able meet them. Despite the apparently equal conditions in the trade agreements, stronger trading partners are thereby better positioned to exploit the trading and development potentials of the agreement unequally. The problem is added to from the view point of the consumer, in that the Central American floodgates are open to so called "substandard goods" from the industrial countries and countries from other regions. In this way a double urgency arises for Central American countries through the principle of non-discrimination, to strongly advance the establishment of a competent quality infrastructure which can be recognized along with conformity assessment procedures. It seems that some political decision makers, as well as parts of the private sector, in Central America are not well-informed about the challenges and responsibilities which derive from the trade agreements. 161 BIBLIOGRAPHY Chapter 1 Cheikhrouhou, H. et al. 2006. “The U.S.-Guatemala Remittance Corridor: Understanding Better the Drivers of Remittances Intermediation.� World Bank Working Paper 86. de Ferranti, D. G.E. Perry, I. Gill et al. 2003. Closing the Gap in Education and Technology. World Bank Latin American and Caribbean Studies. Washington D.C. Desruelle, D. and A. Schipke. 2007. “Economic Growth and Integration in Central America.� IMF Occasional Paper 257. Washington DC. Dollar, D., M. Hallward-Driemeier and T. Mengistae. 2005. “Investment Climate and Firm Performance in Developing Economies.� Economic Development and Cultural Change 54(1): 1-30. Fiess, N. 2007. “Business Cycles Synchronization and Regional Integration: A Case Study for Central America.� World Bank Economic Review 21(1): 49-72. Gould, D. and W. Gruben. 2005. “The Impact of DR-CAFTA on Economic Growth.� Processed report. Central American Country Management Unit. World Bank: Washington D.C. Hausmann, R., J. Hwang and D. Rodrik. 2005. “What You Export Matters.� NBER Working Paper 11905. Jaramillo, C. F. and D. Lederman. 2006. Challenges of DR-CAFTA: Maximizing the Benefits for Central America. Washington DC: World Bank. Kose, M.A. and A. Rebucci. 2005. “How might DR-CAFTA change macroeconomic fluctuations in Central America? Lessons from NAFTA.� Journal of Asian Economics 16(1): 77–104. Loayza, N., P. Fajnzylber and C. Calderón. 2005. Economic Growth in Latin America and the Caribbean: Stylized Facts, Explanations and Forecasts. Washington D.C.: World Bank. Montobbio, F. and F. Rampa. 2005. “The Impact of Technology and Structural Change on Export Performance in Nine Developing Countries.� World Development 33(4): 527-547. Ravallion, M., S. Chen, and P. Sangraula (2008). “Dollar a Day Revisited,� Policy Research Working Paper No.4620, The World Bank. Roache, S.K. 2008. “Central America’s Regional Trends and U.S. Cycles.� IMF Working Paper WP/08/50. Secretaría de Integración Económica Centroamericana (SIECA) (2007). “Central American Report 2004-2006.� Central American Report 3. Buenos Aires: IADB-INTAL. World Bank. 2004. A Better Investment Climate for Everyone. World Development Report 2005. Washington D.C. World Bank. 2006. Review of the Market for Remittances in El Salvador on the Basis of the CPSS-World Bank General Principles for International Remittances Services. Washington D.C. Processed. 162 World Bank. 2007a. Economic Performance in Latin America and the Caribbean: A Microeconomic Perspective. Volume 1: Main Findings. Report No. 40717-LAC. Washington D.C. World Bank. 2007b. Review of the Market for Remittances in Honduras on the Basis of the CPSS-World Bank General Principles for International Remittances Services. Washington D.C. Processed. Yuki, K. 2007. “Determinants of Sectoral Composition in a Small Open Economy: Theoretical and Quantitative Investigations of the Philippines. The Developing Economies 45(4): 391- 436. Chapter 2 Bernard, A.., J. Eaton, J. Jensen, and S. Kortum (2003). “Plants and Productivity in International Trade.� American Economic Review 93(4): 1268-90. De Ferranti, D., G. Perry, D. Lederman, and W. Maloney (2002). “From Natural Resources to the Knowledge Economy: Trade and Job Quality,� The World Bank, Washington, DC. Easterly, W. (1999). “Life during Growth: International Evidence on Quality of Life and Per Capita Income.� World Bank, Policy Research Working Paper 2110, The World Bank, Washington D.C. Escribano, A. and J.L. Guasch (2005). “Assessing the Impact of the Investment Climate on Productivity Using Firm-Level Data,� World Bank Policy Research Working Paper No. 3621. Jaramillo, C., and D. Lederman (2006). Challenges of CAFTA. Maximizing the Benefits for Central America. The World Bank. Kaufmann, D. and Kraay, A. (2007). Governance Indicators: Where Are We, Where Should We Be Going?� Policy Research Working Paper 4370, The World Bank, Washington, D.C. Lederman, D. (2007). “Product Innovation by Incumbent Firms in Developing Economies: The Roles of R&D Expenditures, Trade Policy, and the Investment Climate,� Policy Research Working Paper No.4319, The World Bank. Lederman, D., W. Maloney, and L. Servén (2003). Lessons from NAFTA for Latin America and the Caribbean Countries: A Summary of Research Findings. The World Bank. Loayza, N., P. Fajnzylber and C. Calderón (2005). Economic Growth in Latin America and the Caribbean. Stylized Facts, Explanations, and Forecasts. The World Bank, Washington, DC. Lopez-Cordova, E. (2007). “Labor Regulation and the Allocative Efficiency of International Trade,� mimeo, The World Bank, Washington, DC. Lucas, R. (1988). “On the Mechanics of Economic Development,� Journal of Monetary Economics, 22: 3–22. Melitz, M. (2003). “The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity.� Econometrica 71(6): 1695-1725. 163 Romer, P.M. (1986). “Increasing Returns and Long-run Growth,� Journal of Political Economy, 94(5): 1002–1037. World Bank (2004), El Salvador Investment Climate Assessment, Report No. 32711-SV. World Bank (2004), Guatemala Investment Climate Assessment: Evidence from the Manufacturing Sector, Report 35195. World Bank (2004), Honduras Investment Climate Assessment, Report No. 31458 – HN. World Bank (2005), Ecuador Investment Climate Assessment, Report No. 31900-EC. World Bank (2007), Costa Rica Investment Climate Assessment, Report No. 35424-CR. World Bank (2007), Panama Investment Climate Assessment. World Bank (2007). “Economic Growth in Latin America and the Caribbean: A Microeconomic Perspective,� Report No 40717-LAC, The World Bank, Washington, DC. World Bank (2007). World Development Indicators, Washington, DC. World Bank (2008), Guatemala Investment Climate Assessment. World Bank, (2005). World Development Report: A Better Investment Climate for Everyone, Washington, DC. World Bank, Doing Business, http://www.doingbusiness.org/ World Bank, Enterprise Surveys, http://www.enterprisesurveys.org/ World Bank, Logistics Performance Index, http://go.worldbank.org/4MQIU06Z00 World Economic Forum (2007). The Global Competitiveness Report 2007 – 2008, New York. Chapter 3 Arndt, C. and Oman, C. (2006). The Uses and Abuses of Governance Indicators. Development Centre Studies, OECD, Paris. Available at www.governanceindicators.org. Burnside, C. and Dollar, D. (1997). Aid, Policies, and Growth. World Bank Policy Research Working Paper, No. 1777, The World Bank, Washington, D.C. Chaffee, S., Morduchowicz, R., and H. Galperin (1997). Education for Democracy in Argentina: Effects of a Newspaper-In-School Program. International Journal of Public Opinion Research 9:313-39. Citrin, J. and Muste, C. (1999). Trust in Government. In J. Robinson, P. Shaver and L. Wrightsman (Eds.) Measures of Political Attitudes (pp. 465-532). San Diego: Academic Press. Corporación Latinobarómetro. Informe Latinobarómetro 2007. Santiago, Chile. Available at www.latinobarometro.org. Demombynes, G. and Hincapie, D. (2008) “Crime’s Cost to Firms in Central America: Analysis Using the 2006 Enterprise Surveys,� The World Bank, mimeo. 164 Finkel, S., Sigelman, L., and S. Humphries (1999). Democratic Values and Political Tolerance. In J. Robinson, P. Shaver and L. Wrightsman (Eds.) Measures of Political Attitudes (pp. 203-296). San Diego: Academic Press. Herman, B. (2004), “How Well Do Measurements of An Enabling Domestic Environment For Development Stand Up?� UNCTAD, Geneva. Available at www. g24.org/003gva04.pdf . Kaufmann, D. and Kraay, A. (2007). Governance Indicators: Where Are We, Where Should We Be Going?� Policy Research Working Paper 4370, The World Bank, Washington, D.C. Latin American Public Opinion Project. 2006 country reports for Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and Panama. Available at www.vanderbilt.edu/lapop. Mishler, W. and Bratton, M. (2005?). Independent Evaluation of CAM/C Democracy Surveys. MSI/USAID. Norris, P. (1999). Introduction: The Growth of Critical Citizens? In P. Norris (Ed.), Critical Citizens Global Support for Democratic Governance (pp. 1-27).New York: Oxford University Press. Rodrik, D. (2004). Getting Institutions Right (unpublished paper). John F. Kennedy School of Government, Harvard University, Cambridge, MA. Available at http://ksghome.harvard.edu/~drodrik/ifoinstitutions%20article%20_April%202004.pdf. Rodrik, D. (2008). Second-best Institutions (unpublished paper). John F. Kennedy School of Government, Harvard University, Cambridge, MA. Available at http://ksghome.harvard.edu/~drodrik/Second-best%20institutions%20paper.pdf. Romero, C. and Arndt, C. (2007). Some Comments on the World Governance Indicators 2007 for the Central American Countries (internal memorandum). The World Bank, LCSPS. Seligson, M. (2007). The Rise of Populism and the Left in Latin America. Journal of Democracy Vol. 18:3 pp. 81-95. UNDP (2004) Strategy Note on Governance for Human Development. Weil, Frederick (1989). The Sources and Structure of Legitimation in Western Democracies: A Consolidated Model Tested with Time-Series Data in Six Countries Since World War II. American Sociological Review Vol. 54: 682-706. World Bank, Logistics Performance Index, http://go.worldbank.org/4MQIU06Z00 Chapter 4 Andres, L., Guasch, J.L., Diop, M. and Lopez Azumendi, S. (2007). “Assessing he Governance of Electricity Regulatory Agencies in the Latin American and Caribbean Region: A Benchmarking Analysis,� The World Bank, Policy Research Working Paper 4380. Calderón, C., and L. Servén (2003). “The Output Cost of Latin American’s Infrastructure Gap�, in Easterly and Servén (eds.) The Limits of Stabilization—Infrastructure, Public Deficits and Growth in Latin America. CEPAL: Istmo Centroamericano, Estadísticas de hidrocarburos, August 2006 165 CEPAL: Istmo Centroamericano, Estadísticas del Subsector Eléctrico, October 2007 Consejo de Electrificación de América Central (CEAC), Plan Indicativo Regional de Expansión de la Generación Período 2007-2020, April 2007 Esfahani, H., and M. Ramírez (2003). “Institutions, Infrastructure and Economic Growth,� Journal of Development Economics 70: 443–77. Fernald, J.G. (1999). “Roads to prosperity? Assessing the link between public capital and productivity,� American Economic Review, 89 (3): 619–638. Guasch J.L., (2005). Granting and Renegotiating Infrastructure Concessions: Doing it Right. World Bank Institute Development Studies Report 28816. Honduras- Power Sector Issues and Options, World Bank, July 2007 Jaramillo, C., and D. Lederman (2006). Challenges of CAFTA. Maximizing the Benefits for Central America. The World Bank. Kent, P., and A. Fox (2004). “Is Puerto Lim6n a Real Lemon? Port Inefficiency and its Impacts,� Work Submitted to the Trade Capacity Building Project, U.S. Agency for International Development, Washington, D.C. Krugman, P. (2004). The Great Unraveling: Losing Our Way in the New Century. New York: W.W, Norton & Company. Loayza, N., P. Fajnzylber and C. Calderón (2005). Economic Growth in Latin America and the Caribbean. Stylized Facts, Explanations, and Forecasts. The World Bank, Washington, DC Nicaragua Policy Note, World Bank, June 2007 Roller, L-H. and L. Waverman (2001). “Telecommunications Infrastructure and Economic Development: A Simultaneous Approach,� American Economic Review 91 (4): 909-23. World Bank (2007), Costa Rica Investment Climate Assessment, Report No. 35424-CR. World Bank (2008). World Development Indicators, Washington, DC. World Economic Forum. The Global Competitiveness Report, various years Chapter 5 Banco Interamericano de Desarrollo (2005), Unlocking Credit, Informe para el Progreso Económico y Social. Bebczuk R. (2008), “SME Access to Credit in Guatemala and Nicaragua: Challenging Conventional Wisdom with New Evidence�, Background paper for this report. Bebczuk R. (2007), “Access to Credit in Argentina�, Serie Financiamiento del Desarrollo No. 188, ECLAC, March. Bebczuk R. and L. Garegnani (2006), “Autofinanciamiento y crecimiento económico�, Documento de Trabajo No. 6, Banco Central de la República Argentina. Beck T., A. Demirguc-Kunt, and V. Maksimovic (2004a), “Financial and Legal Constraints to Growth: Does Firm Size Matter?�, World Bank. 166 Beck T., A. Demirguc-Kunt, and V. Maksimovic (2004b), “Financing Patterns around the World: Are Small Firms Different?�, World Bank. Beck, T. and M.S. Martinez Peria (2007), Foreign Bank Acquisitions and Outreach: Evidence from Mexico. The World Bank Beck, T., A. Demirguc-Kunt, and M.S. Martinez Peria. (2006). Reaching Out: Access to and Use of Banking Services Across Countries. World Bank mimeo. Beck, T., A. Demirguc-Kunt, and R. Levine. (2004). Finance, Inequality and Poverty. World Bank Policy Research Working Paper 3338. Demirguc-Kunt, A., T. Beck and P. Honohan (2008), “Finance for All?: Policies and Pitfalls in Expanding Access,� A World Bank Policy Research Report. Djankov, S., McLeish, C., and Shleifer, A., (2005). Private credit in 129 countries. NBER working paper No. 11078. Fitch Ratings (2008), “Central American Banks: New Players in the Field.� Special Report. Levine, R. (2005), “Finance and Growth: Theory and Evidence�, in P. Aghion and S. Durlauf (eds.), Handbook of Economic Growth, Elsevier: Amsterdam. Morales, A. and A. Schipke (2005), “Regional Integration and Financial System Issues.� Peachey, S., and A. Roe (2004). Access to Finance, a Study for the World Savings Banks Institute. Oxford Policy Management. Rajan, R. and L. Zingales (2003). Saving Capitalism from the Capitalists. Crown Business Division of Random House. Rivera, E. and A. Rodríguez (2006), “Competencia y Regulación en la Banco de Centroamérica y México: Un Estudio Comparativo.� Background paper for the Project IDRC/CEPAL “Reforzando la competencia en el Istmo Centroamericano: políticas e instituciones nacionales, coordinación regional y participación en negociaciones internacionales.� Chapter 6 Agapitova, Natalia (2005). “The Role of Social Networks for National Innovation Systems’ Dynamics�. Washington, DC: World Bank. Boone, J. (2000). “Competitive Pressure: The Effects on Investment in Product and Process Innovation.� RAND Journal of Economics 31(3): 549-569. Bosch, Mariano, Daniel Lederman, and William F. Maloney. 2005, “Patenting and Research and Development: A Global View.� Policy Research Working Paper 3739, World Bank, Washington, DC. Canton, Eric, Debby Lanser, Joelle Noailly, Marieke Rensman, and Jeroen van de Ven (2004). “Science to Industry Knowledge Spillovers: Theory and International Experiences.� De Ferranti, D. et al (2003). Closing the Gaps in Education and Technology. World Bank, Washington, DC. 167 Di Gropello, Emanuela (2005). “Barriers to Better Quality Education in Central America.� En Breve Series, No. 64 (March). Washington, DC: World Bank. Gould, David M. and William C. Gruben (1996). “The Role of Intellectual Property Rights in Economic Growth.� Journal of Development Economics, March 1996, v. 48, iss. 2, pp. 323- 50. Griliches, Z. (1992). “The Search for R&D Spillovers.� NBER Working Paper 3768. Cambridge, United States: National Bureau of Economic Research. Guasch, J. Luis, Jean-Louis Racine, Isabel Sanchez, and Makhtar Diop (2007). Quality Systems and Standards for a Competitive Edge. Washington, DC: World Bank. Hall, Robert E, and Charles I. Jones (1999). “Why Do Some Countries Produce So Much More Output than Others?� Quarterly Journal of Economics 114: 83-116. Hausmann R. and D. Rodrik (2003). “Economic Development as Self-Discovery.� Journal of Development Economics, Volume 72. Jaramillo, Carlos Felipe and Daniel Lederman (2006). Challenges of CAFTA: Maximizing the Benefits for Central America. Washington, DC: World Bank. Jones, Charles I. and John C. Williams (2000). "Too Much of a Good Thing? The Economics of Investment in R&D.� Journal of Economic Growth, v. 5, issue 1 (March):65-85 Kim, L. (1997). Imitation to Innovation: The Dynamics of Korea’s Technological Learning. Cambridge, United States: Harvard Business School Press. Klinger, Bailey, and Daniel Lederman (2004). “Discovery and Development: An Empirical Exploration of ‘New’ Products.� Policy Research Working Paper 3450, World Bank, Washington, DC. Klinger, Bailey, and Daniel Lederman (2006). “Diversification, Innovation, and Imitation inside the Global Technological Frontier.� Policy Research Working Paper 3450, World Bank, Washington, DC. Kumar, N. (2002). “Intellectual Property Rights, Technology and Economic Development: Experiences of the Asian Countries.� CIPR Background Paper 1b. London, United Kingdom: Commission on Economic Property Rights. Lederman, Daniel, and Laura Saenz (2005). “Innovation and Development around the World, 1960-2000.� Policy Research Working Paper 3 774, World Bank, Washington, DC. Lederman, Daniel. (2007). "Product Innovation by Incumbent Firms in Developing Countries: The Roles of R&D Expenditures, Trade Policy, and the Investment Climate." World Bank Policy Research Working Paper 4319, The World Bank, Washington D.C. Lederman, D. and W. Maloney (2003). R&D and Development. World Bank Policy Research Working Paper 3024. Washington, DC: World Bank. Link, Albert N. and John T. Scott (2004). “The Role of Public Research Institutions in a National Innovation System: An Economic Perspective�. White Paper for Office of the Chief Economist, Latin America and Caribbean, World Bank Maloney, William. 2005, “Global Patterns of Innovation.� Mimeo. World Bank, Washington, D.C. PRELIMINARY 168 Maskus, K., and C. McDaniel (1999). “Impacts of the Japanese Patent System on Productivity Growth. Japan and the World Economy 11: 557-574. Maloney, W. F. and G. Perry (2005). “Hacia una Política de Innovación Eficiente,� Revista de la Cepal 87, pp. 25-44. OECD (2005). Science, Technology, and Industry Scoreboard. Paris: OECD. Rodríguez-Clare, Andrés (2005). “Innovation and Technology Adoption in Central America.� Inter-American Development Bank, Working Paper # 525. Washington, DC: Inter-America Development Bank. Rodrik, D. (1992). “Closing the Technology Gap: Does Trade Liberalization Really Help?� In: G. Helleiner, editor. Trade Policy, Industrialization and Development: New Perspectives. Oxford, United Kingdom: Clarendon Press. Salazar, S. (2002). “Innovación en Centroamérica: El Papel de las Patentes.� San Jose, Costa Rica: Universidad de Costa Rica. Mimeographed document. Sánchez-Páramo, Carolina and Norbert Schady (2003). “Off and Running? Technology, Trade, and the Rising Demandfor Skilled Workers in Latin America.� World Bank Policy Research Working Paper 3015. Washington, DC: World Bank. Sanetra, Clemens and Rocío Marbán (2007). The Answer to the Global Quality Challenge: A National Quality Infrastructure. Organization of the American States (OAS) and Physikalisch-Technische Bundesanstalt (PTB). SEDLAC (Socio-Economic Database for Latin America and the Caribbean) (2008). Online data based on collaboration between the Centro de Estudios Distributivos, Laborales y Sociales (CEDLAS) of the Universidad Nacional de La Plata, Argentina, and the World Bank. http://www.depeco.econo.unlp.edu.ar/cedlas/sedlac/default.html Vegas, Emiliana and Jenny Petrow (2008). Raising Student Learning in Latin America: The Challenge for the 21st Century. Washington, DC: World Bank. Maskus, Keith E., Tsunehiro Otsuki and John S. Wilson (2005). “The Cost of Compliance with Product Standards for Firms in Developing Countries: An Econometric Study.� World Bank Policy Research Working Paper Series #3590. World Bank (2005a). Central America: Education Strategy Paper. Report No. 29946. Human Development Department, Latin America and the Caribbean Region. Washington, DC: World Bank. World Bank (2006a). Chile Development Policy Review. Volume II: Main Report. Report No. 33501-CL. Washington, DC: World Bank. World Bank (2006b). Costa Rica Country Economic Memorandum: The Challenges for Sustained Growth. Report No. 36180-CR. Washington, DC: World Bank. 169