63261 DIREC TIONS IN DE VELOPMENT Trade Getting the Most Out of Free Trade Agreements in Central America J. Humberto López and Rashmi Shankar, Editors Getting the Most Out of Free Trade Agreements in Central America Getting the Most Out of Free Trade Agreements in Central America J. Humberto López and Rashmi Shankar, Editors © 2011 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org All rights reserved 1 2 3 4 14 13 12 11 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The bound- aries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. ISBN: 978-0-8213-8712-2 eISBN: 978-0-8213-8713-9 DOI: 10.1596/978-0-8213-8712-2 Library of Congress Cataloging-in-Publication Data Getting the most out of free trade agreements in Central America / J. Humberto López and Rashmi Shankar, editors. p. cm. ISBN 978-0-8213-8712-2 — ISBN 978-0-8213-8713-9 (electronic) 1. Free trade—Central America. 2. Central America—Commercial policy. 3. Central America— Foreign economic relations. I. Lopez, J. Humberto. II. Shankar, Rashmi. III. World Bank. HF1782.G48 2011 382'.90972—dc23 2011017660 Cover photos: Crane and cargo containers © Zhang Lianxun/Dreamstime.com; Trucks on a highway parking place © Ginsanders/Dreamstime.com; Kindergarten children learning to use computers © Monkey Business Images/Dreamstime.com. Cover design: Candace Roberts/Quantum Think. Contents Preface xvii Acknowledgments xxi About the Contributors xxiii Abbreviations xxviii Chapter 1 Getting the Most out of Central America’s Free Trade Agreements 1 J. Humberto López and Rashmi Shankar What Is the Expected Impact on Trade Volumes from Central America’s Efforts to Liberalize and Promote Trade? 5 What Is the Expected Impact on Growth from an Increase in Central America’s Trade? 8 What Is the Complementary Agenda for Promoting Trade? 9 What Are the Expected Welfare Effects of Trade Liberalization and Promotion in Central America? 18 Notes 24 References 24 v vi Contents Chapter 2 The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 27 Ana Cristina Molina, Maurizio Bussolo, and Leonardo Iacovone The Data 29 Export Trends 31 Firm-Level Patterns of Extensive Margin 34 Relationship between Tariff Reductions and Exporters’ Behavior: Preliminary Evidence 36 Empirical Strategy and Results 41 Conclusions 49 Notes 51 References 54 Chapter 3 Exports, Wages, and Skills: Implications for CAFTA 57 Irene Brambilla, Lucio Castro, and Guido Porto Firms in International Trade 58 Exporting, Productivity, and Wages: Causality 73 Conclusions and Policy Implications 77 Notes 79 References 79 Chapter 4 Trade and Economic Growth: Evidence on the Role of Complementarities for the DR-CAFTA Countries 83 César Calderón and Virginia Poggio Literature Review 85 The Data 88 Econometric Methodology 90 Empirical Assessment 91 Trade and Growth: The Role of Complementarities 97 Economic Implications of Our Estimates: Discussion for DR-CAFTA 110 Concluding Remarks 117 Notes 118 References 119 Contents vii Chapter 5 Power Integration in Central America: From Hope to Mirage? 123 Juan Miguel Cayo What Does Power Integration Mean? 125 The Political Economy of Integration 132 Power Integration in Central America: The SIEPAC Project 134 The Central America Power Sector in a Nutshell 135 Obstacles to Integration of the Central American Power Sector 140 Reality or Mirage? 144 Conclusions 146 Notes 147 References 149 Chapter 6 Supply Chain Analyses of Exports and Imports of Agricultural Products: Case Studies of Costa Rica, Honduras, and Nicaragua 151 Raquel Fernández, Santiago Flórez Gómez, Francisco Estrázulas de Souza, and Henry Vega Case Studies of Agricultural Trade 153 Methodology and Sources 154 Supply Chain Analysis: Intraregional and Extraregional Trade 156 Main Logistics Challenges 157 Quantitative Results for the Fresh Tomatoes Supply Chain 162 Quantitative Results: Wheat, Rice, and Corn Supply Chains 168 Conclusions 176 Notes 177 References 179 Chapter 7 Logistics Challenges in Central America 181 José A. Barbero The Relevance of Logistics as a Factor in Trade 181 The Impact of Logistics and Trade Facilitation on Trade Costs 182 viii Contents International Logistics Indicators 186 Country Logistics Review 193 Assessing Logistics Performance in Central America 199 Policy Priorities to Enhance Trade Logistics 208 Notes 211 References 212 Chapter 8 Access to Credit and Productivity in Central America 215 Inessa Love, Teresa Molina Millán, and Rashmi Shankar Productivity and Access to Financial Services in Central America 220 The Data 222 Estimating Productivity 227 Cross-Country Differences in the Relationship between TFP and Financial Products 236 Conclusions 239 Note 240 References 241 Chapter 9 Are Food Markets in Central America Integrated with International Markets? An Analysis of Food Price Transmission in Honduras and Nicaragua 245 Mario A. De Franco and Diego Arias Price Transmission of International to Domestic Prices of Food Products 249 Understanding the Price Transmission Results 255 Main Conclusions and Policy Implications 270 Notes 271 References 272 Chapter 10 Intellectual Property Rights and Foreign Direct Investment: Lessons for Central America 275 Walter G. Park Trends in IPRs and FDI in the DR-CAFTA Region 277 Regional Integration and FDI 286 Contents ix Intellectual Property Rights and FDI 291 Implications for DR-CAFTA 300 Concluding Remarks 304 Notes 305 References 306 Chapter 11 Trade Openness and Labor Gender Gaps 309 Maurizio Bussolo, Samuel Freije, Calvin Z. Djiofack, and Melissa Rodríguez Literature Review 310 Recent Trade Patterns in DR-CAFTA Countries 313 Labor Gender Gaps in DR-CAFTA Countries 314 Methodology 326 Results 328 Conclusions 339 Notes 341 References 342 Chapter 12 Trade Liberalization and Welfare Distribution in Central America 345 Maurizio Bussolo, Samuel Freije, Calvin Z. Djiofack, and Melissa Rodríguez Literature Review 347 Evolution of Poverty and Inequality in DR-CAFTA Countries 352 Methodology 356 Data Sources 358 Results 359 Conclusions 369 Notes 370 References 371 Chapter 13 DR-CAFTA and the Environment 375 Muthukumara Mani and Bárbara Cunha DR-CAFTA and the Environment 378 Trade and the Environment: A Review of the Literature 380 x Contents The Empirical Analysis 384 Results 388 Conclusions 397 Annex. The Data 397 Notes 401 References 401 Index 405 Box 1.1 Trade in Central America 2 Figures 1.1 Supply Chain Diagram of the Cost Contributions to the Average Price of Yellow Corn Used for Animal Feed in Nicaragua 12 2.1 Dominican Exports, by Destination, 2002–09 31 2.2 Dominican Exports to CAFTA Members in Select Years, 2002–09 33 2.3 Number of Firms, by Export Status, 2003–09 35 2.4 Decomposition of the Firm-Level Extensive Margin, 2003–09 37 2.5 Extensive Margin, by Tariff Cut, 2003–09 39 4.1 Correlation between Growth and Trade Openness in DR-CAFTA Countries 92 4.2 Correlations between Trade Openness and Growth, by Select Indicators of Economic Development 93 4.3 Growth Response to a 1 Standard Deviation Increase in Trade Openness, by Level of Select Indicators of Development 100 4.4 Growth Response to a 1 Standard Deviation Increase in Trade Openness, by Aggregate Stock of Select Infrastructure 112 4.5 Growth Response to a 1 Standard Deviation Increase in Trade Openness, by Level of R&D Spending and Regulations 114 5.1 Electricity Tariffs in Select Cities 124 5.2 Central American Power Generation, by Type of Power, 1985–2007 137 5.3 Electricity Exports, by Country, 1985–2008 138 6.1 Breakdown of Costs for a Small Exporter of Tomatoes 163 Contents xi 6.2 Breakdown of Costs for a Large Exporter of Tomatoes 165 6.3 Breakdown of Transport Costs for Small and Large Exporters 167 6.4 Cost Components as a Percentage of the Final Price of the Good, by Product and Country 172 6.5 Total In-Country Costs as a Percentage of the Final Price of the Good, by Product and Country 173 6.6 Logistics Costs (Transport and Other Logistics Costs) as a Percentage of the Final Price of the Good, by Product and Country 174 6.7 Other Logistics Costs as a Percentage of the Final Price of the Good, by Product and Country 176 7.1 LPI Rank of Central American and Comparator Countries, 2007 and 2009 187 7.2 Performance on the LPI Subindexes of Central American and Comparator Countries 188 7.3 Performance on Customs and Infrastructure of Central American and Comparator Countries, 2009 189 7.4 Quality of Overall Infrastructure of Central American and Comparator Countries, 2008–10 190 7.5 Performance on the Global Enabling Trade Index of Central American and Comparator Countries, 2009 191 8.1 TFP Growth Rates, by Region, 1981–2005 220 8.2 TFP Growth Rates in Central America, by Country, 1981–2005 221 8.3 Credit to the Private Sector and M2 as a Percentage of GDP in Central America, by Country, 2008 222 9.1 Transmission of Rice and Coffee Prices in Honduras 254 9.2 Transmission of Rice and Coffee Prices in Nicaragua 255 11.1 Gender Gaps in Labor Participation Rates (Women’s Minus Men’s) in DR-CAFTA Countries, 1990–2007 316 11.2 Gender Gaps in Unemployment Rates (Women’s Minus Men’s) in DR-CAFTA Countries, 1990–2007 318 11.3 Employment Shares, by Informal and Formal Status in DR-CAFTA Countries, ca. 1990–2007 319 11.4 Gender Gaps in Employment Shares (Women’s Minus Men’s Shares) in DR-CAFTA Countries, by Sector, ca. 1990–2007 321 xii Contents 11.5 Gender Gaps in Informal Share of Employment (Women’s Minus Men’s Shares) in DR-CAFTA Countries, ca. 1990–2007 322 11.6 Informal Share of Employment in DR-CAFTA Countries, by Gender and Sector, ca. 1990–2007 323 11.7 Monthly Wage Gender Gaps in Urban Areas (Female Minus Male Percentage Difference) in DR-CAFTA Countries 324 11.8 Monthly Wage Gender Gaps in Rural Areas (Female Minus Male Percentage Difference) in DR-CAFTA Countries 325 12.1 Costa Rica: Trade Simulation 363 12.2 Dominican Republic: Trade Simulation 364 12.3 El Salvador: Trade Simulation 365 12.4 Guatemala: Trade Simulation 366 12.5 Honduras: Trade Simulation 367 12.6 Nicaragua: Trade Simulation 368 13.1 Average Pollution per Year, before and after DR–CAFTA Negotiations (pounds, millions) 389 13.2 Decomposition in Total Emissions: Baseline Scenario 390 13.3 Decomposition in Total Emissions: Alternative Scenario 392 Tables 1.1 TFP Premiums of Exporters 7 1.2 Wage Premiums of Exporters 18 1.3 Environmental Regulatory Regime Index 23 2.1 Summary Statistics, Select Years, 2003–09 30 2.2 OLS Estimates of the Number of New Exporters 44 2.3 OLS Estimates of the Number of Exporters Adding New Product-Market Relationships 46 2.4 Estimates of the Probability of Exit 48 3.1 Exporting by U.S. Manufacturing Firms, 2002 59 3.2 Exporter Premiums in U.S. Manufacturing, 2002 60 3.3 Exporting and Importing by U.S. Manufacturing Firms, 1997 61 3.4 Trading Premiums in U.S. Manufacturing, 1997 62 3.5 Productivity and Wage Exporter Premiums in Latin America and the Caribbean 64 3.6 Exporting and Productivity Gains, by Investment Climate 66 3.7 Wage Premiums, by Investment Climate 69 Contents xiii 3.8 Exports and the Industry Skill Premium 72 3.9 Exports, Export Destination, and Wages: Wage Regression with Instrumental Variables 76 4.1 Trade and Growth: Baseline Regression under Different Estimation Techniques 94 4.2 Trade and Growth: Interaction with Structural Factors and Policies 98 4.3 Trade and Growth: The Role of Physical Infrastructure 104 4.4 Trade and Growth: The Role of Innovation 107 4.5 Trade and Growth: The Role of Regulations 109 4.6 Growth Effects due to Changes in Trade Openness 116 6.1 Breakdown of Costs for a Small Exporter of Tomatoes 164 6.2 Breakdown of Costs for a Large Exporter of Tomatoes 166 6.3 Supply Chain Analysis and Cost Contributions to the Average Price of Wheat Flour Sold in San Pedro Sula, Honduras 170 6.4 Supply Chain Analysis and Cost Contributions to the Average Price of Wheat Flour Sold in Managua 171 6.5 Breakdown of Cost Components 173 7.1 Performance on Doing Business Indicators of Trading across Borders of Central American and Comparator Countries 192 7.2 Freight Flows in Costa Rica, by Mode of Transport 194 8.1 Geographic and Demographic Penetration of Branches and ATMs in Central America, by Country 223 8.2 Distribution of Firms in Central America, by Country and Size of Firm 223 8.3 Use of Financial Products in Central America, by Size of Firm 224 8.4 Descriptive Statistics for Survey Variables 226 8.5 Correlations among Financial Indicators 227 8.6 TFP Estimation 229 8.7 Correlations among Estimated TFP and Observed Labor Productivity 229 8.8 Access to Credit and TFP value added 231 8.9 Access to Credit and Labor Productivity 233 8.10 Use of Financial Instruments in Central America, by Country 236 xiv Contents 8.11 Use of Financial Instruments, by Country and Size of Firm 237 8.12 OLS Regression: TFP, by Country 238 8.13 OLS Regression: Labor Productivity, by Country 239 8.14 Probit: TFP and Exports 240 9.1 Weight of Tradables of Select Food Products in Nicaragua, 2005 248 9.2 Growth in Domestic Prices in Nicaragua Given a Permanent Increase of 10 Percent in the International Price 251 9.3 Change of Consumer Prices in Honduras Given a Permanent Increase of 10 Percent in International Prices, by Region 253 9.4 Findings of Estimates of Price Transmission Analysis of Select Food Products in Honduras and Nicaragua 256 9.5 Number and Market Share of Large Agribusiness Companies in Nicaragua, 2005 258 9.6 Number and Export Share of Large Exporters in Nicaragua, 2005 258 9.7 Elasticity of Substitution of Demand for Select Food Products in Nicaragua, by Quintile 265 9.8 Elasticity of Substitution of Demand for Select Food Products in Honduras, by Level of Poverty 266 9.9 Importance of Demand Factors in Explaining Domestic Food Price Deviations in Nicaragua and Honduras 267 9.10 Price Wedges for Select Agricultural Products for Honduras, 2006 and 2007 269 9.11 Price Wedges for Select Agricultural Products in Nicaragua, 2006 and 2007 269 10.1 Strength of Patent Protection in DR-CAFTA Countries and Comparison Groups, 1990–2005 277 10.2 Intellectual Property Provisions in DR-CAFTA Countries, 2005–07 278 10.3 Correlations among Intellectual Property Measures 280 10.4 Flows of Foreign Direct Investment in DR-CAFTA Countries and Comparison on Groups, 1980–2008 282 10.5 Stocks of Foreign Direct Investment in DR-CAFTA Countries and Comparison Groups, 1980–2008 284 Contents xv 10.6 Amount of U.S. Foreign Direct Investment in DR-CAFTA Countries and Comparison Groups, Historical Cost Basis, 2004–08 286 10.7 U.S. Foreign Direct Investment in DR-CAFTA Countries and Comparison Groups, by Industry, 2004–08 Average 287 11.1 Sources of Data Used 315 11.2 Regressions of Coefficients on Trade Variables in the Labor Market Models 332 11.3 Summary of Results on the Impact of Trade Variables in Wage Models 335 11.4 Simulated Impact of Trade Liberalization through Different Indicators 337 11.5 Change in Explanatory Variables in DR-CAFTA Countries, by Sector 338 12.1 Data Sources 359 12.2 Price Changes in DR-CAFTA Countries in Select Periods 361 13.1 Environmental Regulatory Regime Index, 2001 379 13.2 Regression Analysis: CA-4 393 13.3 Regression Analysis: CA-4 Countries 395 13A.1 Regression Variables 398 13A.2 Full Regression Analysis: CA-4 Countries 399 Preface I still remember my first visit to Central America back in 1988. When I reflect on the dramatic changes that the region has gone through since then, it is difficult not to have a sense of optimism about the prospects of the region for the coming years. It cannot be forgotten that in the late 1980s four Central American countries were still involved in armed con- flicts in one way or another, and that it was not until the Esquipulas Agreement (1987) that a framework for peaceful resolution of the con- flicts emerged. Then, it was just fifteen years ago that, in 1996, following the Guatemala Peace Accords, Central America managed to build a durable peace that has lasted until today. For this, all Central Americans need to be congratulated. Peace came accompanied not only by the end to the human drama asso- ciated with the conflicts, but also by a significant economic dividend— a much needed development in a region where per capita GDP had stag- nated between 1970 and 1990 and where two countries (El Salvador and Nicaragua) had been experiencing negative average growth rates for more than two decades. The social dimension of the dismal growth per- formance is well captured in the poverty rates. According to World Bank statistics, in the first half of the 1990s the average poverty rate in the region was close to 60 percent in countries such as Honduras and xvii xviii Preface Nicaragua; almost three-quarters of the population lived on less than US$4 a day. In contrast, between 1990 and 2010 per capita growth averaged 2.4 percent (across countries and time periods) and the poverty rate declined to 44 percent of the population. In fairness, the economic dividend of the last two decades was due not only to the peace process but also to the sig- nificant modernization agenda implemented by all the countries in the region. A key element of these agendas was the promotion of interna- tional trade, an implicit acknowledgement that given the size of the dif- ferent Central American economies in the absence of international trade it would be difficult for domestic firms to specialize in areas of compara- tive advantage and exploit economies of scale associated with bigger mar- kets. As a result, today Central American countries are very open, with volumes of international trade (exports plus imports of goods and serv- ices) ranging from 70 percent of GDP in Guatemala to more than 150 percent of GDP in Panama. And yet, it is evident that the region aspires to further exploit the opportunities created by international trade. In addition to a signifi- cant number of bilateral free trade agreements (FTAs), the past few years have witnessed the conclusion of two important regional agree- ments: the Dominican Republic–Central America Free Trade Agreement (DR-CAFTA) that is now in effect between the United States, Costa Rica, Dominican Republic, El Salvador, Guatemala, Honduras, and Nicaragua; and the Association Agreement reached in May 2010 between the European Union and Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and Panama. I honestly think that the outward orientation showed by the region and the conclusion of the recent FTAs are part of the answer to the devel- opment challenges of Central America. But I also think that, taking into account that 19 million Central Americans still live in poverty, it is not enough to pursue the signature of FTAs. Policy makers in the region have to make an effort to make the best out of them and ensure that their ben- efits permeate to all segments of the population. This reflection is the motivation of this book. Several lessons emerge from Getting the Most Out of Free Trade Agreements in Central America, but I would like to stress three. First, Central America should not take the positive results of signed FTAs as a given. As these authors note, trade agreements create oppor- tunities but do not guarantee results. Indeed, the analysis in this book Preface xix indicates that it is up to the different countries to take the necessary steps to enhance the benefits of the agreements, including improve- ments in education levels that result in a more productive labor force, in public infrastructure so that the extremely high logistics costs in the region can be reduced, and in the energy matrix to improve access to reliable energy at competitive prices. Admittedly, this is a complex agenda that, realistically speaking, cannot be implemented overnight. And yet, if Central America wants to make the best of its trade agree- ments, this is the agenda it will need to progress on. Second, trade promotion needs to be complemented by a strong focus on the poor. In some cases, this focus is because of the challenges brought by additional external competition, which may negatively affect some industries or sectors. Perhaps even more importantly, in other cases, the benefits and opportunities created by an expansion of trade may not be dis- tributed evenly across the population and important groups could be excluded. The evidence in this book indicates that the skill premium may be on the rise in Central America. In other words, ongoing development forces are benefiting the highly educated more than those with fewer skills, therefore potentially contributing to increases in inequality in a region where it is already high. Third, is the need for more competitive markets. Although many of us tend to think about the benefits of growth in terms of quantities (that is, more exports, more employment, and increased access to goods) many of the welfare effects of FTAs are transmitted through prices (such as lower prices for imported goods). But for this transmission channel to operate, it is important that markets are competitive and that no agent has a pre- dominant market position that offers the possibility of capturing the rents created by lower tariffs. This is a complex issue in small economies where fixed costs may result in too many natural monopolies. This book does not come with a conclusive recommendation on this front, yet, I would like to invite policy makers and practitioners alike to explore whether fur- ther regional integration would help address this concern by increasing the market size. Given that at the time of this book’s publication I will have trans- ferred from the position I have held for the past three years, that of World Bank Country Director for Central America, I would like to share one final thought: Central America should, must, and can exploit all the existing opportunities to improve the standard of living of its population. As I mentioned, let’s not forget where the region was xx Preface twenty years ago, and let us not be shy of where the region can be in twenty years. Laura Frigenti Director, Strategy and Operations Latin America and the Caribbean Region The World Bank April 2011 Acknowledgments This report is the result of a collaborative effort of a large team led by J. Humberto López (Poverty Reduction and Economic Management Department) and Rashmi Shankar (Economic Policy Unit) of the Latin America and the Caribbean Region and including Diego Arias, José Barbero, Irene Brambilla, Maurizio Bussolo, César Calderón, Lucio Castro, Juan Miguel Cayo, Bárbara Cunha, Calvin Z. Djiofack, Francisco Estrázulas de Souza, Raquel Fernández, Norbert Fiess, Santiago Flórez Gómez, Mario A. De Franco, Samuel Freije, Leonardo Iacovone, Inessa Love, Muthukumara S. Mani, Teresa Molina Millán, Ana Cristina Molina, Walter G. Park, Virginia Poggio, Guido Porto, Melissa Rodríguez , Patricia Tovar, Riccardo Trezzi, and Henry Vega. Background papers not published in this volume may be found on the World Bank’s web site. Patricia Holt and Santiago Flórez Gómez provided invaluable support in all aspects of the production of this manuscript. We would also like to thank Tammy Lyn Pertillar for help on formatting. We owe a debt of gratitude to our peer reviewers: Chad Bown, Caroline Freund, Carlos Felipe Jaramillo, John Nash, and Jordan Schwartz from the World Bank; David Coady (International Monetary Fund); Pravin Krishna (Johns Hopkins University); and Marcelo Olarreaga, (Département d’Economie Politique Université de Genève) for their suggestions during xxi xxii Acknowledgments the preparation of this report. We also thank Patricia Katayama, acquisi- tions editor, and Janice Tuten, publications production manager in the World Bank’s Office of the Publisher. Finally, we would like to thank Rodrigo A. Chaves (director, Poverty Reduction and Economic Management, Latin America and the Caribbean Region), Laura Frigenti (director, Strategy and Operations, Latin America and the Caribbean Region), and Carlos Felipe Jaramillo (director, Central America Country Management Unit) for overall supervision and guidance to the team. About the Contributors Diego Arias is a senior economist in the Agriculture and Rural Development Division of the Latin America and the Caribbean Region of the World Bank. His recent publications have covered topics such as food policy and managing risks in the agriculture sector in Latin America. José A. Barbero is a senior transport specialist. He has worked in most Latin American countries on freight logistics, urban transportation, infra- structure planning, and transport institutional organization. After several years at the World Bank, he is currently an independent consultant. Maurizio Bussolo is a senior economist in the Economic Policy Unit of the Latin America and the Caribbean Region of the World Bank. He has been working on quantitative analyses of economic policy and development. Bussolo previously worked at the OECD, at the Overseas Development Institute in London, and at Fedesarrollo and Los Andes University in Colombia. He has published in several international journals and his recent publications include a volume on macro-micro modeling, edited with Francois Bourguignon and Luiz Pereira da Silva. He holds a Ph.D. in economics from the University of Warwick. xxiii xxiv About the Contributors César Calderón is a senior economist at the Regional Chief Economist’s Office for the Latin America and the Caribbean Region of the World Bank. Prior to this post, he worked in the Research Department at the Central Bank of Chile and the Central Reserve Bank of Peru and was an invited lecturer at the ILADES-Georgetown University Masters Program in Economics in Santiago, Chile. He was awarded his Ph.D. in economics from the University of Rochester in 2002. Calderón has published in the areas of open macroeconomics, growth, and development. He is currently working in issues such as globalization, economic fluctuations, and finan- cial development. Lucio Castro is the director of International Economics and Productive Development of CIPPEC (Center for the Implementation of Public Policies for Growth and Equity), a think tank based on Argentina. His recent publications have covered topics such as the effects of food prices on development and poverty, price shocks and energy consumption, the determinants of subnational growth and trade, among others. Juan Miguel Cayo is a senior energy specialist in the Latin America and the Caribbean Region of the World Bank. Before joining the Bank, he worked as the counselor for the executive director for Peru and Chile at the Inter-American Development Bank. He has had a long career in the Peruvian public sector. Cayo was the Vice Minister of Economy (2006) at the Ministry of Economy and Finance, responsible of the macroeconomic, social, and sectorial policies. In 2004 he was appointed Vice Minister of Energy, overseeing the subsectors of electricity and hydrocarbons. He was involved in the reform of the electricity sector in Peru of 2006 and in the development of both the Camisea Gas and the Peru LNG projects. Bárbara Cunha is an economist in the Economic Policy unit of the Latin America and the Caribbean Region of the World Bank. Her recent work has covered topics such as informality and development and credit short- ages in Latin America. She received a Ph.D. in economics from the University of Chicago. Calvin Z. Djiofack is consultant for the Economic Policy Unit of the Latin America and the Caribbean Region of the World Bank World Bank. His recent publications have covered topics such as trade in ser- vices, regional integration, migrations, fiscal policy, natural resources, and CGE models. About the Contributors xxv Francisco Estrázulas de Souza is an analyst at Castalia, a strategic advi- sory firm based in Washington, D.C. His latest work has focused on water and energy regulation, public-private partnerships in infrastructure, and management of public utilities. Estrázulas conducted this supply chain analysis as part of his master’s degree thesis at the Harvard Kennedy School of Government. Raquel Fernández is a junior professional associate at the Economics Unit of the Sustainable Development Department of the Latin America and the Caribbean Region of the World Bank. Her work has focused on the linkages between trade logistics and countries’ agricul- tural competitiveness. Santiago Flórez Gómez is a consultant at the Economic Policy Unit of the Latin America and the Caribbean Region of the World Bank. His work has focused on trade facilitation and competitiveness. He previously con- ducted research for the Inter-American Development Bank and the European Commission. Mario A. De Franco is an economic and business consultant for the World Bank and other development agencies. His recent publications have cov- ered the topics of social protection, microfinance, and trade and compet- itiveness in the Latin American region. Samuel Freije is senior economist for the Poverty and Gender Unit of the Latin America and the Caribbean Region of the World Bank. His recent publications refer to microsimulation and impact evaluation of social policies as well as to analysis of welfare and income distribution. He is also an associate editor of Economia, the journal of the Latin American and Caribbean Economic Association. Leonardo Iacovone is an economist currently working for the Private and Financial Sector Department of the Africa Region of the World Bank. Before joining the Bank, he served as economic advisor (ODI fellow) for the government of Mozambique and as a consultant for various international organizations (WTO, USAID, UNIDO, UNDP, DfID, and the European Commission). His research focuses on firm-level responses to challenges and opportunities of globalization, industrial dynamics, exports, commodity prices, and regional trade agreements. xxvi About the Contributors J. Humberto López is the lead economist for the Central America Department of the Latin America and the Caribbean Region of the World Bank. His recent publications have covered topics such as remittances and development, and the investment climate in Latin America. Inessa Love is a senior economist in the Finance and Private Sector Team of the Development Research Group. Since joining the World Bank as a Young Economist in 2001, her research has focused on access to external finance, entrepreneurship, the impact of financial crisis, and development of the domestic financial sector. She holds a Ph.D. in finance and econom- ics from Columbia University Graduate School of Business. Muthukumara S. Mani is a senior environmental economist in the Sustainable Development Department of the South Asia Region of the World Bank, based in Delhi. Currently, his work focuses mainly on cli- mate change mitigation and adaptation issues in India. He has also worked on country environmental assessments, pollution and natural resources management, environmental institutions and governance, cli- mate change and adaptation, trade, and environment issues. Mani has a number of publications in peer reviewed journals and has a Ph.D. in eco- nomics from the University of Maryland. Teresa Molina Millán is a consultant at the Economic Policy Unit of the Latin America and the Caribbean Region of the World Bank and a Ph.D. student in economics at the Paris School of Economics. Her research interests include access to finance at the firm and at the household level and rural to urban migration in developing countries. Ana Cristina Molina is an economist in the WTO’s Regional Trade Agreements Section. She has been working extensively on trade and development issues. Her areas of expertise include regional integration, export survival, trade diversification, and competitiveness. Among her previous assignments, she worked as a consultant for the World Bank and UNCTAD. Molina holds a Ph.D. in economics from the Graduate Institute in Geneva. Walter G. Park is an associate professor of economics at the American University. He researches international intellectual property rights (IPRs), measurement, and effects on innovation and technology diffusion. He has conducted projects on IPRs for the World Bank, OECD, European Patent Office, World Intellectual Property Office, and Industry Canada. About the Contributors xxvii Virginia Poggio is a research fellow at the Office of Evaluation and Oversight at the Inter-American Development Bank. Prior to this post, she worked at the Regional Chief Economist’s Office for the Latin America and the Caribbean Region at the World Bank. She holds a master’s degree in economics from Universidad de San Andrés (Argentina). She is cur- rently working in issues such as infant health, poverty, and economic development. Guido Porto is a professor of economics at the University of La Plata in Argentina. Before joining the University of La Plata, he was an economist in the research department of the World Bank. He received a Ph.D. in economics from Princeton University. Porto’s research focuses on the econometric estimation of the impacts of trade policies in developing countries, including impacts on poverty, household welfare, wages, and the distribution of income, as well as on firm behavior. Melissa Rodríguez is a consultant for the Poverty Reduction and Gender Unit in the Latin America and the Caribbean Region at the World Bank. Rashmi Shankar is a senior economist in the Economic Policy Unit of the Latin America and the Caribbean Region of the World Bank. Her recent work has focused on trade and trade facilitation in Central America. She has published widely in several areas including macroeconomics and international finance, growth, and international trade. Henry Vega is a research fellow at Center for Transportation Policy, Operations, and Logistics at George Mason University. His research has focused on assessing agricultural supply chains and on measuring the effects of air transportation costs on exports of perishables and high-tech goods. He holds a Ph.D. in public policy from George Mason University. Abbreviations ADDAPCA Proyecto de Diseño y Aplicación de Políticas Comunes Centroamericanas ADF augmented Dickey-Fuller ASEAN Association of South East Asian Nations ATM automated teller machine BEA Bureau of Economic Analysis, United States CBTPA Caribbean Basin Trade Partnership Act CCHAC Central America Hydrocarbons Cooperation Committee (Comité de Cooperación de Hidrocarburos de América Central) CEAC Central America Electrification Committee (Comité de Electrificación de America Central) CFE Comision Federal de Electricidad, Mexico COCATRAM Comisión Centroamericana de Transporte Marítimo COMITRANS Comité Técnico Regional Permanente de Transportes CRIE Comisión Regional de Interconexión Eléctrica DR-CAFTA Dominican Republic–Central America Free Trade Agreement EC European Community EKC environmental Kuznets curve xxviii Abbreviations xxix ENDESA Empresa Nacional de Electricidad, Spain ENEE National Electric Power Company, Honduras EOR regional system operator (ente operador regional) EPR owner of the grid (Empresa Propietaria de la Red) EPZ export-processing zone FDI foreign direct investment FTA free trade agreement GDP gross domestic product GMM generalized method of moments HIE high-income exports HS Harmonized System ICRG International Country Risk Guide IPPS Industrial Pollution Projection System IPR intellectual property right ISA Interconexión Eléctrica, Colombia IV instrumental variable kilometer 1,000 meters, 0.62 mile KPSS Kwiatkowski-Phillips-Schmidt-Shin LFTTD Linked-Longitudinal Firm Trade Transaction Database LPI logistics performance index LPM linear probability model LSMS Living Standards Measurement Survey MAGFOR Ministry of Agriculture and Forestry, Nicaragua MER regional electricity market (mercado electrico regional) MPS market price support MWh megawatt-hour NAFTA North American Free Trade Agreement NRCA normalized revealed comparative advantage index OECD Organisation for Economic Co-operation and Development OLS ordinary least squares OIRSA Regional International Organization for Farming and Livestock Sanitation OLI ownership, location, and internalization PPP purchasing power parity RCA revealed competitive advantage R&D research and development RICAM Red Internacional de Carreteras Mesoamericanas SEDLAC Socio-Economic Database for Latin America and the Caribbean xxx Abbreviations SIC Standard Industrial Classification SICA Central American Integration System (Sistema de la Integración Centroamericana) SIEPAC Central American Electrical Interconnection System SLS standardized logistics survey SME small and medium enterprise TFP total factor productivity TIM Procedimiento Mesoamericano para el Tránsito Internacional de Mercancías TO trade openness TRAINS Trade Analysis and Information System TRIPS Trade-Related Aspects of Intellectual Property Rights UNCTAD United Nations Conference on Trade and Development VAR vector autoregression WEF World Economic Forum WITS World Integrated Trade Solut CHAPTER 1 Getting the Most out of Central America’s Free Trade Agreements J. Humberto López and Rashmi Shankar Central America has put trade liberalization and the promotion of inter- national trade at the center of its development agenda. Over the past years, the region has witnessed the successful conclusion of negotiations for a significant number of free trade agreements (FTAs). Some of these FTAs have taken the form of bilateral agreements (for example, Costa Rica with Canada, Chile, Mexico, Panama, China, and Singapore; Honduras with Mexico), whereas others have been negotiated as a block. These include the historic Dominican Republic–Central America Free Trade Agreement (DR-CAFTA)1 between Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and the Dominican Republic with the United States and, more recently, the Association Agreement of the CA-6 (Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and Panama) with the European Union, which has yet to be ratified. Box 1.1 highlights the priority placed on trade by Central American policy mak- ers and the steady progress made on liberalizing tariffs. The priority given to trade liberalization in Central America’s develop- ment strategy is not surprising: trade is generally perceived as being both a benefit for growth and a means of advancement for developing coun- tries. Trade may contribute to faster growth through different channels, all of which are extremely relevant for Central America. First, trade openness 1 2 López and Shankar Box 1.1 Trade in Central America The Central American economies exhibit a high degree of trade openness relative to comparators. The volume of international trade (exports plus imports of goods and services) ranges from 70 percent of gross domestic product (GDP) in Guatemala to more than 150 percent of GDP in Panama. The composition of exports is quite different across the subregion. Nicaragua’s exports are dominated by agriculture and agricultural products, while Costa Rica, the Dominican Republic, and El Salvador largely export manufactures. Guatemala and Honduras export a mix of both agricultural and manufactured goods. The importance of international trade as an engine of growth for Central Amer- ica was recognized by the region well before the CA-5 (Costa Rica, El Salvador, Guatemala, Honduras, and Nicaragua) and the Dominican Republic signed the DR-CAFTA agreement with the United States. The General Treaty on Central American Economic Integration signed in Managua on December 13, 1960, already called for the creation of a common market and a customs union. The Central American countries have also long enjoyed preferential access to the U.S. market—since 1983—with the Caribbean Basin Initiative. By 2000, Central America had been extended the same terms as Mexico for apparel, and duty- free access was given to approximately 75 percent of Central America’s exports to the United States. In retrospect, the common market had succeeded in unifying external tariffs and removing duties on most products being traded among the mem- ber countries, leading to a dramatic increase in trade flows within the member nations. The decrease in tariffs by nearly 50 percent on average between 1995 and 2009 has been accompanied by an increase in trade openness for the region, with exports and imports increasing as a share of GDP for the CAFTA countries by about 30 percentage points. Another noteworthy trend has been the increase in intraregional trade. In 1960, 50 percent of Central American exports flowed toward the United States, and only 7 percent flowed toward other Central American countries. In 2010, the United States was still the main single market of the region, accounting for nearly 40 percent of exports, but the relevance of the region itself as a destination of exports has increased dramatically, and the region now accounts for more than 20 percent of exports. Source: Author calculations based on World Integrated Trade Solution (WITS), World Bank. Getting the Most out of Central America’s Free Trade Agreements 3 may improve a country’s access to foreign markets, allowing domestic firms to take advantage of economies of scale. Second, trade can enhance productivity through technological diffusion and transmission of know- how and managerial practices, thanks to stronger interactions with foreign firms and markets, and may provide innovators with new business oppor- tunities. And third, trade may enhance product market competition, thus reducing anticompetitive practices of domestic firms and leading to higher specialization due to exploitation of the comparative advantages of domestic firms. Beyond the positive impact on growth, trade also con- tributes to a better standard of living for the population. Empirical evi- dence supports the theoretical view that trade liberalization will alleviate poverty. There is also evidence that trade results in higher wages for those employed by exporter companies. Thus, the region should be congratu- lated without reservations for the effort made on this front so far. At the same time, it is important to recognize that the signing of the FTAs is not the end of the road, but rather the beginning. This is so for several reasons. First, the benefits of trade liberalization and promotion should be considered not in isolation, but rather in the context of the gen- eral policy and institutional framework in place in the different countries. As argued by Lederman, Maloney, and Servén (2007) in their evaluation of the North American Free Trade Agreement (NAFTA) after its tenth year of implementation, trade agreements create opportunities but do not guarantee results. Indeed, the benefits associated with trade agreements appear to be related to, among others, the quality of institutions, human capital, infrastructure, and the process of technological upgrading in the country in question. In other words, countries that want to get the most out of trade will have to create an enabling policy and institutional envi- ronment, which will entail structural reforms referred to as the “comple- mentary agenda.” Lederman, Maloney, and Servén (2007) also argue that the implementation of a sound complementary agenda is more profitable and, at the same time, more urgent in the context of an FTA (particularly if it has the relevance of CAFTA or the Association Agreement with the European Union) given the opportunities and also the challenges associ- ated with trade agreements and trade promotion. Second, even within a given country, the effects of trade vary widely across regions, firms, and workers. Trade liberalization may therefore need to be complemented with policy actions to ease the transition for those who benefit the least—or even who might be adversely affected. For example, workers with higher skills and education benefited more from NAFTA than those without—a phenomenon referred to as the skill 4 López and Shankar premium—suggesting that a well-targeted policy of investment in educa- tion and training was needed. Similarly, larger firms benefited more from NAFTA than smaller firms, which could be related to, among other factors, differential access to credit. States with higher initial levels of education, better infrastructure, and stronger local institutions did better during the first years of NAFTA, accelerating their rate of convergence toward the more prosperous North. Beyond NAFTA, Harrison (2007) reviews a series of case studies using firm- and household-level data and concludes that the poor are more likely to share the gains from trade integration in developing countries when complementary policies are in place. In the cases of India (Topalova 2005) and Colombia (Goldberg and Pavcnik 2005), the evidence points to the importance of policies that ensure labor mobility and, more generally, facilitate a smoother adjustment for poorer households. For Zambia, Balat and Porto (2007) conclude that poor farm- ers benefit from an increased exposure to international markets only when they also have access to credit, technical knowledge, and other complementary policies. Similarly, Cadot, Dutoit, and Olarreaga (2009) suggest that supply responses to trade-related changes in prices are more likely for rural producers when accompanied by appropriate complemen- tary factors such as access to inputs, information, credit, productive assets and capital, education, and quality land. In the context of developing countries, therefore, the evidence is that the opening of the economy to international trade normally entails access to higher prices for local exporters and potential exporters. However, the extent to which local firms are able to benefit from trade openness is closely related to struc- tural features of the economy that can be influenced by policy—the extent of market competition, for example, or logistical efficiency. Third, although in principle the overall impact of trade liberalization on the environment is likely to be country specific and could be either positive or negative, even the possibility of trade-related environmental degradation calls for policy attention. This impact may be due to one or a combination of factors. Trade leads to an overall expansion in output and, therefore, to an increase in emissions—also known as the scale effect. Trade liberalization affects resource allocation and production structure by changing the relative prices of goods, leading to either an increase or a decrease in the relative share of output of pollution-intensive sectors— known as a composition effect. Finally, changes in production technolo- gies (including pollution intensity per unit of output) could also affect overall emissions—typically referred to as the technique effect. While trade facilitates the access to and adoption of more efficient (and cleaner) Getting the Most out of Central America’s Free Trade Agreements 5 technologies of production, an increase in competition could trigger a race to the bottom on environmental standards, especially in the short run. Therefore, trade promotion needs to be accompanied by policy attention to environmental regulation and enforcement. This book has been prepared with the main purpose of exploring how the positive impact of trade can be augmented further and how the potential negative effects can be mitigated or offset. It complements and builds on Jaramillo and Lederman (2005), who analyze the chal- lenges of CAFTA and its potential benefits. In our view, the study’s tim- ing is highly appropriate given the new agreement with the European Union, yet to be ratified, and the relatively short implementation of DR-CAFTA.2 The rest of this chapter reviews the main findings and recommenda- tions of the 12 background papers presented in this volume, which ana- lyze the extent to which (a) trade liberalization and promotion will result in more trade; (b) higher trade flows will result in faster growth; and (c) trade-induced growth can be expected to be inclusive and sustainable. What Is the Expected Impact on Trade Volumes from Central America’s Efforts to Liberalize and Promote Trade? FTAs—and more generally, trade liberalization—are a means to expand trade. However, one of the findings of this study is that, unless Central America removes existing structural bottlenecks, this expan- sion will be modest. The work of Molina, Bussolo, and Iacovone in chapter 2 examines the export behavior of Dominican Republic exporters following implemen- tation of the DR-CAFTA in 2007 using a firm-level data set for the 2002–09 period. The study examines the impact on exports of tariff reductions through (a) the number of new exporters that entered the market, (b) the number of existing exporters that added a new product- market relationship to their export mix, and (c) the probability that an exporter would exit a given market. The evidence suggests that tariff reductions had a positive but very small effect on the number of new exporters as well as on the behavior of incumbents. This finding is interpreted as signifying that other trade barriers such as standards, phytosanitary requirements, credit constraints, and transport costs are constraining exporters from taking full advantage of the agreement. The Dominican Association of Exporters has identified three major constraints that undermine the ability of Dominican 6 López and Shankar exporters to compete in export markets. These are high electricity costs, high transport costs, as well as difficult credit access conditions. The removal of bottlenecks affecting firm competitiveness would be a neces- sary policy complement to trade liberalization. The relationship between export survival and tariff reductions is also positive, but modest. Survival among Dominican exporters is very low, with six out of 10 exporters exiting the export market after one year. Molina, Bussolo, and Iacovone use their model to test whether tariff cuts help exporters to consolidate their position in a market and diminish their probability of exiting. They find that tariff cuts do improve survival rates, but only marginally. Other important results concern the probabil- ity of survival for firms located in export-processing zones (EPZs). In general, these firms seem to perform better than their peers in the national territory. The probability of exiting the market after one year is 9 to 18 percent lower for EPZ exporters than for exporters located in the national territory. However, this may be due to self-selection—that is, the fact that better firms choose to locate in EPZs—rather than to the effectiveness of the favorable fiscal regime of EPZs. Trade liberalization and the associated tariff reductions are therefore generating a positive payoff in terms of additional trade flows in Central America. However, other factors are limiting the potential impact of trade liberalization on the expansion of trade flows. An important corol- lary is that these factors should be the focus of policy attention. One reason often used to justify trade policy is that exporters are more productive than nonexporters. For example, table 1.1 reports the esti- mated total factor productivity (TFP) premiums of exporters (that is, the difference in the productivity level enjoyed by an exporter over a non- exporter) for several Latin American countries. As the table indicates, all estimated premiums are positive. For the South American countries and Mexico, the average premiums are on the order of 39 percent. Although lower for Central America, suggesting scope for improve- ment in incumbent exporters, the estimated premium is still significant at about 13 percent. This, however, is just a correlation and tells us little about the direction of causality. In fact, the interesting question is whether exporters become good firms or whether instead good firms become exporters. This issue is discussed by Brambilla, Castro, and Porto in chapter 3. Under the first hypothesis (exporters become good firms), exporting improves productivity. The most common explanation, known as “learn- ing by exporting,” is that exporters acquire information from foreign Getting the Most out of Central America’s Free Trade Agreements 7 Table 1.1 TFP Premiums of Exporters Region and country Premium Latin America and the Caribbean 0.39 Argentina 0.60 Brazil 0.35 Chile 0.48 Colombia 0.30 Ecuador 0.09 Mexico 0.00 Peru 0.52 Uruguay 0.68 Central America 0.13 Costa Rica 0.01 El Salvador 0.36 Guatemala 0.15 Honduras 0.22 Nicaragua 0.02 Panama 0.02 Source: Casacuberta and others 2007. customers on how to improve the product design, the manufacturing process, or the quality of the good.3 Foreign demand also allows domes- tic firms—particularly in small countries—to take advantage of unex- ploited economies of scale. Under the second hypothesis (that is, good firms become exporters), the best firms self-select into export markets. One rationale for this self-selection is that important entry barriers exist in export markets because of the higher costs associated with selling in foreign markets (transport, but also distribution, marketing, and even production costs when firms need to adapt their product to foreign stan- dards). Thus only the more productive firms can enter foreign markets, and the observed differences between exporters and nonexporters can then be explained by preexisting differences. These two hypotheses are not mutually exclusive and are likely to be relevant to a different extent, creating a virtuous circle. But depending on which is the most important force, the policy implications can be very different. On the one hand, export promotion activities such as those already discussed are often justified on the basis of the learning-by- exporting explanation. On the other hand, the self-selection explanation would suggest that policy makers should focus on the internal determi- nants of productivity growth. The existing literature offers no clear-cut answer regarding the relative strength of the self-selection hypothesis 8 López and Shankar versus the learning-by-exporting hypothesis. Moreover, by nature, this lit- erature is country specific, and, depending on the country examined, studies seem to reach different conclusions. As discussed by Brambilla, Castro, and Porto in chapter 3, the work of Casacuberta and others (2007) surveys 54 studies (covering 70 countries) that look at the productivity premium associated with export activity, finding that the main reason behind the positive correlation of exports and productivity is self-selection: good firms become exporters, suggest- ing that penetrating foreign markets may require higher productivity. Casacuberta and others (2007) also find something relevant to the Central American region, namely that exporting is more likely to create productivity premiums in small countries, perhaps because of an economies-of-scale rationale for productivity premiums. Once again, the message that emerges from chapter 3 is that FTAs are potentially benefi- cial for the Central American countries, but they need to be comple- mented with actions aimed at improving firm productivity (such as an enabling business environment). What Is the Expected Impact on Growth from an Increase in Central America’s Trade? Given that trade has a positive, though modest, impact on growth, how can this impact be enhanced? The empirical literature on trade and growth has typically argued that growth is positively correlated with higher trade volumes, even after accounting for a variety of growth deter- minants. Edwards (1992), Dollar (1992), Ben-David (1993), Sachs and Warner (1995), Ades and Glaeser (1999), Frankel and Romer (1999), and Alesina, Spolaore, and Wacziarg (2000) are examples of this sort. However, a recent paper by Chang, Kaltani, and Loayza (2009) finds that, although trade stimulates growth, this effect appears to be dependent on the reforms undertaken in the economy. The authors specifically find that interactions among trade and structural factors such as human capital, financial depth, infrastructure, and economic regulations are statistically and economically significant and robust to changes in specification, econometric method, and openness measure.4 Calderón and Poggio further explore these issues in chapter 4 with a focus on Central America. These authors rely on an empirical growth model that is estimated using a sample of 136 countries over the period 1960–2010. Beyond allowing for an estimate of the impact of trade flows on the growth of Central American countries, the model also allows the Getting the Most out of Central America’s Free Trade Agreements 9 estimation of interactions between trade and other key structural factors. In practice, this is equivalent to an assessment of how the growth impact of trade changes when the policy environment changes—or, in other words, when countries make progress on what could be termed the “com- plementary agenda.” What are the main findings of chapter 4? As in the discussion in the previous section, Calderón and Poggio find that increases in Central American trade flows are associated with faster growth. Yet the estimates are quite modest. For example, they estimate that a 30 percent increase in the volume of regional international trade in Central America could result, on average, in a 0.16 percentage point increase in the annual growth rate of the region—hardly a dramatic increase. However, when Calderón and Poggio explore the impact on growth of a trade expansion that is accompanied by progress in, for example, improving infrastructure or raising educational attainment to the level of the benchmark countries, the model predicts that growth projections could be twice as large. What Is the Complementary Agenda for Promoting Trade? In this section, we review the findings presented in this book regarding the need to pay attention to key areas in the complementary agenda. These include infrastructure (differentiating between energy and logistics and transportation), human capital, access to finance, competition policy, and enforcement of intellectual property rights. Infrastructure Since the need to improve infrastructure may be too broad to implement in practice, chapters 5 to 7 dig deeper into three main areas: energy and logistics and transport-related infrastructure. Energy. Energy is a key input for production; as such, high energy prices put firms at a competitive disadvantage. The issue is particularly relevant in Central America, where energy prices tend to be high by regional stan- dards. In chapter 5, Cayo explores what Central American countries can do to address this problem and argues that several structural problems need to be overcome in regional energy markets. These include (a) a tight balance between power generation and demand, which adversely affects the reliability of supply and its quality; (b) significant exposure to oil price volatility and shocks due to excessive dependence on oil imports, which have increased with the growing reliance on thermal power sta- tions; (c) significant inefficiencies in the institutional and regulatory 10 López and Shankar framework of several countries, which affect the financial sustainability of power utilities and their operations; and (d) relatively low levels of access in certain countries, affecting rural areas in particular. Facilitation of the energy trade—and there are several integration initiatives, particularly in electricity—is perceived in Central America as being potentially effective for achieving cost competitiveness. International exchange of electricity could bring three major advantages: (a) lower operating costs through the use of the most economic energy resources, particularly through coordinated management of hydropower and thermal systems, (b) the possibility of balancing generation with cur- rent needs and accounting for seasonal variations through exporting or importing and pooling of reserve capacity, thereby improving efficiency and avoiding the extra cost of emergency power contracts, and (c) better risk management once the market is no longer constrained by the size of individual domestic economies, thereby improving investment in the sec- tor and supporting lower prices. The Central American Electrical Interconnection System project—without doubt the most ambitious integration initiative so far—physically links the six Central American countries (including Panama). However, for the intraregional power trade to increase significantly, there are four preconditions: political will, regional institutional capacity, harmonized regulatory frameworks, and investment in generation and transmission capacity. On the whole, the potential for improvement in energy collaboration in Central America is clear, as are the prospective benefits in terms of energy cost competitiveness. The main message is that strengthening institutionality and regulation are as important as attracting more invest- ment, and none of these actions is possible without significant political will and consensus. Strengthening domestic markets is part of the complementary agenda required to generate a strong, integrated power system. Logistics and Transport. Global competition has intensified the need for efficiency in transport and logistics systems, from the point of manufac- ture to delivery to the customer. Studies on the share of logistics costs in the final price of delivered goods reveal that these costs represent a greater barrier to trade than import tariffs, especially in the light of free trade agreements such as the DR-CAFTA. In fact, the World Bank has estimated that on average, ad valorem tariffs for food imports have decreased in the Latin America and Caribbean region from 2005 to 2008 Getting the Most out of Central America’s Free Trade Agreements 11 and currently range from 3 to 12 percent of product value. Transport and logistics costs, in contrast, measured in this case by the international maritime and road haulage components alone, can total about 20 percent of the free-on-board value of goods. These issues are explored by Fernández, Flórez Gómez, Estrázulas de Souza, and Vega in chapter 6, where the authors describe the results of an analysis of eight supply chains that follow the entire dis- tribution process to shed light on the share of logistics in total costs for a few products in Central America. This analysis was undertaken for Costa Rica’s exports of tomatoes to Nicaragua by a big and a small exporter and U.S. exports of rice, wheat, and corn to Nicaragua and Honduras. Trade in tomatoes presents an opportunity to study the difficulties in distributing perishable goods that require refrigeration. Moreover, by differentiating between a large and a small firm, one can observe whether small firms are particularly affected by logistics costs. Rice, wheat, and corn dominate the Central American food basket and are imported into Nicaragua and Honduras largely from the United States. The analyses in chapter 6 show that overall high domestic transporta- tion costs, along with bottlenecks at land border crossings, present the biggest hurdle to intraregional trade, such as between Costa Rica and Nicaragua, and to extraregional imports, such as grain shipments from the United States. The surveyed Central American exporters point to the lack of good-quality paved secondary roads, especially for linking farms with cities, which impedes intraregional commerce notwithstanding the rela- tively good condition of the major transit arteries. The poor road quality, in turn, causes direct losses from delays in shipments and breakage of 8 to 12 percent of the sales value of exported goods and is seen by a large share of local firms as presenting a severe obstacle to growth. Additionally, the costs related to the reception of the grains at the port of entry, including those stemming from phytosanitary and sanitary revi- sions, are important in some cases and are a potential source of cost saving through trade facilitation. On the whole, estimates of the logistics costs for the eight value chains in chapter 6 would range from 17 percent for a big Costa Rican tomato exporter to Nicaragua to 48 percent for yellow corn imports to Nicaragua, with typical logistics costs in the range of 30 percent or more for the other value chains (see figure 1.1). This compares negatively with Latin America, where logistics costs are estimated at about 25 percent, and with the Organisation for Economic Co-operation and Development 12 Figure 1.1 av er ag e fa rm pr ice average price (US$ per kilogram) Source: Authors’ calculations. tru in M 0 0.10 0.20 0.30 0.40 0.50 0.60 ck in in ba tra ne su rg ns so ra e po ta nc to rt re e, po to ce oc rt ba pt ea of rg io n Ne e n fre w la at ig Or nd po ht le tra rt ,b an ns an ro s po d ke rta cu r’s st pr tio om of n sc it to other logistics sil fe le ag ed ar an e m ce an an d sa ck w uf ac ar tu in eh g, lo re pa ss ou sin r ck es in ot du g g, he rin co rm g st de ex s liv illi profit margins er ng tra yt lo ct o ss io tra es n ns an ot he po d rta co rf st ixe tio s Note: The lines on each bar refer to the addition to costs from each category listed on the horizontal axis. d n an ve fe d hi cle ed va m ria s operating costs an bl uf e ac co tra st ns tu s po re r’s rta lo pr tio ad of n in it to g w in ho w tru le ho ck sa le s le sa le transportation rs rc or os po ts ul try w pl ho an le sa t le rp ro farm gate to fit ta lc os ts 7% 29% 41% 13% 10% Supply Chain Diagram of the Cost Contributions to the Average Price of Yellow Corn Used for Animal Feed in Nicaragua Getting the Most out of Central America’s Free Trade Agreements 13 (OECD), where logistics costs appear to be below 10 percent. It also suggests that the improvement of regional competitiveness may require a strong emphasis on policy interventions that target the distribution of goods after they leave the firm and that factor productivity at the firm level may offer just a very partial picture of the competitiveness problem- atic in the region. Against this background, Barbero argues in chapter 7 that a strategy to address logistics costs in Central America should consider a combination of elements including ports, transport services, freight security, and cus- toms modernization. More specifically, the most serious deficiencies are found in the quality and connectivity of roads and ports, domestic and regional surface transportation (mainly trucking), and security of surface freight. Border management and border-crossing facilities also reveal gaps, as do airports, international transport services (air and maritime), carriers’ ability to manage their supply chain efficiently, and logistics operators and intermediaries. Human Capital Firm productivity and exports are closely related, as discussed earlier in this chapter. A related question is how trade affects returns to individual skill endowments. This is a critical question for Central America since one potential channel through which trade liberalization can lead to impor- tant changes in the structure of an economy is by affecting factor rewards. For example, trade can result in shifts in production technology that favor goods with a higher component of skilled labor over unskilled labor. This is an important issue from a policy perspective because to the extent that trade liberalization is skilled biased—that is, favors skilled labor—policy makers may need to think of investing in people as a critical complement to trade liberalization. Brambilla, Castro, and Porto also explore in chapter 3 whether the implementation of CAFTA has resulted in any change in the skill pre- mium (that is, the ratio of skilled to unskilled wages) so far. They find empirical evidence consistent with an increase in the skill premium in Central America in recent years, which is not observed in other Latin American non-CAFTA countries. Chapter 3 also takes the additional step of exploring whether the estimated increase in the skill premium is driven by within-sector changes or between-sector compositional shifts in the skilled labor force, concluding that the former factor dominates. The authors also find that the export share of GDP seems to affect the skill premium positively and that both per capita GDP and skill composition 14 López and Shankar are significant determinants of the industry-skill premiums: richer countries seem to have greater disparities between skilled and unskilled wages, and, as expected, countries with a greater fraction (supply) of skilled workers pay smaller skill premiums. On the whole, the results call for investment in human capital if the region is to be in a better position to exploit the income-generating opportunities created by trade. This message is fully aligned with that of Calderón and Poggio in chapter 4 regarding the role of education as a nec- essary complement to using trade to boost growth in Central America. Additionally, although not uniform across the six countries of Central America or across all years, there is evidence of high and sometimes ris- ing returns to higher education in the subregion. At the same time, the coverage of tertiary education systems in Central America has increased, albeit starting from a comparatively low base, leading to an increasing supply of skilled labor. A rise in the skill premium in the presence of an increase in the supply of skilled workers is seen as evidence of skill-biased technical change. An increase in the supply of higher education graduates seems to be called for, especially as many middle-income Central American countries have lower educational attainment than other middle- income comparator countries. However, there are other factors to consider. Increasing returns may reflect the fact that there is a shortage of higher education grad- uates of good quality; hence, while coverage has expanded, not all higher education graduates have the necessary skills and knowledge. Much of the recent expansion of tertiary education in Central America has been through the private sector, and in the absence of rigorous accreditation standards the quality of courses is known to be variable. The majority of courses in the public universities are also not accredited. A second issue is that there can be imbalances in demand for and supply of specific disciplines, such as technology and engineer- ing. Enrollments in Central American universities are highly skewed toward disciplines such as law, literature, and the arts, with enrollment in applied science, engineering, and technology courses representing only a small fraction of the total. The content, teaching methodology, and assessment system in these courses are not aligned with interna- tional standards, leading to low quality. Reorganizing and modernizing higher education is thus a priority even if expanding coverage is nec- essary. More detailed follow-up of graduates of different disciplines and their labor market performance is also required to better under- stand the demand for skills. Getting the Most out of Central America’s Free Trade Agreements 15 Access to Finance Access to credit remains limited in Central America, and this is a criti- cal issue, as there is now a growing literature that emphasizes the role of the financial sector in enhancing the positive impacts of trade. Love, Molina Millán, and Shankar explore in chapter 8 the relationship between productivity, access to finance, and exports, noting the grow- ing evidence of the comparative advantage that financial development provides to exporters or firms entering foreign markets. For example, trade finance is found to be a critical part of the institutions that coun- tries need if they are to take full advantage of trade-related opportuni- ties. The work of Rajan and Zingales (1998), Demirgüç-Kunt and Maksimovic (1998), and Beck, Demirgüç-Kunt, and Levine (2001) shows that a well-developed financial sector helps countries to secure access to external finance for investment projects and puts them in a bet- ter position to execute new ideas and therefore innovate. More recently, Beck, Demirgüç-Kunt, and Levine (2003) and Svaleryd and Vlachos (2005) find a positive relationship between financial sector development and the specialization pattern of international trade and comparative advantage. The main findings of Love, Molina Millán, and Shankar are that in the Central American context (a) access to credit and productivity are posi- tively associated and (b) productivity and exports are also positively associated. Although it must be admitted that due to data limitations, chapter 8 does not allow interpretation of these correlations as signify- ing unidirectional causality from financial access to firm productivity, the analysis is suggestive of a close relationship between these variables. Moreover, there is now considerable empirical evidence justifying a focus on the financial sector, especially given that access to credit remains limited in Central America. For example, the composite indica- tor developed by Beck, Demirgüç-Kunt, and Martínez Peria (2007), which measures the percentage of the adult population with access to an account with a financial intermediary, indicates that the entire Central American region, especially Nicaragua, is well below the median of Latin America, suggesting that measures to improve financial access, especially of smaller firms, could facilitate trade. From a policy perspective, the work by Love (2009) gives two clear ideas of how to expand access levels. One is to make progress on the judi- cial front. Love presents empirical evidence indicating that the quality of courts is positively correlated with a number of measures of access (having a checking account, using credit, and so forth) and concludes 16 López and Shankar that court reform should promote wider use of the courts for resolving disputes and improve outcomes in terms of the percentage of cases that result in court judgments. Second, even though financial development cannot be equated with financial access, Love presents evidence indicating that higher levels of financial development are associated with longer loan maturity, larger size of loans relative to sales, and more likely use of land and buildings as col- lateral and that this trend appears to be driven by the access of small and median firms. In other words, measures aimed at modernizing the finan- cial sector can not only have a positive impact on overall access, but also be biased toward small and medium firms (which, as noted, have less access than larger firms). Competition Policy Trade allows improved resource allocation, enabled by the transmission of price signals to producers and consumers, which sets into motion a chain of adjustments—in production and consumption baskets and therefore in labor markets. As noted earlier, the labor market adjust- ments allow workers to realize the benefits of a skill premium, provided the institutional environment permits the necessary reorientation in human capital formation. To understand the potential supply and demand responses, De Franco and Arias estimate in chapter 9 the degree of transmission of international prices to domestic prices of key agricul- tural commodities in Nicaragua and Honduras, analyzing the extent to which a change in the international price of a given food product affects the domestic price of that same good, at the level of the consumer and producer as well as in different regions within each country. De Franco and Arias examine markets for key agricultural commodities (sugar, cof- fee, meat, beans, maize, rice, and vegetable oil) in Nicaragua and Honduras. The goods in the analysis represent a mix of imported and exported products and are highly traded in the subregion. The main finding is that price transmission is imperfect and has been persistently weak, with little improvement over time. The weakness of price signals suggests that welfare impacts of trade liberalization may not be fully realized and that the failure of price signals to transmit can lead to sluggish growth and reduce the gains to consumers, who would nor- mally benefit from the competition offered by cheaper imports. Imperfect price transmission can be explained by several factors, but the two most plausible explanations for Central America are (a) the existence of noncompetitive market structures, where one agent has significant Getting the Most out of Central America’s Free Trade Agreements 17 market power, or (b) the costs of price adjustments at some point within the supply chain. This would, in turn, suggest a need not only to improve logistics (in line with the discussion above), but also to introduce compe- tition in relevant markets. Enforcement of Property Rights So far the discussion has centered on the benefits of trade. FTAs can also have a positive impact by encouraging foreign direct investment (FDI). FDI can not only enhance Central America’s opportunities to secure financing for private sector development, but also be a key instrument to transfer technology to the region. Yet this will be unlikely unless firms perceive that their intellectual property rights (IPRs) are protected. IPRs have traditionally been regarded as a means to encourage research and development. However, they are increasingly seen as a means to encour- age technology transfers through an expansion of FDI, particularly through licensing, which may have positive implications for knowledge spillover. De Ferranti and others (2003) report findings associating improved IPRs with higher FDI through licensing from the United States, for example. How can Central America use its FTAs to attract FDI? The negoti- ated FTAs already devote significant attention to IPRs, but legal copy- right protection can be meaningless unless investors perceive that enforcement is effective. In chapter 10, Park examines intellectual pro- tection in CAFTA countries and finds that the Dominican Republic and El Salvador have the strongest protection in the region and are the only CAFTA members above the Latin America mean (admittedly a poor benchmark). Both countries rank well below Chile and the United States, for example. However, investor survey results published in the Global Competitiveness Reports suggest that enforcement is perceived as being weak in El Salvador. Estimates by the Business Software Alliance suggest that piracy rates in 2008 ranged from 59 percent in Costa Rica to 74 percent in Honduras and about 80 percent in all the remaining DR-CAFTA members, implying that legal protection in the Dominican Republic and El Salvador is virtually meaningless in the software mar- ket segment. A useful benchmark is the United States, where piracy rates are 20 percent. The study also finds that the substantive IPR reforms envisaged under the DR-CAFTA are likely on balance to encourage investment and technology transfer in the longer run, but only if institutions evolve so that they are strong enough to enforce the new regulations. 18 López and Shankar What Are the Expected Welfare Effects of Trade Liberalization and Promotion in Central America? As already discussed, exporters pay higher wages (see table 1.2), and therefore an expansion of trade should be welcomed from a welfare per- spective. However, not everybody benefits equally from these higher wages. The existing evidence highlights that these higher wages are asso- ciated with a skill premium (that is, exporters hire a more skilled labor force), suggesting that the beneficiaries of higher wages are those with higher levels of human capital (who are not likely to be at the bottom of the income distribution to start with). This, however, should not be understood as implying that trade is likely to create jobs only in sectors where the poor are disproportionately underrepresented. In the Central American context, changes in wages due to tariff cuts appear to lead to welfare gains that are distributionally progressive in Guatemala and Honduras. In contrast, Costa Rica, El Salvador, and Nicaragua seem to generate regressive welfare effects, suggesting adverse impacts for poorer unskilled and rural workers. Higher trade may therefore result in higher wages but also may cause greater inequality in countries that already have inequitable income distributions. These findings reinforce the need to increase the levels of human capital, but in this case stressing the need Table 1.2 Wage Premiums of Exporters Region and country Premium Latin America and the Caribbean 0.20 Argentina 0.09 Brazil 0.27 Chile 0.26 Colombia 0.27 Ecuador 0.02 Mexico 0.12 Paraguay 0.05 Peru 0.30 Uruguay 0.46 Central America 0.27 Costa Rica 0.54 El Salvador 0.30 Guatemala 0.35 Honduras 0.21 Nicaragua –0.02 Panama 0.26 Source: Casacuberta and others 2007. Getting the Most out of Central America’s Free Trade Agreements 19 to expand access to opportunities for all rather than just to increase average skills. The welfare impact of lower prices caused by tariff cuts varies depend- ing on whether we consider net producers or net consumers. This is there- fore more an empirical than a theoretical question. Estimates of the welfare impact of tariff cuts in Central America indicate that for all coun- tries (with the exception of Nicaragua), the resulting price changes lead to positive welfare gains. These welfare gains are either distributionally neutral or slightly progressive in the sense that they are marginally larger for households at the bottom than at the top of the income distribution. The welfare impact of wage changes shows considerably more variation across the subregion: it is progressive in the Dominican Republic, Guatemala, and Honduras and regressive elsewhere; in fact, for Costa Rica and Nicaragua, these effects are actually negative. Since the skill con- tent of net exports is a key determinant of these changes, the bottom line for policy makers is that investing in people and in flexible labor markets is a priority. Trade may also affect male and female workers differently. Due to social norms and discrimination outside as well as inside the household, women and men differ not only in terms of education, but also in terms of access to labor markets, remuneration, sectoral employment, control over resources, and roles within the household. While price changes are the main mechanism by which trade affects gender outcomes, many fac- tors mediate this impact—resource endowments, labor market institu- tions, systems of property rights, access to markets and information, and other socioeconomic characteristics. This implies that the impact of trade on gender gaps can be highly heterogeneous across countries. Because of these gender differentials, which tend to be higher in poorer households, men and women may not be uniformly able to take advantage of the opportunities created by trade liberalization. These issues are explored by Bussolo, Freije, Djiofack, and Rodríguez in chapter 11 for Central America. They find two counterbalancing forces in the countries under analysis. On the one hand, the impact of trade on returns to employment in the tradable sector implies that lower tariffs in the United States (that is, increased market access and poten- tially increased export orientation) would reduce the gender wage gap. On the other hand, a reduction of own tariffs imposed on exports to the United States would increase the gender wage gap for skilled workers in all sectors and for skilled and unskilled workers in tradable sectors. Reducing own tariffs induces wage increases for skilled and unskilled 20 López and Shankar workers in the tradable sector. But it does so more among males than females, particularly among the skilled, hence increasing the gender wage gap. Besides, exposure to external competition (import shares from the United States) can also cause a larger relative decline in male wages in the sector. An increase of 1 percentage point in the share of imports within sector GDP also brings about a fall in the gender wage gap of about 1 percentage point. In the end, the final impact of trade openness is an empirical matter and one that will likely be country specific. The simulations show very different patterns depending on the coun- try. Trade would narrow the gender wage gap in Costa Rica and widen it in the Dominican Republic. For example, trade would have reduced the gender wage gap by 1.1 and 4.6 percentage points in urban and rural areas, respectively, in Costa Rica. This reduction would have been through a decline of about 3 percentage points in the gender wage gap between unskilled workers in urban areas and even larger reductions in the gap for rural workers. This simulation does not match the actual trend of increasing wage-gender gaps in Costa Rica in recent years but high- lights that trade can have a beneficial impact in the future as trade lib- eralization and promotion continue. Trade expansion, however, would have slowed the decline in the gender wage gap actually observed in El Salvador, Guatemala, and Honduras in recent years (that is, it would have contributed ceteris paribus to a higher gender wage gap). In Nicaragua, as in Costa Rica, trade is potentially beneficial in this context. On the whole, these differential impacts would call for policies that enhance female participation in the tradable sector, while promoting the acquisition of skills through schooling and training. In addition, the effec- tive implementation of institutional measures that prevent discrimination in the workplace is a priority—as industries become profitable, the best jobs should not be male biased. Also, to the extent that some of these effects are associated with the penetration of imports and hence may be associated with job losses or other competitive pressures on the labor market of workers in the tradable sector, remedial policies, which could include mechanisms to facilitate reallocation from declining sectors unable to face external competition and investment in human capital— education, training, and retraining of the labor force, particularly in areas associated with tradables—are clearly necessary. One additional finding of chapter 11 is the puzzling absence of a statistically significant association between trade liberalization and job reallocation, for both males and females. This is both a surprise and a cause for concern, given that the benefits of trade to individual workers Getting the Most out of Central America’s Free Trade Agreements 21 depend to a large extent on their ability to exploit new opportunities. Excessive informality in the labor market and lack of employment cre- ation in trade-related activities are perhaps signs of a labor market that is not responding to the incentives of trade. In the countries of the region, this lack of responsiveness may be associated with rigidities in labor mar- kets, leading to lack of mobility from one activity to another, lack of retraining opportunities, and inability to reallocate labor services. In addition to the welfare effects that emerge from the creation and destruction of jobs, tariff reductions can also have welfare implications by leading to changes in prices that affect both producers and consumers. Thus a natural question that emerges in this context is related to the likely overall welfare impact in the Central American context of price and wage effects associated with FTAs. Bussolo, Freije, Djiofack, and Rodríguez address this question in chapter 12 by simulating the impact of trade openness on household welfare—through both consumption prices and labor markets for the Central American countries. In principle, the consumption price–related welfare gain is positive for all countries (except Nicaragua, where prices increased after DR-CAFTA), suggesting that all population groups benefit from falling prices in consumer goods due to trade openness. The net gain varies between 2 and 6 percent. These gains, however, can be easily lost if macroeconomic mismanagement leads to inflation and depreciation of the exchange rate, all of which nullifies the gains of lower prices through lower tariffs. This part of the analysis assumes perfect pass-through, which, as discussed in chapter 9, does not hold, at least for Nicaragua and Honduras. The welfare impact of wage changes shows considerably more varia- tion across the subregion. These welfare gains are progressive in the Dominican Republic, Guatemala, and Honduras and are the result of the structure of exports and the kind of demand for labor that arises from higher trade given this structure. In these countries, exports favor unskilled workers, who tend to be members of poorer households, and are therefore distributionally progressive. The other three countries (Costa Rica, El Salvador, and Nicaragua) demonstrate regressive changes in labor income; in fact, for Costa Rica and Nicaragua, these effects are actually negative. Again this is a result of the trade shock that, in these cases, produces adverse effects for poorer unskilled and rural workers. In the case of El Salvador, unskilled labor earnings rise, but trade favors skilled and urban workers. The percentage net gain by the average house- hold varies by country as well: more than 15 percent for the Dominican 22 López and Shankar Republic, 10 percent for Guatemala, and more than 13 percent for Honduras compared to 5 percent for El Salvador, a marginal loss of about 0.5 percent for Costa Rica, and a larger loss of 2.4 percent for Nicaragua. These findings support the message emphasized in this vol- ume that investing in people and promoting labor mobility through train- ing and flexible market policies can help households to realize potential gains from trade. A final issue that needs particular attention in this context is the potential impact of trade liberalization and promotion on the environ- ment. The seminal work of Grossman and Krueger (1993) ignited the debate on the impact of international trade on the environment. The existing empirical studies on the relationship between trade and the envi- ronment have found varying results. For example, Grossman and Krueger (1993), who examine the environmental impacts of NAFTA, find no evidence that a comparative advantage was being created by lax environ- mental regulations in Mexico. This result is confirmed by Stern (2005), who finds only small pollution effects of NAFTA on Mexico shortly after the agreement, followed by an improvement in environmental quality. In a related work, Gamper-Rabindran and Jha (2004) analyze the empirical relationship between trade liberalization and the environment in the Indian context. Their findings indicate that exports and FDI grew in the more polluting sectors relative to the less polluting sectors between the periods before and after liberalization. Mani and Jha (2006) find similar results for Vietnam and Turkey, respectively. This evidence provides some support for concerns raised about the environmental impact of trade liberalization. Mani and Cunha review these issues in chapter 13, noticing that there is significant heterogeneity in the environmental regulatory regimes of the region (see table 1.3). In fact, while the United States and Costa Rica appear to be above the international average, Guatemala and Honduras are at the bottom of the international classifications. These differences could potentially favor the rise of pollution havens within the region. In this regard, the analysis by Mani and Cunha assesses the pollution effects related to implementation of the DR-CAFTA along different dimensions or channels. One of these channels is the scale effect associated with trade-related acceleration of growth causing an expansion in produc- tion. The idea is that scaling up (holding constant the mix of goods pro- duced and production techniques) would lead to an increase in pollution emissions. A second channel is the composition effect that may result from the impact that changes in relative prices may have on the structure of Getting the Most out of Central America’s Free Trade Agreements 23 Table 1.3 Environmental Regulatory Regime Index Ranka Country Index 1 Finland 2.303 14 United States 1.184 36 Costa Rica –0.078 60 Dominican Republic –1.014 62 Nicaragua –1.164 63 El Salvador –1.215 66 Honduras –1.300 69 Guatemala –1.532 Source: Esty and Porter 2005. a. Out of 71 countries. production. This effect can either increase or decrease relative output in pollution-intensive sectors, depending on the changes in relative prices. Finally, changes in production technologies (including pollution intensity by unit of output) tend to follow trade liberalization. This technique effect can result from different forces: while trade facilitates access to and adop- tion of more efficient (and cleaner) technologies of production, stronger competition can trigger a race to the bottom of environmental standards, favoring the adoption of cheaper, dirtier technology in the short run. Nevertheless, as income grows, the demand for environmental quality tends to increase. By adopting both tighter environmental policies and more advanced, cleaner technologies, countries can afford to reduce emis- sions after they attain a certain level of income. By comparing average annual emissions of different types of pollutants before and after implementation of the agreement, chapter 13 concludes that the scale effect outweighs the composition and technique effects, with most of the variation in pollution resulting from a scaling up in pro- duction. Composition effects appear to be small and vary across member countries with similar regulatory frameworks, mainly since implementa- tion varies considerably. A policy implication is that all countries could benefit from closing the gaps in their environmental regulatory frame- work in terms of actual regulations, capacity, and monitoring. Countries such as El Salvador, where there has been a relative expansion of more polluting industries, should go beyond the legal requirements of the envi- ronmental agenda of the DR-CAFTA, for example, and work on strength- ening implementation in the short run. For poorer countries such as Nicaragua and Honduras, environmental regulation does not seem to play an important role in the current allocation of production. Nevertheless, as 24 López and Shankar these economies grow, this situation will change. For this reason, these countries should start planning and implementing a medium-term envi- ronmental agenda that will mitigate risks down the road. Notes 1. Throughout the book, references to the “region” or to Central America are used loosely to include Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and Panama (not a DR-CAFTA signatory) and, in some cases, the Dominican Republic (not an Association Agreement signatory). 2. The agreement has been effective only for a year and half in Costa Rica and for two to three years in the remaining countries—years that have been com- plicated by the food and fuel crisis (2007–08) and the global financial crisis. 3. The expected effects from learning by exporting could occur either at the time of entry into exporting (a one-time effect) or every year after entry (a continuous effect). 4. The empirical growth literature offers some related examples of nonlinear specifications considering interaction effects. Borensztein and others (1998) and Alfaro, Chanda, and Kalemli-Ozcan (2006) find that the growth benefits from foreign direct investment are attained when the host country has sufficiently high levels of human capital and financial develop- ment, respectively. References Ades, A. F., and E. L. Glaeser. 1999. “Evidence on Growth, Increasing Returns, and the Extent of the Market.” Quarterly Journal of Economics 114 (3): 1025–45. Alesina, A., E. Spolaore, and R. Wacziarg. 2000. “Economic Integration and Political Disintegration.” American Economic Review 90 (5): 1276–96. Alfaro, L., A. Chanda, and S. Kalemli-Ozcan. 2006. “How Does Foreign Direct Investment Promote Economic Growth? Exploring the Effects of Financial Markets on Linkages.” NBER Working Paper 12522, National Bureau of Economic Research, Cambridge, MA. Balat, J., and G. Porto. 2007. “Globalization and Complementary Policies: Poverty Impacts in Rural Zambia.” In Globalization and Poverty, ed. A. Harrison, 373–416 Chicago: University of Chicago Press. Beck, T., A. Demirgüç-Kunt, and R. Levine. 2001. “Legal Theories of Financial Development.” Oxford Review of Economic Policy 17 (4): 483–501. ———. 2003. “Law, Endowments, and Finance.” Journal of Financial Economics 70 (2): 137–81. Getting the Most out of Central America’s Free Trade Agreements 25 Beck, T., A. Demirgüç-Kunt, and M. S. Martínez Peria. 2007. “Reaching Out: Access to and Use of Banking Services across Countries.” Journal of Financial Economics 85 (1): 234–66. Ben-David, D. 1993. “Equalizing Exchange: Trade Liberalization and Income Convergence.” Quarterly Journal of Economics 108 (3): 653–79. Borensztein, E., J. De Gregorio, and J. W. Lee. 1998. “How Does Foreign Direct Investment Affect Economic Growth?” Journal of International Economics 45 (1): 115–35. Cadot, O., L. Dutoit, and M. Olarreaga. 2009. “Barriers to Exit from Subsistence Agriculture.” World Bank, Washington, DC. Casacuberta, C., N. Gandelman, M. Olarreaga, G. Porto, and E. Rubiano. 2007. “Exporter Premiums.’’ In The Latin American and Caribbean Regional Study on Microdeterminants of Growth, ch. 7. Washington, DC: World Bank. Chang, R., L. Kaltani, and N. Loayza. 2009. “Openness Can Be Good for Growth: The Role of Policy Complementarities.” Journal of International Economics 90 (1): 33–49. De Ferranti, D., G. Perry, I. Gill, L. Guasch, W. F. Maloney, C. Sánchez-Páramo, and N. Schady. 2003. Closing the Gap in Education and Technology, The World Bank, Washington, DC. Demirgüç-Kunt, A., and V. Maksimovic. 1998. “Law, Finance, and Firm Growth.” Journal of Finance 53 (6): 2107–137. Dollar, D. 1992. “Outward-Oriented Developing Economies Really Do Grow More Rapidly: Evidence from 95 LDCs, 1976–85.” Economic Development and Cultural Change 40 (3): 523–44. Edwards, S. 1992. “Trade Orientation, Distortions, and Growth in Developing Countries.” Journal of Development Economics 39 (1): 31–57. Esty, D., and M. Porter. 2005. “National Environmental Performance: An Empirical Analysis of Policy Results and Determinants.” Environment and Development Economics 10 (4): 391–434. Frankel, J. A., and D. Romer. 1999. “Does Trade Cause Growth?” American Economic Review 89 (3): 379–99. Gamper-Rabindran, S., and S. Jha. 2004. “Environmental Impact of India’s Trade Liberalization.” Working Paper, University of North Carolina at Chapel Hill, Department of Public Policy. Goldberg, P. K., and N. Pavcnik. 2005. “The Effects of the Colombian Trade Liberalization on Urban Poverty.” NBER Working Paper 11081, National Bureau of Economic Research, Cambridge, MA. Grossman, G. M., and A. B. Krueger. 1993. “Environmental Impacts of a North American Free Trade Agreement.” In The Mexico–U.S. Free Trade Agreement, ed. P. Garber, 13–56. Cambridge, MA: MIT Press. 26 López and Shankar Harrison, A., ed. 2007. Globalization and Poverty. Chicago, IL: University of Chicago Press. Jaramillo, C. F., and D. Lederman. 2005. “Challenges of CAFTA.” World Bank, Washington, DC. Lederman, D., W. Maloney, and L. Servén. 2007. Lessons from NAFTA for Latin America and the Caribbean. Palo Alto, CA: Stanford University Press. Love, I. 2009. “What are the Determinants of Financial Access in Latin America’’ Chapter 5 in Does the Investment Climate Matter?, The World Bank, Washington, DC. Mani, M., and S. Jha. 2006. “Trade Liberalization and the Environment in Vietnam.” Policy Research Working Paper 3879, World Bank, Washington, DC. Rajan, R., and L. Zingales. 1998. “Financial Dependence and Growth.” American Economic Review 88 (3): 559–86. Sachs, J. D., and A. M. Warner. 1995. “Economic Reform and the Process of Global Integration.” Brookings Papers on Economic Activity 1: 1–118. Stern, D. I. 2005. “The Effect of NAFTA on Energy and Environmental Efficiency in Mexico.” Working Paper in Economics 0511, Rensselaer Polytechnic Institute, Troy, NY. Svaleryd, H., and J. Vlachos. 2005. “Financial Markets, the Pattern of Specialization, and Comparative Advantage: Evidence from OECD Countries.” European Economic Review 49 (1): 113–44. Topalova, P. 2005. “Trade Liberalization, Poverty, and Inequality: Evidence from Indian Districts.” NBER Working Paper 11614, National Bureau of Economic Research, Cambridge, MA. World Economic Forum. Various years. Global Competitiveness Report. Geneva: World Economic Forum. CHAPTER 2 The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis Ana Cristina Molina, Maurizio Bussolo, and Leonardo Iacovone Regional trade agreements have spread rapidly around the world since the 1990s, and Latin America is no exception. Today all countries in the region are signatories of at least one regional trade agreement. This new economic environment has created new challenges and opportunities for exporters in the region. Yet despite its importance to policy makers, the dynamics of exporters following a regional trade agreement have been poorly documented. Previous studies have looked at the effects of such agreements on export flows and especially at the trade creation and diver- sion effects (for a theoretical and empirical review, see Freund and Ornelas 2009). These studies use product, sectoral, or country-level data and thus provide little insight as to the effects on exporters’ activities and performance. Using a novel firm-level data set of trade transactions for the Dominican Republic covering the period 2002–09, we examine exporters’ responses to the Dominican Republic–Central America Free Trade Agreement (DR-CAFTA). The agreement entered into force in 20071 and includes the United States—the Dominican Republic’s main trade partner—Costa Rica, El Salvador, Guatemala, Honduras, and Nicaragua. We test whether DR-CAFTA has had a positive impact on the extensive margin—that is, whether it has increased exports through the 27 28 Molina, Bussolo, and Iacovone entry of new exporters and the introduction of new product-market rela- tionships. Finally, we analyze whether the agreement has improved exporters’ survival in foreign markets. The adjustments across the four margins (that is, firms, products, mar- kets, and survival) are particularly important. First, exports generated by the entry of new firms and the introduction of new products or markets are a source not only of growth, but also of export diversification, both key elements of a country’s development strategy. Second, exporters’ sur- vival in foreign markets is crucial for sustained export growth. Finally, it is essential to understand the effects of market access on exporters in order to design policies aiming to help exporters reap all the benefits from trade liberalization. For our analysis, we follow the theoretical frameworks outlined in Melitz (2003) and Bernard, Redding, and Schott (2009). Melitz’s influen- tial paper describes firms’ dynamics following trade liberalization. In his framework, firms differ in their productivity (that is, marginal costs) and have to pay a fixed cost to enter the export market, thus implying that only the most productive firms will export. One of the main implications of the model is that a reduction in trade costs (that is, tariffs) will lead to the entry of new exporters, as more firms will be able to afford to enter the export market. An indirect implication of the model is that lower tariffs will raise the profits of incumbent exporters, thus increasing their likelihood of staying in the market and decreasing their exit rates. Bernard, Redding, and Schott (2009) refine the Melitz model to account for firms with multiple products and destinations.2 In their framework, a decrease in variable trade costs induces surviving exporters to start selling products that were not profitable before, thus increasing the number of goods exported by each firm.3 We test these predictions in the case of the Dominican Republic exporters. Using the export firm–level data set provided by the Dominican Republic Customs Agency for the period 2002–09, we estimate the effect of tariff reductions on the number of new exporters and on the number of incum- bents that introduce a new product or enter a market within the CAFTA area.4 In addition, we evaluate the effect of tariff cuts on the probability of exiting a market. We find that tariff preferences have a positive effect on the number of new exporters and on the number of exporters that introduce new product-market combinations. The effect is, however, very small, which would suggest that other factors may be preventing exporters from taking advantage of improved market access. Finally, we find that tariff cuts reduce the probability of exiting a CAFTA market. The effect is also very small. The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 29 This chapter contributes to the burgeoning literature on firm-level export dynamics. Most studies have focused on the relationship between export participation and firm productivity (that is, firm selection and learning by exporting). Some others have explored the effects of trade lib- eralization on the productivity of import-competing plants (see, for instance, Pavcnik 2002; Trefler 2004), but only very few have looked at the effects of tariffs on firms’ export behavior. Bernard, Jensen, and Schott (2003) analyze the effect of tariff reductions on the participation of firms in export markets using survey data for U.S. manufacturing plants for three years over the period 1987–97. They find that firms in industries with declining export costs face lower probabilities of death and higher probabilities of becoming exporters. Baldwin and Gu (2004) study Canadian manufacturing firms and find that the tariff cuts following the U.S.-Canada Free Trade Agreement (of 1988) promoted the entry of Canadian plants into export markets. Their data, however, do not include a destination dimension. Their estimates are therefore “aggregate” estimates—that is, they are not market specific. This chapter also relates to studies exploring the relationship between tariff cuts and the extensive margin. These studies use trade data at the product level and suggest that lower tariffs have a positive effect on the exports of new goods (see, among others, Kehoe and Ruhl 2009; Gómez and Volpe 2008; Debaere and Mostashari 2005). Given the level of dis- aggregation of the data, such studies can investigate the effect of tariffs only on the range of products, not on firms’ dynamics. The rest of the chapter is organized as follows. It begins by presenting the data and then describes the pattern of aggregate exports and the firm- level extensive margin of the Dominican Republic. It then looks at tariffs and exporters’ behavior and presents the empirical exercise and the results. A final section concludes. The Data Our study employs a unique, very detailed export firm–level data set pro- vided by the Dominican Republic Customs Agency. The data contain all transactions (that is, amount and quantity) by product at the Harmonized System (HS) 12-digit level and by destination for all exporters for the period 2002–09.5 The sample includes 135,016 exporter-destination- product relationships. The universe of firms during this period consists of 8,706 firms, among which not all export in every year.6 Table 2.1 reports the number of exporting firms, products, and markets for select years. 30 Molina, Bussolo, and Iacovone Table 2.1 Summary Statistics, Select Years, 2003–09 Indicator 2003 2005 2007 2009 Exports by firm (US$, millions) Mean 1.4 1.5 1.4 1.5 Standard deviation 9.4 9.9 9.0 9.8 Products by firm Mean 6.3 5.2 6.5 6.2 Standard deviation 10.7 8.9 13.1 13.1 Markets by firm Mean 2.1 2.1 2.1 2.2 Standard deviation 2.7 2.9 2.9 3.1 Number of firms 2,660 2,622 3,237 3,031 Number of products 2,035 2,049 2,937 2,812 Number of markets 133 131 147 147 Total exports (US$, millions) 3,742 3,969 4,474 4,563 Source: Authors’ calculations based on data provided by the Dominican Republic Customs Agency. The data show that the number of exporters, products, and markets increased until 2007, before declining in the next years (except for the number of markets). The number of exporters rose from 2,660 in 2003 to 3,237 in 2007, which is a 21.6 percent increase. In addition, the num- ber of products rose 44.3 percent for the same period, from 2,035 in 2003 to 2,937 in 2007. The change in the number of both products and firms was more significant during 2005–07 than during 2003–05. As for the number of markets, it rose 10.5 percent between 2003 and 2007. Each year about 60 percent of firms are multiproduct exporters (that is, they export more than one product). Half of them export to a single market, while the other half export to multiple markets.7 For the remain- ing firms, about 37 percent export only one product to one destination and 3 percent export one product to multiple destinations. Almost all exports come from multiproduct firms. These findings are similar to those in Bernard, Redding, and Schott (2009) and show that exports are con- centrated within multiproduct exporters. Another important characteristic of our data set is that it distinguishes between exporters in export-processing zones (EPZs) and exporters in the national territory. This is an important feature, as it allows us to control for the firms’ location in our empirical exercise. In terms of value, EPZ exporters are particularly important: their exports account for 74 percent (2009) to 84 percent (2002) of total exports. About 20 percent (545) of the total number of exporters operate in an EPZ each year. The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 31 Finally, we exclude from our analysis reexports, ferronickel exports,8 occasional exporters, couriers, as well as firms whose partners are not identified. These exclusions reduce our sample by 18 percent to 110,702 trade relationships. Export Trends With the purpose of identifying any major change in export patterns fol- lowing the DR-CAFTA, we briefly review in this section the evolution of exports from Dominican firms to the United States and the other CAFTA members (that is, El Salvador, Honduras, Guatemala, and Nicaragua).9 Figure 2.1 reports the exports from the Dominican Republic by desti- nation from 2002 to 2009 (in millions of U.S. dollars). The United States is by far the largest partner of the Dominican Republic, with average exports amounting to US$2.6 billion each year. Other main destinations for the shipments of Dominican exporters are Puerto Rico (10.2 percent Figure 2.1 Dominican Exports, by Destination, 2002–09 5,000 4,000 exports (US$, millions) 3,000 2,000 1,000 0 02 03 04 05 06 07 08 09 20 20 20 20 20 20 20 20 United States other CAFTA members Central America and the Caribbean OECD rest of world Source: Authors’ calculations. 32 Molina, Bussolo, and Iacovone in 2008), Haiti (9.9 percent), Belgium (2.1 percent), Spain (2 percent), and the Netherlands (1.9 percent). Exports to the other CAFTA mem- bers account for only 1 percent of total exports from the Dominican Republic in 2008, and this share has been relatively stable since 2002. Two main features of the Dominican trade with the United States can be highlighted. First, the share of exports to the United States has declined since 2002 as a result of market diversification and a slowdown in exports to this market. Between 2002 and 2008, exports to other coun- tries in the Caribbean (for example, Haiti and Jamaica) as well as to other developed countries (for example, Spain and Belgium) expanded rapidly, growing some 150 percent, while exports to the United States expanded a modest 10 percent. Second, exports to the United States are relatively volatile. They exhibit alternating periods of modest growth and decline. Following a decrease in 2004, exports picked up, before falling again in 2007 and in 2009. As a result, the years after the DR-CAFTA have been characterized by slow export growth. In 2009, the demand for Dominican exports fell across all regions (except for exports to the developed countries), and exports to the United States were almost at the same level as in 2002. However, this does not necessarily reflect the effectiveness of the agree- ment, but rather the difficult business conditions in the U.S. market dur- ing this period. In a recent study, Swiston (2010) shows that the 2008 recession in the United States reduced growth by 4–5 percent in Central America and that the transmission channels were mainly the financial conditions and the fluctuations in export demand. At the same time, exports to the other CAFTA members (that is, El Salvador, Honduras, Guatemala, and Nicaragua) not only expanded sig- nificantly after 2005, but also accelerated in 2007, before declining in 2009. Figure 2.2, panel a, shows that exports to the CAFTA area went from US$21.5 million in 2005 to US$57.7 million in 2008. This remark- able performance is mainly due to the surge in exports to Guatemala and Honduras (see figure 2.2, panel b). Between 2005 and 2008, the value of exports to Guatemala doubled to US$11.8 million, while exports to Honduras trebled from US$10.9 million to US$32.8 million. Exports to El Salvador and Nicaragua also exhibited strong growth: the value of their exports more than doubled from 2005 to 2008.10 This suggests that the tariff preferences were effective in boosting exports to the CAFTA members.11 Moreover, Dominican exporters seem to have antic- ipated the agreement and started to expand their exports to the CAFTA area in 2006. The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 33 Figure 2.2 Dominican Exports to CAFTA Members in Select Years, 2002–09 a. Exports to the CAFTA area 60 50 exports (US$, millions) 40 30 20 10 0 02 03 04 05 06 07 08 09 20 20 20 20 20 20 20 20 b. Exports to CAFTA members 40 exports (US$, millions) 30 20 10 0 or a s a ra al gu ad em u ra nd lv ca at Sa Ho Ni Gu El 2003 2005 2007 2008 Source: Authors’ calculations. 34 Molina, Bussolo, and Iacovone To summarize, following the DR-CAFTA, exports to the United States stagnated in absolute terms, while exports to the other CAFTA members exhibited strong growth, but their share still represents only 1 percent. This highlights the potential for further trade expansion among Central American countries. However, in both cases, exports dropped in 2009, probably as a consequence of the global economic crisis at that time. Firm-Level Patterns of Extensive Margin Having analyzed the aggregate pattern of exports, we now examine the exporters’ dynamics that lie behind the aggregate movement of exports. In particular, we look at firms’ export extensive margin to the United States and the CAFTA area to assess whether there were significant changes in exporters’ behavior subsequent to the agreement. The exten- sive margin refers to all new trade relationships and can be decomposed into four components12: • Exports by existing firms of a new product to a new market • Exports by existing firms of a new product to an existing market • Exports by existing firms of an existing product to a new market • Exports by new firms. Using this decomposition, we sketch exporters’ behavior during this period and can determine whether they reacted as predicted in Melitz (2003) and Bernard, Redding, and Schott (2009). To identify each of the components, we classify firms, products, and markets according to three different statuses: new, existing, and exit. A new firm is a firm that exports in t, but not in t – 1. An existing firm is a firm that exports in t and in t – 1. Finally, an exiting firm is a firm that exports in t – 1, but not in t. The same definition applies to firm-product relationships and firm- destination relationships. For example, a new firm-product relationship refers to a product exported by firm i in t, but not in t – 1. Figure 2.3 reports the number of firms by year according to their export status. The number of newly exporting firms surged in the year when DR-CAFTA took effect. As for exiting firms, about 1,000 firms stopped exporting in any given year during the 2002–09 period, except in 2007 and 2008. From 2006 to 2007, the number of exits dropped significantly to 766, before reaching their high (that is, 1,595) in 2008. Two events may explain this result. First, the large number of entries during the DR-CAFTA year may have boosted the number of exits in 2008. Short The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 35 Figure 2.3 Number of Firms, by Export Status, 2003–09 4,000 2,000 number of firms 0 –2,000 03 04 05 06 07 08 09 20 20 20 20 20 20 20 existing new exiting Source: Authors’ calculations. survival among new exporters has been well documented in the literature (see, among others, Besedes and Prusa 2006, 2007; Eaton and others 2007; Cadot and others 2010) and recently modeled in Albornoz and others (2009). In their framework, firms discover their export profitabil- ity only once they become exporters. Experienced exporters entering a new market are therefore better informed than new exporters about their skills and their chances of success. As a consequence, new exporters are more likely to exit than experienced ones. It could be that, drawn by the more attractive conditions created by the agreement, many exporters entered the market in 2007, only to realize their true export perform- ance, which forced them to exit in mass the next year. But in addition to the inherent low survival of exporters, the global economic crisis that started at the end of 2007 could have amplified this phenomenon, increasing even more the likelihood to exit of both experienced and new exporters. The weight of the extensive margin in total exports varies across the United States and the CAFTA area.13 In the case of exports to the United States, the extensive margin is volatile and accounts for a small share of total 36 Molina, Bussolo, and Iacovone trade to this country, about 5 to 13 percent of total exports. As for exports to the CAFTA area, their extensive margin has been growing since 2005, and so has their share in total exports. In 2007, the extensive margin tre- bled, accounting for about 55 percent of total exports to the region. In both cases, the data suggest that the extensive margin expanded during the year the DR-CAFTA came into force, but declined in the following years. To understand what drives the extensive margin, see figure 2.4, which decomposes the extensive margin for each market. First, the data show that the composition of the extensive margin is very different in each market. In the case of U.S. exports, the extensive margin is driven mainly by the generation of exports by new entrants and by the introduction of new products by incumbents. In the case of the other CAFTA members, the extensive margin is driven mainly by the exports of incumbent exporters—in particular since 2007, when there is a clear jump in the extensive margin. In this year and in the next ones, the main actors were incumbent exporters who either introduced a new product in the CAFTA area or exported for the first time to the CAFTA area. With the marked decline in the extensive mar- gin in 2008 and 2009, only exports generated by the introduction of a new product remained (approximately) at the level of 2007. These figures suggest that new exporters played a more important role in the expansion of the extensive margin in the U.S. market than in other CAFTA markets. This could reflect the preferences of new exporters for the United States, which see this destination as a market with more opportunities, given its size and wealth. Yet this could also reflect the existence of nontariff barriers to entry into the CAFTA markets. Relationship between Tariff Reductions and Exporters’ Behavior: Preliminary Evidence In this section, we describe the tariffs and the tariff reductions that fol- lowed the introduction of DR-CAFTA. We then examine the relationship between the extensive margin and tariff cuts. Tariffs and Tariff Cuts The data on tariffs come from the tariff schedules that were negotiated in 2004 by each CAFTA member. Each country has its own tariff schedule in which goods are classified in eight to 12 different categories14 accord- ing to the time frame over which tariffs will be eliminated. While most The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 37 Figure 2.4 Decomposition of the Firm-Level Extensive Margin, 2003–09 a. U.S. market 400 exports (US$, millions) 300 200 100 0 03 04 05 06 07 08 09 20 20 20 20 20 20 20 b. CAFTA market 30 exports (US$, millions) 20 10 0 03 04 05 06 07 08 09 20 20 20 20 20 20 20 existing firm, existing product, new market existing firm, new product, existing market existing firm, new product, new market new firm Source: Authors’ calculations. Dominican products could enter the U.S. market duty free immediately after the agreement, between 15 percent (Guatemala) to 40 percent (El Salvador) of the total number of products in the other CAFTA markets were subject to a duty. Products whose tariffs were not eliminated immediately after the agreement are being phased out progressively over a five- to 20-year period.15 38 Molina, Bussolo, and Iacovone For the Dominican Republic–U.S. trade, the agreement consolidates the existing preferences and grants duty-free treatment to almost all products entering the U.S. market,16 except for some agricultural prod- ucts that are subject to quotas.17 In the case of textile and apparel exports, the DR-CAFTA provisions are more flexible than those in the Caribbean Basin Trade Partnership Act (CBTPA), as they allow for cumulative rules of origin.18 This allows apparel exporters in Central America and the Dominican Republic to use inputs from any member, without losing their duty-free access to the U.S. market (Hornbeck 2008). For the Dominican–Central American trade,19 the DR-CAFTA extends duty-free treatment to goods produced in export-processing zones. As for the tariff levels, the median tariff faced by Dominican exporters before the agreement was 3.2 percent in the United States and 5 percent in the other CAFTA members.20 To assess the size of the tariff preferences granted by the DR-CAFTA, we compute the tariff cuts by product for the United States21 and the other CAFTA mem- bers. The median tariff cut is 3.1 percent in the case of the United States and 1 percent in the case of the CAFTA members. In the United States, sectors that experienced high tariff cuts include agriculture, prepared foodstuffs, footwear, textiles, and clothing. In the case of the other CAFTA members, important tariff cuts also took place in agricul- ture, textiles, and clothing, prepared foodstuffs, as well as in machinery and appliances. Did Tariff Reductions Affect the Extensive Margin? To explore the relationship between the extensive margin and the tariff reduction, we look at the extensive margin according to the size of the tariff cut. If the tariffs had an effect on the exporters’ behavior, one expects the sectors with the larger tariff cuts to exhibit a relatively bigger increase in exports. We first classify sectors (HS two-digit) according to whether they had a low or a high tariff cut. A sector is a high-cut sector if its median tariff cut is larger than the overall median tariff (that is, 3.1 percent in the case of the United States and 1 percent in the case of the other CAFTA mem- bers). We use tariffs at the sector level (HS two-digit) rather than at the product level, to have the largest product concordance between exports of the Dominican Republic and those of its partners. Exports by tariff cuts are reported in figure 2.5. In the case of exports to the Central American countries, the extensive margin behaves as expected. The growth in the extensive margin between The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 39 Figure 2.5 Extensive Margin, by Tariff Cut, 2003–09 a. U.S. market 200 extensive margin (US$, millions) 150 100 50 0 gh gh gh gh gh gh gh w w w w w w w lo lo lo lo lo lo lo hi hi hi hi hi hi hi 2003 2004 2005 2006 2007 2008 2009 b. CAFTA market 25 extensive margin (US$, millions) 20 15 10 5 0 gh gh gh gh h gh gh w w w w w w w g lo lo lo lo lo lo lo hi hi hi hi hi hi hi 2003 2004 2005 2006 2007 2008 2009 existing firm, existing product, new market existing firm, new product, existing market existing firm, new product, new market new firm Source: Authors’ calculations. 2007 and 2009 is driven by the export growth in sectors that experienced high tariff cuts. This suggests that the DR-CAFTA did promote new exports to the CAFTA members, and incumbent exporters seem to have benefited the most. These sectors include cotton, tobacco, and apparel. However, the effect seems to vanish with time, as the extensive margin in 40 Molina, Bussolo, and Iacovone both low- and high-cut sectors exhibits a progressive decline from 2008 onward. This could reflect the effects of the 2008 economic crisis rather than a decline in the exporters’ enthusiasm for the agreement. In the case of the U.S. market, the dynamics are less clear. From 2005 to 2007 exports in high-cut sectors grew rapidly, suggesting that exporters may have anticipated the agreement and started exporting before the agreement entered into force. New exports in sectors with high tariff cuts include apparel, footwear, plastic, and tobacco. In the case of textiles and apparel, such anticipation effect is very likely, as textiles and apparel exports were subject to a retroactivity rule provided by the DR-CAFTA. Under this rule, importers can apply for refunds of duties when DR-CAFTA’s rules of origin have been met. As a result this could have created a major incentive to start buying immediately from Dominican exporters. In addition to the increase in exports in high-cut sectors, we also observe in 2007 an impressive expansion of exports in low-cut sectors. This is driven mainly by the sales from new exporters. This should not come as a surprise, as it could suggest that even if the tariff cut was low, it was big enough to modify the decision to export in that year. The extensive margin in both low- and high-cut sectors dropped in 2008, but picked up again in 2009. Did Tariff Reductions Affect Exporters’ Survival? In this subsection, we look at the exporters’ survival probabilities before and after the DR-CAFTA agreement. A tariff reduction will raise the profits of incumbent exporters, thus improving firms’ position in the export market and increasing their likelihood of surviving. We expect the survival probabilities to be larger in high-cut sectors.22 To check whether this is the case, we compute the survival probabilities by cohort and by sector type (that is, high tariff cut and low tariff cut) for firms exporting to the United States, as well as for firms exporting to the other CAFTA members. First, the number of firms entering the U.S. market in any given year is considerably larger than the number of firms entering any other CAFTA market. This is not necessarily surprising considering the size of the U.S. market relative to the other markets. Second, the number of firms enter- ing sectors with high tariff cuts between 2006 and 2009 is larger than the number entering sectors with low tariff cuts. This is true for the U.S. mar- ket, except for 2008. Third, the survival probabilities in 2007 and 2008 are higher for exporters in sectors with high tariff cuts than for exporters in sectors with low tariff cuts, for all markets, regardless of the year of The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 41 entry. There were only very few cases, in both the United States and the other CAFTA members, where this is not observed. Finally, the survival probabilities before 2006 and in 2009 do not exhibit any particular pat- tern in any of the sectors, unlike in subsequent years. This constitutes preliminary evidence that tariffs do affect the pat- tern of firms in terms of survival, especially in sectors with high tariff cuts, where survival seems to be higher in the first years of the DR-CAFTA. Empirical Strategy and Results In this section, we formally test the relationship between tariff cuts and exporters’ dynamics. We are particularly interested in examining whether the tariff reductions implemented by the DR-CAFTA promoted (a) the participation of firms in the export market and (b) the introduction of new product-market relationships. In the third part of this section, we also look at whether the tariff cuts improved (c) the survival of Dominican exporters by preventing firms from exiting the export market. Before proceeding to the empirical exercise, a few caveats related to our data set need to be mentioned. First, the observations in our data set are likely to be subject to left and right censoring. In the case of left cen- soring we cannot determine whether a firm with a positive trade value in 2002 started exporting in 2002 or before (that is, whether it is a new exporter or not). So we only consider firms that started exporting strictly after 2002 when estimating the effects of tariffs (a) on the number of new exporters and (b) on the decision to exit the export market. Similarly for right-censored observations, we cannot determine whether exporters reporting a positive trade in 2009 exited the next year or not. Only the exits that took place before 2009 can be assessed. A second caveat concerns the period covered by our study. We observe only three years after the agreement, namely 2007, 2008, and 2009, and thus our empirical exercise considers only the short-term adjustments of the DR-CAFTA. Moreover, this period coincides with the economic cri- sis that broke in the United States, which could have undermined the effects of the agreement on Dominican exports. Finally, given the nature of our data (that is, customs data) we can observe only the firms that exported at least once during our period of observation and not the total number of firms that potentially could have exported but didn’t. This selection problem could be a major handicap if one would like to estimate the probability of entering the export market 42 Molina, Bussolo, and Iacovone (that is, becoming an exporter). However, this is not the empirical strat- egy adopted in the present study. New Exporters One of the main implications of the Melitz model is that lower trade costs will induce the entry of new firms into the export market. To test this, we estimate the effect of tariff cuts on the number of new exporters according to the following equation: New Exportersjpt = b1Δtariffsjpt + aControls + hk + dj + gt + ejpt. (2.1) The dependent variable is the number of new entrants in the national territory and in export-processing zones exporting product p (at the HS eight-digit level) to country j in time t. Δtariffsjpt is the duty reduction (in percentage points) in product p implemented in period t by country j.23 The coefficient b1 will be positive if the change in tariffs has a positive effect on the number of new entrants. We also include two proxies for information spillovers: the number of firms that exported product p in t – 1 and the number of exporters that served market j with product p in t – 1. Potential exporters may see the participation of other Dominican firms in foreign markets as a signal of profitability. We therefore expect these two covariates to have a positive effect on the decision of new entrants to export. Additionally, we control for the exporters’ location by adding a dummy EPZ that equals 1 for firms operating in an export-processing zone, and 0 otherwise. To account for the benefits of a reduction in import costs due to the bilateral feature of the agreement, we introduce the cut of the import- weighted tariff. This variable was calculated at the sector level (HS two- digit) using as weights the coefficients observed in the 2005 input-output table of the Dominican Republic.24 The rationale is that following imple- mentation of the DR-CAFTA, firms in the Dominican Republic could have access to cheaper inputs. If this is the case, implementation of the DR-CAFTA would reduce their costs, improve their competitiveness, and therefore increase their propensity to export. We also add a measure of comparative advantage, namely the normal- ized revealed comparative advantage index (NRCA) suggested by Laursen (1998) to assess whether new exporters start in sectors with a comparative advantage. The NRCA is based on the revealed competitive advantage (RCA)25 and was computed for each year at the HS four-digit level according to NRCAkt = (RCAkt–1) / (RCAkt+1). The main advantage The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 43 of the NRCA is its symmetry: it ranges between –1 and 1. An NRCAkt between –1 and 0 suggests a comparative disadvantage in sector k in period t, while an NRCAkt between 0 and 1 indicates a comparative advantage. We also add an interaction term between the NRCA and the tariff cuts to determine whether the tariff preferences promoted the entry into sectors of comparative advantage to the Dominican Republic more than the entry into nontraditional sectors. Finally, fixed effects for time (gt), country (dj), and sectors (hk), at the HS two-digit level, are included. By their inclusion, we expect to control for any market, sector, and year characteristic that could affect our results, such as the difficult business environment of 2008 and 2009. ejpt is the usual idiosyncratic error term. We also estimate equation 2.1 using the actual applied tariffs enjoyed by Dominican exporters in foreign markets instead of the tariff cuts. By including the tariff levels, we can distinguish between the short-term effects (that is, tariff reductions or adjustments) and the long-run effects (that is, levels) of tariffs. Tariff levels should negatively affect the number of new exporters. The data for applied tariffs (HS six-digit) comes from TRAINS/WITS and is incomplete for the period 2002–09, which reduces the number of observations for our empirical exercise.26 Table 2.2 presents the results. The three first columns contain the esti- mations using the tariff cuts, while columns 4 to 6 show the results using tariff levels. In each case there are three specifications. In the first one, we include only the market access proxy. In the second specification, we con- trol for information spillovers and exporters’ location. In the last one, we add the weighted import tariff cut, the comparative advantage index, and its interaction term. As shown in table 2.2, the coefficient of the tariff cuts is positive and statistically significant in all three specifications. Ten additional percentage points in the tariff cut will increase by 1 the number of new exporters of product p. This implies that sectors with large tariff cuts attract a larger number of new exporters, but the effect is small. The effect of the tariff levels is also significant and negative, as expected. This suggests that sectors with relatively low tariffs attract more exporters. The effect is, however, almost negligible. As for the effect of the information spillovers, the fact that other firms export product p does not affect the behavior of potential exporters in the same sector. In contrast, the number of new entrants selling product p in a given destination rises with the number of Dominican firms in the same market. The behavior of other exporters in a market seems 44 Molina, Bussolo, and Iacovone Table 2.2 OLS Estimates of the Number of New Exporters Variable (1) (2) (3) (4) (5) (6) Tariff cut 0.100** 0.097*** 0.081*** (0.040) (0.036) (0.027) Tariff –0.017*** –0.018*** –0.016*** (0.006) (0.005) (0.005) Number of exporters, same 0.001 0.000 –0.000 –0.001 product (t – 1) (0.002) (0.002) (0.003) (0.003) Number of exporters, same 0.126*** 0.122*** 0.126*** 0.125*** product and market (t – 1) (0.012) (0.011) (0.013) (0.013) EPZ –0.918*** –0.963*** –1.337*** –1.378*** (0.177) (0.189) (0.302) (0.310) Dominican Republic –0.149*** –0.123** import-weighted tariff cut (0.054) (0.061) Comparative advantage 0.204*** 0.464*** (0.070) (0.130) Comparative advantage x 0.126*** tariff cut (0.039) Comparative advantage x tariffs –0.009 (0.006) Number of observations 14,187 14,187 14,179 5,730 5,730 5,729 R2 0.301 0.437 0.452 0.281 0.402 0.403 Source: Authors’ calculations. Note: Time, country, and sector fixed effects are used in all specifications, but are not reported. Errors are clustered by product. Robust standard errors are in parentheses. ** p < .05, *** p < .01. therefore to affect the decision of new entrants. As for the exporters’ location, there is a negative and statistically significant relationship between the number of new exporters and export zones. This suggests that the entry of new exporters takes place mainly in the national ter- ritory. This should not come as a surprise since the requirements needed to start a firm in an export zone are more demanding. The effect of the import tariff cuts is negative and statistically significant. Having access to cheaper inputs does not seem to benefit new exporters. One possible reason for this could be that sectors with large import tariff cuts were highly protected sectors in the past, and not necessarily competitive, thus explaining the lack of new entries. Finally, our results suggest that the number of new exporters is larger in sectors where the Dominican Republic has a comparative advan- tage. As for the interaction term, only the interaction between the NCRA and the tariff cut is significant. Its effect is positive, thus The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 45 implying that the tariff cuts had an additional and positive effect on the number of new exporters in a comparative advantage sector. The results suggest that both tariff reductions (that is, short-run adjustment) and lower tariffs levels (that is, long-run equilibrium) do induce the entry of new exporters.27 However, both effects seem to be very small. This result could reflect the existence of other factors including, but not limited to, high transport costs and phytosanitary and standards requirements. Introduction of New Product-Market Relationships in the Export Mix The second effect that we are interested in is whether a tariff decline encourages incumbent firms to export an additional product to the CAFTA area. For this exercise, we consider only existing firms and test the effect of tariffs and tariff cuts on the number of exporters that intro- duce a new product in a given market (that is, a new product-market combination). The equation to be estimated is the following: Addjpt = b1Δtariffsjpt + aControls + hk + dj + gt + ejpt, (2.2) where Addjpt is the number of incumbent exporters in the national terri- tory and in export-processing zones that start shipping product p to mar- ket j in year t. Δtariffsjpt is the tariff cut (in percentage points) in product p implemented in period t. We expect that the number of firms adding product p to their export mix is the largest for products exhibiting large tariff reductions. Table 2.3 reports the results. The first three specifica- tions use the tariff cuts (columns 1 to 3), while columns 4 to 6 show the results using tariff levels. As for the control variables we include the same covariates as for equation 2.1, namely, information covariates, exporters’ location, import-weighted tariff cuts, and a comparative advantage meas- ure. Time (gt), country (dj), and sector (hk) fixed effects are also included. ejpt is an idiosyncratic error term. The effect of the tariff cut is positive and statistically significant, but very small (that is, 0.007 to 0.009). The effect of tariffs is negative and significant, but also small. This suggests that in the short and long run, low tariffs can promote the participation of existing exporters by inducing them to add new product-market relationships in their export mix. The effect is, however, very small, which could indicate the existence of other nontariff barriers. As for the information spillovers, both the number of exporters of a given product and the number of exporters of a given product in a given destination have a positive effect on the number of new trade relationships 46 Molina, Bussolo, and Iacovone Table 2.3 OLS Estimates of the Number of Exporters Adding New Product-Market Relationships Variable (1) (2) (3) (4) (5) (6) Tariff cut 0.009* 0.009** 0.007* (0.005) (0.004) (0.004) Tariff –0.007** –0.007** –0.005** (0.003) (0.003) (0.003) Number of exporters, same 0.003*** 0.002*** 0.005*** 0.004*** product (t – 1) (0.000) (0.000) (0.001) (0.001) Number of exporters, same 0.111*** 0.110*** 0.097*** 0.096*** product and market (t – 1) (0.007) (0.007) (0.006) (0.006) EPZ –0.078 –0.093 0.117 0.094 (0.076) (0.077) (0.141) (0.142) Dominican Republic 0.038*** 0.076** import-weighted tariff cut (0.015) (0.033) Comparative advantage 0.161*** 0.359*** (0.024) (0.055) Comparative advantage x 0.019*** tariff cut (0.006) Comparative advantage x tariffs –0.009*** (0.003) Number of observations 37,486 37,486 37,435 13,629 13,629 13,621 R2 0.487 0.627 0.629 0.480 0.603 0.606 Source: Authors’ calculations. Note: Time, country, and sector fixed effects are used in all specifications, but are not reported. Errors are clustered by product. Robust standard errors are in parentheses. * p < .1, ** p < .05, *** p < .01. introduced by existing exporters. Compared to the effect for new exporters, the effect of a cut in the import tariff is also positive for exist- ing exporters. The exporters’ location seems not to have an effect on the number of new product-market relationships. Finally, more new trade rela- tionships are created in comparative advantage sectors. The interaction term between NRCA and tariff cuts is statistically significant and posi- tive, which indicates that the effect of a larger tariff cut is amplified when the product belongs to a comparative advantage sector. A similar result is found in the case of the interaction term between NRCA and the tariff levels.28 Exporter Exit According to Melitz (2003), we could expect that, as tariffs fall, the profits of incumbent exporters rise, thus improving exporters’ chances of survival. This implies that lower tariffs could help incumbent The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 47 exporters to consolidate their market position and prevent them from exiting foreign markets. We examine this implication by estimating the effect of tariffs and tar- iff cuts on the probability that a firm will exit the export market. As men- tioned at the beginning of this section, we consider only firms that start exporting and then exit during the 2002–09 period. The probability29 for firm i to stop exporting product p to country j in year t is given by: Pr(Exitijpt = 1) = b1Δtariffsjpt + aControls + hk + dj + gt + ejpt, (2.3) where the dependent variable Exitijpt is a dummy that equals 1 if firm i stops exporting product p to market j in t and 0 otherwise. As before, Δtar- iffsjpt refers to the tariff reductions in product p in time t applied by coun- try j. We also include three measures of export experience. Our first measure is the number of years a firm has been an exporter. Market expe- rience is proxied by the number of products firm i exports to country j in t – 1. Product experience is given by the number of markets firm i serves with product p in t – 1. To account for the weight of product p in the exports of firm i, we also add the share of product p in the sales of firm i in t – 1. We expect that, as the share of a product increases, the probabil- ity that the firm will stop exporting the product declines. We also intro- duce a dummy that takes 1 if the firm is a multiproduct firm in t – 1 and 0 otherwise. Finally, as in the previous estimations, we control for the exporters’ location and add a measure of comparative advantage and an interaction term. We estimate three specifications. The first one is the baseline regression, which includes the tariff cuts and the share of prod- uct p in total sales. The second one includes measures of the exporters’ experience. The last one controls for the exporters’ location as well as for sector characteristics. We estimate each specification using the tariff cuts as well as the tariff levels. Results are reported in table 2.4. While the tariff levels do not seem to affect the probability of exiting export markets, the coefficient of the tariff cuts is negative and statisti- cally significant. However, its effect is very small (0.001). The coefficients of the other covariates are very similar and statisti- cally significant across all specifications. The size of previous sales has a negative effect on the probability of stopping exports of product p. But the effect is very small (that is, 0.001 to 0.002). Products with a small share in the export mix have a higher probability of exit. This is in line with the model of Bernard, Redding, and Schott (2009), which predicts that firms will stop producing or selling products that are not in their core competencies. 48 Molina, Bussolo, and Iacovone Table 2.4 Estimates of the Probability of Exit Variable (1) (2) (3) (4) (5) (6) Tariff cut –0.001*** –0.001*** –0.001*** (0.000) (0.000) (0.000) Tariff 0.000 0.000 0.000 (0.000) (0.000) (0.000) Share in sales (t – 1) –0.001*** –0.002*** –0.002*** –0.000*** –0.002*** –0.002*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Market experience (t – 1) –0.003*** –0.003*** –0.005*** –0.004*** (0.000) (0.000) (0.000) (0.000) Product experience (t – 1) –0.018*** –0.016*** –0.020*** –0.018*** (0.001) (0.001) (0.001) (0.001) Years as an exporter –0.014*** –0.014*** –0.010*** –0.009*** (0.001) (0.001) (0.002) (0.002) Multiproduct –0.206*** –0.187*** –0.232*** –0.208*** exporter (t – 1) (0.010) (0.009) (0.015) (0.014) EPZ –0.093*** –0.134*** (0.006) (0.009) Comparative advantage –0.052*** –0.045*** (0.004) (0.006) Comparative advantage x –0.000 tariff cut (0.000) Comparative advantage 0.000 x tariffs (0.000) Number of observations 90,261 90,223 90,149 47,145 47,113 47,102 R2 0.720 0.736 0.738 0.693 0.711 0.715 Source: Authors’ calculations. Note: Time, country, and sector fixed effects are used in all specifications, but are not reported. Errors are clustered by product. Robust standard errors are in parentheses. *** p < .01. Market and product experience decreases the probability of exiting foreign markets. The more experienced a firm is, the lower is the probability of stopping the sale of product p in market j. But product experience seems to matter more than market experience when it comes to exit. As for the number of years as an exporter, the probability of stopping the sale of a product in a given market decreases 1 percentage point with an additional year of export experience. Also being a multi- product exporter in t – 1 decreases the probability of dropping a product- market combination in t by 18.7 to 23.2 percent. This is the covariate with the largest effect on the probability of stopping the sale of a prod- uct in a given market. The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 49 Firms located in an export-processing zone show a lower probability of exiting from market j. This is not surprising, as the requirements to be located in an EPZ are more stringent than those to be located in the national territory. Larger fixed costs in the EPZ could therefore explain the better survival of EPZ firms through a hysteresis mechanism (Baldwin and Krugman 1989; Dixit 1989).30 Another explanation could be that the firms in export zones are the most productive ones (that is, firms self- select into export zones) and therefore also exhibit higher survival rates. Finally, exporters operating in a comparative advantage sector have a lower probability of discontinuing the sale of a product in a given market. Yet no significant effect is found for the interaction term between the comparative advantage measure and the tariff cuts (tariff levels). Recent studies have documented the low survival of trade relation- ships in their first years of activity. Eaton and others (2007) show that during the 1996–2005 period, most Colombian exporters survived only one year. In another recent study, Cadot and others (2010) also document the short duration of the export activity for exporters in four African countries and look at its determinants. These firm-level analyses as well as other studies at the product level (Besedes and Prusa 2006, 2007) high- light the importance of surviving in those first years. Dominican exporters are no exception. On average, 63 percent of the new exporters last only one year (that is, exit after one year). We perform the same exercise as before, but this time we consider only new entrants. We test whether the probability of stopping the export of product p to market j in t of firm i that started exporting in t – 1 diminishes with a decline in tariffs.31 We find that only tariff cuts affect the probability of exiting after one year. However, its effect is very small (0.001), and it disappears once we con- trol for the exporters’ location and sector characteristics. Young exporters located in an EPZ show an exit probability that is 18 percent lower than those in the national territory. As for the remaining variables, they are in most cases similar to those in our previous exercise. Conclusions Thanks to the implementation of the DR-CAFTA in 2007, Dominican exporters face better market access not only in the United States, but also in El Salvador, Guatemala, Honduras, and Nicaragua. Using an original firm-level data set with exports by product and destination for the 2002–09 period, this chapter looks at the Dominican exporters’ responses following the agreement and analyzes whether increased market access 50 Molina, Bussolo, and Iacovone supported the expansion of the extensive margin and improved exporters’ survival. Based on the theoretical findings of Melitz (2003) and Bernard, Redding, and Schott (2009), we test the effect of tariff reduc- tions on (a) the number of new exporters, (b) the number of existing exporters that added a new product-market relationship to their export mix, and (c) the probability of exiting a given market. Our results suggest that tariff reductions have had a positive, but very small, effect on the number of new exporters. A similar result is found in the case of incumbent exporters. Tariff preferences seem to affect their behavior, but the effect is also fairly small. Such results could suggest that other trade barriers such as standards, phytosanitary requirements, credit constraints, and transport costs, among others, could be preventing exporters from taking full advantage of the agreement. This implies that beyond tariffs, further efforts must be undertaken to identify the factors that are constraining exporters from benefiting from the agreement. This is also essential for the design of complementary policies aiming to stim- ulate export participation. Finally, we also look at the relationship between export survival and tariff preferences. Survival among Dominican exporters is very low: six out of 10 firms exit the export market after one year. We test whether tar- iff cuts help exporters to consolidate their position in a market and diminish their probability of exit. We find that tariff cuts improve survival rates, especially among experienced exporters, although the effect is very small. Our findings also highlight the great challenge that export survival represents for young exporters and therefore the need for policies that help them to develop and grow in foreign markets. Other important results concern the exporters located in EPZs. In gen- eral, these seem to perform better than their peers in the national terri- tory when it comes to surviving in export markets. Our results suggest that the probability of exiting the market is 9 to 18 percent lower for EPZ exporters than for exporters located in the national territory; however, this may be due to self-selection—that is, better firms may choose to locate in EPZs—rather than to the effectiveness of the favorable fiscal regime for EPZs. Our findings provide some preliminary insight into the effects of the DR-CAFTA on Dominican exporters, in particular into the effects of improved market access on the extensive margin in the Dominican Republic. However, trade liberalization may also affect exporters’ per- formance through other channels such as access to cheaper inputs. Our future research will look at these effects, as well as those on the The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 51 exporters from other CAFTA members. The purpose is to get a more complete assessment of the effects of the DR-CAFTA on exporters in the region. Notes 1. The agreement was signed in 2004, but it was ratified only in 2007. Compared to earlier agreements with the United States, the Caribbean Basin Initiative, and the CBTPA, the DR-CAFTA covers almost all products and, unlike its predecessors, is based on reciprocal trade preferences. In 2008, the Dominican trade policy with the European Union also changed. The Cotonou Agreement, which granted nonreciprocal trade preferences to the Dominican Republic, was replaced by Economic Partnership Agreements, a reciprocal trade regime. 2. Their setup generates not only firm selection, but also product selection by including a product-market fixed cost and by taking into account (product- country) demand heterogeneity. 3. A similar result is also predicted in Eckel and others (2009). 4. In our study, we do not classify Costa Rica as part of the group of countries that are signatories of the DR-CAFTA, since this country ratified the agree- ment only in 2009. 5. Exporters are identified through their names. 6. From this universe, only 459 export in every year. 7. Calculations are available from the authors on request. 8. We exclude ferronickel exports because they involve only one exporter and account on average for 13 percent of the Dominican Republic’s total exports. Moreover, their export value is heavily dependent on international price fluc- tuations, which could also bias our results. Ferronickel exports accounted for 26 percent of total exports in 2007, but only 9 percent in 2008 as conse- quence of a price decline. 9. We exclude Costa Rica from the CAFTA country group, as it ratified the treaty only in 2009. 10. Exports to Costa Rica also experienced an important expansion since 2005, rising from US$6 million to US$31 million. This suggests that Costa Rican exporters may have anticipated the ratification of the DR-CAFTA. 11. Sectors (HS two-digit) with the largest exports to the CAFTA area in 2008 were plastics (22 percent), cotton (20 percent), and tobacco (19 percent). In the case of the United States, the main export sectors were apparel (20 per- cent), medical and surgical instruments (16 percent), machinery and electri- cal appliances (14 percent), and jewelry (14 percent). 52 Molina, Bussolo, and Iacovone 12. A trade relationship is defined as the combination of a firm, a product, and a destination. A new trade relationship can therefore be generated by the par- ticipation of new firms, the introduction of new products, or the introduction of new markets. Our unique data set contains information at the firm, prod- uct, and destination level, which allows for such decomposition. 13. Calculations are available from the authors on request. 14. The United States has eight categories. El Salvador, Nicaragua, and Honduras have 12 categories, and Guatemala has 11 categories. 15. There are also cases in which duty-free treatment is delayed and will not begin until seven or 12 years after the agreement enters into force. All tariffs will be eliminated in 20 years. 16. According to the U.S. International Trade Commission, before the agreement about 80 percent of exports from the Dominican Republic had preferential access in the U.S. market. 17. Sugar in the case of the Dominican Republic. 18. The provision was retroactive to January 1, 2004. 19. The Central American countries (that is, Costa Rica, El Salvador, Guatemala, Honduras, and Nicaragua) and the Dominican Republic signed a free trade agreement in 1998, which entered into force in all countries between 2001 and 2002. The agreement guaranteed duty-free entry to almost all goods that comply with the rules of origins and transformation criteria, excluding goods from export-processing zones. 20. To compute the median tariff in the case of the Central American members, we first compute the median tariff by product across the five countries and then calculate the overall median tariff. The median tariff cut by product and the overall median tariff cut are computed in the same way. These com- putations are possible using eight-digit data since countries in Central America share the same product classification at this level of disaggregation, namely the Sistema Arancelario Centroamericano. 21. U.S. tariff figures exclude products whose tariff scheme depends on the char- acteristics (weight, length, and so forth) of the good, as well as some tobacco products whose tariffs are equal to 350 percent. In total, we exclude 963 products. 22. The effect of tariffs on export survival could have an ambiguous lagged effect if we take into account the survival behavior of new entrants. Consider a decline in tariffs in t; according to the theoretical evidence this would increase unambiguously the survival of incumbent exporters in t and in subsequent years, but also increase the number of new exporters. Yet the empirical evidence shows that most new exporters live only for one year; that is, most of the new exporters will exit in t + 1. If the number of new entrants is very large in t and so is the number of exits in t + 1, the export The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 53 survival could be lower than in previous years, unless the survival of new exporters also improves. 23. To have the largest product concordance between tariff data and exports, tar- iffs are averaged at the HS six-digit level. To compute the tariff cuts, we fur- ther assume that there are no changes in the trade policy of other countries vis-à-vis the Dominican Republic during the period under consideration. 24. The input-output table comes from http://www.bancentral.gov.do/ publicaciones_economicas.asp. 25. The RCA index for a given year is given by RCAjk = (xik/Xi) / (xwk/Xw), where xik and xwk are the values of country i’s exports of product k and world exports of product k and where Xi and Xw refer to the country’s total exports and the world’s total exports. The RCA ranges between zero and infinity. An RCA lower than 1 suggests that the country has a revealed comparative disadvan- tage in the product. Similarly, if the index exceeds unity, this implies that the country has a revealed comparative advantage in the product. Moreover, the RCA is a static concept and does not allow for comparison across time. One way to deal with this is to demean the RCA using the average RCA in each year. This is not necessary in the present study, as we have incorporated time fixed effects in the regression. 26. TRAINS/WITS is the Trade Analysis and Information System/World Integrated Trade Solution, developed by the World Bank and United National Conference on Trade and Development. 27. The effect of tariffs seems to be larger in the short run than in the long run. 28. We also estimate equations 2.1 and 2.2 in their log-linear version (that is, dependent variable in logs); results remain very similar for tariffs and tar- iff cuts, although not always significant. Another possibility would have been to employ a fixed-effect poisson estimator. However, this model does not allow us to evaluate the effect of time-invariant variables such as the firm location, EPZ (Cameron and Trivedi 2005, ch. 23). As a robustness check, we run equations 2.1 and 2.2 using the poisson estimator (ML and QLM), but excluding the variable EPZ. The results are very similar to those of the ordinary least squares (OLS) model: the effects of tariffs are very small (in terms of incidence ratios). We also estimate equations 2.1 and 2.2 using the negative binomial model, but the latter is subject to stronger distributional assumptions and does not converge for all specifications. 29. The main drawback of the linear probability model (LPM) is that the pre- dicted probabilities can be negative and larger than 1. But despite this, the LPM estimator remains a good indicator of the size of the effect. Moreover, as a robustness check, we test the model using a logit estimator with fixed effects—that is, conditional logit (not shown here). In this type of model, only the signs of the coefficients can be interpreted (see Wooldridge 2001, ch. 15). 54 Molina, Bussolo, and Iacovone The results on the signs of the coefficients are similar to those obtained with the LPM model. 30. Hysteresis refers to the persistency of a firm’s export participation as a conse- quence of the sunk costs associated with entry into new markets (Baldwin 1988). Entry into new markets is generally costly, so if a firm enters a market following a shock (that is, an exchange rate depreciation) it will not necessar- ily exit once the shock disappears. 31. Results are not shown due to space considerations but are available on request. References Albornoz, F., H. Calvo Pardo, G. Corcos, and E. Ornelas. 2009. “Sequential Exporting.” CEP Discussion Paper dp0974, London School of Economics, Centre for Economic Performance, London. Baldwin, J., and W. Gu. 2004. “Trade Liberalization: Export-Market Participation, Productivity Growth, and Innovation.” Oxford Review of Economic Policy 20 (3, Autumn): 372–92. Baldwin, R. 1988. “Hysteresis in Import Prices: The Beachhead Effect.” American Economic Review 78 (4): 773–85. Baldwin, R., and P. Krugman. 1989. “Persistent Trade Effects of Large Exchange Rate Shocks.” Quarterly Journal of Economics 104 (4): 635–54. Bernard, A., J. B. Jensen, and P. Schott. 2003. “Falling Trade Costs, Heterogeneous Firms, and Industry Dynamics.” NBER Working Paper 9639, National Bureau of Economic Research, Cambridge, MA. Bernard A., S. J. Redding, and P. Schott. 2009. “Multi-Product Firms and Trade Liberalization.” Working Paper 09-21, U.S. Census Bureau, Center for Economic Studies, Washington, DC. Besedes, T., and T. J. Prusa. 2006. “Ins, Outs, and the Duration of Trade.” Canadian Journal of Economics 39 (1): 266–95. ———. 2007. “The Role of Extensive and Intensive Margins and Export Growth.” NBER Working Paper 13628, National Bureau of Economic Research, Cambridge, MA. Cadot, O., L. Iacovone, F. Rauch, and D. Pierola. 2010. “Success and Failure of African Exporters.” World Bank, Washington, DC. Cameron, A. C., and P. Trivedi. 2005. Microeconometrics: Methods and Applications. Cambridge, U.K.: Cambridge University Press. Debaere, P., and S. Mostashari. 2005. “Do Tariffs Matter for the Extensive Margin of International Trade? An Empirical Analysis.” CEPR Discussion Paper 5260, Centre for Economic Policy Research, London. The DR-CAFTA and the Extensive Margin: A Firm-Level Analysis 55 Dixit, A. 1989. “Entry and Exit Decisions under Uncertainty.” Journal of Political Economy 97 (3): 620–38. Eaton, J., M. Eslava, M. Kugler, and J. Tybout. 2007. “Export Dynamics in Colombia: Firm-Level Evidence.” NBER Working Paper 13531, National Bureau of Economic Research, Cambridge, MA. Eckel, C., L. Iacovone, B. Javorcik, and P. Neary. 2009. “Multi-product Firms at Home and Away.” Centre for Economic Policy Research, London. http://www .cepr.org/meets/wkcn/1/1721/papers/Neary.pdf. Freund, C., and E. Ornelas. 2009. “Regional Trade Agreements.” CEP Discussion Paper dp0961, London School of Economics, Centre for Economic Performance, London. Gómez, S., and C. Volpe. 2008. “Trade Policy and Export Diversification: What Should Colombia Expect from the FTA and the United States?” Working Paper 05, Inter-American Development Bank, Washington, DC. Hornbeck, J. F. 2008. “The Dominican Republic-Central America-United States Free Trade Agreement (CAFTA-DR).” Congressional Research Service Report for the U.S. Congress, Washington, DC. http://www.nationalaglawcenter.org/ assets/crs/RL31870.pdf. Kehoe, T. J., and K. J. Ruhl. 2009. “How Important Is the New Goods Margin in International Trade?” Staff Report 324, Federal Reserve Bank of Minneapolis. Laursen, K. 1998. “Revealed Comparative Advantage and the Alternatives as Measures of International Specialisation.” DRUID Working Paper 98-30, DRUID, Copenhagen Business School, Department of Industrial Economics and Strategy, Copenhagen; Aalborg University, Department of Business Studies, Aalborg. Melitz, M. 2003. “The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity.” Econometrica 71 (6): 1695–725. Pavcnik, N. 2002. “Trade Liberalization, Exit, and Productivity Improvement: Evidence from Chilean Plants.” Review of Economic Studies 69 (1): 245–76. Swiston, A. 2010. “Spillovers to Central America in Light of the Crisis: What a Difference a Year Makes.” IMF Working Paper WP35, International Monetary Fund, Washington, DC. Trefler, D. 2004. “The Long and Short of the Canada-U. S. Free Trade Agreement.” American Economic Review 94 (4): 870–95. Wooldridge, J. 2001. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press. CHAPTER 3 Exports, Wages, and Skills: Implications for CAFTA Irene Brambilla, Lucio Castro, and Guido Porto This chapter explores whether implementation of the Dominican Republic–Central America Free Trade Agreement (DR-CAFTA) has as yet resulted in changes in the skill premium (that is, the ratio of skilled to unskilled wages). We find that the evidence is consistent with an increase in the skill premium in Central America in recent years, but not in other Latin American non-CAFTA countries. We also explore whether the estimated increase in the skill premium is driven by within-sector changes or by between-sector compositional shifts in the skilled labor force. We find that the former factor dominates, and, in fact, the empirical evidence indicates that the service sector, which is the largest employer of skilled workers, has increased its share of this type of worker. In contrast, half of the changes observed in the skill premium in the non-CAFTA comparison group are explained by shifts in the composition of skilled labor across sectors. Given these trends, this chapter seeks to provide evidence on the overall link between exports, wages, and skill utilization to illustrate the likely impacts of the DR-CAFTA on workers, employment, and skill composition (as well as on the skill premium). This evidence can help guide policy makers in designing a set of policies to boost the gains from the CAFTA agreement. Our analysis is based on a detailed review of the literature, including results from recent research on this topic. We first review a paper by 57 58 Brambilla, Castro, and Porto Bernard and others (2007), who use different types of data to character- ize exporters and importers in international trade. Their paper provides a summary of the literature and is thus a natural starting point in our analy- sis. We then describe some of the available results for Latin America. To this end, we review a paper by Casacuberta and others (2007). This paper focuses on exporting firms and provides a full characterization of the exporter premium in the region, including a productivity premium and a wage premium. We then discuss recent research by Brambilla and others (2010) and Brambilla, Lederman, and Porto (2010), who look at the over- all link between exports and wages as well as between export destina- tions, wages, and skill utilization. This research is relevant for CAFTA because it suggests that exporting to high-income countries typically requires higher skills than either exporting to low- or middle-income countries or producing for the local market. Since DR-CAFTA implies access to U.S. markets, which are high-income markets, the agreement may have implications for skill utilization and the skill premium that are worth discussing in detail. Firms in International Trade This section is based on Bernard and others (2007), who provide a nice overview of the main features of exporting firms vis-à-vis firms devoted to the local market. The authors also review some of the characteristics of importing firms. A basic feature of the data is how rare firm exporting is. Table 3.1 illustrates this point with data from the 2002 U.S. Census of Manufactures. Column 1 reports the distribution of firms across three-digit industries; col- umn 2 displays the share of firms in each industry that actually do some exporting. Two conclusions emerge: while the overall share of U.S. manufac- turing firms that export is only 18 percent, there are wide differences within industry categories. At the top, in sectors like computer and electronic prod- ucts and electrical equipment and appliances, 38 percent of firms export; at the other end, only 5 percent of firms export in printing and 8 percent export in apparel manufacturing or wood product manufacturing. Moreover, even among those exporting, export sales are only a small fraction of the firms’ activities. This information is in the last column of table 3.1. The average share of exports is 14 percent (which is lower than the 18 percent share of exporting firms). There is also a lot of heterogene- ity across sectors, with the highest shares observed in computer and elec- tronic products (21 percent) and the lowest shares in beverages and tobacco (7 percent) and paper manufacturing (9 percent). Exports, Wages, and Skills: Implications for CAFTA 59 Table 3.1 Exporting by U.S. Manufacturing Firms, 2002 % of firms Mean exports as NAICS industry % of firms that export % of total shipments 311 Food manufacturing 6.8 12 15 312 Beverage and tobacco products 0.7 23 7 313 Textile mills 1.0 25 13 314 Textile product mills 1.9 12 12 315 Apparel manufacturing 3.2 8 14 316 Leather and allied products 0.4 24 13 321 Wood product manufacturing 5.5 8 19 322 Paper manufacturing 1.4 24 9 323 Printing and related support 11.9 5 14 324 Petroleum and coal products 0.4 18 12 325 Chemical manufacturing 3.1 36 14 326 Plastics and rubber products 4.4 28 10 327 Nonmetallic mineral products 4.0 9 12 331 Primary metal manufacturing 1.5 30 10 332 Fabricated metal products 19.9 14 12 333 Machinery manufacturing 9.0 33 16 334 Computer and electronic products 4.5 38 21 335 Electrical equipment and appliances 1.7 38 13 336 Transportation equipment 3.4 28 13 337 Furniture and related products 6.4 7 10 339 Miscellaneous manufacturing 9.1 2 15 Aggregate manufacturing 100 18 14 Source: Bernard and others 2007. Note: The first column of numbers summarizes the distribution of manufacturing firms across three-digit NAICS (North American Industry Classification System, U.S. Census Bureau) manufacturing industries. The second reports the share of firms in each industry that export. The final column reports mean exports as a percentage of total shipments across all firms that export in the noted industry. Clearly, exporting is a relatively rare phenomenon, and it is thus not sur- prising to learn that exporters are very different from nonexporters in var- ious characteristics. Using U.S. data, Bernard and others (2007) calculate the export premiums in U.S. manufacturing in 2002 for different firm char- acteristics, and we report those in table 3.2. If we focus on unconditional differences (column 1), we see that exporting firms have 119 percent more employment, 148 percent more shipments, 26 percent higher value added per worker, 2 percent higher productivity, 17 percent higher wages, 32 per- cent higher capital-labor ratios, and 19 percent higher skill per worker. Column 2 reports results from a regression model that includes indus- try fixed effects in the explanatory variables. This allows us to control for 60 Brambilla, Castro, and Porto Table 3.2 Exporter Premiums in U.S. Manufacturing, 2002 Variable (1) (2) (3) Log employment 1.19 0.97 n.a. Log shipments 1.48 1.08 0.08 Log value added per worker 0.26 0.11 0.10 Log TFP 0.02 0.03 0.05 Log wage 0.17 0.06 0.06 Log capital per worker 0.32 0.12 0.04 Log skill per worker 0.19 0.11 0.19 Source: Bernard and others 2007. Note: TFP = total factor productivity. All results are from bivariate ordinary least squares (OLS) regressions of the firm characteristic in the first column on a dummy variable indicating the firm’s export status. Regressions in col- umn 2 include industry fixed effects. Regressions in column 3 include industry fixed effects and log firm employ- ment as controls. TFP is computed as in Caves, Christensen, and Diewert (1982). Capital per worker refers to capi- tal stock per worker. Skill per worker is the share of nonproduction workers in total employment. All results are significant at the 1 percent level. n.a. = Not applicable. the basic inherent heterogeneity of firms across industries. The same dif- ferences between exporters and nonexporters can still be seen (although the differences are now smaller because export participation is positively correlated with industry characteristics). Within industries, exporters are larger than nonexporters: employment is 97 percent higher, and ship- ments are 108 percent higher. Exporters are also more productive by 11 percent in value added per worker and by 3 percent in total factor pro- ductivity (TFP). Exporters also pay higher wages, by about 6 percent, and are more cap- ital intensive (12 percent) and skill intensive (11 percent). While these results are based on U.S. data, they are representative of the literature. We show below similar results for select outcomes (namely, wages and pro- ductivity) for Latin American countries. The evidence discussed so far has focused on exporters, mainly because of data limitations. It is, however, interesting to ask whether importers also have special characteristics. Bernard and others (2007) are in a unique position to shed some light on this matter; they have information on U.S. importers from the Linked-Longitudinal Firm Trade Transaction Database (LFTTD), which is based on data collected by the U.S. Census Bureau and the U.S. Customs Bureau. This data set captures all U.S. inter- national trade transactions between 1992 and 2002. The main results are reported in table 3.3, which shows that importers and exporters have basically the same characteristics. First, the act of importing is rare, and it is even rarer than the act of exporting. In the Table 3.3 Exporting and Importing by U.S. Manufacturing Firms, 1997 % of firms that import NAICS industry % of all firms % of firms that export % of firms that import and export 311 Food manufacturing 7 17 10 7 312 Beverage and tobacco products 1 28 19 13 313 Textile mills 1 47 31 24 314 Textile product mills 2 19 13 9 315 Apparel manufacturing 6 16 15 9 316 Leather and allied products 0 43 43 30 321 Wood product manufacturing 5 15 5 3 322 Paper manufacturing 1 42 18 15 323 Printing and related support 13 10 3 2 324 Petroleum and coal products 0 32 17 14 325 Chemical manufacturing 3 56 30 26 326 Plastics and rubber products 5 42 20 16 327 Nonmetallic mineral products 4 16 11 7 331 Primary metal manufacturing 1 51 23 21 332 Fabricated metal products 20 21 8 6 333 Machinery manufacturing 9 47 22 19 334 Computer and electronic products 4 65 40 37 335 Electrical equipment and appliances 2 58 35 30 336 Transportation equipment 3 40 22 18 337 Furniture and related products 6 13 8 5 339 Miscellaneous manufacturing 7 31 19 15 Aggregate manufacturing 100 27 14 11 Source: Bernard and others 2007. Note: The first column of numbers summarizes the distribution of manufacturing firms across three-digit NAICS industries. Remaining columns report the percentage of firms in each 61 industry that export, import, and do both. 62 Brambilla, Castro, and Porto LFTTD data, about 27 percent of firms export (this figure is higher than the census data), while only 14 percent import. As before, these shares vary significantly across industries. There is a strong correlation (0.87) between industries with a high share of importers and those with a high share of exporters. Moreover, 41 percent of exporting firms also import, while 79 percent of importers also export. Table 3.4 reports the “trading premium’’ for both exporters and importers. As shown, exporters and importers share a variety of positive attributes: they are both bigger and more productive, they pay higher wages, and they are more skill and capital intensive than nonexporters and nonimporters. This is consistent with all the previously available evidence. The Latin American Experience The data presented in Bernard and others (2007) are based on U.S. data. Many of the patterns found here are representative of the behavior of firms worldwide. In particular, we are interested in characterizing the export premiums in Latin America. To do this, we review results first reported in Casacuberta and others (2007), who focus on exports and wages and exports and productivity. Casacuberta and others (2007) estimate productivity and wage exporter premiums using the enterprise surveys available for Latin American and Caribbean countries. They regress measures of TFP and average wages on firms’ age, size (number of employees), foreign ownership, unique estab- lishment, log of capital per worker, and region and industry dummies. They carry out two exercises. First, they run one regression for each of the Table 3.4 Trading Premiums in U.S. Manufacturing, 1997 Exporter Importer Exporter and premium premium importer premium Variable (1) (2) (3) Log employment 1.50 1.40 1.75 Log shipments 0.29 0.26 0.31 Log value added per worker 0.23 0.23 0.25 Log TFP 0.07 0.12 0.07 Log wage 0.29 0.23 0.33 Log capital per worker 0.17 0.13 0.20 Log skill per worker 0.04 0.06 0.03 Source: Bernard and others 2007. Note: All results are from bivariate OLS regressions of the firm characteristic listed on the left on a dummy variable noted at the top of each column as well as industry fixed effects and firm employment as additional controls. Employment regressions omit firm employment as a covariate. TFP is computed as in Caves, Christensen, and Diewert (1982). Exports, Wages, and Skills: Implications for CAFTA 63 16 countries for which data are available. Second, they run a regression with all countries pooled together and country dummies to account for country heterogeneity. Results are reported in table 3.5. Exporters are indeed better than nonexporters in Latin America, too. First, exporters are more productive. In 14 of the 16 Latin American countries in the sample, there is a positive TFP exporter premium. The exporter premium also shows up in the pooled regression. Second, exporters pay higher wages. Again, in 15 of the countries, the regressions capture a positive wage exporter premium. This wage exporter premium also shows up in the pooled regressions. Notice that only about half of these positive exporter premiums are statistically different from 0 (eight in the case of TFP and 12 in the case of wages), which indicates that they are not estimated very precisely. However, they tend to be very large. For Latin America and the Caribbean as a whole, the productivity exporter premium is about 34 percent, and exporters tend to pay wages that are about 20 percent higher. An Exploration into the Exporter Premium The exporter premiums documented above indicate that firms that export are much more productive and pay much higher wages to their workers than other firms. This, however, is just a correlation and tells lit- tle about the direction of causality: do exporters become “good” firms, or do “good” firms become exporters? Under the first hypothesis (that is, exporters become good firms), exporting improves productivity. The most common explanation, known as “learning by exporting,” is that exporters acquire information from for- eign customers on how to improve the product design, the manufactur- ing process, or the quality of the good (Westphal, Rhee, and Pursell 1984).1 Foreign demand also allows domestic firms—particularly in small countries—to take advantage of unexploited economies of scale. Under the second hypothesis (that is, good firms become exporters), the best firms self-select into export markets. One rationale for this self- selection is that important entry barriers exist in export markets because of the higher costs associated with selling in foreign markets (transport, but also distribution, marketing, and even production costs when firms need to adapt their product to foreign standards). Thus, only the more productive firms can enter foreign markets, and the observed differences between exporters and nonexporters can then be explained by preexist- ing differences. These two hypotheses are obviously not mutually exclusive, but depending on which is the most important force, the policy implications 64 Brambilla, Castro, and Porto Table 3.5 Productivity and Wage Exporter Premiums in Latin America and the Caribbean Premium country TFP Log of wages Latin America and 0.339*** 0.203*** the Caribbean (0.035) (0.026) Argentina 0.597*** 0.09 (0.111) (0.074) Bolivia –0.353* 0.084 (0.193) (0.163) Brazil 0.478*** 0.274*** (0.070) (0.054) Chile 0.427*** 0.262*** (0.135) (0.090) Colombia 0.304*** 0.275*** (0.083) (0.071) Costa Rica 0.011 0.538*** (0.009) (0.177) Ecuador 0.091 0.017 (0.259) (0.173) El Salvador 0.357*** 0.305*** (0.126) (0.092) Guatemala 0.153 0.347** (0.148) (0.143) Honduras 0.22 0.208* (0.163) (0.126) Mexico 0.001 0.124 (0.132) (0.096) Nicaragua –0.02 –0.019 (0.109) (0.080) Panama 0.024 0.257 (0.362) (0.125) Paraguay 0.485 0.052 (0.297) (0.179) Peru 0.520*** 0.297** (0.160) (0.121) Uruguay 0.678*** 0.465*** (0.185) (0.105) Source: Casacuberta and others 2007. Note: Standard errors are in parentheses. An OLS regression is run for each country, as well as a pooled regression labeled for Latin America and the Caribbean. Only the coefficient on the exporter dummy is reported. All regres- sions include as control variables firm-, region-, and industry-level characteristics (see text for more details). * p < .10, ** p < .05, *** p < .01. Exports, Wages, and Skills: Implications for CAFTA 65 can be very different. On the one hand, export promotion activities, which are quite common in Latin America, are often justified on the basis of the learning-by-exporting explanation. On the other hand, the self- selection explanation suggests that policy makers should focus their efforts on the internal determinants of productivity growth. The existing literature offers no clear-cut answer regarding the relative strength of the self-selection hypothesis versus the learning-by-exporting hypothesis. By nature, this literature is country specific, and depending on the country examined, studies seem to reach different conclusions. A sur- vey follows. We begin with the productivity premium, and we then turn to the wage premium. In their paper, Casacuberta and others (2007) survey 54 studies (cov- ering 70 countries) that look at the productivity premium associated with export activity. In 86 percent of these studies, exporters are found to be more productive than nonexporters. Most of these studies—with rare exceptions (see, for example, the results for the Republic of Korea in Aw, Chung, and Roberts 2000)—find evidence of self-selection: good firms become exporters, suggesting that penetrating foreign markets may require higher productivity. About 60 percent of the studies test the learning-by-exporting hypothesis, but the evidence is mixed. Half of the studies find support for it, and the other half find no evidence of differ- ences in productivity growth between firms that just became exporters and nonexporters. Thus, the general messages coming from the literature are that exporters are indeed more productive that nonexporters, that firms do self-select into the export market, but that exporting does not always improve productivity (or does so only half the time). These findings suggest that a lot of heterogeneity exists across studies in terms of the learning-by-exporting hypothesis. To illustrate in which types of countries exporting leads to productiv- ity gains at the firm level, Casacuberta and others (2007) run probit regressions in which the explained variable is a dummy that takes the value of 1 when the study finds that in a particular country exports cause pro- ductivity gains and 0 when the study finds that exports do not cause pro- ductivity gains (any study in which the question of causality is not addressed is excluded from this regression). A regression is also run that explores in which types of countries learning by exporting is more likely. The dummy is regressed on country characteristics, such as the degree of trade openness of the country, its level of development, its size, and vari- ables capturing the investment climate. 66 Brambilla, Castro, and Porto The purpose of this exercise is to illustrate the type of countries in which causality running from exports to productivity is most likely to be found. One would expect exports to be more likely to cause productivity increases in poorer countries with a smaller domestic market. Indeed, poorer countries tend to be further away from the technological frontier, and they have potentially much more to learn from foreign buyers. Similarly, in small countries, firms may count on foreign demand to take advantage of unexploited economies of scale. Openness to trade and a good investment climate may have ambiguous signs. A better investment climate allows firms to take advantage of business opportunities more freely, but may make it more difficult for exporters to appropriate these productivity gains when barriers to enter or exit are small. Obviously, this possibility does not necessarily mean that there are no productivity pre- miums or that exporting does not allow firms to become productive. The point is that it is difficult or impossible for the statistician to identify this effect if the benefits created by the exporter are easily captured by all other firms in the economy. Table 3.6 reports results from these probit regressions. Each column is run with a different investment climate variable taken from the World Bank’s Doing Business database. They are not all included simultaneously because they tend to be highly collinear. The clear message of table 3.6 is that exporting is more likely to create productivity premiums in small Table 3.6 Exporting and Productivity Gains, by Investment Climate Difficulty Difficulty of entry of closing a Difficulty of Difficulty of Variable procedures business paying taxes firing workers Openness ([M + X ])/GDP) –0.47 –0.64 –1.73 –1.25 (0.97) (1.01) (1.32) (1.19) Level of development 0.05 0.01 0.05 0.20 (GDP per capita) (0.16) (0.15) (0.16) (0.2) Size (GDP) –0.31** –0.41** –0.57** –0.38** (0.15) (0.17) (0.23) (0.19) Investment climatea 0.13 –0.41 –0.98** 0.40 (0.73) (0.36) (0.50) (0.31) Pseudo R2 0.22 0.15 0.25 0.22 Source: Casacuberta and others 2007. Note: Standard errors are in brackets. All regressions are estimated using probit where the left-hand-side variable takes the value of 1 when a study finds that export causes growth and 0 when it does not find any evidence of causality. Each column runs this regression using a different variable to capture the investment climate. Each regression has 34 observations. All regressions include a dummy equal to 1 when the period under examination is in the 1990s or later. This dummy is never significant. a. See the variables in the top row. ** p < .05. Exports, Wages, and Skills: Implications for CAFTA 67 countries. This finding somehow gives more prominence to the economies-of-scale rationale for productivity premiums than to the knowledge-acquisition hypothesis, although this specification clearly does not allow disentanglement of these two forces. Probably because of the conflicting forces, the level of development does not seem to be an important determinant of the causality between exports and productivity: poorer countries have much more to learn from foreign buyers, but absorbing this knowledge may be more difficult. The degree of trade openness is also always statistically insignificant. The investment climate variables give an ambiguous picture, but the only result that is statistically significant tends to suggest that a cumbersome business environment is not likely to help exporters to take advantage of some of the potential benefits in foreign markets. Finally, two important points should be kept in mind. First, the determi- nants of self-selection remain an open question, and little work has been done to explain the sources of productivity growth before entry into the export market. As Yeaple (2005) argues, productivity is likely to be an endogenous decision, and trade opportunities may induce some firms to adopt new technologies. The expected future entry into export markets may well encourage firms to invest in new technology and product design and to benefit from the experience and know-how of potential foreign buyers. Thus, the increase in productivity observed before entering export markets may well be due to this export potential. Alvarez and López (2005) call this conscious self-selection, and they find evidence using plant- level data from Chile that self-selection is indeed a conscious process, where firms increase productivity with the objective of becoming exporters.2 Second, regardless of whether exports cause firm-level productivity gains, the fact that exporters are more productive and larger (perhaps exclusively because of self-selection) suggests that, as countries increase their export orientation, larger and more productive firms will produce a larger share of national output. This reallocation of resources from less productive and smaller firms to more productive and larger firms in itself is a source of aggregate gross domestic product (GDP) growth. Turning now to the wage premiums, Casacuberta and others (2007) review 30 studies that explore wage premiums associated with export activity. In two-thirds of the studies analyzed, there is evidence of an over- all wage premium. In all but two studies, evidence exists of large skilled- wage premiums,3 whereas unskilled workers in the export sector benefit from a premium in only 45 percent of the studies. Thus, the big-picture message emerging from this review is that exporters pay higher wages to skilled workers and sometimes pay higher wages to less-skilled workers. 68 Brambilla, Castro, and Porto The issue of causality is seldom addressed in the literature, however. Although many of the papers are based on panel data that allow for controls of fixed effects and unobserved heterogeneity at the firm level, the issue of causality from exports to wages remains largely unsolved. One exception is Feenstra and Hanson (1997), who set up an instrumental variables estimator of the effects of foreign direct investment and outsourcing on wages and wage inequality. Another exception is Verhoogen (2008), who uses the Mexico peso devaluation to establish that exporting leads to higher wage payments in Mexican manufacturing. Below, we review recent research that addresses the issue of causality. Before turning to the causality problem, we again explore the hetero- geneity in the findings regarding overall wage premiums and unskilled wage premiums to see whether some country characteristics are associ- ated with a higher likelihood of finding a wage premium associated with exports (regardless of whether self-selection or causality is involved). Again, for the same reasons as before (the productivity channel), one may expect the wage premium to be especially likely in small and poor coun- tries. A better and more competitive investment climate should in princi- ple help transmit these gains from firms to workers, but the same caveat discussed earlier applies. One would also expect unskilled wages to have a higher premium in countries with an abundant supply of skilled work- ers. Indeed, because unskilled workers are the rare factor in skill-abundant countries, one would expect them to benefit later from these wage premiums. Table 3.7 shows the results of the probit regressions. The first two columns explain the presence of an overall wage premium using two investment climate variables; the second two columns explain the pres- ence of an unskilled wage premium, again using two different investment climate variables. The regression could not be run for the skilled wage premium because almost all the studies surveyed find that a skilled wage premium exists. The clear message coming out of table 3.7 is that wage premiums are more likely to be observed in small countries. This finding again provides tangible evidence of the importance for firms in small countries of being able to take advantage of unexploited economies of scale in world mar- kets. Unskilled wage premiums are also more likely to be observed in skill-abundant countries. This finding may be because exporters may need to pay higher wages to attract the rare factor (unskilled workers) to their firms in countries where skilled workers are relatively abundant. Table 3.7 Wage Premiums, by Investment Climate Overall wages Unskilled wages Difficulty of entry Difficulty of Difficulty of entry Difficulty of firing Variable procedures firing workers procedures workers Openness ([M + X]/GDP) –0.40 –0.23 –1.36 –1.53 (0.96) (1.04) (0.86) (1.14) Level of development 1.49 0.64 –0.90 –0.90 (GDP per capita) (1.37) (0.65) (0.67) (0.68) Size (GDP) –2.43* –0.69 –1.45** –1.23** (1.39) (0.54) (0.57) (0.61) Unskilled abundance (unskilled workers to 0.92 –0.24 –4.16** –4.51** skilled workers) (0.95) (0.77) (1.45) (2.07) Investment climatea –2.76* –0.36 –0.55 0.05 (1.46) (0.56) (0.39) (0.35) Pseudo R2 0.36 0.13 0.38 0.35 Source: Casacuberta and others 2007. Note: Standard errors are in parentheses. All regressions are estimated using probit where the left-hand-side variable takes the value of 1 when a study finds that there are wage premiums (overall wage premiums for the first two columns and unskilled wage premiums for the last two columns). Each of the two columns runs the regression using a different variable to capture investment climate: either the difficulty of entry procedures or the difficulty of firing workers. Only 17 observations in the overall wage regression and 26 in the unskilled wage regression reflect the number of studies surveyed that tried to answer these questions. All regressions include a dummy equal to 1 when the period under examination is in the 1990s or later. This dummy is never significant. a. See the variables in the top row. * p < .10, ** p < .05. 69 70 Brambilla, Castro, and Porto The Skill Premium in Exporting To complete our discussion of the wage premium due to exporting, this section explores the wage-skill premium. We discuss scenarios in which exporting may lead to the existence of a skill premium, and we review evidence on this for Latin America. To this end, we summarize some of the key findings in a recent paper by Brambilla and others (2010). The authors work with 64 household surveys for 16 countries covering more than 5 million workers in the region. The countries included in the study are Argentina, Brazil, Chile, Colombia, Costa Rica, the Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, and Uruguay. Following the literature on industry wage differentials (Dickens, Katz, and Lang 1986; Dickens and Lang 1988; Gibbons and Katz 1992), the authors allow the skill pre- miums to vary across industries, as in Galiani and Porto (2010). Then they study econometrically the relationship between the industry skill premi- ums and the level of sectoral exports. Once the superior performance of exporting firms (as well as importing firms) has been established (see above), the analysis in Brambilla and others (2010) is useful to illustrate the existence of an exporter skill premium. Two leading theories explain this potential link between industry exports and skill premiums. One argues that the act of exporting requires activities that are skill intensive, although the production of the good may require unskilled labor. Exporting firms, and therefore industries with more exports in general, will thus demand higher skills and pay a higher skill premium. The alternative theory argues that exporting is associated with higher profits (because more productive firms self-select into exports) and these higher profits are shared with the workers via profit- sharing rules. The theory focusing on the need to engage in skill-intensive activities to export a product is based on Brambilla, Lederman, and Porto (2010). If skilled labor is imperfectly mobile and unskilled labor is perfectly mobile, unskilled labor earns an economywide competitive wage, while industries using skilled workers pay more. Exporting requires both the production of the physical units of the product and the provision of export services. These services include labeling, marketing, technical sup- port, and consumer support (web page, e-mail, warranty) and are assumed to be skill intensive.4 The high-export sector pays higher wages to their skilled workers, therefore, since the wage offered to the unskilled workers is assumed to be the same across industries (given the competi- tive national market for unskilled labor). An alternative theory is based on Exports, Wages, and Skills: Implications for CAFTA 71 profit-sharing mechanisms. Skilled workers demand a wage premium to exert the necessary effort because it is considered fair to share the profits of the firms. In consequence, while marginal firms pay the competitive outside wage, more profitable firms pay increasingly higher wages. In equilibrium, if high-export firms are high-profit firms, they offer higher wages to their skilled workers. Under both hypotheses, the industry- specific skill premium is an increasing function of the level of sectoral exports. Country and Industry Effects Brambilla and others (2010) estimate two-digit industry skill premiums for 16 Latin American countries and exploit these estimates to provide evidence in support of the claim that the premiums depend positively on sectoral exports. As a first step, we assess the role of country and industry dummies. More specifically, the industry skill premium is explained by (a) country dummies alone, (b) industry dummies alone, and (c) country and industry dummies. For each of these models, we calculate the R2 (adjusted) and the F-test of joint significance of each set of dummies. We do this for all sectors, for the manufacturing sectors, and for the nontrad- able (and services) sectors. If we include all sectors, country dummies alone account for 20 percent of the variance in the skill premium, while industry dummies alone account for almost 48 percent. Both sets of dum- mies jointly explain about 69.2 percent of the variation in the industry skill premium. The dummies are always jointly statistically significant. In this case, it appears that the industry dummies play a more important role than the country dummies. It should be kept in mind, however, that the comparison of R2 is a descriptive assessment of the role of the dummies in explaining the variance in the dependent variable. Exports and the Skill Premium As mentioned above, sectoral exports could be an important determinant of the industry-specific skill premiums. To assess this claim, Brambilla and others (2010) regress the skill premium for sector j in country c on the log of the ratio of sectoral exports to GDP together with country and industry dummies. The model is estimated with weighted least squares to account for the fact that the industry-specific skill premiums are esti- mated. The weights are thus the inverse of the standard errors. Naturally, these estimates do not provide any causal evidence; instead, they suggest a clear reduced-form interpretation to illustrate whether the data support any link between sectoral exports and sectoral skill premiums. 72 Brambilla, Castro, and Porto Table 3.8 presents the results. Column 1 shows the estimate of the model when the skill premiums are regressed on a constant and the log of the ratio of exports over GDP. The estimate is positive and significant, suggesting that the skill premium rises with exports. The estimate in col- umn 1 implies that doubling a sector’s share of exports over GDP (a change in the log of exports over GDP equal to 1) is associated with an increase of 0.0028 in the skill premium, that is, the wage differential between skilled and unskilled workers rises by 0.28 percentage point. Notice that the simulated shock of a change of 1 in the log of exports over GDP is reasonable because the standard deviation of the variable in our sample is about 2.1. Thus this association is positive and significant, but it is not very large. In columns 2 to 5 of table 3.8, we perform several robustness tests. The incidence of industry exports remains significant, with a similar magni- tude as in column 1. Column 3 includes country dummies only, and the link between exports and the skill premium disappears. In column 4, we include both sets of dummies, and the link disappears, too. Controlling for both country and industry dummies might be too restrictive, however. Country fixed effects explain about a third of the variation in the skill premium, and both country and industry dummies account for about 60 percent. This leaves little room for exports to explain the skill premium because much of the variation in the dependent variable is attenuated by the dummies. To learn more about the role of sectoral exports, we work with a more parsimonious version of the regression model where, instead Table 3.8 Exports and the Industry Skill Premium Variable (1) (2) (3) (4) (5) Log exports to GDP 0.0028*** 0.0033*** 0.0004 –0.0002 0.0027** (0.001) (0.0011) (0.0011) (0.0015) (0.001) Log GDP_pc 0.0284*** (0.004) Log skilled to unskilled –0.014*** (0.004) Country dummies No No Yes Yes No Industry dummies No Yes No Yes Yes Number of observations 273 273 273 273 273 R2 0.03 0.31 0.43 0.58 0.46 Source: Authors’ calculations. Note: Standard errors are in parentheses. ** p < .05, *** p < .01. Exports, Wages, and Skills: Implications for CAFTA 73 of country dummies, we control for country characteristics—namely, the log of per capita GDP and the ratio of skilled (completion of high school) over unskilled labor. These results are reported in column 5 of table 3.8. Both per capita GDP and the skill composition are statistically significant determinants of the industry skill premiums with the expected signs: richer countries seem to have greater disparities between skilled and unskilled wages, and, as expected, countries with a greater fraction (sup- ply) of skilled workers pay smaller skill premiums. The significance of these variables supports their use in lieu of country fixed effects. Also the R2 of the model remains high at 0.46, which is higher than the R2 from the model with country dummies. In these models, the coefficient of exports as a fraction of GDP is positive and statistically significant (col- umn 5), and the estimate is of similar magnitude as the one reported in columns 1 and 2. Exporting, Productivity, and Wages: Causality The main goal of this section, which closely follows Porto (2007), is to generate evidence on the causality of exports to wages with an applica- tion to Argentina. The empirical strategy exploits the Brazilian devalu- ation of 1999. Argentina and Brazil are major trade partners, and the Brazilian devaluation greatly affected Argentine exports. Having an exogenous shock to exports is crucial in identifying a causal relationship between exports and wages. The combination of the panel data set and the devaluation shock is an important instrument with which to address this problem because the same firm can be considered before and after the devaluation to see how the wages paid by this firm changed when an exogenous change occurred in exporting opportunities. This situation presents the opportunity to determine causality from exports to wages. Regression Model: Exports and Export Destinations This section examines whether wages and the share of nonproduction workers (an approximation of the percentage of skilled workers) depend on exports and on the country of destination of exports. Some of the fol- lowing hypotheses are tested: (a) whether exporting firms pay higher wages and have higher ratios of nonproduction workers than firms pro- ducing for the domestic market; (b) whether the composition of export destinations of a firm matters (that is, whether firms that export to rich countries pay higher wages and have higher ratios of nonproduction workers than firms that either produce for the domestic market or export 74 Brambilla, Castro, and Porto to low-income countries); and (c) whether product quality is one of the factors behind the wage-employment effect of exports. The following regression models are set up to test these hypotheses: lnwit = α1Expit + x ′ β1 + φ tw + φiw + ε it it w (3.1) lnsit = α12 Expit + x ′ β2 + φ ts + φ is + εit , it s where w is average wage paid by firm i at time t, and s is the share of skilled workers (nonproduction workers). Controls in x are industry dum- mies; location dummies; year dummies; indicators of whether the firm is foreign; the percentage of foreign ownership; the firm size, as measured by total number of workers and, alternatively, by sales; materials con- sumption as a proxy for productivity shocks; and age of the plant. The error terms have a firm fixed effect φi and φi The variable Exp captures w s exports and export destinations. Concretely, we include the share of exports in sales (to account for export status) and the share of exports to high-income destinations in total exports (to account for the role of export destinations). The model in equation 3.1 includes firm fixed effects that control for time-invariant unobserved heterogeneity. However, these regressions may still suffer from endogeneity or omitted variable biases. An instrumental variable approach is followed to address these issues and the issue of causality. The strategy is to explain both the level of exports of the firms and their composition of countries of destination by the exogenous expo- sure to the Brazilian devaluation of 1999. Notice that heterogeneity exists in the exposure to this shock because firms and industries that exported more to Brazil before the devaluation were more likely to be affected by the shock. Two endogenous variables are used in the model. One is the HIE vari- able—the share of exports to high-income countries—which is related to the export destinations of the firm. The other is the share of exports in sales, which is related more closely to export status. The instrument for the HIE share is built by interacting a post-devaluation dummy with the share of the industry’s exports that were destined for Brazil in 1998. More specifically, two specifications are adopted. In the nonparametric model with dummies, the impacts of the devaluation are allowed to vary from one year to the other (as the economy adjusted, exposure in 1999 was dif- ferent from exposure in 2000). Consequently, this instrument is built by interacting the level of exposure to Brazil before the devaluation—that is, Exports, Wages, and Skills: Implications for CAFTA 75 the share of exports to Brazil in 1998—with a 1999 dummy variable and a 2000 dummy. In the second specification, an alternative instrument is built, which is the interaction of the pre-devaluation share of exports to Brazil (at the firm level) with the exchange rate of the Brazilian currency in 1999 and 2000. This instrument is a parametric model of the exposure to the shock. Support for these instruments comes from the preview of patterns of exports in Argentina. The argument is that, following the devaluation, firms that were most exposed to the Brazilian devaluation had to adjust and move away from this market, exploring new markets in high-income countries. A similar strategy is followed to deal with the endogeneity of the ratio of exports to sales. More concretely, the share of exports to Brazil in total sales is used as an instrument for the share of total exports in total sales. Two arguments support this instrument. One claim is that firms with a larger share of exports to Brazil in total sales had smaller shares of exports in sales because part of the effect of the devaluation was to make them retrench into local markets. Another argument is that, conditional on the share of exports to Brazil in total exports, firms with a higher ratio of exports to Brazil in total sales had a lower base from which to divert exports to high-income countries. As before, exposure is measured non- parametrically with a dummy for 1999 and another for 2000, using the Brazilian exchange rates. First-stage regressions reveal that the instruments work well (see Porto 2007 for more details). They have substantial explanatory power and are statistically significant in all the regressions. Furthermore, the results imply that, in fact, following the Brazilian devaluation, firms that were more exposed to it switched to high-income destination countries but faced lower export-to-sales ratios, as expected. Also, the correlations per- sist even after including year effects to account for the macroeconomic impacts of the devaluation (and other time effects that affected all firms in the same fashion). Now this chapter turns to the main results: the instrumental variable coefficients of exports on wages. These results are reported in table 3.9, which lists the two potential endogenous variables and has nine columns. These columns correspond to the three models: only exports-to-sales ratio, only share of exports to high-income countries, and both export to sales and HIE together. We work with different specifications: without year effects and nonparametric dummy instruments (columns 1 to 3), with year effects and nonparametric dummy instruments (columns 4 to 6), and with year effects and parametric instruments using exchange rates 76 Table 3.9 Exports, Export Destination, and Wages: Wage Regression with Instrumental Variables dependent variable: log average wage Variable (1) (2) (3) (4) (5) (6) (7) (8) (9) Exports to sales –0.979 –0.343 –0.735 –0.259 –0.543 –0.075 (0.638) (0.478) (0.594) (0.513) (0.546) (0.495) High-income 0.365*** 0.357*** 0.317*** 0.305*** 0.296*** 0.293*** exports (0.106) (0.111) (0.108) (0.107) (0.107) (0.110) Log sales 0.054*** 0.054*** 0.055*** 0.064*** 0.057*** 0.058*** 0.063*** 0.057*** 0.058*** (0.019) (0.018) (0.018) (0.020) (0.020) (0.020) (0.021) (0.020) (0.020) Number of firms 901 901 901 901 901 901 901 901 901 Number of observations 2,544 2,544 2,544 2,544 2,544 2,544 2,544 2,544 2,544 Source: Authors’ calculations. Note: All regressions include firm fixed effects. Robust standard errors are in parentheses. Year effects are included in columns 4–9, but not in columns 1–3. The instruments are dummies for columns 1–6 and the exchange rate for columns 7–9. *** p < .01. Exports, Wages, and Skills: Implications for CAFTA 77 (columns 7 to 9). The major conclusion of this work is that, although exporting to high-income countries improves wages, the ratio of exports to sales does not affect them. Thus exporting per se is not a significant channel toward higher wages, but exporting to high-income countries is. That is, what appears to matter is the composition of exports. The magnitudes are important, too: firms with average shares of exports to high-income countries pay wages between 8.79 (29.3 × 0.30) and 9.51 (31.7 × 0.30) higher than firms with no exports to high-income countries. It is difficult not to overemphasize this fact: the results are very robust and survive the inclusion of firm fixed effects (using the panel data) and the use of instrumental variables. Conclusions and Policy Implications Export firms are more productive and pay higher wages than other firms. There is also evidence that exporters pay a higher skill premium at the industry level. These exporter premiums have two explanations that are not mutually exclusive. Either more productive firms self-select into export activities, because only highly productive firms could face the entry costs associated with selling in foreign markets, or learning by exporting occurs, in which case participation in the export market allows firms to become more productive. The existing empirical evidence strongly supports the self-selection hypothesis; learning by exporting is observed in only some countries. More than 80 percent of the studies surveyed for this chapter find evi- dence that exporting firms were more productive than other firms before entering the export market, whereas only half the studies find that the productivity of exporting firms grew faster than the productivity of non- exporting firms after the former entered the export market. Learning by exporting is more likely to be observed in small coun- tries, suggesting that the ability to exploit economies of scale in foreign markets may be part of the explanation behind learning by exporting. In contrast, trade openness, the investment climate, and the level of devel- opment do not explain why learning by exporting occurs, suggesting that they are not important determinants of learning by exporting. Wage export premiums (that is, the fact that exporters pay higher wages than nonexporters) are more likely in small economies, whereas trade openness, the investment climate, and the level of development do not seem to matter. Recent research shows that exporting is also associ- ated with an industry skill premium. In the presence of some sort of labor 78 Brambilla, Castro, and Porto immobility, so that skilled workers earn a premium that can vary by industry, this premium is larger in those industries that are more oriented toward export markets in general. This evidence is very strong for Latin American countries. The existing literature on exports and wages stops at identifying whether a wage export premium is present and does not address the issue of causality. More recent studies, including Verhoogen (2008), Porto (2007), and Brambilla, Lederman, and Porto (2010), provide stronger evi- dence of a causal effect of exports on wages. Furthermore, the evidence in the last paper also suggests that export destination matters and that, in particular, exporting to high-income countries matters. This paper finds, in a panel of manufacturing firms in Argentina, that exporting to high- income countries is associated with higher wages and higher skill utiliza- tion. This may be due to two major reasons. First, the act of exporting (rather than the act of producing) may require skills. Second, accessing high-income destinations may also require quality upgrades that are, in turn, skill intensive. The results are consistent with both a quality and a profit-sharing story. High-income countries tend to demand higher-quality goods, a sit- uation that allows firms to pay higher wages. However, Porto (2007) can- not rule out the existence of profit-shifting mechanisms, whereby firms that export to high-income countries tend to share part of the excess profits with their workers. But what determines the rapid productivity growth observed by firms before they enter the market? Little is known about this growth, and some authors, such as Alvarez and López (2005), argue that an important part of the explanation—at least among Chilean manufacturing plants— is conscious self-selection. Firms become more productive with the objec- tive of becoming exporters. So export activity may be the cause of the jump in productivity before firms enter the export market. An important point to keep in mind is that, regardless of whether exports cause firm-level productivity gains, the fact that exporters are more productive and larger (perhaps exclusively because of self-selection) suggests that, as countries increase their export orientation, larger and more productive firms will produce a larger share of national output. This reallocation of resources from less productive and smaller firms to more productive and larger firms in itself is a source of aggregate GDP growth. Depending on which is the most important force, however, the policy implications can be very different. On the one hand, offshore export Exports, Wages, and Skills: Implications for CAFTA 79 promotion activities, which are quite common in Latin America, are often justified on the basis of learning by exporting. On the other hand, the self-selection explanation suggests that policy makers should focus their efforts on the internal determinants of productivity growth. Evidence elsewhere indirectly suggests that the latter is probably more important in developing countries. In a recent paper, Lederman, Olarreaga, and Payton (2006) show, with the help of a recent survey on export promotion activities, that in developing countries the returns to onshore export promotion activities, such as technical assistance and training for (large) domestic firms on how to enter foreign markets, are much larger than the returns to offshore export promotion activities, such as country image, fair participation, and other marketing activities abroad, including foreign offices. They also find that, to maximize the effect on aggregate exports, export promotion activities should focus on large domestic firms that are not yet exporting rather than on established exporters. These findings are consistent with the idea that self-selection plays an important role in explaining export premiums. Notes 1. The expected effects from learning by exporting could occur either at the time of entry into exporting (a one-time effect) or every year after entry (a continuous effect). 2. Hallward-Driemeier, Iarossi, and Sokoloff (2002) show that part of what self- selection may be capturing is the idea that exporters invest in retooling for foreign markets in advance of entering the market and that therefore self- selection may be associated with export activity. 3. These two studies (Breau and Rigby 2006; Schank, Schnabel, and Wagner 2007) match employer-employee data, which allows them to control for workers’ characteristics, such as education, age, and experience, that are unob- servable when working with firm (employer) data only. It is probably too early to conclude anything, but a large share of what is captured as an exporter skill premium seems to result from the “superior” characteristics of workers in the export sector. 4. In Verhoogen (2008), exporting requires quality upgrades. References Alvarez, R., and R. López. 2005. “Exporting and Performance: Evidence from Chilean Plants.” Canadian Journal of Economics 38 (4): 1384–400. 80 Brambilla, Castro, and Porto Aw, B., S. Chung, and M. Roberts. 2000. “Productivity and Turnover in the Export Market: Micro-level Evidence from the Republic of Korea and Taiwan (China).” World Bank Economic Review 14 (1): 65–90. Bernard, A., B. Jensen, S. Redding, and P. Schott. 2007. “Firms in International Trade.” Journal of Economic Perspectives 21 (3, Summer): 105–30. Brambilla, I., R. Carneiro, D. Lederman, and G. Porto. 2010. “Skills, Exports, and the Wages of Five Million Latin American Workers.’’ Policy Research Working Paper 5246, World Bank, Washington, DC. Brambilla, I., D. Lederman, and G. Porto. 2010. “Exports, Export Destinations, and Skills.” World Bank, Washington, DC. Breau, S., and D. Rigby. 2006. “Is There Really an Export Wage Premium? Case Study of Los Angeles Using Matched Employee-Employer Data.” Working Paper 06-06, U.S. Census Bureau, Washington, DC. Casacuberta, C., N. Gandelman, M. Olarreaga, G. Porto, and E. Rubiano. 2007. “Exporter Premiums.” In Latin American and Caribbean Regional Study on Microdeterminants of Growth, ch. 7. Washington, DC: World Bank. Caves, D. W., L. R. Christensen, and W. E. Diewert. 1982. “The Economic Theory of Index Numbers and the Measurement of Input, Output, and Productivity.” Econometrica 50 (6): 1393–1414. Dickens, W. T., L. F. Katz, and K. Lang. 1986. “Are Efficiency Wages Efficient?” NBER Working Paper Series w1935, National Bureau of Economic Research, Cambridge, MA. Dickens, W., and K. Lang. 1988. “Labor Market Segmentation and the Union Wage Premium.” Review of Economics and Statistics 70 (3): 527–30. Feenstra, R., and G. Hanson. 1997. “Foreign Direct Investment and Relative Wages: Evidence from Mexico’s Maquiladoras.” Journal of International Economics 42 (3–4): 371–93. Galiani, S., and G. Porto. 2010. “Trends in Tariff Reforms and Trends in the Structure of Wages.” Review of Economics and Statistics 92 (3, August): 482–94. Gibbons, R., and L. F. Katz. 1992. “Does Unmeasured Ability Explain Inter- Industry Wage Differences?” Review of Economic Studies 59 (3): 515–35. Hallward-Driemeir, M., G. Iarossi, and K. Sokoloff. 2002. “Exports and Manufacturing Productivity in East Asia: A Comparative Analysis with Firm- Level Data.” NBER Working Paper 8894, National Bureau of Economic Research, Washington, DC. Lederman, D., M. Olarreaga, and L. Payton. 2006. “Export Promotion Agencies: What Works and What Doesn’t.” Policy Research Working Paper 4044, World Bank, Washington, DC. Porto, G. 2007. “From Exports to Wages: Evidence from Panel Data in Argentina.” Background paper for Latin American and Caribbean Regional Study on Microdeterminants of Growth, World Bank, Washington, DC. Exports, Wages, and Skills: Implications for CAFTA 81 Schank, T., C. Schnabel, and J. Wagner. 2007. “Do Exporters Really Pay Higher Wages? First Evidence from German Linked Employer-Employee Data.” Journal of International Economics 72 (1): 52–74. Verhoogen, E. 2008. “Trade, Quality Upgrading, and Wage Inequality in the Mexican Manufacturing Sector.” Quarterly Journal of Economics 123 (2): 489–530. Westphal, L., Y. Rhee, and G. Pursell. 1984. “Sources of Technological Capability in South Korea.” In Technological Capability in the Third World, ed. M. Fransman and K. King, 279–300. London: Macmillan. Yeaple, S. 2005. “A Simple Model of Firm Heterogeneity, International Trade, and Wages.” Journal of International Economics 65 (1): 1–20 CHAPTER 4 Trade and Economic Growth: Evidence on the Role of Complementarities for the DR-CAFTA Countries César Calderón and Virginia Poggio One of the salient features of the world economy has been the important surge in trade and financial globalization in the past two decades. Multiple free trade agreements and regional integration agreements are being celebrated—with more than 400 regional trade agreements in force by December 2008 according to the World Trade Organization. In addition, world trade has grown at least twice as fast as world output over the past two decades, thus deepening economic integration. Theoretically, it has long been argued in the literature that trade stim- ulates long-term growth and that it can do so through multiple channels. International trade would allow countries to specialize in areas where they possess comparative advantage, expand potential markets, allow firms to exploit economies of scale, enable the diffusion of technological innovation and frontier managerial practices, and reduce incentives for The authors would like to thank their peer reviewers, J. Humberto López and Rashmi Shankar, for invaluable comments. 83 84 Calderón and Poggio firms to conduct rent-seeking activities through higher market competi- tion. Empirically, earlier works find evidence in support of the growth- enhancing effects of trade. However, Rodríguez and Rodrik (2001) suggest that most of the evidence is not robust due to measurement issues of trade openness and trade policy as well as econometric prob- lems (that is, endogeneity of trade measures and collinearity of trade and institutions). Rodrik (2005) also argues that policies toward trade open- ness may not render the same results for all countries since there is no unique mapping from economic principles to economic packages. Most of these criticisms have been tackled in recent empirical efforts by devel- oping new identification strategies (Frankel and Romer 1999), develop- ing new trade indicators (Wacziarg 2001), examining the trade-growth correlation around episodes of policy changes (Wacziarg and Welch 2008), and addressing the issue of mapping from principles to policies by assessing the role of complementarities between trade and other structural reforms in stimulating growth (Calderón, Loayza, and Schmidt-Hebbel 2006; Calderón and Fuentes 2009; Chang, Kaltani, and Loayza 2009). The goal of this chapter is to assess the growth effects of trade among Dominican Republic–Central America Free Trade Agreement (DR-CAFTA) countries and, more specifically, to evaluate the structural areas that might become a constraint to reaping the benefits from growth. In this context, the chapter argues that policy complementarities are a cornerstone of growth. Pro-growth policies should mutually reinforce— for example, trade openness will have positive and substantial effects on growth in countries with higher levels of human capital. At the same time, policy complementarities may also impose severe restrictions on the design of an optimal growth strategy—especially among countries with less favorable initial conditions. To accomplish this task, we gather annual information for a sample of 136 countries over the period 1960–2009 and construct a panel database of five-year nonoverlapping observations. We run our cross-country regressions using econometric techniques suitable for dynamic panel data models that account not only for the presence of unobserved compo- nents, but also for the likely endogeneity or reverse causality of the growth determinants. Our results find that trade has indeed promoted growth, and our result is robust to the specification and technique used. However, the growth benefits of rising trade openness are conditional on the level of progress in structural areas such as education, innovation, infrastructure, institutions, regulatory framework, financial development, Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 85 and international financial integration. Indeed, we find that lack of progress in these areas can restrict the potential benefits of trade. We discuss the implications of our regression analysis for the DR- CAFTA nexus, putting emphasis on the impact of trade openness on growth per capita and identifying the structural areas that may represent a constraint to growth. To do so, we calculate the impact of trade on growth among DR-CAFTA countries over the past 15 years and the potential growth gains from raising trade openness to the levels of a benchmark country or region (in our case, the East Asian tigers, EAP-7). In both cases, we find that there is room for trade to stimulate growth, but special attention should be placed on reforms in structural areas that are complementary to the trade reform efforts launched by DR-CAFTA countries, mainly in the areas of education, institutional quality, and infrastructure. This chapter is divided into five sections. It begins with a brief review of the literature on trade and growth with some emphasis on the chan- nels of transmission, the problems in the empirical literature, and the complementarities between trade and other structural factors in driving growth. It then describes the data and outlines the econometric method- ology used to estimate our cross-country growth regressions. This is fol- lowed by a presentation of the empirical evidence on trade and growth and a test of whether the impact of the former on the latter is enhanced by advances in structural areas such as education, domestic financial mar- ket development, institutional quality, infrastructure, financial integra- tion, innovation, and regulatory framework. We also discuss the economic implications of our statistical analysis for DR-CAFTA countries. A final section concludes. Literature Review The classical paradigm of international trade argues that trade promotes growth by increasing the relative price of the good that is intensive in the relatively abundant factor (see, for example, Deardorff 1974). The stan- dard theory predicts an effect of trade openness on the long-run level rather than on the long-run growth of GDP (Lucas 1988; Young 1991). The new trade literature, in contrast, argues that long-term growth from trade can be channeled through more intense research and development (R&D) activity (see Romer 1990; Grossman and Helpman 1991). In this context, trade promotes long-term growth by raising the availability of resources for R&D and, thus, increasing the availability of specialized 86 Calderón and Poggio inputs and the size of the market, among other things. More broadly speaking, the theoretical literature is ambiguous about the impact of trade on long-run growth. A strand of the literature suggests that the growth effects are positive when trade specializes in increasing returns- to-scale activities (Grossman and Helpman 1991; Young 1991). Others suggest that the effect is either negligible or negative whenever there are market or institutional imperfections (Rodríguez and Rodrik 2001), underutilization of human or capital resources, focus on extrac- tive activities (Sachs and Warner 1995), or specialization away from technologically intensive sectors with increasing returns to scale (Matsuyama 1992). The literature suggests that trade may affect economic growth through different channels. First, trade openness may increase a country’s market size and, thus, may provide innovators with new business opportunities and allow domestic firms to take advantage of scale economies (Alesina, Spolaore, and Wacziarg 2005). Second, trade can enhance technological diffusion and transmit know-how and managerial practices thanks to stronger interactions with foreign firms and markets (Coe and Helpman 1995; Sachs and Warner 1995; Coe, Helpman, and Hoffmaister 1997). Relatedly, Lewer and van den Berg (2003) find that the strength of trade as an engine of growth depends on the composition of trade—that is, countries that import mostly capital goods and export consumer goods tend to grow faster than those that export capital goods. Third, trade may enhance product market competition, thus reducing the anticompetitive practices of domestic firms and leading to higher specialization due to exploitation of the comparative advantages of domestic firms (Trefler 2004; Aghion and others 2008). Finally, the literature on the effects of trade liberalization can also be classified into two strands: (a) the long-run productivity benefits of free trade policies (for example, Tybout, de Melo, and Corbo 1991; Levinsohn 1993; Krishna and Mitra 1998; Head and Ries 1999; Pavcnik 2002) and (b) the impact of freer trade on short-run worker displacement and earnings (for example, Gaston and Trefler 1995; Levinsohn 1999; Krishna, Mitra, and Chinoy 2001). The empirical literature on trade and growth typically argues that growth is positively correlated with higher trade volumes, even after accounting for a variety of growth determinants. Dollar (1992), Edwards (1992), Sachs and Warner (1995), Ades and Glaeser (1999), and Alesina, Spolaore, and Wacziarg (2000) are examples of this sort. However, Rodríguez and Rodrik (2000) argue that most of these findings are less robust than claimed due to (a) difficulties in measuring openness and Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 87 especially trade policy, and (b) the statistical sensitivity of the specifica- tions and other econometric difficulties—among them, collinearity of protectionist policies with other bad policies and likely endogeneity of trade openness. These authors argue that the literature focuses on the growth effects of trade volumes rather than trade policy and that the for- mer is plagued by severe endogeneity problems. Moreover, indicators of trade openness are deemed as controversial proxies for trade barriers. Finally, a critical assessment is issued on the inadequacies of the liter- ature for addressing endogeneity as well as controlling for other structural factors—notably institutions. Frankel and Romer (1999) tackle the issue of endogeneity by using the gravity model to instrument for trade openness. Here, trade flows between countries depend on geographic and cultural characteristics of trading partners (say, distance, remoteness, common border, landlocked, among others) as well as their size. Using gravitational variables, they attempt to establish a causal link between trade and growth and find that the impact of the former on the latter is positive and statistically signifi- cant. Wacziarg and Welch (2008), in contrast, study the contingent rela- tionship between trade policy and growth by examining the evolution of growth, investment, and openness around episodes of trade liberalization. Growth rates in countries that liberalized their trade regimes were, on average, 1.5 percentage points higher after than before liberalization, whereas investment rates were almost 2 percentage points higher. Finally, the ratio of trade to GDP rose 5 percentage points due to liberalization. In sum, trade and growth have a positive co-movement, with investment being a channel of transmission. Although it is suggested that, on average, trade openness appears to be beneficial to economic growth, its effect may vary considerably across countries. It is argued that the growth benefits from open trade may kick in after the country surpasses a “minimum critical threshold” associated with the level of development (Helleiner 1986) or the structure of trade (Kohli and Singh 1989). Recently, Chang, Kaltani, and Loayza (2009) find that, although trade stimulates growth, this effect can be enhanced by complementary reforms undertaken in the economy—especially in the areas of education, financial development, infrastructure, and regu- latory framework. Finally, Bolaky and Freund (2004) find that trade openness is effective in promoting an expansion of income in countries that are not excessively regulated—that is, resource allocation toward the most productive sectors and companies is prevented in highly regu- lated countries. 88 Calderón and Poggio The Data We have collected a panel data set of 136 countries organized in five-year nonoverlapping observations over the period 1970–2010, with each country having at most eight observations. Given that the availability of data is different across variables, we have an effective sample of 99 coun- tries with at least four consecutive observations for all variables involved in our analysis. This subsection describes the construction and sources of the data used in our empirical analysis. Our dependent variable is the average annual growth rate in real GDP per capita within the five-year period, which is computed as the simple average of log differences in real GDP per capita over the five-year period. Real GDP per capita is expressed in 2005 international dollars (adjusted by purchasing power parity, PPP) from Heston, Summers, and Aten (2009). Our set of control variables includes the (log) level of real GDP per capita at the beginning of the five-year period to test for the existence of transitional convergence. Our set of long-run growth determinants fol- lows Loayza, Fajnzylber, and Calderón (2005): human capital, financial depth, institutional quality, lack of price stability, infrastructure, financial openness, and our variable of interest, trade openness. Human capital is approximated by the initial gross rate of secondary schooling (in logs), and the data are obtained from Barro and Lee (2001).1 Financial development is measured by the ratio of domestic credit to the private sector to GDP, and the data are collected from Beck, Demirgüç-Kunt, and Levine (2000) and from Beck and Demirgüç-Kunt (2009) and are updated using data from the International Monetary Fund’s International Financial Statistics and the World Bank’s World Development Indicators. For the sake of robustness, we use other proxies of financial development: domestic credit provided by domestic money banks and liquid liabilities of the financial sector. Both variables are expressed as a percentage of GDP and in logs. Institutional quality com- prises different dimensions such as absence of corruption, rule of law, enforcement of contracts, quality of the bureaucracy, and democratic accountability, among others. We use the International Country Risk Guide (ICRG) index of political risk as our indicators of institutional quality, and the data are published in the ICRG by the Political Risk Services Group. The lack of price stability is approximated by the average consumer price index inflation rate. This variable typically reflects the quality of monetary and fiscal policies and is directly related to other indi- cators of poor macroeconomic management. The data on the inflation Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 89 rate are gathered from the International Monetary Fund’s International Financial Statistics. Infrastructure is a multidimensional concept. To account for this, we use principal component analysis to build synthetic indexes summarizing information on the quantity of different types of infrastructure assets (see Calderón and Servén 2004). These synthetic indexes combine informa- tion on three core infrastructure sectors—telecommunications, power, and roads—and help address the problem of high collinearity among their individual indicators.2 We denote IK the synthetic quantity indexes that result from this procedure. The indexes can be expressed as linear com- binations of the underlying sector-specific indicators, and hence their use in a regression context is equivalent to imposing linear restrictions on the coefficients of the individual infrastructure indicators. We define the syn- thetic infrastructure quantity index IK1 as the first principal component of three variables: total telephone lines (fixed and mobile) per 1,000 peo- ple (Z1/L), electric power installed capacity expressed in megawatts per 1,000 people (Z2/L), and the length of the road network in kilometers per 1,000 people (Z3/L). Each of these variables is expressed in logs and stan- dardized by subtracting its mean and dividing by its standard deviation. All three infrastructure stocks enter the first principal component with roughly similar weights: ⎛Z ⎞ ⎛Z ⎞ ⎛Z ⎞ IK1 = 0.603 * ln ⎜ 1 ⎟ + 0.613 * ln ⎜ 2 ⎟ + 0.510 * ln ⎜ 3 ⎟ . (4.1) ⎝ L⎠ ⎝ L ⎠ ⎝ A⎠ The index accounts for almost 80 percent of the overall variance in the three underlying indicators. As a robustness check, we compute an alter- native index IK2, which uses main telephone lines instead of the com- bined main lines and mobile phones employed in the first index.3 Financial openness is approximated by the data on holdings of foreign assets and liabilities from Lane and Milesi-Ferretti (2001, 2007). Specifically, we use summary measures of financial openness FOit = (FAit + FLit) / GDPit and FO(L)it = FLit / GDPit, where FA and FL refer to the stock of foreign assets and liabilities—expressed as a ratio to GDP. Note that FA and FL include the stock of assets and liabilities in foreign direct invest- ment, portfolio equity, financial derivatives, and debt (portfolio debt, bank debt, and trade-related lending). Our variable of interest, trade openness, is measured as the ratio of real exports and imports to real GDP (all these magnitudes are expressed in 90 Calderón and Poggio local currency at constant prices), and the data are collected from the World Bank’s World Development Indicators. We also use an alternative measure of openness that adjusts the volume of trade over GDP for the size (area and population) of the country and for whether the country is landlocked or an oil exporter.4 Loayza, Fajnzylber, and Calderón (2005) argue that this structure-adjusted volume of trade may be preferable to the unadjusted ratio given that the econometric analysis is based on cross-country comparisons. Unadjusted measures of trade volume may unfairly attribute to trade policy what is merely the result of structural country characteristics—for example, smaller countries are more dependent on foreign trade than larger countries, oil exporters may have large trade volumes and also impose high import tariffs, and landlocked countries tend to trade less than other countries due to higher transport and trading costs. Finally, we describe two sources of data for which we lack extensive time series, but have good cross-country coverage: R&D and economic regulations. We argue that positive complementarities between trade and innovation can trigger higher and sustained growth. Our proxies for inno- vation are R&D spending as a percentage of GDP, R&D scientists (per 1 million people), and R&D technicians (per 1 million people). We sum- marize these three measures in an aggregate R&D index. In addition, we use the share of high-tech exports to manufacturing exports as a proxy for innovation. Econometric Methodology Having an effective cross-country and time-series data set for 99 countries over the period 1970–2010 requires us to use an estimation method that accounts for the dynamic specification of our growth equation, for unob- served time- and country-specific effects, and for likely endogeneity or reverse causality among the explanatory variables. In short, we use the generalized method of moments (GMM) for dynamic panel data models developed by Arellano and Bond (1991), Arellano and Bover (1995), and Blundell and Bond (1998). We regress the growth in real output per capita on a standard set of growth determinants that includes our variable of interest, trade openness. Our basic set of control variables comprises information on the level of human capital, domestic financial depth, institutional quality, lack of price stability, financial openness, and infrastructure stocks. In addition to our baseline regression, we explore the role of complementarities between Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 91 trade and structural factors in driving growth. In short, our dynamic regres- sion equation can be specified as follows: yit – yit–1 = ayit–1 + f'Kit + g'Zit + mt + hi + eit (4.2) = ayit–1 + b'Xit + mt + hi + eit, where y denotes the real GDP per worker (in logs), K is a set of standard growth or inequality determinants, and Z is our variable of interest: trade openness. The terms mt and hi, respectively, denote an unobserved com- mon factor affecting all countries and a country effect capturing unob- served country characteristics. The second equality follows from defining Xit = (K'it, Z'it)’ and b = (f',γ ')’. The econometric challenges posed by our growth equation are tackled as follows. First, we control for unobserved time effects by including period-specific dummies in our regressions, while accounting for unob- served country effects by differencing and instrumentation. Second, we address joint endogeneity by instrumentation. More specifically, the assumption of strong exogeneity of the explanatory variables is lifted by allowing them to be correlated with current and previous realizations of the error term, e. Since no obviously exogenous instruments are available, the methodology relies primarily on internal instruments—that is, suitable lags of the explanatory variables (Arellano and Bond 1991). Rather than using internal instruments for our variable of interest, we use external instruments. We are concerned that future shocks to growth may lead to an expansion in foreign trade, thus invalidating our moment conditions. To do so, we follow Loayza, Fajnzylber, and Calderón (2005) and Chang, Kaltani, and Loayza (2009) and consider measures of size and geography as instruments of trade openness—that is, (actual and lagged values of) population, surface area of the country, and dummies for oil-exporting countries and landlocked countries. The consistency of the GMM-IV (instrumental variable) estimator relies on the validity of these moment conditions.5 Empirical Assessment We first present the basic panel correlations and then the baseline regression. Simple panel correlations between trade openness and growth are calculated for a sample of 136 countries with five-year nonoverlapping observations spanning the period 1960–2010. The positive correlation, 92 Calderón and Poggio +0.08, is significant and larger in countries with higher levels of income per capita, human capital, infrastructure, and financial openness.6 The cor- relation between growth and trade openness is larger in the 2000s, 0.21 compared with 0.08 for the full sample of countries. Zooming in on the DR-CAFTA countries, we observe that most of these countries are close to or below the medians of both trade openness and growth, and they have a flatter relationship than the rest of the sample (see figure 4.1). Plotting the trade-growth nexus in countries with high versus low lev- els of structural policies (say, human capital, financial development, insti- tutions, financial openness, infrastructure, and regulations) shows that the trade-growth correlation is stronger in countries with more educated peo- ple, stronger institutions, an improved infrastructure network, and more flexible regulations (see figure 4.2).7 Table 4.1 reports the coefficient estimates for our baseline regres- sions using different estimation techniques. The coefficient estimate of our variable of interest, trade openness, is positive and significant (at least at the 10 percent level) regardless of the technique used. In col- umn 1, we run a pooled ordinary least squares (OLS) regression, while Figure 4.1 Correlation between Growth and Trade Openness in DR-CAFTA Countries 15 10 GDP growth (%) 5 DOM CRI HND GTM SLV NIC 0 –5 3 4 5 6 trade openness CR = Costa Rica DOM = Dominican Republic GTM = Guatemala HND = Honduras NIC = Nicaragua SLV = El Salvador Source: Authors’ calculations. Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 93 Figure 4.2 Correlations between Trade Openness and Growth, by Select Indicators of Economic Development a. Correlation to levels of human capital b. Correlation to levels of financial development 15 15 10 10 GDP growth (%) GDP growth (%) 5 5 0 0 –5 –5 3 4 5 6 3 4 5 6 trade openness trade openness c. Correlation to levels of institutions d. Correlation to levels of financial openness 15 15 10 10 GDP growth (%) GDP growth (%) 5 5 0 0 –5 –5 3 4 5 6 3 4 5 6 trade openness trade openness e. Correlation to levels of infrastructure f. Correlation to levels of economic regulation 15 15 10 10 GDP growth (%) GDP growth (%) 5 5 0 0 –5 –5 3 4 5 6 3 4 5 6 trade openness trade openness Source: Authors’ calculations. column 2 controls for time dummies, and column 3 controls only for country dummies. We apply the Arellano and Bover (1995) GMM-IV difference estimator in column 4, thus controlling for unobserved com- ponents and endogeneity and instrumenting the differences in the explanatory variables using their lagged levels. However, the GMM-IV difference estimator may face the problem of weak instruments if the explanatory levels are highly persistent. Hence, columns 5 and 6 esti- mate our baseline regression using the GMM-IV system estimator (Arellano and Bover 1995; Blundell and Bond 1998). While the estimation 94 Table 4.1 Trade and Growth: Baseline Regression under Different Estimation Techniques dependent variable: growth in real GDP per capita (annual average, %) Pooled OLS time Within GMM-IV GMM-IV GMM-IV OLS dummies group difference system system Variable (1) (2) (3) (4) (5) (6) Variable of interest Trade openness (exports and imports as 0.5756* 0.4860* 1.7448** 5.8530** 0.3614** 0.6245** % of GDP, log) (0.319) (0.322) (0.757) (1.053) (0.134) (0.143) Control variable Transitional convergence (initial GDP per capita, log) –1.7788** –1.9137** –6.4594** –7.4420** –2.1768** –2.1263** (0.394) (0.437) (0.852) (0.760) (0.343) (0.218) Human capital (gross secondary enrollment rate, log) 0.7783** 0.9918** –1.3719** 1.0505* 1.8700** 1.5336** (0.350) (0.327) (0.575) (0.597) (0.285) (0.207) Financial depth (credit to private sector, % GDP, log) 0.2492 0.1963 0.4415 0.6054** 0.2939* 0.6229** (0.299) (0.305) (0.378) (0.281) (0.158) (0.148) Institutional quality (ICRG political risk index, log) 0.6914 0.9657 0.4272 0.6872 1.0118** 1.5695** (0.725) (0.713) (0.854) (0.732) (0.345) (0.418) Lack of price stability (CPI inflation rate, log) –2.5916** –2.4343** –2.9280** –3.4301** –3.6547** –3.7073** (0.610) (0.642) (0.526) (0.823) (0.134) (0.184) Infrastructure stock (principal component)a 0.6284** 0.5882** 1.2285** –0.2651 0.4335** 0.2987** (0.169) (0.187) (0.244) (0.293) (0.139) (0.146) Financial openness (foreign assets and liabilities, –0.6767** –0.4241 –0.2343 –1.2307** –0.3706** –0.5876** % GDP, log) (0.281) (0.303) (0.279) (0.380) (0.123) (0.129) Time dummy Dummy: 1976–80 period .. 0.0138 .. .. –0.1739 –0.2339 Dummy: 1981–85 period .. –2.4998** .. –1.4681** –2.6141** –2.5612** Dummy: 1986–2000 period .. –1.2370** .. 0.7218** –1.4532** –1.3186** Dummy: 1991–95 period .. –1.6349** .. –0.5876* –1.8917** –1.6285** Dummy: 1996–2000 period .. –1.7096** .. –0.5278 –1.9168** –1.5804** Dummy: 2001–05 period .. –1.5078** .. 0.3271 –1.9714** –1.5529** Dummy: 2006–09 period .. –0.6260 .. 1.0000** –1.0186** –0.5994* Number of countries 99 99 99 99 99 99 Number of observations 646 646 547 547 646 646 Country effects No No Diff Diff Diff Diff Time effects No Yes No Yes Yes Yes Instrumentsb No No No Internal Internal External Specification tests (p-value) Sargan test (overidentifying restrictions) .. .. .. (0.072) (0.310) (0.256) Second-order serial correlation (0.082) (0.044) (0.273) (0.181) (0.182) (0.211) Source: Authors’ calculations. Note: Numbers in parentheses are robust standard errors. Regression includes constant. .. = negligible. a. The aggregate stock of infrastructure is computed as the first principal component of (a) main telephone lines and mobiles, (b) electric power installed capacity (in megawatts), and (c) length of the road network (in kilometers). All these physical indicators of infrastructure are expressed in their corresponding units per 1,000 people. b. The set of “internal instruments” corresponds to lagged levels and differences of the corresponding explanatory variables in our regression analysis. In contrast, “external instruments” include variables that instrument for trade openness such as lagged population, surface area of the country, dummy for landlocked countries, and oil-exporting countries. * p < .10, ** p < .05. 95 96 Calderón and Poggio in column 5 uses internal instruments, the estimation in column 6 uses external instruments to account for the likely endogeneity of trade openness. As pointed out earlier, those external instruments are the (actual and lagged levels of) population (in logs), the surface area of the country (in logs), and dummies for landlocked and oil-exporting coun- tries. Our preferred estimation is the one reported in column 6, and we discuss these results for our baseline regressions. We find a negative and significant coefficient for the initial (log level of) GDP per capita, thus providing evidence of conditional convergence. Growth is enhanced by a faster accumulation of human capital (as prox- ied by rising gross rates of secondary schooling), deeper domestic financial markets (as measured by higher ratios of domestic credit to the private sector to GDP), and better institutions (as approximated by higher levels of the ICRG political risk index). Lack of price stability, measured by higher rates of consumer price inflation, hinders growth. A faster accumu- lation of infrastructure stocks (as proxied by deeper penetration of telecommunications, larger installed capacity for electricity, and a longer road network) promotes long-term growth. Financial openness, however, seems to have an adverse effect on growth rate. Our variable of interest, trade openness, has a positive and significant coefficient. This result implies that long-run growth is enhanced by a more outward orientation in goods markets. Our coefficient estimates suggest that doubling trade openness would raise the growth rate by 43 basis points a year—that is, more than 4 percentage points over a decade. Finally, the coefficient estimate of trade openness may vary according to the extent of the outward orientation of the country and over time. Some extensions of the sensitivity of growth to trade are explored in Calderón and Poggio (2010). First, they investigate whether the effect of trade openness on growth depends on the extent of integration with world goods markets. That is, they test whether the growth effects of openness increase as the extent of integration increases. Their results show that trade openness exerts a positive impact on growth despite the extent of outward orientation of the country; however, it is statistically significant only for countries with deeper trade integration (that is, coun- tries with trade openness beyond the sixty-seventh percentile of the world distribution). Second, they test whether the impact of trade open- ness on growth has changed over time. The authors find that, while the coefficient estimate for the 1980s is negative and significant in most cases, it is positive and significant for the 2000s.8 Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 97 Trade and Growth: The Role of Complementarities Since the growth elasticity of trade appears to vary over time and across countries, we proceed to estimate the growth regression in equation 4.3: yit – yit–1 = a yit–1 + f'Kit + git'Zit + mt + hi + eit, (4.3) where the trade openness (TO) coefficient, git, is allowed to vary across countries and time. Thus, the existence of complementarities between trade and structural factors (F) is modeled as git = g0 + g1Fit, where the coef- ficient of trade openness depends directly on TO as well as its interaction with structural factors. In this section we consider the complementarities between trade openness and the following factors: human capital, finan- cial development, institutions, infrastructure, financial openness, R&D, and certain aspects of the regulatory framework. Complementarities between Trade and Structural Factors Table 4.2 presents regression estimates that test for the significance of complementarities between trade openness and human capital (regres- sion 2), trade openness and financial development (regressions 3 to 5), and trade openness and institutional quality (regression 6). The impact of trade openness on growth now depends on the level of the specific struc- tural factor in each country at a determined period of time. Regression 1 in table 4.2 includes the interaction between trade open- ness and the level of income per capita in our baseline regression. While the TO coefficient (g0) is negative and significant, its interaction with income per capita is positive and significant (g1). This finding suggests that opening up the current account would require a minimum develop- ment threshold to generate positive growth effects. Economically speak- ing, our regression suggests that a 1 standard deviation increase in trade openness—that is, the ratio increases approximately 75 percent—would lead to a decline in the growth rate of 30 basis points a year for countries with lower levels of income per capita (approximately US$2,500 at 2005 PPP prices—Mongolia in 2005), while it would raise growth of output per capita by almost 1 percentage point (more precisely, 97 basis points) in countries with higher levels of income per capita (US$22,000— Republic of Korea in 2005). The first panel of figure 4.3 reports the growth effects of rising trade openness for different levels of income per capita—that is, selected percentiles of the distribution (tenth, twenty- fifth, thirty-third, the median, sixty-seventh, seventy-fifth, ninetieth per- centiles), regions (CAFTA, Latin America and the Caribbean excluding 98 Table 4.2 Trade and Growth: Interaction with Structural Factors and Policies dependent variable: growth in real GDP per capita (annual average, %) Ancillary regressions Variable Baseline regression (1) (2) (3) (4) (5) (6) Variable of interest Trade openness, TO (exports and 0.6245** –8.2487** –9.8907** –0.6676 –1.2141 –2.5225** –10.0006** imports as % of GDP, log) (0.143) (1.627) (1.105) (0.474) (0.873) (1.251) (2.054) TO * ypc .. 0.9916** (0.183) TO * human .. .. 2.6520** .. .. .. .. (0.282) TO * findev1 .. .. .. 0.4230** .. .. .. (0.129) TO * findev2 .. .. .. .. 0.4382** .. .. (0.216) TO * findev3 .. .. .. .. .. 0.8046** .. (0.312) TO * instq .. .. .. .. .. .. 2.5798** (0.492) Control variable Transitional convergence, ypc –2.1263** –7.7486** –3.8154 –2.7008** –2.2621** –1.8939** –3.3400** (initial GDP per capita, log) (0.218) (0.864) (0.213) (0.239) (0.261) (0.301) (0.186) Human capital, human (gross 1.5336** 1.5093** –9.0224 1.6259** 2.1980** 1.9878** 1.5617** secondary enrollment rate, log) (0.207) (0.205) (1.116) (0.212) (0.198) (0.232) (0.181) Financial depth, findev1, (domestic 0.6229** 0.7660** 0.7449 –1.2707** .. .. 0.6589** credit to private sector, % GDP, log) (0.148) (0.128) (0.120) (0.549) (0.127) Financial depth, findev2, (banking .. .. .. .. –2.0794** .. .. credit private sector, % GDP, log) (0.823) Financial depth, findev3, (liquid .. .. .. .. .. –3.0230** .. liabilities—M3—% GDP, log) (1.260) Institutional quality, instq (ICRG 1.5695** 1.3471** 1.4749 0.2013** 0.1840 1.8149 –8.7044** political risk index, log) (0.418) (0.303) (0.274) (0.334) (0.356) (0.424) (1.942) Number of countries 99 99 99 99 99 99 99 Number of observations 646 646 646 646 646 646 646 Specification test (p-value) Sargan test (overidentifying restrictions) (0.256) (0.243) (0.196) (0.226) (0.280) (0.299) (0.190) Second-order serial correlation (0.211) (0.213) (0.181) (0.201) (0.193) (0.261) (0.214) Source: Authors’ calculations. Note: Numbers in parentheses correspond to robust standard errors. The full regression includes as control variables: the initial GDP per capita (log), gross secondary enrollment rate (log), domestic credit to the private sector as a percentage of GDP (log), ICRG political risk index (log), consumer price index inflation rate, the aggregate index of infrastructure stock (in logs, see definition in note a of table 4.1), foreign assets and liabilities as a percentage of GDP (log). The regression also includes constant and time (five-year period) dummies. We control for endo- geneity using lagged levels and differences for all the variables other than trade openness. The latter variable, in turn, is instrumented using lagged population, surface area of the country, and dummies for landlocked and oil-exporting countries. .. = Negligible. * p < .10, ** p < .05. 99 growth response (%) Un growth response (%) 100 ite 90 d –1.0 –0.5 0 0.5 1.0 1.5 2.0 –1.0 –0.5 0 0.5 1.0 1.5 2.0 th O 90 th St pe ECD at 75 rc pe es th en rc en pe tile til r La 75 e Un cen tin th OE ite tile pe C La 67 d Am 67 r D th Sta er th cen tin pe te ica pe til Am rc s rc e an e Calderón and Poggio er en ica t (e d t C nti an m ile xc he os le lu C ta ed di a R (e d t Co ia xc he s n ng rib ica t Do CA be Doludi Ca a Ri FT an m ng rib ca m in C be in m A) ica AF an ica e n TA n dia Re ) Re n p p Ni ub Gu ubl ca lic at ic r e E ag El ma 33 l Sa ua Sa la rd lv lv ad pe ado o Openness, by Level of Select Indicators of Development rc r 33 CA r en rd til pe FTA e rc CA en Gu FT 25 at A 25 Hon tile th em th du pe ra a. Conditional on the level of income per capita pe al rc s b. Conditional on the level of secondary schooling rc a e en Ho tile Ni ntil Figure 4.3 Growth Response to a 1 Standard Deviation Increase in Trade 10 10 ca e th ndu th rag pe ra pe ua rc s rc en en til til e e (continued next page) 90 th growth response (%) Un growth response (%) pe ite d –0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 –0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 rc 90 en th St til pe at Figure 4.3 e rc es Un O en ite EC til 75 d D 75 e th Sta th OE p t pe CD 67 erc es 67 r Source: Authors’ calculations. th en th cen (continued) pe til pe til rc e rc e e en Co ntil Ho tile st e nd aR i Co ura m ca st s Do e aR m El dia El Sa ica La in Sa n ica lv lv tin n ado ad Am Re p r La m or er tin ed ica Ni ubli ia an ca c Am CA n ra er (e d t gu ica Ni FT xc he ca A lu C CA a di a F an r ng rib TA (e d t Gu agu xc he at a CA be lu C em Do d a G F an m ing rib ala column 1; for panel b, column 2; for panel c, column 3; and for panel d, column 6. 33 ua TA) in C be rd te ica AF a p m n T n 25 erc ala 33 Re A) th en rd pu pe til p b rc e d. Conditional on the level of institutional quality e 25 erc lic th en Ho ntile 10 pe tile 10 nd th c. Conditional on the level of domestic financial development th rce pe ura pe nt Trade and Economic Growth: Complementarities for the DR-CAFTA Countries rc s rc ile en en til til e e Note: The computed responses were obtained using the estimated coefficients in table 4.4 as follows: for panel a, 101 102 Calderón and Poggio CAFTA countries, the Organisation of Economic Co-operation and Development [OECD] countries), and specific countries (DR-CAFTA countries, the United States). Our evidence shows that countries with higher income per capita reap the largest growth benefits from rising trade openness. We also find that all DR-CAFTA countries (with the exception of Costa Rica) have levels of income per capita below the sam- ple median for 2005 and, hence, a growth effect that is lower than the median response—that is, an increase that is smaller than 40 basis points a year in the growth rate. The higher growth in Costa Rica is approxi- mately 55 basis points a year (higher than the median), and the smallest response among DR-CAFTA countries is registered in Nicaragua. Here, the increase in trade openness leads to an annual decline in growth of out- put per capita of 38 basis points. Regression 2 of table 4.2 adds to our baseline regression the interac- tion between TO and the enrollment rate of secondary schooling (our proxy of human capital). Again, we find that the coefficient of trade openness is negative and significant, while that of the interaction is pos- itive and statistically significant. Hence, growth benefits from trade are positive and larger in countries with higher levels of human capital. More specifically, rising trade openness in countries with low rates of enrollment in secondary schooling (43 percent—for example, Bangladesh and Ghana, in the twenty-fifth percentile of the 2005 sam- ple) would have negligible effects on growth (almost 5 basis points a year). However, a 1 standard deviation increase in TO would raise the growth rate almost 1.3 percentage points a year in countries with higher levels of secondary schooling (96 percent—for example, the Slovak Republic and Slovenia in the seventy-fifth percentile). Regressions 3 through 5 report the interaction between trade openness and measures of financial development such as domestic credit to the pri- vate sector, domestic credit provided by domestic money banks, and liq- uid liabilities, respectively. All of these variables are expressed as a percentage of GDP and in logs, and they are interacted with trade open- ness. Regardless of the indicators of financial development used in our analysis, we find that the coefficient of TO is negative and not statistically significant in most cases, but the interaction with financial development is robustly positive. Our estimation suggests that countries with deeper domestic financial markets may reap the largest growth benefits from trade. Economically, countries with low financial development (say, with domestic credit to the private sector of 20 percent of GDP—for example, the average for the 2006–08 period in Paraguay and Botswana at the Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 103 twenty-fifth percentile of the distribution for that period) would raise their growth per capita by 35 basis points if trade openness were to increase by a 1 standard deviation. An analogous increase in TO would raise growth by 72 basis points in countries with high financial develop- ment (for example, Israel, with average domestic credit of 90 percent of GDP in 2006–08). Finally, regression 6 interacts trade openness and the level of institu- tional quality. We use the ICRG index of political risk (in logs) as our indicator of institutional quality. Again, the interaction between TO and institutions is positive and significant, while the coefficient of TO is neg- ative (although statistically significant). This implies that a minimum institutional threshold is required to reap the benefits from trade. It is consistent with the finding that trade reforms may lead to higher growth per capita in countries with stronger institutional quality (Calderón and Fuentes 2006, 2009; Chang, Kaltani, and Loayza 2009). Economically speaking, growth per capita would increase by only 30 basis points a year in countries with weak institutions (say, Bolivia and Honduras, at the twenty-fifth percentile of the sample distribution for the 2006–09 period), whereas the annual growth per capita benefit for a country with strong institutions (Poland and the Slovak Republic, at the seventy-fifth percentile) is approximately 72 basis points a year. Complementarities between Trade and Infrastructure There is ample evidence in the literature that an enlarged and more effi- cient infrastructure network will promote long-term growth (Sánchez- Robles 1998; Calderón and Servén 2004, 2010), while improved access to this network may help to reduce income inequality (Calderón and Chong 2004; Calderón and Servén 2004; Galiani, Gertler, and Schargrodsky 2005). Recent work also finds that the efficient provision of infrastructure is crucial for the success of trade liberalization strategies aimed at optimal resource allocation and export growth (Lederman, Maloney, and Servén 2007). Table 4.3 includes the interactions between trade openness and a bat- tery of infrastructure indicators (either at the aggregate level or by sec- tor). We have constructed two aggregate indexes of infrastructure, IK1 and IK2, that summarize information on telecommunications, electricity, and roads. Regressions 1 and 2 in table 4.3 include the interaction between trade openness and the aggregate indexes of infrastructure, IK1 and IK2, respectively. In both cases, we find that the coefficient of TO is negative and significant, whereas that of the interaction between TO and Table 4.3 Trade and Growth: The Role of Physical Infrastructure 104 dependent variable: growth in real GDP per capita (annual average, %) Ancillary regressions Baseline Variable regression (1) (2) (3) (4) (5) (6) (7) (8) (9) Variable of interest Trade openness. TO (exports and 0.6245** –1.7379** –1.1129** 0.0076 1.2848** –0.5840** –1.6612** 1.9481** –2.3922** –2.4914** imports as % of GDP, log) (0.143) (0.225) (0.186) (0.236) (0.140) (0.123) (0.389) (0.216) (0.271) (0.454) TO * IK1 .. 0.7038** .. .. .. .. .. .. .. .. (0.066) TO * IK2 .. .. 0.4733** .. .. .. .. .. .. .. (0.049) TO * TC1 .. .. .. 0.1188** .. .. 0.5174** .. .. .. (0.048) (0.072) TO * EGC .. .. .. .. 0.4285** .. .. 1.2333** .. .. (0.038) (0.132) TO * RD .. .. .. .. .. 0.0246** .. .. 1.8601** .. (0.012) (0.149) TO * TC2 .. .. .. .. .. .. .. .. .. 0.7101** (0.096) Control variable Index of aggregate infrastructure, IK1 0.2987** –1.6361** .. 0.3848** 0.0341 –1.2645** .. .. .. .. (first principal component: TC, EGC, RD) (0.146) (0.228) (0.158) (0.138) (0.117) Index of aggregate infrastructure, IK2 .. .. –0.2592* .. .. .. .. .. .. .. (first principal component: TC, EGC, RD) (0.166) Telecommunications 1, TC1 (main lines .. .. .. .. .. .. –1.5394** .. .. .. and mobiles per 1,000 people, log) (0.285) Electric power, EGC (installed capacity, .. .. .. .. .. .. .. –4.5521** .. .. in megawatts per 1,000 people, log) (0.508) Roads, RD (length of total network, in .. .. .. .. .. .. .. .. –7.3920** .. kilometers per 1,000 people, log) (0.608) Telecommunications 2. TC2 (main .. .. .. .. .. .. .. .. .. –2.0130** telephone line per 1.000 people, log) (0.374) Number of countries 99 99 99 99 99 99 99 99 99 99 Number of observations 646 646 646 646 646 646 646 646 646 646 Specification test (p-value) Sargan test (overidentifying restrictions) (0.256) (0.222) (0.377) (0.221) (0.252) (0.246) (0.206) (0.283) (0.214) (0.192) Second-order serial correlation (0.211) (0.177) (0.162) (0.173) (0.184) (0.180) (0.195) (0.163) (0.172) (0.142) Source: Authors’ calculations. Note: Numbers in parentheses correspond to robust standard errors. .. = Negligible. a. The full regression includes as control variables: the initial GDP per capita (log), gross secondary enrollment rate (log), domestic credit to the private sector as a percentage of GDP (log), ICRG political risk index (log), consumer price index inflation rate, the aggregate index of infrastructure stocks (in logs, see definition in note a of table 4.1), foreign assets and liabilities as a percentage of GDP (log). The regression also includes constant and time (five-year period) dummies. We control for endogeneity using lagged levels and differences for all the variables other than trade openness. The latter variable, in turn, is instrumented using lagged population, surface area of the country, and dummies for landlocked and oil-exporting countries. * p < .10, ** p < .05. 105 106 Calderón and Poggio infrastructure is positive and significant. Our evidence suggests that a bet- ter infrastructure network would enhance the impact of trade on growth. Using the estimates of regression 1 we find that a 1 standard deviation increase in trade openness would increase the growth rate per capita by 16 basis points in countries with a poor infrastructure network (that is, India and Pakistan, with an average index of infrastructure at the twenty- fifth percentile of the distribution for the 2006–08 period), while growth per capita would be higher by 1.4 percentage points for countries with a better infrastructure network (both for Taiwan, China, and for Singapore, with levels of infrastructure provision in the seventy-fifth percentile of the distribution). Note that when sectoral measures of infrastructure are included sepa- rately—rather than the aggregate measures—the results hold (see regres- sions 3–5 and 7–10 in table 4.3). Hence, an adequate supply of telecommunications and electricity and an improved road network may help raise the growth benefits from trade. Complementarities between Trade Openness and R&D Table 4.4 further investigates the complementarities between trade and human capital by examining the interaction between trade openness and innovation. Among proxies of innovation, we consider R&D spending (as a percentage of GDP), number of R&D scientists (per 1 million people), number of R&D technicians (per 1 million people), and high-technology exports (as a percentage of manufacturing exports). We also explore the interaction between trade and an index of R&D that summarizes infor- mation on R&D spending, R&D technicians, and R&D scientists. Regression 1 incorporates the interaction between trade openness and the index of innovation—the latter being measured as the first principal component of spending and the number of scientists and technicians in R&D. Higher values of this index indicate more resources devoted to R&D. We find that the coefficient of TO and its interaction are positive and significant, thus implying that trade openness enhances growth and that this effect is larger in countries with higher levels of innovation—as proxied by more resources devoted to R&D. When assessing the individual impact of the components of our index, we find that the interaction between TO and R&D is positive and sig- nificant only for R&D spending (regression 2) and R&D scientists (regression 3). Our estimates suggest that a 1 standard deviation increase in TO would lead to growth per capita higher by 90 basis points a year in countries with low R&D (that is, Colombia and Thailand, with 1.2 percent Table 4.4 Trade and Growth: The Role of Innovation dependent variable: growth in real GDP per capita (annual average, %) Ancillary regression Variable Baseline regression (1) (2) (3) (4) (5) Variable of interest Trade openness, TO (exports and 0.6245** 2.4385** 1.5130** 3.5576** 1.5577** 0.7394 imports as % of GDP, log) (0.143) (0.874) (0.216) (0.055) (0.748) (0.161) TO * R&D index (R&D aggregate index) .. 0.0002* .. .. .. .. (0.000) TO * R&D spending (R&D spending as % of GDP) .. .. 0.1989** .. .. .. (0.041) TO * R&D scientists (scientists in R&D .. .. .. 0.0004** .. .. per 1 million people) (0.000) TO * R&D technicians (technicians in R&D .. .. .. .. 0.0001 .. per 1 million people) (0.000) TO * high-tech exports (high-tech exports, % .. .. .. .. .. –0.0062 manufacturing exports) 0.013 Number of countries 99 67 82 78 72 98 Number of observations 646 446 545 519 472 641 Specification test (p-value) Sargan test (overidentifying restrictions) (0.256) (0.693) (0.311) (0.318) (0.638) (0.164) Second-order serial correlation (0.211) (0.394) (0.219) (0.263) (0.641) (0.188) Source: Authors’ calculations, using the GMM-IV System Estimator from Arellano and Bover 1995; Blundell and Bond 1998. Note: Numbers in parentheses correspond to robust standard errors. The full regression includes as control variables: the initial GDP per capita (log), gross secondary enrollment rate (log), domestic credit to the private sector as a percentage of GDP (log), ICRG political risk index (log), CPI inflation rate, the aggregate index of infrastructure stock (in logs, see definition in note a of table 4.1), foreign assets and liabilities as a percent of GDP (log). The regression also includes constant and time (five-year period) dummies. We control for endogeneity using lagged lev- els and differences for all the variables other than trade openness. The latter variable, in turn, is instrumented using lagged population, surface area of the country, and dummies for land- 107 locked and oil-exporting countries. The aggregate index of R&D is calculated as the first principal component of the following variables: R&D spending as a percentage of GDP, scientists in R&D per 1 million people, and technicians in R&D per 1 million people. .. = Negligible. * p < .10, ** p < .05. 108 Calderón and Poggio of GDP at the twenty-fifth percentile of the sample distribution in 2000–09). And the impact is 101 basis points a year in countries with higher spending (say, Ireland and New Zealand, with 3.2 percent of GDP at the seventy-fifth percentile). Complementarities between Trade Openness and Regulations Table 4.5 presents evidence on the complementarities between trade openness and economic regulations—that is, firm entry regulations and labor market regulations. Previous research shows that trade openness is unable to promote growth in heavily regulated economies (Bolaky and Freund 2004). We have constructed an index of economic regulations that comprises two subcategories: firm-entry regulations (the number of procedures to start a business, the number of days to start that business, and its cost) and labor regulations (difficulty of hiring, difficulty of firing, and rigidity of hours). Each subindex includes three variables. We construct this index and its two subindexes using either simple averages or principal compo- nents. The results are robust to either method of aggregation. Hence, for the sake of brevity, we discuss the results using simple averages, that is, regressions 1 through 3. In these regressions, we find that the coefficient of TO is positive and significantly different from 0, whereas that of the interaction between TO and regulations is negative and significant. This confirms existing evidence that more stringent regulations in the econ- omy may hinder economies from reaping the growth benefits of rising trade openness. Our estimates suggest that rising trade (a 1 standard devi- ation increase in TO) in countries with more flexible regulations (for example, Colombia, in the twenty-fifth percentile of the sample distribu- tion) would lead to higher growth per capita by almost 50 basis points. The increase in growth is lower (30 basis points) in countries that are heavily regulated (for example, France, in the seventy-fifth percentile of the sample distribution). Are Trade and Financial Openness Complementary in the Growth Process? Calderón and Poggio (2010) explore the complementarities between trade and financial openness in the growth process. The authors find that the interaction between trade openness and financial openness is positive and significant and that this effect may be driven by the higher accumulation of equity-related foreign assets and liabilities. These results suggest that the structure of external assets and liabilities may have a role in catalyzing the Table 4.5 Trade and Growth: The Role of Regulations dependent variable: growth in real GDP per capita (annual average, %) Ancillary regressions Aggregation method: Aggregation method: simple averages principal components Variable Baseline regression (1) (2) (3) (4) (5) (6) Variable of interest Trade openness, TO (exports 0.6245** 0.7914** 1.0219** 0.6772** 0.6950** 0.7087** 0.8693** and imports as % of GDP, log) (0.143) (0.144) (0.171) (0.192) (0.135) (0.184) (0.230) TO * index of regulations .. –0.5878** –0.3190** (0.224) (0.055) TO * index of firm entry regulations –1.6636** –0.4504** (0.276) (0.178) TO * index of labor regulations –0.6731** –0.3388** (0.135) (0.096) Number of countries 99 99 99 99 99 99 99 Number of observations 646 646 646 646 646 646 646 Specification test (p-value) Sargan test (overidentifying restrictions) (0.256) (0.250) (0.194) (0.321) (0.201) (0.282) (0.211) First-order serial correlation (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Second-order serial correlation (0.211) (0.181) (0.158) (0.311) (0.194) (0.251) (0.192) Source: Authors’ calculations using the GMM-IV System Estimator from Arellano and Bover 1995; Blundell and Bond 1998. Note: Numbers in parentheses correspond to robust standard errors. The full regression includes as control variables the initial GDP per capita (log), gross secondary enrollment rate (log), domestic credit to the private sector as a percentage of GDP (log), the ICRG political risk index (log), consumer price index inflation rate, the aggregate index of infrastructure stock (in logs, see definition in note a of table 4.1), foreign assets and liabilities as a percentage of GDP (log). The regression also includes constant and time (five-year period) dummies. We control for endogeneity using lagged levels and differences for all the variables other than trade openness. The latter variable, in turn, is instrumented using lagged population, surface area of the country, and dummies for landlocked and oil-exporting countries. Our indexes of regulations comprise information on the following dimensions: (a) firm entry regulations: number of pro- cedures to start a business, time to start (in days), and its cost (as a percentage of income per capita), and (b) labor market regulations: difficulty of hiring, rigidity of hours and difficulty of firing. All these indexes are constructed such that higher values indicate more obstacles to entry and industry and more rigidities in the labor market. Our index of regulations comprises information for all six indicators, and it is aggregated either using simple averages or the principal components analysis (that is, we take the first principal components). Analogously, we 109 compute the aggregate index of regulation for firm entry regulations and labor market regulations by either taking simple averages or the first principal component of the three indicators in each category. .. = Negligible. * p < .10, ** p < .05. 110 Calderón and Poggio effects of trade on growth. Hence growth benefits from trade openness may be larger in countries that accumulate more equity than debt assets and liabilities. The results are largely discussed in Calderón and Poggio (2010) and are available from the authors on request. Economic Implications of Our Estimates: Discussion for DR-CAFTA We now discuss the economic implications of the regression analysis for the DR-CAFTA countries. We conduct this analysis along three dimen- sions: (a) plot the growth response to a 1 standard deviation increase in trade openness conditional on the DR-CAFTA country’s level of deter- mined structural factors, (b) calculate the growth effects of an increase in trade openness in 2006–10 vis-à-vis 1991–95, and (c) assess potential growth benefits of trade openness if DR-CAFTA countries reach the extent of trade openness in a benchmark region (EAP-7). Growth Implications of Rising Trade in DR-CAFTA Figure 4.3 depicts the growth response to a 1 standard deviation increase in trade openness (that is, an increase in the trade ratio of approximately 75 percent during the period 2006–09) conditional on the level of income per capita (panel a), human capital (panel b), financial develop- ment (panel c), and institutional quality (panel d). We calculate the response for all DR-CAFTA countries (Costa Rica, Dominican Republic, El Salvador, Guatemala, Honduras, and Nicaragua), select regions and countries (CAFTA, Latin America and the Caribbean excluding CAFTA, OECD, and United States), and select percentiles of the sample distribu- tion in 2006–09. The bars represent the growth response (in percentage points), and the lines represent the 95 percent level of confidence inter- val. Growth benefits from trade vary greatly across DR-CAFTA coun- tries. For instance, the growth benefits of DR-CAFTA countries conditional on the level of secondary schooling are below the median of our sample distribution (that is, below 1.1 percentage points a year), with Costa Rica close to the median, Honduras below the twenty-fifth percentile of the distribution, and the model predicting a contraction in growth per capita of 19 basis points. However, the growth benefits of ris- ing trade conditional on the depth of domestic financial markets among DR-CAFTA countries cannot surpass those of the sixty-seventh per- centile of the sample distribution (66 basis points a year). Indeed, growth in Costa Rica, El Salvador, and Honduras rises between 53 and Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 111 57 basis points a year. The lowest benefits from trade are registered by the Dominican Republic (42 basis points), which is closer to that of countries in the thirty-third percentile of the sample distribution. Finally, the growth effects of trade openness conditional on institutional quality are also below that of the sixty-seventh percentile of the sample distribution for DR-CAFTA countries. Figure 4.4 plots the growth response to rising trade openness condi- tional on the level of infrastructure. We depict the response conditional on the aggregate index of infrastructure stock IK1 (panel a) and the stock of telecommunications (panel b), electricity (panel c), and roads (panel d). For the sake of brevity, we focus on the results for the aggre- gate index of infrastructure. The growth effect of rising trade in DR-CAFTA countries ranks below that of the country with the median level of infrastructure. Costa Rica and the Dominican Republic enjoy the largest benefits from trade (with increases in growth per capita of 95 and 84 basis points, respectively), thanks to their relatively better infrastructure network than other DR-CAFTA countries. Nicaragua is the country with the lowest gains from growth (below 50 basis points) among DR-CAFTA countries. Finally, figure 4.5 displays the growth response to trade openness con- ditional on spending on research and development (panel a) and regula- tory indexes for firm entry (panel b) and labor markets (panel c). Although growth response to rising trade is always positive, we fail to find significant differences between countries with low levels of innovation and regulation vis-à-vis those with high levels—especially in the case of labor market regulations. For instance, differences in the growth response to trade openness of Costa Rica (close to the R&D sample median) and El Salvador (below the tenth percentile in R&D spending) is not signifi- cant (91 and 88 basis points, respectively). The same can be said for reg- ulations. On average, DR-CAFTA countries have regulations that are more restrictive than those of the representative country in our sample. However, growth effects are not large when comparing the countries with the most stringent regulations and those with the most flexible regula- tions among DR-CAFTA countries (see panels b and c). Assessing the Growth Benefits from Open Trade in DR-CAFTA We evaluate the growth effects of a 1 standard deviation increase in trade openness for DR-CAFTA countries conditional on structural factors and policies. If we denote a0 as the coefficient estimate of trade openness (TO) and a1 as the interaction between TO and a determined structural number of main lines and mobiles 112 90 th per 1 million people (logs) Un IK1 index pe ite d 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 –1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 rc 90 en th St til pe at 75 e rc es th OE en til pe CD e Un rce 75 it n th OE 67 ed tile CD th Sta La 67 perc La tin th en tin pe te Calderón and Poggio rc s Am pe til Am e rc e er El nti er en ica Sa le ica ti an lv an m le ad (e d C edi (e d t xc the os an xc he m or lu C ed l t di a ia Do udi Ca a R ng rib n m ng ribb ica CA be in CA e a ica F an Gu FTA n n TA at ) Re ) Do p m C ema El ub in os la Sa lic ica ta lv n R ad Re ica 33 o pu rd CA r bl 33 ic pe FTA rd CA rc en pe FTA Trade Openness, by Aggregate Stock of Select Infrastructure Ho tile rc nd en Ho tile Gu ur nd at as em Figure 4.4 Growth Response to a 1 Standard Deviation Increase in N ur Ni a 25 ica as 25 ca la th rag th rag p u p u a. Conditional on level of aggregate stock of infrastructure 10 erc a 10 erc a th en b. Conditional on level of aggregate stock of telecommunications th en pe tile pe til rc rc e en en til e til e (continued next page) length of the total road network electricity installed capacity Un (kilometers) per 1,000 people Un (megawatts) per 1 million people i i –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 –2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 90 ted 90 ted Figure 4.4 th St th St pe ate pe ate rc s rc s en en til til 75 e e th O 75 La th OE tin pe EC Source: Authors’ calculations. pe CD rc D e 6 r (continued) Am er Co ntil Do 7th cen ica 6 st e m pe tile La in an 7th a R ica rcen (e d t pe ica tin n t xc he rc Am Re ile lu C en er pu di a t ng rib ile ica b CA be an m lic ed FT an (e d t Co ia m A) xc he s n e lu C ta N di di a Ri ng rib ca 33 ica an CA be rd ra pe gua FT an rc A) Do en C m til column 7 for panel b, column 8 for panel c, and column 9 for panel d. in e Ho AFT nd A ica CA n FT El ura 25 Rep A 33 Sa s th u rd lvad pe bli rc c pe o e rc r e Ho ntile Gu nti nd at le El ur em N a d. Conditional on level of aggregate stock of roads 10 Sal as 25 ica la th vad th rag pe o p u c. Conditional on level of aggregate stock of electric power rc r 10 erc a Gu en th en Trade and Economic Growth: Complementarities for the DR-CAFTA Countries at tile pe til em rc e al en a til e Note: The computed responses were obtained using the estimated coefficients from table 4.3: column 2 for panel a; 113 114 Calderón and Poggio Figure 4.5 Growth Response to a 1 Standard Deviation Increase in Trade Openness, by Level of R&D Spending and Regulations a. Conditional on level of R&D spending 1.4 1.2 growth response (%) 1.0 0.8 0.6 0.4 0.2 0 s e CD ile ile Co an ica ng an le th FT e e A e or a as a e gu al il 25 CA d th til til FT i ad ur at nt nt nt lu ric ent i aR pe A) em OE ed en en ra CA St nd ce ce ce lv m (e Am erc rc rc st ca at Sa d r r r Ho pe pe pe pe ite Ni Gu di a an in rd p El Un th th th th xc e 90 75 67 10 rib be Lat 33 Ca 0.8 b. Conditional on level of firm entry regulations 0.7 growth response (%) 0.6 0.5 0.4 0.3 0.2 0.1 0 e es CD e e a n ic e A or a e as F he ica e gu al til til til til til til ia FT bl ad ric dur at CA d t aR em Co TA) OE ed en en en en en en pu ra CA St lv ng an (e Am on m rc rc rc rc rc rc st ca at Re Sa d pe pe pe pe pe pe ite Ni Gu di a H El n Un th th rd ica th th th xc e 10 25 67 75 90 33 in m lu an in Do be Lat rib Ca 0.7 c. Conditional on level of labor market regulations 0.6 growth response (%) 0.5 0.4 0.3 0.2 0.1 0 –0.1 –0.2 es e e e ic or D n a ua FT he pe TA Co tile ica e as e al til til til til til ia C bl ad ur at ag CA d t F aR em A) OE ed en en en en en en pu CA St nd lv (e Am icar ng an m rc rc rc rc rc rc st at Re Sa d Ho pe pe pe pe pe ite Gu di a N El n lu ric Un th th rd ica th th th xc e 10 25 67 75 90 33 in m an in Do be Lat rib Ca Source: Authors’ calculations. Note: The computed responses were obtained using the estimated coefficients from column 2 of table 4.4 for panel a and from column 2 of table 4.5 for panel b. R&D spending is the average ratio of R&D expenditure as a percentage of GDP for the 2000–09 period. Firm entry regulations are calculated as the simple average of the following measures: number of procedures, time, and cost. Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 115 factor SF, then the response of growth to a change in trade openness is as follows: ( ) dg = α 0 + α1 ⋅ SF ⋅ dTO, (4.4) where dTO is the standard deviation increase in the ratio of trade to GDP for each country over the sample period and SF is the level of the struc- tural factor. Table 4.6 reports the growth response to rising trade open- ness conditional on the following structural factors: human capital, financial development, institutional quality, infrastructure stock, financial openness, innovation, and regulations. Panel a of table 4.6 assumes that SF is the level of the structural policy in the DR-CAFTA country in the period 2006–10. Panel b calculates the growth benefits that DR-CAFTA countries could potentially obtain if their structural policies were at the level of the seventy-fifth percentile of the sample distribution in 2006–10. Finally, panel c presents the growth gains for DR-CAFTA of shifting their structural policies to those of the leaders (seventy-fifth per- centile) in the face of rising trade openness. The first column of table 4.6 shows the growth effects of a 1 standard deviation increase in trade openness for DR-CAFTA countries. The coun- tries in the region reaping the largest benefit in our baseline model (with- out interactions) are Costa Rica (25 basis points), Nicaragua (25 basis points), and El Salvador (22 basis points), whereas the Dominican Republic and Honduras obtain the lowest benefits (7 basis points). Next we report the contribution in the models with interactions. If the growth benefits are higher than those reported in the baseline model, then the complementarities at work enhance rather than hinder the impact of trade openness on growth. Conditional on the level of human capital in the corresponding DR- CAFTA countries, we find that Costa Rica obtains the largest benefits (67 basis points a year), while human capital in Honduras offsets the effect of rising trade integration on growth. Financial development, in contrast, does not amplify the growth effect of openness as much as human capital. Again, Costa Rica is the winner, with a growth rate higher by 37 basis points, while the Dominican Republic obtains the lowest ben- efits from growth. The same is found for infrastructure; in Costa Rica, the large network of infrastructure allows the country to raise the growth rate by 41 basis points a year, while in Honduras the growth rate increases by a meager 6 basis points. Table 4.6 Growth Effects due to Changes in Trade Openness 116 Trade openness interacted with Baseline Human Financial Institutional Infrastructure Financial Research & Economic model capital development quality stock openness development regulations a. Conditional on the structural factors of the DR-CAFTA country in 2006–10 Costa Rica 25 67 37 41 65 30 62 22 Dominican Republic 7 17 8 10 17 8 .. 8 El Salvador 22 41 32 29 39 28 53 22 Guatemala 10 8 12 11 13 5 24 9 Honduras 7 –4 12 6 10 10 18 6 Nicaragua 25 49 34 31 28 33 60 25 b. Conditional on the structural factors of the 75th percentile of the world distribution in 2006–10 Costa Rica 25 88 49 49 98 54 69 23 Dominican Republic 7 26 15 15 29 16 21 7 El Salvador 22 77 43 43 86 47 61 20 Guatemala 10 35 19 19 39 21 27 9 Honduras 7 26 15 15 29 16 21 7 Nicaragua 25 88 49 49 98 54 69 23 c. Growth gains due to a1 standard deviation increase in trade if the structural factors improve to 75th percentile of distribution Costa Rica 20 12 8 34 24 7 1 Dominican Republic 10 6 5 12 8 0 El Salvador 36 11 14 47 19 7 0 Guatemala 26 7 9 25 16 4 0 Honduras 30 3 8 19 6 3 0 Nicaragua 39 15 18 70 21 9 0 Source: Authors’ calculations. Note: .. = Negligible. Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 117 Panel b presents the growth effects of trade evaluated in the seventy- fifth percentile of the structural policies in the sample distribution. In most cases, given the distance to the frontier by DR-CAFTA models, the growth effects are larger. Panel c summarizes these differences by calcu- lating the potential growth gains of a 1 standard deviation increase in trade openness if DR-CAFTA countries were to reach the level of struc- tural policies of the top quartile of the sample distribution. The results show that human capital and infrastructure are the sectors with the largest potential to realize the growth effects from trade open- ness. Whereas the growth rate would increase between 10 and 39 basis points a year if the level of human capital were to increase in DR-CAFTA countries, it would increase between 12 and 70 basis points if the improvement would occur in infrastructure. Of course, reaching these levels implies a large amount of investment that is likely to be implausi- ble to undertake in a short time horizon. Finally, other exercises are presented in Calderón and Poggio (2010): (a) growth effect of the change in trade openness in 2006–10 vis-à-vis 1991–95, conditional on the structural policies of DR-CAFTA countries; and (b) potential growth benefits of DR-CAFTA countries of attaining the trade integration levels of the East Asian tigers. The results, although not reported, are available from the authors on request. Concluding Remarks The goal of this study is to evaluate the growth effects of trade openness among DR-CAFTA countries and, more specifically, to examine whether these growth effects are stimulated or hindered by advances in structural policies and institutions. Following recent empirical literature, we evalu- ate the role of complementarities between trade openness and the follow- ing factors: human capital, financial development, institutional quality, infrastructure, financial openness, innovation, and economic regulations. Using our effective regression sample of 99 countries with five-year nonoverlapping observations over the period 1960–2010, we find the fol- lowing results. First, there is a robust causal link between trade and growth. Regardless of the set of instruments used in our regression analysis, we find that trade openness stimulates growth. In fact, our estimates are not only statistically but also economically significant: a 1 standard deviation increase in the ratio of trade to GDP (that is, an increase of roughly 75 percent in the ratio) would lead to an increase in the rate of growth 118 Calderón and Poggio per capita of 35 basis points a year (and an accumulated increase of 5.5 percentage points over 15 years). Second, we find strong evidence that the impact of trade openness on growth depends on country-specific conditions in structural areas such as education, financial development, institutional quality, infrastructure, financial openness, innovation, and regulations. In general, we find that growth benefits from trade openness will be larger in countries that sur- pass a certain threshold in the structural areas mentioned above. Third, trade stimulates growth in countries with higher levels of human capital, deeper domestic financial markets, stronger institutions, more developed infrastructure networks, high integration with world financial markets, higher intensity in R&D investment, and less stringent regulations. Fourth, although our baseline model (without) interactions predicts growth benefits from trade for DR-CAFTA countries, we find that not accounting for complementarities between trade openness and structural factors may overstate these results. In fact, we find that human capital, infrastructure development, and institutional quality may play an impor- tant role in enhancing the growth benefits from trade. Finally, there is ample room among DR-CAFTA countries for reaping the growth benefits from trade. However, a larger role should be played by further reforms in areas such as education, infrastructure, international financial integration, and the development of domestic financial markets. Notes 1. This “flow” measure more closely captures current policies on schooling and human capital investment than “stock” measures related to educational attain- ment of the adult population or life expectancy (Loayza, Fajnzylber, and Calderón 2005). 2. The sector-specific indicators of infrastructure quantity and quality employed below, while standard in the literature, are subject to caveats regarding their homogeneity and international comparability. For example, the quality and condition of a “paved road” can vary substantially across countries—even within the same country. More homogeneous measures of infrastructure per- formance would clearly be preferable, but unfortunately they do not exist, at least with any significant coverage across countries and time periods. 3. The correlation between the two synthetic quantity indexes is over 0.996. This is not surprising given the similarly high correlation between the two indicators of telephone density underlying the respective synthetic indicators. Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 119 4. A similar adjustment is presented in Pritchett (1996). 5. For details, see Arellano and Bover (1995); Blundell and Bond (1998). 6. The results are presented in Calderón and Poggio (2010) and are available from the authors on request. 7. Although these scatter plots are not reported, they are available from the authors on request. 8. A more detailed presentation of these results is provided by Calderón and Poggio (2010). References Ades, A. F., and E. L. Glaeser. 1999. “Evidence on Growth, Increasing Returns, and the Extent of the Market.” Quarterly Journal of Economics 114 (3): 1025–45. Aghion, P., J. Fedderke, P. Howitt, C. Kularatne, and N. Viegi. 2008. “Testing Creative Destruction in an Opening Economy: The Case of the South African Manufacturing Industries.” Documents de Travail 2008-23, Observatoire Français des Conjonctures Economiques, July. Alesina, A., E. Spolaore, and R. Wacziarg. 2000. “Economic Integration and Political Disintegration.” American Economic Review 90 (5): 1276–96. ———. 2005. “Trade, Growth, and the Size of Countries.” In Handbook of Economic Growth, ed. P. Aghion and S. N. Durlauf vol. 1, part 2, 1499–542. Amsterdam: Elsevier North-Holland. Arellano, M., and S. Bond. 1991. “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations.” Review of Economic Studies 58 (2): 277–97. Arellano, M., and O. Bover. 1995. “Another Look at the Instrumental-Variable Estimation of Error-Components Models.” Journal of Econometrics 68 (1): 29–52. Barro, R. J., and J. W. Lee. 2001. “International Data on Educational Attainment: Updates and Implications.” Oxford Economic Papers 53 (3): 541–63. Beck, T., and A. Demirgüç-Kunt. 2009. “Financial Institutions and Markets across Countries and over Time: Data and Analysis.” Policy Research Working Paper 4943, World Bank, Washington, DC, June. Beck, T., A. Demirgüç-Kunt, and R. Levine. 2000. “A New Database on Financial Development and Structure.” World Bank Economic Review 14 (3): 597–605. Blundell, R., and S. R. Bond. 1998. “Initial Conditions and Moment Restrictions in Dynamic Panel Data Models.” Journal of Econometrics 87 (1): 115–43. Bolaky, B., and C. Freund. 2004. “Trade, Regulations, and Growth.” Policy Research Working Paper 3255, World Bank, Washington, DC, March. 120 Calderón and Poggio Calderón, C., and A. Chong. 2004. “Volume and Quality of Infrastructure and the Distribution of Income: An Empirical Investigation.” Review of Income and Wealth 50 (1): 87–105. Calderón, C., and R. Fuentes. 2006. “Complementarities between Institutions and Openness in Economic Development: Evidence for a Panel of Countries.” Cuadernos de Economía 43 (127): 49–80. ———. 2009. “Removing the Constraints from Growth: Some Guidelines.” Working Paper 366, Instituto de Economía, December, Pontificia Universidad Católica de Chile, Santiago. Calderón, C., N. Loayza, and K. Schmidt-Hebbel. 2006. “External Conditions and Growth Performance.” In External Vulnerability and Preventive Policies, ed. R. Caballero, C. Calderón, and L. F. Cespedes, 41–70. Series on Banking, Analysis and Economic Policies. Santiago: Central Bank of Chile. Calderón, C., and V. Poggio. 2010. “Trade and Economic Growth: Evidence on the Role of Complementarities for CAFTA-DR Countries.” Policy Research Working Paper 5426, World Bank, Washington, DC, September. Calderón, C., and L. Servén. 2004. “The Effects of Infrastructure Development on Growth and Income Distribution.” Policy Research Working Paper 3400, World Bank, Washington, DC, September. ———. 2010. “Infrastructure in Latin America.” Policy Research Working Paper 5317, World Bank, Washington DC. Chang, R., L. Kaltani, and N. Loayza. 2009. “Openness Can Be Good for Growth: The Role of Policy Complementarities.” Journal of International Economics 90 (1): 33–49. Coe, D. T., and E. Helpman. 1995. “International R&D Spillovers.” European Economic Review 39 (5): 859–87. Coe, D. T., E. Helpman, and A. W. Hoffmaister. 1997. “North-South R&D Spillovers.” Economic Journal 107 (440): 134–49. Deardorff, A. V. 1974. “Factor Proportions and Comparative Advantage in the Long Run: Comment.” Journal of Political Economy 82 (4): 829–33. Dollar, D. 1992. “Outward-Oriented Developing Economies Really Do Grow More Rapidly: Evidence from 95 LDCs, 1976–85.” Economic Development and Cultural Change 40 (3): 523–44. Edwards, S. 1992. “Trade Orientation, Distortions, and Growth in Developing Countries.” Journal of Development Economics 39 (1): 31–57. Frankel, J. A., and D. Romer. 1999. “Does Trade Cause Growth?” American Economic Review 89 (3): 379–99. Galiani, S., P. Gertler, and E. Schargrodsky. 2005. “Water for Life: The Impact of the Privatization of Water Services on Child Mortality.” Journal of Political Economy 113 (1): 83–120. Trade and Economic Growth: Complementarities for the DR-CAFTA Countries 121 Gaston, N., and D. Trefler. 1995. “Union Wage Sensitivity to Trade and Protection: Theory and Evidence.” Journal of International Economics 39 (1–2): 1–25. Grossman, G. M., and E. Helpman. 1991. Innovation and Growth in the Global Economy. Cambridge, U.K.: MIT Press. Head, K., and J. Ries. 1999. “Can Small-Country Manufacturing Survive Trade Liberalization? Evidence from the Canada-U.S. Free Trade Agreement.” Perspectives on North American Free Research Publication 1, Industry Canada, April. Helleiner, G. 1986. “Outward Orientation, Import Instability, and African Economic Growth: An Empirical Investigation.” In Theory and Reality in Development: Essays in Honor of Paul Streeten, ed. S. Lall and F. Stewart, 139–53. London: Macmillan. Heston, A., R. Summers, and B. Aten. 2009. “Penn World Table Version 6.3.” Center for International Comparisons of Production, Income and Prices, University of Pennsylvania, Philadelphia. Kohli, I., and N. Singh. 1989. “Exports and Growth: Critical Minimum Effort and Diminishing Returns.” Journal of Development Economics 30 (2): 391–400. Krishna, P., and D. Mitra. 1998. “Trade Liberalization, Market Discipline, and Productivity Growth: New Evidence from India.” Journal of Development Economics 56 (2): 447–62. Krishna, P., D. Mitra, and S. Chinoy. 2001. “Trade Liberalization and Labor Demand Elasticities: Evidence from Turkey.” Journal of International Economics 55 (2): 391–409. Lane, P. R., and G.-M. Milesi-Ferretti. 2001. “The External Wealth of Nations: Measures of Foreign Assets and Liabilities for Industrial and Developing Countries.” Journal of International Economics 55 (1): 263–94. ———. 2007. “The External Wealth of Nations Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities, 1970–2004.” Journal of International Economics 73 (2): 223–50. Lederman, D., W. F. Maloney, and L. Servén. 2007. Lessons from NAFTA for Latin America and the Caribbean. Latin American Development Forum Series. Palo Alto, CA: Stanford University Press, Stanford Economics and Finance; Washington, DC: World Bank Levinsohn, J. 1993. “Testing the Imports-as-Market-Discipline Hypothesis.” Journal of International Economics 35 (1–2): 1–22. ———. 1999. “Employment Responses to International Liberalization in Chile.” Journal of International Economics 47 (2): 321–44. Lewer, J., and H. van den Berg. 2003. “Does Trade Composition Influence Economic Growth? Time Series Evidence for 28 OECD and Developing Countries.” Journal of International Trade and Economic Development 12 (1): 39–96. 122 Calderón and Poggio Loayza, N., P. Fajnzylber, and C. Calderón. 2005. “Economic Growth in Latin America and the Caribbean: Stylized Facts, Explanations, and Forecasts.” Latin American and the Caribbean Studies. Washington, DC: World Bank, April. Lucas, R. E. Jr. 1988. “On the Mechanics of Economic Development.” Journal of Monetary Economics 22 (1): 3–42. Matsuyama, K. 1992. “Agricultural Productivity, Comparative Advantage, and Economic Growth.” Journal of Economic Theory 58 (2): 317–34. Pavcnik, N. 2002. “Trade Liberalization, Exit, and Productivity Improvement: Evidence from Chilean Plants.” Review of Economic Studies 69 (1): 245–76. Pritchett, L. 1996. “Measuring Outward Orientation in LDCs: Can It Be Done?” Journal of Development Economics 49 (2): 307–35. Rodríguez, F., and D. Rodrik. 2001. “Trade Policy and Economic Growth: A Skeptic’s Guide to the Cross-National Evidence.” In NBER Macroeconomics Annual 2000, ed. B. Bernanke and K. Rogoff, 261–335. Cambridge, MA: MIT Press. Rodrik, D. 2005. “Growth Strategies.” In Handbook of Economic Growth, vol. 1, ed. P. Aghion and S. Durlauf, 967–1014. Amsterdam: Elsevier. Romer, P. M., 1990. “Endogenous Technological Change.” Journal of Political Economy 98 (5, pt. 2): S71–S102. Sachs, J. D., and A. M. Warner. 1995. “Economic Reform and the Process of Global Integration.” Brookings Papers on Economic Activity 1: 1–118. Sánchez-Robles, B. 1998. “Infrastructure Investment and Growth: Some Empirical Evidence.” Contemporary Economic Policy 16 (1): 98–108. Trefler, D. 2004. “The Long and Short of the Canada-U.S. Free Trade Agreement.” American Economic Review 94 (4): 870–95. Tybout, J. R., J. de Melo, and V. Corbo. 1991. “The Effects of Trade Reforms on Scale and Technical Efficiency.” Journal of International Economics 31 (3–4): 231–50. Wacziarg, R. 2001. “Measuring the Dynamic Gains from Trade.” World Bank Economic Review 15 (3): 393–429. Wacziarg, R., and I. Welch. 2008. “Trade Liberalization and Growth: New Evidence.” World Bank Economic Review 22 (2): 187–231. Young, A. 1991. “Learning by Doing and the Dynamic Effects of International Trade.” Quarterly Journal of Economics 106 (2): 369–405. CHAPTER 5 Power Integration in Central America: From Hope to Mirage? Juan Miguel Cayo Energy is a key input for production, and as such high energy prices put firms at a competitive disadvantage. The question then is, why are energy prices so high in Central America (see figure 5.1), and what can be done to address this problem? According to a recent study by the World Bank (2010), several obstacles remain in the path to energy security in Central America. These include (a) a tight balance between power generation and demand, which adversely affects the reliability of supply and its quality; (b) significant exposure to oil price volatility and shocks due to overde- pendence on oil imports, which have increased with the region’s growing reliance on thermal power stations; (c) significant inefficiencies in the institutional and regulatory framework of several countries, which affect the financial sustainability of power utilities and their operations; and (d) relatively low levels of access in certain countries, which affect rural areas in particular. The six Central American countries of Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and Panama share a long tradition of regional integration, including a common market, substantial intrare- gional trade, as well as coordinated commercial policies, such as the Central America Free Trade Agreement (CAFTA) with the United States. In the electricity subsector, the most significant example of regional 123 124 Cayo Figure 5.1 Electricity Tariffs in Select Citiesa 30 tariff (US$ per kilowatt-hour) 25 20 15 10 5 0 s ) go ta ito n a s ) r go o zil ca do re ca Lim de io go ia Qu in ra Ai Ri ra nc lva vi nt m Bo (B Ca ta os te u Sa Sa Do As na os on en n (C ra o M Bu Sa nt Pa sé Sa Jo n Sa Source: CIER 2010. a. Industries using less than 50,000 kilowatt-hours a month. integration consists of the SIEPAC (Central American Electrical Interconnection System) interconnection line, which is expected to fully link the six countries by 2011; Costa Rica and Panama were connected in November 2010. The interconnection has been a long-term effort, start- ing in the early 1990s with the support of the Inter-American Development Bank and the government of Spain. SIEPAC was designed to bring the benefits of integration to the six countries and to improve their national power systems. Due to the rela- tively small size of the power system in each of the region’s nations, the opening of the regional market was seen as a means for creating a larger market to enhance competition among power producers and for provid- ing a secure supply of power to all individual countries at the same time. The goal is for the regional market to allow qualified agents to buy or sell electricity no matter where they are located in the Central American region. Additionally, a regional market with clear and uniform rules is expected to offer incentives for building larger and more efficient power plants, sparking investments that would help to reduce the costs and increase the reliability of electricity systems in the region. However, as the national markets evolve toward integration and increased trade, important barriers still hinder the full implementation of the regional electricity market (MER). This chapter identifies some of the barriers to the development of a truly enhanced regional electricity market in Central America. As this Power Integration in Central America: From Hope to Mirage? 125 book is aimed at a nonenergy public, the chapter begins by explaining some of the complexities of modern electricity markets and the chal- lenges of power trade and integration. Then it presents some of the key political economy considerations around power integration. In particular, the trade-off between energy integration and sovereignty is a fact that governments participating in integration initiatives will have to deal with and solve. Next the paper presents a brief history of the SIEPAC project and the harmonization efforts to date. Finally, it identifies the remaining obstacles to power integration that the governments of Central America will have to overcome for success in this important effort. The year 2010 may be crucial to determining the success or failure of the SIEPAC project. The measure of how far the region is able to advance in the near future will determine whether power integration in Central America will fulfill the expectations of its designers or will convert into a wasteful effort. What Does Power Integration Mean? The physical properties of electricity production, transmission, and distri- bution make the challenge of matching supply and demand at every moment especially difficult. Because storing electricity is virtually impos- sible (at least in economical terms) and capacity constraints on produc- tion from a plant cannot be breached for significant periods without incurring extreme risks, the amount of energy that can be delivered at any particular moment is essentially fixed. Any failure to equate demand and supply endangers the stability, not only of the market participants that caused the imbalance, but of the system as a whole. Moreover, an action that could be profitable to one market participant but simultaneously degrades the system’s reliability can negatively affect the ability of other buyers or suppliers to fulfill their contractual obligations (that is, there are important externalities as a result of being interconnected). For this rea- son, modern electricity markets usually have a system operator that con- trols the operation of the flows generated (dispatch) to preserve adequate functioning—and avoid major oscillations in tension—throughout the transmission network. One major feature of modern electricity markets is that, while elec- trons flow according to the laws of physics, energy payments flow accord- ing to the terms of financial contracts.1 This means that the laws governing the flow of energy are totally independent of the financial flows. When consumer A signs a contract to purchase energy from 126 Cayo generator B, it does not necessarily mean that the energy received by con- sumer A was physically produced by generator B. On the contrary, this generally is not the case. As Wolak (2004) clearly states, Contrary to common perception, a buyer of electricity is not purchasing megawatt-hours (MWhs) of energy produced by a specific generation unit. A buyer is only purchasing the right to withdraw that quantity of MWhs from a specific location in the network, and a seller is paid for injecting a cer- tain quantity of MWhs into the grid at a specified location in the network. In the spot or wholesale market, generators can buy or sell energy among themselves as long as the system is ultimately balanced. For exam- ple, if the system operator who administers system dispatch instructs one generator to generate less energy than the amount agreed in its contracts, it becomes a “deficit generator” and has to buy the rest of the energy needed to complete its contracts in the spot market. On the contrary, if a generator is made to generate more energy than it is contracted to pro- vide, it becomes a “surplus generator,” so it has to sell surplus energy in the spot market. The spot market price is given by the marginal cost of the system, in other words, by the variable cost of the most expensive generator that is dispatched. The supply curve of the industry is—as usual—the aggregation of all of the marginal cost curves, which is represented by a stairway profile because the marginal cost of the system jumps up every time a new power plant enters the dispatch merit order. The market price is deter- mined by the marginal cost of the last plant that is generating power to meet the demand for energy in that precise moment in the very short term. In peak hours (typically in the evenings) when demand increases, the spot price jumps because the more expensive plants are required to generate. Originally, electric power systems were developed as small isolated monopolies with a few generating units under central control. Gradually these isolated systems were interconnected to allow power trading and reserve sharing. As interconnections grew, so did the scope and complex- ity of the control system. No real-time pricing of grid interactions was established, but there was little reason to do so because neighboring gen- erators (monopolies) essentially bartered energy and reserves and cooper- ated to maintain system reliability. Eventually, simple contract trading developed among neighboring monopolies, but this trading did not include any actual pricing of network interactions. Pressure for competitive generation developed only in the Power Integration in Central America: From Hope to Mirage? 127 1970s and 1980s, when independent power producers could also be paid for delivering energy to the local monopoly utility. But real competition, in which a generator could compete to sell directly to a consumer or to a distributor not affiliated with the local monopoly utility, did not develop until the 1990s (see Hunt 2002).2 In modern electricity markets, there are multiple participants; a large number of generators; and a large number of consumers, whether regu- lated (that is, small household and commercial consumers that are served by distribution utilities) or not regulated (that is, big industrial plants, mines, and so forth that can purchase energy directly from suppliers at negotiated prices). As mentioned, the control of this complex system relies on a centralized system operator, which manages the dispatch of several power plants according to their marginal costs, either “real” costs or “bid prices,” depending on the regulatory framework, in the best manner to ensure the stability and proper functioning of the trans- mission grid. The financial flows linked to contracts among participants in the market are independent of the physical flows, so a market administrator is needed to function as a clearinghouse for net energy flows so that payments can clear accordingly.3 Interconnection Is Not Equal to Integration There are many examples of electricity interconnections between two or more national power systems in the Latin American region. For example, the Colombia-Ecuador interconnection started operations in late 1998, but due to the physical configuration of the transmission line, it is just a “radial” interconnection—meaning that it is impossible to attain a syn- chronized operation of both systems.4 Under these circumstances, it would be difficult to assert that Colombia and Ecuador have an inte- grated power system; they just trade power. Another example is the interconnection of Brazil and Paraguay through the co-owned Itaipu power plant. Itaipu is owned on a 50-50 basis by Brazil and Paraguay, but 90 percent of its production goes to Brazil and only 10 percent goes to Paraguay. The Paraguayan energy surplus—that is, the difference between its right to access 50 percent of the plant’s energy and domestic consumption—is obligated to be sold to Brazil through the Brazilian electricity monopoly, which acts as a sin- gle buyer.5 Again, it would be fair to say that the Paraguayan power system is linked to Itaipu (almost 90 percent of Paraguay’s domestic demand is provided by this plant), but not necessarily to the Brazilian power system. The reverse is even more true: the Brazilian power system 128 Cayo is not really integrated with the Paraguayan system; they just share a com- mon power plant. Hence, the difference between interconnection and integration is analogous to the difference between trading goods and hav- ing a common economic market. One important feature of power integration is that it is a very long process with multiple stages. The first stage is, of course, to create the physical interconnection through a transmission line across the border. But from that point onward, power integration can take different modal- ities or degrees of “integration.” In its most basic form, a generator in one country can provide electricity to consumers in another country at a con- tract price. Itaipu is an example of this basic form of trade, which would not be referred to as “integration.” Neither system is working in coordina- tion nor does the dispatch of different plants on either side of the border affect the energy flows of this plant. The next stage is to integrate the two systems so that they work with a coordinated dispatch that allows participants in both markets to know the spot prices in each market at every moment.6 The higher-price coun- try will always import from the lower-price country and vice versa. These are the so-called short-term international transactions; contracts are not between specific suppliers or consumers but rather between the two wholesale markets (the Colombia-Ecuador trade operates through this modality). In the case of the existence of a long-term contract between a supplier in country A and a consumer in country B (for instance, a distri- bution company), the fact that a contract exists does not mean that the supplier will necessarily generate its own energy to comply with the con- tract. It may well honor its contract by buying energy in the domestic spot market (if PA < PB) or in the foreign spot market (if PA > PB). One important feature of this level of integration is that, whenever the low-price country is exporting, the domestic spot price will rise as a con- sequence of having additional demand that will have to be met with its installed capacity. As a result of export activity, higher-cost plants will have to be ordered to dispatch power, as the new demand (from the import country) exerts pressure on the domestic system. Conversely, the spot market price in the import country will decline as the new supply (coming from across the border) partially meets the domestic demand, thereby displacing high-cost plants in the local market. This is a basic idea that is important to have in mind when promoting power trade: prices will rise in the export country as a consequence of the power trade. This is why some countries find it politically difficult to con- vince domestic consumers that exporting power is beneficial. By analogy, Power Integration in Central America: From Hope to Mirage? 129 power generators in the import country are prone to lobby the govern- ment, as imports of cheap energy from abroad can erode their profits. The third stage of integration is the integrated market—or power pool—in which the two systems behave as one, with a single centralized dispatch and no difference between market participants, whether nation- als or foreigners. In this integrated market, a unique supply curve is deter- mined by the marginal costs of all the power plants in both systems,7 and the aggregate demand for power is the sum of all consumption on both sides of the frontier. Consequently, there is only one price in the spot mar- ket for this integrated system. This is the last stage and what most people have in mind when they talk about “integration.” Of course, this level of integration (attained, for example, in the Nordpool system of the Nordic countries) requires an impressive amount of political will, substantial investments in information technology systems, and an advanced level of regulatory harmonization. Potential Benefits of Power Integration In theory, international exchange of electricity brings four major advan- tages: (a) trade allows countries to make better use of complementary resources—for example, exchanging hydropower for thermal power when individual countries do not have both resources; (b) international interconnection allows countries to balance variations in annual demand—for example, Ecuador and Colombia have asynchronic rainy seasons, so they can exploit water resources efficiently throughout the year through trading; (c) power trade allows countries to balance genera- tion with current needs, exporting or importing to match their require- ments without incurring emergency power contracts; and (d) international trade allows countries to pool their reserve capacity, thereby reducing costs for extra power stations and limiting the inefficient dispatch of power plants required for the provision of spinning reserves. In a developing-country context, as is the case of the Central American MER, the creation of a regional power system by a group of smaller mar- ket economies can reduce the risks and help the region to match supply and demand more efficiently. The existence of an enlarged power system enhances a project developer’s ability to finance and construct regional power-generating facilities that would be impossible if relying exclusively on the domestic demand of smaller market economies. Consequently, it can make the development of a country’s or a subregion’s capital-intensive power projects more attractive to both domestic and international investors and lenders, reducing risks by creating a broader demand pool 130 Cayo of utilities and off-takers of potential generating facilities. The result would benefit all consumers in the region by lowering prices and improv- ing the quality and safety of the power supply and would eventually result in a lower environmental impact relative to power development (World Bank 2001). In the medium to long term, the main benefits would be (a) lower operating costs from using the most economically favorable energy resources, particularly through the integration and coor- dination of hydropower and thermal systems, which reduce operating costs by generating hydropower in off-peak periods, and (b) lower invest- ment costs in the long term from using integrated planning on a multisys- tem basis, realizing economies of scale, and reducing total reserve requirements (USAID 2008). The benefits arising from cross-border interconnection facilities, once built and put into operation, are derived primarily from the multiplica- tion of energy exchanges among national power utilities. In economic terms, such growth in cross-border energy exchanges should increase until the marginal benefits from displacing more expensive capacity or from making additional sales equal the marginal costs of transmission across the interconnected networks. The same applies to expansion of an interconnection, for which the costs of new generation and transmission must be taken into account (World Bank 2008). However, as Robinson (2009) asserts, much of the literature on the benefits of regional power pools is more advocacy than serious analysis. Establishing a causality relationship between power integration and higher investments based on least-cost projects is problematic in prac- tice. Despite regional generation and transmission optimization, exer- cises show significant gains over the sum of national plans. The problem is that politicians equate energy security with having domestic generation capacity (that is, there is a bias toward national power development plans). Countries remain unwilling to surrender sovereignty to regional bodies or to depend on other countries’ ability or willingness to provide the power needed to supply domestic demand. The recent epsiodes in Europe with regard to the Russian gas supply and the problems between Argentina and Chile with regard to gas contract compliance illustrate the difficulties that energy integration (whether gas or elec- tricity) can entail and their implications for energy security. As a gen- eral proposition, Robinson (2009) states that, even though the benefits from power trade are clear and accountable with regard to electricity exchanges, most of the theoretical benefits from power integration are yet to be demonstrated. Power Integration in Central America: From Hope to Mirage? 131 Power Trade: Contracts and Spot Transactions When two electricity systems are interconnected, energy is transmitted from the low-price country (zone) to the high-price country (zone). In the exporting country, prices increase because additional, more expensive generators are required to dispatch, whereas in the importing zone, prices decrease because expensive plants are no longer required to generate. In equilibrium and assuming infinite transmission capacity, price equalizes between the two zones, creating potential savings in the importing coun- try and potential costs in the exporting country. However, in practice, most of the transmission interconnections are subject to capacity constraints, so there are no conditions for price equal- ization because the demand for exchange constitutes only a relatively small portion of total demand in the import zone system. As a result, the price gap between the two power zones does not vanish. There is, in other words, a price differential between the importing and exporting coun- tries. This price differential, multiplied by the total energy traded, gener- ates “congestion rents.” There has been a lot of debate on how to share these congestion rents between the two zones. In the case of the Colombia-Ecuador power exchanges, the rents go to the export country to compensate for the additional costs incurred by its consumers.8 In other international power exchanges, these rents have been shared equally between both countries. In fact, the lack of agreement about how to share these congestion rents has been the main obstacle to starting operation of the Peru-Ecuador interconnection.9 In the case of Central America, the sharing of congestion rents has not posed an issue because the 50-50 formula has been the rule in the region’s power trade during the past few years. In markets that show high price volatility—such as the power spot markets—it is common for buyers and sellers to smooth out their trans- action prices through long-term contracts. Under forward power pur- chase contracts, the buyer is obliged to purchase a certain amount of power from the seller under a predefined price and for a sufficiently long- term tenure (typically, 10 to 15 years). International experience has shown that the main supply-side benefit of industry restructuring is the competitive procurement of long-term power purchase contracts that have sufficient magnitude and duration to allow suppliers to fund the construction of new generation facilities. These power purchase agree- ments are very important to making the construction of new power facil- ities “bankable.” This fact is of paramount importance when discussing the feasibility of regional power plants in Central America.10 132 Cayo The Political Economy of Integration National policies toward energy security have a significant impact on how countries approach power integration. The perception of whether regional energy trade is seen as contributing to energy security through diversification and cost reduction or, on the contrary, as reducing energy security through the creation of dependencies and disruption risks is crit- ical to the way that power trade develops. If the latter view is true, coun- tries would need to overcome the inclination to equate self-sufficiency (and the bias toward national expansion plans) with energy security, at the expense of trade. The key point here is to recognize that the trade-off between integration and sovereignty is real. For starters, power integration means that governments are not allowed to do a few things anymore, such as (a) prioritize national con- sumers versus foreigners, (b) cut supply under conditions of domestic scarcity when exchanges occur under long-term contracts, or (c) manip- ulate domestic power prices, as this could lead to the awkward situation of having to export even in scarcity conditions (under spot exchange arrangements). Most integration regulatory frameworks establish the nondiscrimination principle, under which every consumer must be treated equally without considerations of nationality. Therefore, in case of a sup- ply disruption or power scarcity, the export country should not be allowed to cut the supply to consumers across the border just because they are not nationals. Accordingly, integration regulatory frameworks contain rules about how to cut supply in case of abrupt shortfalls; these rules generally obligate rationing based on the categories of consumers (for example, first industry, second commerce, then public buildings, and finally domestic consumers), but not by nationality. Long-term contracts to sell electricity to consumers in other countries are usually based on the principle of “firm power,” meaning that contracts are not interruptible or opportunistic, but predictable, secure, and reli- able. Under these circumstances, the export country government cannot force the domestic power generator to relinquish its obligation or to use the allocated energy for another purpose. The regulatory framework or the supranational regulatory body (if one exists) should be strict enough to severely punish this kind of behavior.11 Prices are important for electricity, as with any commodity or traded good. Government manipulation of domestic prices (spot or retail) can cause severe distortions not only in the domestic market but also in trade. Suppose that two countries have wholesale (spot market) Power Integration in Central America: From Hope to Mirage? 133 transactions—like Ecuador and Colombia—but country A puts price caps on spot prices to favor domestic consumers, while country B allows prices to reflect the actual costs of service. It may happen that, because of this arti- ficially low price, demand in A may expand too much, creating scarcity and power shortfalls. But if both countries trade power based on their spot market price differential, country A may be left in the awkward position of having to export even in circumstances of scarcity, which is equivalent to an export subsidy. Integration also entails renouncing some degree of sovereignty and giving power to supranational institutions (regulatory bodies and sys- tem operators). In the case of Central America, the regional electricity market has a supranational regulatory body (Comisión Regional de Interconexión Eléctrica—CRIE) and a regional system operator (ente operador regional—EOR) that have governing power over the regional transmission network and the international power trade. Finally, integra- tion means that governments must postpone or abandon national plans to increase their domestic power supply and instead rely on trade for energy sufficiency purposes, based on least-cost alternatives under a regional or subregional perspective. Domestic consumers in the export country could be opposed to export because domestic prices will increase, reflecting the additional demand coming from abroad. Therefore, export countries need to gener- ate the mechanisms to compensate or alleviate the additional costs incurred by domestic consumers; otherwise, internal discontent may jeopardize the power integration efforts. Suppliers in import countries could have concerns that power trade will reduce their profits. In small importing markets with few participants, suppliers could exert pressure on politicians to limit power trade and avoid profit cutbacks. Energy integration—and power integration in particular—is a complex process whose success will ultimately depend on four types of conditions: • Political. Governments need to take bold actions to reduce their sover- eign power to some degree (depending on the model of integration chosen) and to create the institutional safeguards to avoid relinquish- ing their commitments. • Institutional. Strong institutions have to be put in place to manage the complexities of power trade, including coordinating national dis- patches, operating the common infrastructure, solving commercial disputes, and overseeing competitive behavior. 134 Cayo • Regulatory. Regulatory frameworks need to be harmonized among all the participants or be created ad hoc to make the power transactions possible. This is a complex and difficult process, and regulations in each country have to compromise to facilitate international exchanges. • Physical. There has to be enough actual generation and transmission capacity. Power Integration in Central America: The SIEPAC Project Central America is a region with a long tradition of integration.12 Integration initiatives in the region are channeled through the Central American Integration System (Sistema de la Integración Centroamericana—SICA), created in 1991, which manages different organizations under it. In the energy sector, two regional organizations are part of SICA: the Central America Electrification Committee (Comité de Electrificación de América Central—CEAC) and the Central America Hydrocarbons Cooperation Committee (Comité de Cooperación de Hidrocarburos de América Central—CCHAC), which were organized more than 15 years ago. Since its creation, CEAC’s prominence in the electrical integration of the region has progressively grown. It comprises representatives from the energy authorities of the different countries, and it has provided a forum for supporting initiatives such as the regional power market, the SIEPAC project, and the interconnections with Mexico and Colombia. In 1996, the six Central American countries agreed to create MER, the regional electricity market. The Framework Treaty for the Central American Electricity Market was ratified by the governments in 1998. To support the MER, the treaty created the regional regulatory commission (CRIE), the regional system operator (EOR), and the company owner of the grid (empresa propietaria de la red—EPR). The SIEPAC project con- sists of two interdependent subprojects13: • The development of a regional electricity market based on a standard set of trading rules at the regional (supranational) level. Part of the MER initiative is the creation of a regional institutional structure, including a regional regulator and a regional transmission operator. • The development and completion of a new 1,800-kilometer interna- tional transmission line, running from Panama in the south to Guatemala in the north, that will expand transfer capacity at all bor- ders in the region to 300 megawatts.14 Power Integration in Central America: From Hope to Mirage? 135 The SIEPAC project is an initiative to create an integrated regional electricity market among the six Central American countries. The stated objectives are to (a) improve security of supply by widening reserve mar- gins, (b) reduce the problem of electricity rationing in countries with capacity deficits, (c) improve operating efficiency and reduce fuel con- sumption, (d) introduce greater competition into the domestic markets, (e) lower end-user electricity costs, (f) attract foreign investment to the region’s energy sector, and (g) contribute to the economic development of the region. The project costs approximately US$405 million, financed primarily by the Inter-American Development Bank (US$240 million), the Central American Bank for Economic Integration (US$100 million), and equity contributions from the nine shareholders of the EPR (US$50 million): the six Central American countries, ENDESA (Empresa Nacional de Electricidad) of Spain, ISA (Interconexión Eléctrica) of Colombia, and CFE (Comisión Federal de Electricidad) of Mexico.15 SIEPAC is part of a broader regional initiative under the Mesoamerica project (formerly known as Plan Puebla-Panama). The Mesoamerica proj- ect aims to develop and integrate energy, communications, and transport infrastructure across nine countries, including the six SIEPAC countries plus Mexico, Belize, and Colombia. The Plan Puebla-Panama was pro- posed in 2001 and formally institutionalized in 2004. The Central America Power Sector in a Nutshell The region generated around 38 terawatt-hours in 2007, equivalent to about 70 percent of the annual electricity supply of a medium-size coun- try in Latin America, such as Chile or Colombia. Generation as a whole has grown at a rate of about 6 percent a year since 1990. Generation capacity is on the order of 9,700 megawatts, again similar to 70 percent in Colombia or Chile. The composition of installed capacity varies widely among countries (for example, 70 percent hydropower in Costa Rica and only 13 percent in Nicaragua), mainly as a result of institutional develop- ments that took place in the middle and late 1990s. In those years, several countries implemented vertical unbundling, and only two (Costa Rica and Honduras) retained a vertically integrated state-owned monopoly. However, all countries allowed the entry of private sector enterprises to different degrees, either by selling assets or by purchasing power from new companies via the state-owned utility. Restructuring of the Central American national power sectors has yielded differing sector structures. In the 1990s, the countries approved 136 Cayo new laws and regulations that initiated restructuring processes in their power sectors. Those reforms aimed to promote private participation in a sector that had traditionally been controlled by fully integrated state- owned companies. Reforms in Costa Rica and Honduras were limited to opening the generation segment to private participation. However, signif- icant reforms to liberalize electricity markets were implemented in Guatemala, El Salvador, Nicaragua, and Panama. These countries imple- mented vertical and horizontal unbundling of generation, transmission, and distribution activities, creating specialized companies in the electric- ity sector, as well as permitting retail competition for large consumers. In general, the role of the state was restricted—totally or partially—to the formulation of policies, the promotion of regulatory functions, and the administration of concessions. In all cases, economic dispatch was central- ized and based on audited variable costs (except in El Salvador, where it was based on prices, but is now changing to variable costs). The participation of new private generation enterprises has had both positive and negative consequences. Private investors installed thermal plants, which required less capital and could eventually be moved out of the country if necessary because of system size and lower perceived risk in comparison to renewables. Thermal power was also the least-cost option during the late 1990s due to the high efficiencies associated with heavy fuel oil and diesel plants and the prevalence of low oil prices. In fact, some of the initial investors (in Guatemala) chose to install mobile barge-mounted plants. Private sector investments provided much needed relief to former public sector companies with little access to capital. However, it also made the region increasingly dependent on oil products and on the volatility of the oil market, which resulted in extreme finan- cial consequences in 2006–08, when the price of oil skyrocketed. The share of oil-based power generation grew from almost 0 in 1990 to more than 30 percent in 2007, and several countries have hydrocarbon shares in excess of 50 percent (see figure 5.2). Until 1990, Central American countries used their considerable hydrological resources to generate most of their electricity, and renewables represented 91 percent of the power generated. With demand for electricity growing rapidly, capacity grew from 4,009 megawatts in 1990 to 9,486 megawatts in 2007. However, over this period of almost 20 years, twice as much new gen- eration capacity relying on fossil fuels was built than capacity derived from renewable resources. Indicators point to difficulties associated with meeting the growing demand for electricity in Central America. The precarious balance of Power Integration in Central America: From Hope to Mirage? 137 Figure 5.2 Central American Power Generation, by Type of Power, 1985–2007 100 14 90 12 24 28 80 11 29 32 31 31 30 30 70 10 8 share of mix (%) 60 8 8 8 50 8 8 86 40 76 30 59 57 49 49 46 49 49 49 46 20 10 0 1985 1990 1995 2000 2001 2002 2003 2004 2005 2006 2007 hydrocarbon geothermal wind cogeneration steam diesel gas turbine coal Source: ECLAC 2007. supply and demand is a common threat to all nations in the subregion. Nicaragua experienced severe blackouts in 2006–07, and Costa Rica saw shortages in 2007. As a consequence of this deterioration, power trade among Central American countries has been declining over the past decade. In the past, electricity trade in the Central America region was limited mostly to bilateral transactions in the spot market. As per the current MER regulations, all the contracts are “nonfirm” and must comply with the national legal and regulatory framework. Consequently, the regional market contracts are for import or export of electricity between agents represented by their respective national operators. The main objective of these transactions is to take advantage of energy surpluses and differ- ences in marginal generation costs. As shown in figure 5.3, trade was active in the early 2000s, although restricted by the capacity of exist- ing transmission links. However, trade has dwindled in recent years due mainly to a tight supply-demand balance as most of the countries in the region try first to meet their own internal supply-demand needs. 138 Cayo Figure 5.3 Electricity Exports, by Country, 1985–2008 1,600 1,400 exports (gigawatt-hours) 1,200 1,000 800 600 400 200 0 1985 1990 1995 2000 2002 2003 2004 2005 2006 2007 2008 Costa Rica El Salvador Guatemala Honduras Nicaragua Panama Source: ECLAC 2008. Less than 300 gigawatt-hours of electricity is currently exported collec- tively between the six countries. Currently, Central America faces a series of important challenges in the energy sector: (a) a tight balance between power supply and demand, which casts doubt on the security and reliability of the region’s power sector and raises concern regarding the quality of supply; (b) significant exposure to oil price volatility and shocks due to the overdependence on oil imports, which have increased significantly for power generation pur- poses; (c) significant inefficiencies in the institutional and regulatory framework of several countries, which affect the financial sustainability of power utilities and their operations; and (d) relatively low levels of access in certain countries, in particular in rural areas. The creation of an enlarged, well-functioning market would gradu- ally help Central America to address some of the shortcomings of the electricity sector. Countries in the region are expected to benefit from increased security and reliability of electricity supply due to the enhanced interconnection. An improved investment environment that facilitates the financing of larger projects (for example, regional plants) is expected to flourish.16 Savings from lower operating and investment costs will be realized in the medium to long term, as the regional market con- solidates and eventually evolves into more advanced pool arrangements. Power Integration in Central America: From Hope to Mirage? 139 According to the regional organization CEAC, SIEPAC would produce savings in operational costs on the order of 4 percent and fuel savings of about 3 percent after 8–10 years based on indicative expansion planning exercises. In addition, preliminary estimates show that SIEPAC would result in 1 million tons of avoided carbon dioxide equivalent a year. MER: An Independent Seventh Market Signed in 1996 and ratified in 1998, the treaty creating MER is based on the principles of competition, gradualism, and reciprocity. The treaty establishes that the regional market will include a spot market, based on regional generation dispatch, and a medium- and long-term contract mar- ket and that the governments will establish adequate conditions for the future development of regional power plants. The treaty established a scheme of protocols for future adjustments and clarifications. The first protocol, agreed to in 1998, consisted of several clarifications of and cor- rections to the text of the treaty. The regional electricity market established in the treaty and developed in the regulations is not an integrated regional electricity market, but a seventh market superimposed on the six national markets. As such, the MER was designed as a “loose pool” arrangement in which dispatch will be coordinated but not centralized as in more sophisticated pool designs (see note 7). The MER has its own rules and operates based on the fol- lowing premises: (a) regional electricity trade can take place in a regional contract market and a spot market; (b) all MER agents with the excep- tion of the transmission companies can purchase and sell electricity freely and have open access to the transmission system; (c) MER generation agents can install power plants in any of the member countries and sell energy at the regional level; (d) the MER is a market with its own rules, independent of the national markets, that makes energy transactions using the regional transmission grid and the national networks. The second protocol, which was agreed to in 2007, includes the fol- lowing relevant adjustments to the MER: (a) all agents of the national markets (that is, generation, transmission, distribution, and commercial- ization companies as well as large consumers) as ratified by the legislation of each country, are MER agents and can participate in regional elec- tricity trading; if a country permits the existence of companies with integrated activities, they must separate their business units and employ independent accounting; (b) national interconnection systems and lines that make possible the regional energy transfers are part of the regional transmission grid, whose availability and use include charges 140 Cayo that encompass variable transmission charges, the toll, and the comple- mentary charge; and (c) the governments will carry out the necessary actions to harmonize the national with the regional regulations, permit- ting the normative coexistence of the regional and national markets. Harmonization of national regulations is expected to happen gradually, allowing for firm energy interchanges in which the contracted energy will be prioritized to supply demand in the country where the buyer is located. Currently, the regulatory frameworks of all electricity markets foresee actions to guarantee local self-sufficiency in electricity supply (the bias toward national power sufficiency, mentioned above). One of the main agreements included in the second MER protocol refers to the gradual harmonization of regulations for the regional market. It is understood that this will allow energy trading by firms in the MER, which, in turn, will facilitate the financing of regional plants. Obstacles to Integration of the Central American Power Sector Political, institutional, regulatory, and physical obstacles exist to integra- tion in the Central American power sector. This section deals with these in turn. Political Increasing prices of electricity in exporting countries and the availability of cheaper electricity in importing countries can spur opposition both from consumers in exporting countries and from existing generators in importing countries. If MER energy trading is included in the national economic dispatches, prices may be higher in electricity-exporting coun- tries, while they would be lower in importing countries. This market rule is needed to guarantee nondiscrimination among the national markets (that is, agents in an exporting national market face the same spot price as for occasionally exported energy). However, this does not favor con- sumers in the exporting countries or generators in the importing coun- tries. It is important for governments to hold strong positions against potential pressure from generators and consumer lobbies, while CRIE’s role is to design appropriate and transparent mechanisms to address the effects of power interchanges in domestic markets, including the defini- tion of congestion charges.17 Costa Rica is a key player in this integration agenda. Not only is it the richest country in the region, but it also has a strong power sector with Power Integration in Central America: From Hope to Mirage? 141 few vulnerabilities as a result of having leveraged its renewable poten- tial (primarily hydropower). However, at a political level Costa Rica lags the rest of the region: it is the only country that still has not ratified the second protocol (which has to be cleared at the congressional level). As it was not ratified before the end of 2009, the MER regulations will not enter into full application during 2010, and this may cause delays in the repayment of construction loans for the line. Honduras, Nicaragua, and Panama are urged to modify the legal frame- work that gives priority to domestic demand in the supply of power. The failure to act may pose a definitive barrier to allowing long-term “firm power” contracts in the MER. The main idea is quite simple: successful integration among the six small countries requires the development of scale-efficient regional plants; regional plants require long-term “firm power” contracts; and to have these kinds of contracts, Central America still has to remove several remaining obstacles. These barriers include the priority given to domestic demand in some national regulations, the short tenure of transmission rights, the incomplete commercial framework, the weak- ness of the regulatory body, and the lack of a standarized power pur- chase agreement, among other issues. Institutional At present, there is still limited capacity and resources at CRIE (the regional regulator), which makes it vulnerable to national interests. Addressing the more substantial harmonization problems and preparing a strategy that takes into account national views and interests require addi- tional analysis. However, there is a lack of technical staff and information technology resources in the CRIE, and the commissioners only meet about four times a year. Under these circumstances, the role of CRIE could become very weak and face the risk of allowing national interests to prevail over regional ones. The need to strengthen CRIE is urgent if it is to prepare an adequate foundation for the initial operations of MER. Regulatory Regulatory harmonization is needed to facilitate market operations and regional long-term firm power contracts between qualified agents. There is currently a lack of harmonization of national and regional regulations at the operational and commercial levels. This issue should be dealt with to implement the MER regulations (in substitution of the transitory reg- ulations) and the appropriate interfaces so that MER regulations can work harmoniously with the corresponding regulations in each country. 142 Cayo To advance the harmonization agenda, CRIE will need to focus on two specific areas: (a) standardization of terms and clauses in long-term regional “firm energy” contracts and (b) institutionalization of regional competitive processes and mechanisms for the consolidation of regional coordinated contracts by multiple agents. Asymmetry in the national markets can lead to a lack of reciprocity in the treatment of market agents, as is the case with the vertically inte- grated national electricity markets prevailing in two Central American countries (Costa Rica and Honduras) and the more open electricity mar- kets already structured in the other four countries (Panama, Nicaragua, El Salvador, and Guatemala). The vertically integrated market structure would not allow regional generators (and national generators in the last four countries) to contract electricity directly, with potential distribution, commercialization, and large consumers in Honduras and Costa Rica, because every power operation in these two countries needs to pass through the state monopoly (National Electric Power Company [ENEE] and Costa Rican Electricity Institution [ICE], respectively). In addition, potential regional generators in these two countries would not have clear rules permitting them access to the national transmission grids. However, both ENEE and ICE will have the opportunity to sell to distribution and commercialization companies and large consumers in Panama, Nicaragua, El Salvador, and Guatemala. To correct this lack of reciprocity, significant time and resources (technical and financial) will be required to imple- ment the necessary market reforms in Costa Rica and Honduras, which will have to develop clear rules for agents other than public utilities to participate in the MER. In most countries, domestic demand is still prioritized in the case of power shortages, which creates an obstacle for firm contracts in the regional market. The regional market was designed to allow all SIEPAC members to benefit by using the surplus of one country to cover deficits in another country, a win-win situation. However, to ensure that all coun- tries benefit equally from the regional interconnection, the priority given to national supply during power shortages will have to be eliminated. Price controls lead to misallocation of resources and can imperil the success of a regional market. During the reform processes in the power sector, the stated objective was to achieve a situation where electricity would respond to market supply and demand signals, avoiding distortions in the wholesale price. Political considerations and influences, however, have affected regulatory decisions, for example, by setting ceilings for market prices. Avoiding the introduction of price controls in the supply Power Integration in Central America: From Hope to Mirage? 143 of electricity to domestic markets is needed to support regional invest- ments based on true marginal costs. This issue can prove particularly problematic in interconnected spot markets during shortages: if prices in the spot market are not allowed to reflect such shortages, a country can be required to export electricity to a higher-cost neighbor in spite of hav- ing no surplus to export. Lack of long-term transmission rights will hinder the signature of long- term contracts. Regional long-term firm energy contracts for the develop- ment of new regional power plants would have to be agreed for periods of 10 to 15 years. The EOR will forecast nodal prices periodically for only two-year horizons, while transmission planning is expected to be done for 10-year horizons. To support the regional long-term firm energy contracts associated with new regional power plants, the MER regulations will have to be adjusted to provide longer terms for transmission rights, and com- prehensive methodologies will have to be developed that allow for clear forecasts of transmission charges. Physical The precarious balance of supply and demand makes Central America vulnerable to an electricity crisis. In general, the system is not reliable due to insufficient generation capacity and insufficient transmission infrastructure. Operating costs are disproportionately high because individual country markets are small. With the sole exception of Costa Rica, which suffered some stress in the last few years but was able to invest in new infrastructure and return to a comfortable equilibrium, the other countries in Central America have important vulnerabilities and insufficient capacity to deal with a succesful integration agenda—at least in the short run: • El Salvador adopted a comprehensive set of reforms in the 1990s, but the power sector is still extremely weak. The country is largely exposed to oil price volatility shocks because 50 percent of its capac- ity is based on fossil fuels–based generation. Moreover, the tight bal- ance between supply and demand in the recent past has compelled El Salvador to develop more diesel-based generation. • Guatemala, the largest economy in Central America, made important reforms in the 1990s to modernize its power sector. However, there is a growing concern about the stability of the system, as power outages have become more frequent over the past three years. In addition, there have been problems meeting peak demand. 144 Cayo • Honduras has an immediate challenge in providing access to modern electricity services, in particular to the poorest population. National electricity coverage reached 71 percent in 2006, but only 44 percent in rural areas, where most of the poor are concentrated. However, inefficiency is very high: the state-owned ENEE had financial losses estimated at more than 2 percent of GDP in 2007, and these are contin- uing to rise. • Nicaragua is the second poorest country in Latin America after Haiti; 46 percent of the population is living below the poverty line. Its power sector has several problems and weaknesses, such as high vulnerability to oil price shocks, large inefficiencies, unrealistic tariffs, poorly tar- geted subsidies, and the worst access index in the region. • Panama has had problems attracting new investments into the power sector in recent years, which has produced a tight supply-demand bal- ance that has put a lot of stress on the system and raised concerns about its reliability. In 2008–09, Panama had to contract emergency generation to meet its power needs. In summary, of the four necessary conditions for success enumer- ated in this chapter, the Central American case has—under the current circumstances—the following assessment: decisive political actions still need to be taken; weak institutions need to be strengthened; the regula- tory agenda is incomplete, with important obstacles still to be addressed; and the physical system is inadequate. Reality or Mirage? The backbone of SIEPAC transmission was completed in 2010 and should be ready to commission and initiate commercial operations by 2011. If one assumes that Costa Rica is able to obtain the congressional approval needed and that there are no technical impediments to trans- porting energy from one extreme of the backbone to the other, the key question becomes, will Central America experience a significant increase in the intraregional power trade? Our educated guess is no, at least in the short term. First, there are no surpluses to export in any country. The tight balance between supply and demand that produced the dwindling trade of the last years has not changed significantly.18 Second, ambiguities in the reg- ulatory framework still need to be clarified, in particular, the transmission tariff. Participants and potential investors need to know how much they Power Integration in Central America: From Hope to Mirage? 145 are going to be charged for using the SIEPAC grid. If the amount of energy to be transported through the line is small (as will probably be the case initially), then unit costs to other potential participants may be pro- hibitively high. Determining the tariffs for line use is fundamental, as are the potential complementary subsidies of the six countries necessary to cover revenue shortfalls from transported energy, because compensation payments will be due to line owners immediately after commissioning. If—as we expect—the initial operation of the SIEPAC line does not translate into significant incremental power trade within the region and countries are simultaneously forced to pay fees to the owner of the line and to repay construction loans, the integration agenda may run into opposition. Given a pessimistic cost-benefit evaluation based on the short-term assessment only—benefits may be minuscule initially—this may discourage advances in the required reforms and completion of the regulatory framework that are necessary to attain the true benefits of the enhanced integrated market. If Central American authorities surren- der to a negative sentiment of wastefulness and ineffectiveness because of a slow initial startup, failure in the integration process will be assured. On the contrary, Central America needs to renew its commitment to integra- tion and to accelerate the integration process by completing the MER reg- ulatory framework, strengthening their national markets, promoting investments in regional plants, and opening a new chapter for mutual coop- eration, trust, and trade. The major challenge faced by the regional market is to exploit the potential offered by the transmission line and the MER regulatory and institutional framework by attracting energy projects of a regional scale. Achieving this goal will be a clear test of the options for long-term suc- cess of the market. For this to happen, the regulatory framework and the regional institutions must demonstrate their credibility to investors. As described above, several important barriers could impede the mate- rialization of the full benefits from SIEPAC. These obstacles include weak national power markets reflected in tight supply-demand balances, regu- latory and commercial barriers that hinder long-term contracting, and even political hurdles that will make reliable subregional power trade dif- ficult to achieve (for example, the “national priority” issue). Addressing these obstacles should be an urgent priority for the Central American power authorities. The interconnections between Mexico and Guatemala and between Colombia and Panama, if integrated with the regional transmission backbone, have the potential to provide enough power to address the 146 Cayo precarious balance of supply and demand affecting all countries in the Central American region. A commonly agreed regional expansion strategy that takes into account potential imports from Mexico and Colombia (in the future) is urgently needed. This would also be a test of the rules of the regional market, which need to be flexible to accommodate an evolv- ing reality and to benefit more fully from the opportunities offered by an enlarged market. In summary, the integration process is a long and difficult road that countries need to transit with one eye on the national agenda (strengthening the domestic market in each country) and the other eye on the common integration agenda (strengthening the supranational institutions and removing the obstacles at both the regulatory and commercial levels). Doing so requires a massive and long-lasting dose of political will. Conclusions Power integration is much more than simply building transmission lines and interconnection facilities. Power integration is not a binary category, but rather a long process with different stages and degrees of complexity. Its success will depend on several conditions: political, institutional, regu- latory, and physical. A hefty dose of political will is decisive because integrating means relinquishing some degree of sovereignty. Hence, inte- gration is a two-sided coin, with enhanced trade on one side and more dependence and less sovereignty on the other. Regional power integration makes a lot of sense in Central America, where the electric power market comprises six small markets. However, power integration is not a panacea, and the long path to this goal is fraught with technical and political hurdles. Energy efficiency, electricity access for the poor, optimal power pricing, efficient incentives, and the development of nontraditional renewable sources, among others, are clear examples of issues that will have to be dealt with at a national level, simultaneously with advancing in the integration agenda. While Central America should pursue its power integration, this process is a long and difficult path. Each country needs to focus on strengthening its domestic market, while simultaneously working to strengthen the supranational institutions and remove obstacles, at both the regulatory and commercial levels. The consolidation of the regional regu- latory and institutional framework and the creation of a strong regional power market will not succeed if they are based on weak, inefficient, and Power Integration in Central America: From Hope to Mirage? 147 vulnerable national power sectors. Achieving a strong integrated market based on weak national markets is a mirage. We do not foresee any significant change in the power trade after the commissioning of the SIEPAC line in 2011. First, there are no surpluses to trade. Second, there are still important issues to resolve in the regula- tory and commercial framework. In the medium term, it is critical to pro- mote the construction of regional power plants based on long-term firm power contracts. Central America needs to remove the remaining obsta- cles to making long-term contracts feasible. The conditions for this are not there yet. A source of concern is, however, that modest initial impacts of the SIEPAC project may generate a negative sentiment toward the integra- tion agenda and discourage the reforms and regulations that are neces- sary to attain the benefits of an enhanced integrated market. The Central American integration process could be in jeopardy if this hap- pens. If, on the contrary, Central America accelerates the reform process and begins a new chapter of its development based on mutual cooper- ation, trust, and trade, SIEPAC may bring important benefits, as expected by its designers. Notes 1. Once produced, electricity travels along the transmission grid according to Kirchhoff’s law—that is, following the path of least resistance—and at the speed of light. 2. The exception is Chile, which reformed its electricity market in 1982, intro- ducing competition in generation even before England’s reform of 1990. 3. In many markets, the system operator is the same as the market administrator. However, some market regulations have preferred to separate both entities. 4. In a “radial” interconnection, the demand physically disconnects from its national system and is treated by the export country as if it were part of its own system. But there is no real interconnection of both systems at the same time. 5. In 2009, Brazil agreed to share electricity with Paraguay according to a fairer formula and allowed Paraguay to sell excess power directly to Brazilian com- panies instead of solely through Eletrobras. 6. These markets are known as “loose pools.” 7. Sometimes power pools are divided into “tight pools,” in which a centralized least-cost merit order dispatch is put in place, and “new pools,” in which dis- patching is not based on costs, but rather on the bid price of each generator (that is, on a competitive basis). 148 Cayo 8. In Colombia, 80 percent of the congestion rents are destined for the Fondo de Energía Social, which finances rural electrification infrastructure, while the remaining 20 percent is used to alleviate the higher prices that Colombian consumers have to pay due to the electricity export. 9. The Peru-Ecuador interconnection was physically finished in 2004, and since then it has been used only in two instances, both in response to emergencies in Ecuador. The sharing of congestion rents has been the main obstacle to reaching commercial accords. Peru pushed for the same treatment as in the Colombian example (100 percent to the export country), while Ecuador insisted on the 50-50 scheme. 10. Wolak (2003) points out that the spread of wholesale forward contract mar- kets throughout the United States during the early 1980s led to investments in new generation capacity. 11. The crisis between Chile and Argentina over natural gas exports is a clear example of how governments can intervene in private contracts and oblige domestic suppliers to relinquish their contracts in situations of growing scarcity. There was no clear regulation or safeguard for this kind of episode or any supranational body. This example highlights the difficulty associated with energy integration in South America because it shows that long-term firm contracts are difficult to enforce; future initiatives would need to have stricter rules of compliance and compensation. 12. This section is based on a World Bank study of the Central America regional electricity market (World Bank 2010). 13. See ECA (2009) for a complete overview of the SIEPAC project. 14. The SIEPAC line capacity is equivalent to only 3 percent of existing regional capacity and less than 5 percent of peak demand. 15. The remaining US$15 million is financed by loans from the Andean Development Corporation. 16. A generation project is considered regional when part of its generation is assigned to cover the demand of another country. A regional plant will have long-term contracts with neighboring countries. A merchant plant that oper- ates exclusively in the spot market (without long-term contracts) eventually will be considered regional generation if the neighboring countries can rely on its supply to balance their supply-demand equation. 17. Incumbent generators in Guatemala have begun to exert opposition to the interconnection with Mexico, as cheaper imported electricity would affect their future profits. 18. However, the interconnection with Mexico could prove critical in overcom- ing the supply-demand imbalances in the region, because Mexico has a large idle generation capacity. Notwithstanding, the Mexico-Guatemala intercon- nection was possible due to a bilateral agreement, and it does not respond to a multilateral arrangement with SIEPAC or any subregional authority. Power Integration in Central America: From Hope to Mirage? 149 References CIER (Comisión de Integración Energética Regional). 2010. Informe Trimestral 1 (March). Montevideo. ECA (Economic Consulting Associates). 2009. “Regional Power Sector Integration: SIEPAC Case Study.” ECA, London. ECLAC (Economic Commission for Latin America and the Caribbean). 2007. Estadísticas subsector eléctrico: Statistical Yearbook 2007. Santiago: ECLAC. ———. 2008. Estadísticas Subsector Eléctrico: Statistical Yearbook 2008. Santiago: ECLAC. Hunt, Sally. 2002. Making Competition Work in Electricity. Hoboken, NJ: John Wiley and Sons. Robinson, Peter. 2009. “International Power Integration: Early Findings from an ESMAP Regional Power Study.” Report for the World Bank by Economic Consulting Associates, London. USAID (U.S. Agency for International Development). 2008. “Sub-Saharan Africa’s Power Pools: Development Framework.” White Paper, USAID, Washington, DC. Wolak, Frank. 2003. “Designing Competitive Electricity Markets for Latin America.” Inter-American Development Bank, Washington, DC. ———. 2004. “Lessons from International Experience with Electricity Market Monitoring.” Policy Research Working Paper 3692, World Bank, Washington, DC. World Bank. 2001. “Regional Electricity Markets Interconnections—Phase I. Identification of Issues for the Development of Regional Power Markets in South America.” Technical Paper, World Bank, Energy Sector Management Assistance Program, Washington, DC. ———. 2008. “Building Regional Power Pools: A Toolkit.” World Bank, Washington, DC. ———. 2010. “Central America Regional Electricity Study.” World Bank, Washington, DC. CHAPTER 6 Supply Chain Analyses of Exports and Imports of Agricultural Products: Case Studies of Costa Rica, Honduras, and Nicaragua Raquel Fernández, Santiago Flórez Gómez, Francisco Estrázulas de Souza, and Henry Vega The signing and initial implementation of the Dominican Republic–Central America Free Trade Agreement (DR-CAFTA) represents an important step toward regional trade integration. However, for member countries to reap the potential benefits of DR-CAFTA, a complementary agenda is needed to establish a comprehensive mix of policy priorities that address key challenges of the region. Among these challenges is the weak logistics performance of Central American nations, which hinders their ability to integrate, not only with each other in the context of DR-CAFTA, but also with the rest of the global economy. Studies on the impact of logistics costs on the final price of deliv- ered goods reveal that (a) these costs represent a greater barrier to trade than import tariffs and (b) their impacts become increasingly rel- evant with the prevalence of free trade agreements such as DR-CAFTA (see Baier and Bergstrand 2001; Hummels 2001; Blyde, Moreira, and Volpe 2008). In fact, the World Bank has estimated that, on average, 151 152 Fernández, Gómez, de Souza, and Vega ad valorem tariffs for food imports declined in the Latin America and Caribbean region from 2005 to 2008 and currently range from 3 to 12 percent of product value (Schwartz, Guasch, and Wilmsmeir 2009; World Bank 2009). Transport and logistics costs, in contrast, measured in this case by the international maritime and road haulage compo- nents alone, can total about 20 percent of the free-on-board value of goods. By the time other costs, such as handling, storage, and distribu- tion costs, are accounted for, logistics costs can add up to more than 50 percent of the final price of the good. Quantifying these costs and understanding what factors affect logistics performance are crucial to pinpointing areas for potential policy action. This chapter is intended to address this need, particularly in the context of DR-CAFTA, by presenting the results of eight supply chain analyses pertaining to agricultural products moving within the Central American region: fresh tomato exports from Costa Rica to Nicaragua and wheat, rice, and corn imports from the United States to Nicaragua and Honduras. The analyses performed in this study intend to shed light on the logistics bottlenecks affecting both intraregional and extraregional trade. In terms of intraregional trade, namely between Costa Rica and Nicaragua, the chapter’s findings suggest that the biggest burdens are (a) high domestic transportation costs and (b) bottlenecks at the region’s border crossings mostly attributed to customs delays, which are particu- larly relevant in the trade of perishable goods. For extraregional trade—more precisely for grain imports from the United States—the supply chain analyses show that the most relevant logistics challenges are (a) high domestic transport costs, (b) bottlenecks at land border crossings that prevent countries from utilizing ports in neighboring countries and then bringing the product in by land, and (c) lack of harmonization of sanitary and phytosanitary standards within DR-CAFTA members. The chapter is structured as follows. First, it explains the rationale for choosing tomato exports from Costa Rica to Nicaragua and wheat, rice, and corn imports from the United States to Nicaragua and Honduras and then presents the methodology of the standardized logistics survey (SLS) and the sources used for each of the products. Second, it presents the supply chain analyses, divided into the themes of road infrastruc- ture, customs, sanitary and phytosanitary controls, and port use opti- mization and other port of entry issues. Third, this discussion is followed by specific cost breakdowns for each of the products. A final section concludes. Supply Chain Analyses of Exports and Imports of Agricultural Products 153 Case Studies of Agricultural Trade This section explains the rationale for choosing to study tomato exports from Costa Rica into Nicaragua and wheat, rice, and corn imports from the United States into Nicaragua and Honduras. Intraregional Trade: Fresh Tomatoes from Costa Rica to Nicaragua While the United States is the region’s major trade partner with respect to the trade of agricultural products, Central American countries them- selves also represent increasingly important trade partners for the region. In 2008, agricultural exports from Central American countries totaled US$9.8 billion, out of which 35 percent were exported to the United States, 23 percent to the European Union, and 21 percent to Central American countries.1 Several facts are relevant to the decision to study trade between Costa Rica and Nicaragua: (a) Costa Rica is a high-performing country according to the 2010 logistics performance index (LPI); (b) out of all of Costa Rica’s DR-CAFTA trade partners, the highest percentage of total exports goes to Nicaragua2 due in great part to its geographic proximity; and (c) Nicaragua is the lowest-performing country, as ranked in the LPI. For these reasons, the supply chain analysis evalu- ates the logistics challenges faced by a higher-performing country exporting to a lower-performing country. Furthermore, given that it is common for neighboring countries to trade agricultural goods throughout the year, the supply chain analysis takes tomatoes, the most important vegetable exported to Nicaragua in value terms, as the object of study. Moreover, since tomatoes are perish- able goods, the exercise allows for an analysis of the critical bottlenecks faced in a supply chain requiring expedited and efficient delivery as well as transport in refrigerated containers. International Trade: Rice, Wheat, and Corn from the United States into Nicaragua and Honduras Rice, wheat, and corn are major agricultural commodities imported into DR-CAFTA member countries and represent important components of Central America’s food basket. These commodities are imported mostly from the United States, through the port of New Orleans in Louisiana and into Puerto Corinto on the Pacific coast of Nicaragua and Puerto Cortés on the Atlantic coast of Honduras. 154 Fernández, Gómez, de Souza, and Vega The 2009 World Integrated Trade Solution (WITS) trade data for Honduras and 2007 WITS data for Nicaragua show the following: • 88 percent of the US$60 million of rice imported into Honduras and 97 percent of the US$58 million imported into Nicaragua come from the United States. • 100 percent of the US$53 million of wheat imported into Honduras and 99.9 percent of the US$35 million come from the United States. • 99 percent of the US$90 million of corn imported into Honduras and 98 percent of the US$29 million into Nicaragua come from the United States. Methodology and Sources For the supply chain analysis, we conducted a standardized logistics sur- vey that collected cost data through primary interviews with several actors along the supply chain. These data were complemented by publicly available trade data compiled by government entities. Standardized Logistics Survey (SLS) The SLS was the primary tool used to collect data and was critical in ensuring the consistency and quality of data obtained and allowing us to compare the different supply chains described in this chapter. The SLS dictated the cost components to be compiled during field inter- views, including farm gate prices, maritime and domestic transport costs, warehousing and storage costs, retail prices, and so forth. Other costs, such as milling costs for the grain supply chains as well as time costs for the fresh tomato supply chains, were exclusive to these particular products. Information was collected through primary field interviews with key actors within the supply chains in Costa Rica and Nicaragua for tomato exports and the supply chains in Nicaragua and Honduras for rice, wheat, and corn imports. These interviews provided information that allowed us to identify the costs involved in the various steps of the supply chain, as well as qualitative information that allowed us to assess these countries’ logistics challenges.3 The interviews followed a conversation format and were guided by the SLS template. Interviewees included independent producers and producer associa- tions, exporters and importers, freight forwarders, customs agents and government agencies, retailers, and millers for the analysis of grain Supply Chain Analyses of Exports and Imports of Agricultural Products 155 imports.4 Producers provided data on farm prices and domestic ship- ping costs; exporters on shipping and handling, profit margins, and cus- toms costs and procedures; importers on ocean shipping and port reception costs as well as profit margins; millers on operating costs and extraction losses; freight forwarders on transport and customs costs; customs agents on disaggregated customs costs; and retailers on final retail prices, profits, and distribution costs. The SLS methodology allowed us to harmonize the data collected. By doing so, it enabled us to compare the supply chains of particular products. Furthermore, the methodology calls for a standard unit of measurement, in this case U.S. dollars per kilogram. However, while the SLS calls for the collection of data for similar logistics components (that is, shipping, handling, and customs costs), supply chain analyses for different products are not completely comparable due to their unique characteristics and the nature of their logistics requirements. It would be inaccurate, for example, to compare the costs involved in the movement of perishable tomatoes in refrigerated containers directly with the costs involved in the movement of grains transported in bulk via ocean shipping. Aside from this challenge, the SLS also has limitations. Since the cost data are gathered primarily through conversational interviews, data points must be carefully understood and verified with as many reporters as pos- sible. For example, in reviewing domestic transport costs for the tomato supply chain analysis, it was understood and confirmed that, as stated by both exporters and cargo carriers, these costs include fuel and labor costs, but do not include vehicle amortization or depreciation. Finally, concerning the representativeness of the analyses presented in this chapter, it should be noted that the large tomato exporters and grain importers chosen for the study dominate their respective mar- kets, so that the chains are representative of the logistics costs involved in the trade of these goods. The challenge lies in generalizing the con- clusions to the rest of the Central American economies, as each coun- try has its own market structures and areas of improvement in matters of logistics performance. Therefore, the supply chain analyses provide a snapshot of intraregional and extraregional trade for Costa Rica, Honduras, and Nicaragua, and an attempt is made to analyze how these results fit into country-level data provided by other efforts, like the Doing Business initiative from the World Bank, the Global Competitiveness Report from the World Economic Forum, and the logis- tics performance index from the World Bank. 156 Fernández, Gómez, de Souza, and Vega Complementary Data and Information In addition to data gathered during field visits in Costa Rica, Honduras, and Nicaragua, information was collected from complementary online data sources. These included trade statistics databases provided by the U.S. Department of Agriculture, the U.S. International Trade Commission, the World Integrated Trade Solutions, and national statistics institutes (in-country). Supply Chain Analysis: Intraregional and Extraregional Trade In the trade of fresh tomatoes from Costa Rica to Nicaragua, the small exporter receives the fresh product in 18-kilogram plastic boxes from a small producer in Cartago, located about 36 kilometers from San José, who transports it in a small, unrefrigerated truck to the exporter’s distri- bution center located just 4 kilometers away from the farm gate. Once at the distribution center, the product is transferred to 23-kilogram card- board boxes5 and loaded into a 40-foot container, which then travels toward Peñas Blancas, the border town between Costa Rica and Nicaragua, located about 319 kilometers from the distribution center. The truck travels up the Pan-American Highway at a speed of 60 to 80 kilometers an hour and arrives in approximately seven hours, traveling nonstop. Once at the border, the shipment passes through Costa Rican and Nicaraguan customs and travels three to four hours from the bor- der until it arrives at the Mercado Oriental, where the product is sold to both big wholesalers and retailers. Overall, the large exporter’s chain has a structure similar to that of the small exporter. The large exporter purchases products all year round from a large independent producer who controls market prices due to its overwhelming share of the country’s tomato production. The large producer then transports the product in a 20-foot truck with a capacity of up to 700 boxes of 13 kilograms each to the large exporter’s distri- bution center, located approximately 60 kilometers away. Once at the distribution center, the boxes are loaded into a 45-foot container that can carry up to approximately 1,200 boxes. After the container is fully loaded, the truck travels toward Peñas Blancas, crosses the border, and arrives at the distribution center in Managua, located 149 kilometers from the border. Finally, the product is consolidated with other goods at the distribution center and transported in refrigerated trucks that can carry up to 6,800 kilograms to different supermarket points in Managua.6 Supply Chain Analyses of Exports and Imports of Agricultural Products 157 In the case of grain imports from the United States, the supply chains analyzed in this study assume that the grains originate at the farm gate in the states of North Dakota, Arkansas, and Minnesota, respectively. For all cases, the grain is loaded onto vessels at the port of New Orleans and transported into Puerto Corinto in Nicaragua and Puerto Cortés in Honduras. To reach the port of New Orleans, the grain travels by truck or rail, either directly to New Orleans or on barges. The maritime costs incurred for transportation to Puerto Cortés include freight and insurance for a three-day trip along the Atlantic Ocean, while ocean transportation to Puerto Corinto includes freight, insurance, and Panama Canal charges for a trip that takes 12 days and ends on Nicaragua’s Pacific coast. Once in Honduras, the grain is transported via truck for about 40 to 100 kilometers to the mills in or around San Pedro Sula or to animal feed plants located throughout Honduras. In Nicaragua the grain is transported to mills in Managua or Granada or to animal feed plants in other areas, about 150 to 300 kilometers from Puerto Corinto. Aside from maritime shipping costs, other costs in the analysis include customs clearance costs, which can be disaggregated into unloading expenses, sanitation costs, document processing—that is, cus- toms broker charges—and inventory costs incurred in port or at the border. Milling costs can be disaggregated into extraction losses, unload- ing and loading, other fixed and variable costs, as well as the profits earned by the millers. The wholesale and retail costs are disaggregated into costs and profits. Main Logistics Challenges The information collected through the on-site interviews and secondary sources provides insight into four specific areas of concern in the countries studied: customs, road infrastructure, sanitary and phytosanitary controls, and optimization of port use and other port of entry issues. Customs According to the interviewees, specifically with respect to the trade of fresh tomatoes, time spent on the Costa Rican side of the border averages three hours for both large and small exporters, assuming that shipments have to go through customs, phytosanitary, and narcotics inspections. On the Nicaraguan side of the border, waiting time equals five hours for the same inspections.7 158 Fernández, Gómez, de Souza, and Vega Apart from waiting times at the border, interviewees also cited the lack of technical skills of customs agents and customs personnel at the border as well as the lack of adequate physical infrastructure for carrying out inspections. With respect to the first challenge, interviewees mentioned that customs agents are often misinformed about customs procedures and the documentation required for both imports and exports, which can cre- ate delays in the processing of cargo movements. With respect to the lack of adequate physical infrastructure, tomato exporters highlighted the fact that when customs calls for a thorough inspection of shipments,8 the cargo must be unloaded in open air areas that expose perishable products to warm temperatures and outside conditions. Additionally, in the border crossing between Nicaragua and Costa Rica, inconsistencies in the working hours of customs agents and customs offices seem to be relevant in facilitating or deterring the trade of perish- able products. Interviewees reported that customs agents often shut down operations after 5 p.m. and on weekends. Therefore, the paperwork and procedures related to shipments that arrive at the border on Saturday or Sunday get postponed to Monday, causing significant delays particularly on that day. Furthermore, a large importer on the Nicaraguan side said that if there are delays on the Costa Rican side of the border for ship- ments arriving in the afternoon, trucks must wait along the 1-kilometer road between Costa Rican customs and Nicaraguan customs because Nicaraguan customs has limited working hours. This implies extra fuel costs, as the containers must be kept refrigerated, extra hiring costs for the driver, delays in distributing the product to the retail point, underutiliza- tion of distribution trucks, associated costs for extra fleets, reduced shelf time for the product, and the exporter’s loss of reputation, among others. Moreover, interviewees also mentioned that the electronic system, TICA, which was implemented at Peñas Blancas starting in 2006, shuts down constantly, causing significant delays at both the Costa Rican and Nicaraguan sides of the border. This is in great part due to deficiencies in telecommunications services.9 Another common challenge with respect to customs in Central American countries is corruption. Direct exporters, freight forwarders, and customs agents all said that bribes are a necessary evil to expedite the customs clearance of perishable products and that the cost of bribes is included in the cost of the services provided. Furthermore, bribes to transit police are particularly common during highly congested periods, when shippers pay extra to bypass the long lines of trucks that form at the border. Supply Chain Analyses of Exports and Imports of Agricultural Products 159 Road Infrastructure Although not cited as frequently as bottlenecks at customs, several inter- viewees mentioned deficient road infrastructure at the country level. Most of these interviewees mentioned that, although the Pan-American Highway is in relatively good condition and trucks can travel at accept- able speeds of up to 80 kilometers an hour, secondary roads, especially those connecting farms to cities, are often unpaved. These findings are consistent with empirical studies conducted at the country level. The World Bank has estimated that about 42 percent of Costa Rican firms, for example, have specifically identified road quality as a major or severe obstacle to growth (World Bank 2007). Additionally, the poor and worsening quality of the country’s road net- work has been shown to cause direct losses from delays in shipments and breakage equal to 8 to 12 percent of the sales value of exported goods (World Bank 2006). In Nicaragua, 75 percent of the total road network was considered to be in poor condition in 2006, and nearly 50 percent of the network becomes unusable during the rainy season (see chapter 7 of this volume). Sanitary and Phytosanitary Regulations The findings of the research performed for this study suggest that sani- tary and phytosanitary regulations can restrict the free flow of trade of perishable as well as nonperishable agricultural goods. Perishable prod- ucts must undergo phytosanitary inspections at both the Costa Rican and the Nicaraguan sides of the border. As interviewees explained, on the Costa Rican side, the Ministry of Agriculture and Livestock obtains a sam- ple for each of the products being shipped and clears the cargo. On the Nicaraguan side, phytosanitary inspections are more thorough, as the products are being imported into the country. The Nicaraguan Ministry of Agriculture and Forestry (MAGFOR) obtains a laboratory sample for each of the imported products and sends it to the nearby town of Rivas, where the sample is analyzed.10 For tomatoes specifically, the shipper does not have to wait for the laboratory results and can continue on its way to Managua. Should laboratory results confirm the existence of harmful agents in the product, MAGFOR sends personnel to track down the shipment and prevent importers from selling the product in the local market. With respect to import procedures for nonperishable products such as grains, importers interviewed in Honduras and Nicaragua expressed dis- satisfaction with the way in which phytosanitary controls are managed in 160 Fernández, Gómez, de Souza, and Vega the region. They complained that even though the product has been sanitized and certified in the United States, OIRSA (Regional International Organization for Farming and Livestock Sanitation), an intergovernmental organization in charge of integrating sanitary regula- tions in the region, often does not accept U.S. certifications and forces importers to sanitize their products again. The sanitation process (fumi- gation) can take place on board the vessel, at the port, or at the mills and represents substantial additional costs, delays, and operational uncertainties. These additional costs include, for example, the purchase of the fumigant (US$11.25 per metric ton), delays in the unloading process and in the time that the vessel must remain in port,11 and unpredictability in the timing of shipments, which may result in higher storage costs for the mills.12 Port Use Optimization and Other Port of Entry Issues To understand bottlenecks at the port level, we examine the supply chain for wheat, rice, and corn being imported into Puerto Corinto on the Atlantic coast of Nicaragua and into Puerto Cortés on the Pacific coast of Honduras. Puerto Corinto, located 160 kilometers from Managua, han- dles all grain imports into Nicaragua. In 2008, it handled 1,919 metric tons of cargo, a significantly lower tonnage than that handled by other ports on the Pacific coast such as Balboa in Panama and Acajutla in El Salvador. Nevertheless, Puerto Cortés is the most important port in Honduras and the fourth busiest port in Central America in terms of vol- ume handled. Consistently throughout the interview process, Honduran grain importers indicated that the trip from New Orleans to Puerto Cortés takes an average of three days. The trip to Puerto Corinto in Nicaragua, which passes through the Panama Canal, can take up to 12 days. Since grains are not highly perishable goods, extra days in transit do not add up to higher costs. Nevertheless, as the level of trade—for grains and other products alike—continues to rise with the implementation of DR-CAFTA, the potential exists for efficiency gains due to economies of scale and shorter transit times. A possible solution to facilitate trade into Nicaragua would be to bring grains to Puerto Corinto and then transport them by land into Nicaragua through the border post known as El Guasaule. According to intervie- wees, some importers have used this route, but not very often. The expe- rience of current importers of milled rice, who transport the product in bulk vessels through this path, sheds light on the numerous difficulties of Supply Chain Analyses of Exports and Imports of Agricultural Products 161 the operation. The following are the most common complaints raised by milled rice importers: • Too much time needed for customs procedures. Maritime companies take too long in delivering the bill of lading,13 the document needed to initiate customs procedures in Nicaragua. Therefore, the product reaches El Guasaule within a period of three days, but importers can- not initiate the import authorization procedures with MAGFOR and the customs agency (Dirección General de Aduanas) because they do not have the bill of lading. Upon receipt of the bill of lading, clearance procedures take up to five working days. • Lack of influence by Nicaraguan customs brokers at Puerto Cortés. Brokers are located in Managua, and completing the documentation and proce- dures required for imports can take between five and seven days. The current low volume of products imported through Cortés on their way to Nicaragua does not provide sufficiently strong incentives for brokers to open subsidiary offices in Puerto Cortés. • Lengthy phytosanitary controls at El Guasaule. Upon arriving at El Guasaule, the grain samples have to be shipped to Managua to be inspected at the Universidad Centro-Americana and then sent back to El Guasaule, which results in a two-day delay of shipments at the border. The costs of having the product sit at El Guasaule average US$110 per day per truck. • Different documentation required at El Guasaule and Puerto Corinto. More documents are needed to clear customs in El Guasaule than in Puerto Corinto. The challenges faced by importers that move milled rice into Nicaragua via Honduras provide valuable insight into why countries are unable to take advantage of neighboring ports, such as Puerto Cortés, and create more efficient supply chains. Given that Nicaragua does not have a port with the capacity to handle bulk vessels on the Atlantic side, bulk shipments are forced to travel southbound and through the Panama Canal, incurring additional charges of US$3.59 per metric ton of prod- uct.14 To reach Managua, the country’s major consumption center located closer to the Pacific side, shipments do not travel through Nicaragua itself or through any other Central American country. In theory, if barriers to regional integration were eliminated, importers could use not only Puerto Cortés in Honduras, but also Puerto Limón in Costa Rica and then trans- port their shipments overland to Nicaragua. 162 Fernández, Gómez, de Souza, and Vega If the challenges described in this chapter are illustrative of trade pat- terns not only between Honduras and Nicaragua, but also more broadly between Central American nations, countries are unable to take advan- tage of alternative routes partly because of logistics bottlenecks, including the lack of coordination of import processes and procedures between countries as well as inefficiencies and delays at border points. Quantitative Results for the Fresh Tomatoes Supply Chain The weight of logistics costs within a firm’s cost structure has been shown to be sensitive to the size of the firm. In Latin America, total logistics costs are almost three times as high for the smallest firms in the sample, with a volume of less than US$5 million in sales, as for the largest firms, with more than US$500 million in yearly sales. In an effort to explore the true drivers of the higher costs faced by smaller exporting firms, this section presents the quantitative results of two supply chain analyses of tomato exports from Costa Rica to Nicaragua, one for a small exporter and one for a large exporter.15 Price Breakdown The cost breakdown for the price of tomatoes as they move from the farm gate until they arrive at the final retail point in Managua reveals differences between small and large exporters. As shown in figure 6.1, for the small exporter, the largest component of trade and logistics costs is transport (23 percent), followed by customs (11 percent) and duties (6 percent). Outside of trade and logistics costs, the farm gate price represents the largest component of costs as a percentage of the final price of tomatoes, sold at US$1.86 per kilo at the final retail point (see table 6.1). The cost breakdown for the large exporter displays similar results, with customs (10 percent) and transport (6 percent) the two most important components of trade and logistics costs. The component “other” costs, which mostly refers to profits and margins and to administrative and other retail costs, occupies third place because (a) large supermarket chains have higher operational costs than an independent retailer with a stand at an open-air market, as in the small exporter’s chain, and because (b) due to lower trade and logistics costs, mainly in transport, large super- market chains make higher profits. Additionally, with respect to the farm gate link of the chain, the producer receives 24 percent of the final price of the good, 7 percent less than the small exporter receives (see figure 6.2 and table 6.2). 163 US$ per kilogram Figure 6.1 0 0.20 0.40 0.60 0.80 1.00 1.20 1.40 tra ns fa po rm rt pr ga fro od te m uc Source: Authors’ calculations. fa in er rm te rm pr ga ed of te ia it to ry tra ha di m ns st ar nd gi po lin rib ut n rt g io fro at n m di ce di st nt st rib er cu rib ut ut io st io n om n ce sa ce nt ge nt er nc er ys to er bo vi rd cu ce er farm gate st sC om os st ta im Ri cu e ca st Co om st aR sa others ica Breakdown of Costs for a Small Exporter of Tomatoes ge nc tra ys ns er du po vi ce ty rt cu sN fro st m om ica ra Note: The lines on each bar refer to the addition to costs from each category listed on the horizontal axis. tra bo st gu transport ns rd im a po er e rt to Ni fro ha di ca m nd st ra rib gu di lin a st rib g ut io at n duties ut di ce io st n rib nt er ce ut nt io er n to ce op nt en er ai customs rm ha ar ke nd t lin g lo ss es st or ag e to ta lc os ts 9% 6% 31% 23% 11% 164 Fernández, Gómez, de Souza, and Vega Table 6.1 Breakdown of Costs for a Small Exporter of Tomatoes Additional Tomatoes as a cost (US$ per Cumulative % of cumula- % of each Supply chain element kilogram) cost tive cost element Farm gate 0.47 n.a. 100.0 31.2 Producer profit 0.05 0.52 90.0 3.5 Intermediary margin 0.08 0.60 78.1 5.3 Transport from farm gate to distribution center 0.01 0.61 76.8 0.7 Handling at distribution center 0.00 0.61 76.3 0.3 Transport from distribution center to border 0.14 0.75 62.2 9.3 Customs agency services, Costa Rica 0.02 0.77 60.6 1.3 Customs time, Costa Rica 0.06 0.83 56.2 4.0 Duty 0.10 0.93 50.4 6.5 Customs agency services, Nicaragua 0.04 0.97 48.3 2.7 Customs time, Nicaragua 0.06 1.03 45.3 4.3 Transport from border to center of distribution 0.08 1.11 42.1 5.2 Handling at distribution center 0.03 1.14 41.0 2.0 Transport from distribution center to open-air market 0.03 1.17 39.9 2.0 Handling losses 0.05 1.23 38.2 3.6 Storage 0.00 1.23 38.1 0.1 Retail cost 0.05 1.27 36.7 3.2 Retail profit 0.23 1.50 31.2 15.0 Source: Authors’ calculations based on interviews. Note: n.a. = Not applicable. Finally, the small exporter has the equivalent of a US$0.275 extra cost per kilogram of tomatoes due to logistical inefficiencies. Transport Costs The transport burden for the small exporter is almost four times (17 percent higher than) the burden for the large exporter. The analysis attributes this difference to the small exporter’s lower carrying capacity, as the small exporter carries 5.7 tons of tomatoes, while the large exporter carries 15.6 tons. Transport costs consist of two components: shipping and handling. Shipping costs refer to the service (fuel and salary costs) of moving a 165 Figure 6.2 tra ns po rt US$ per kilogram Source: Authors’ calculations. fro 0 0.2 0.4 0.6 0.8 1 1.2 1.4 m fa fa rm rm ga ga pr te te od to uc tra ha di er ns nd st pr po rib of rt lin ut it fro g io m at n di ce di st nt st rib er cu rib ut st ut io io n om n ce sa ce nt ge nt er nc er ys to er bo vi rd ce er cu sC st os om ta st Ri farm gate im ca cu e st Co om st sa aR ge ica Breakdown of Costs for a Large Exporter of Tomatoes nc others tra ys ns er du po vi ty ce rt cu sN fro st ica m om ra bo st gu rd im a e Note: The lines on each bar refer to the addition to costs from each category listed on the horizontal axis. transport er to Ni ca di ra st gu rib ut a io n ce duties nt tra er ns in in po ha te su rt nd rm ra fro nc m lin ed e di g ia ry at customs st rib di m st ar ut rib gi io n n ut io ce nt n er ce nt to er su pe rm ak et to ta lc os ts 8% 6% 5% 24% 10% 166 Fernández, Gómez, de Souza, and Vega Table 6.2 Breakdown of Costs for a Large Exporter of Tomatoes Additional Tomatoes as a cost (US$ per Cumulative % of cumula- % of each Supply chain element kilogram) cost tive cost element Farm gate 0.44 100.0 24.3 Producer profit 0.10 0.54 4.7 5.5 Transport from farm gate to distribution center 0.01 0.55 4.6 0.6 Handling at distribution center 0.01 0.56 4.6 0.6 Transport from distribution center to border 0.05 0.61 4.2 2.6 Customs agency services, Costa Rica 0.01 0.62 4.1 0.6 Customs time, Costa Rica 0.07 0.69 3.7 3.7 Duty 0.09 0.78 3.3 5.0 Customs agency services, Nicaragua 0.02 0.80 3.2 1.2 Customs time, Nicaragua 0.08 0.88 2.9 4.2 Transport from border to center of distribution 0.03 0.90 2.8 1.4 Insurance 0.00 0.90 2.8 0.0 Intermediary margin 0.00 0.90 2.8 0.0 Handling at distribution center 0.01 0.91 2.8 0.4 Transport from distribution center to supermarket 0.02 0.92 2.8 0.8 Retail cost 0.60 1.52 1.7 33.0 Retail profit 0.31 1.83 1.4 17.1 Source: Authors’ calculations based on interviews. refrigerated container from point A to point B and do not include the truck’s depreciation or amortization. Furthermore, transport costs are disaggregated into four stages: farm gate to distribution center in Costa Rica, distribution center to Costa Rican customs, Nicaraguan customs to distribution center, and distribution center to final sales point in Nicaragua. The breakdown of shipping costs suggests that small and large exporters pay the same costs to transport their product from the farm gate to the distribution center, with both paying 1 percent of the final price of the good (see figure 6.3). The same holds for the costs of trans- porting the product from the distribution center in Nicaragua to the final retail point: costs for the small exporter represent 2 percent of the final Supply Chain Analyses of Exports and Imports of Agricultural Products 167 Figure 6.3 Breakdown of Transport Costs for Small and Large Exporters a. Small b. Large 14% total handling (1%) 25% total handling (6%) from distribution 12% center to open-air transport 23% from distribution market (1%) 9% center to open-air transport 6% from border to market (2%) 20% distribution center: from border to customs 10% mercado oriental (1%) 23% distribution center: customs 11% mercado oriental (5%) duties 5% from distribution duties 6% 38% center to border (2%) 40% from distribution center to border (9%) 16% from farm gate to from farm gate to distribution center (1%) 3% distribution center (1%) Source: Authors’ calculations based on interviews. Note: Data in parentheses are the share of the component in final price. Other percentages are the share of the component in total transport costs (right-hand bar) or the share of the component in total trade and logistics costs (left-hand bar). price of the good, while costs for the large exporter equal only 1 percent. However, when looking at transport from the distribution center in Costa Rica to the distribution center in Managua, costs for the small exporter are higher. In fact, the transport costs from the distribution center in Costa Rica to the distribution center in Nicaragua are 14 percent of the final price of the good for the small exporter and only 3 percent for the large. These differences cannot be attributed to distance, as the differ- ence is only 4 kilometers; rather, they are due to the small exporter’s lower carrying capacity. Additionally, handling costs are almost twice as high for the small exporter as for the large exporter. As a percentage of the final price of 1 kilogram of tomatoes, handling costs are six times as high for the small exporter. Moreover, just the handling costs for the small exporter are equal to the entire transport costs for the large exporter as a share of the final price of the tomatoes, both at 6 percent. Customs Costs Customs costs are the second most important trade and logistics cost for the small exporter (11 percent of the final price of a kilogram of tomatoes), but the most important cost for the large exporter, at 10 percent. 168 Fernández, Gómez, de Souza, and Vega Customs costs are disaggregated into four components: customs agency service fees on the Costa Rican side, customs agency service fees on the Nicaraguan side, waiting time at Costa Rican customs, and waiting time at Nicaraguan customs. The largest component of customs costs for both the small and the large exporter is time spent in customs in Nicaragua, followed by time spent in customs in Costa Rica. Therefore, the findings suggest that delays at customs, not regulated payments per se, are the primary logistical challenge in the export of perishable prod- ucts from Costa Rica to Nicaragua. To illustrate the impact of delays on the cost of a kilogram of tomatoes, we calculate the cost difference between a normally congested period and a highly congested period assuming that the cargo must wait an entire day at customs (9.5 hours on the Costa Rican side and 14.5 hours on the Nicaraguan side). The results reveal that congestion delays at the border translate into an additional cost of US$0.22 per kilogram for the large exporter and US$0.20 per kilogram for the small exporter. In summary, the findings show that a higher share of a small exporter’s cost structure in the export of tomatoes from Costa Rica to Nicaragua can be attributed to logistics costs when compared to the cost structure for a large exporter. Trade and logistics costs, defined as transport (shipping and handling), customs, and duties for the small producer are approximately two times those of the large exporter (23 and 44 percent, respectively). Quantitative Results: Wheat, Rice, and Corn Supply Chains Understanding the logistics challenges involved in the importation of wheat, rice, and corn from the United States to Honduras and Nicaragua is important given these countries’ overwhelming dependence on the U.S. market for these basic grains. For example, in value terms in 2009, 99 percent of yellow corn imports to Honduras originated in the United States, as did 89 percent of yellow corn imports to Nicaragua. This section estimates the costs incurred in the process of transporting wheat from farms in Minnesota, paddy rice from Arkansas,16 and yellow corn from Minnesota to wholesale markets in San Pedro Sula in Honduras and Managua in Nicaragua during December 2009 and January 2010. The initial transport costs associated with these three supply chains are related to the truck and rail freight charges incurred from the originating farm to the port of New Orleans. Once the grains arrive at the port of New Orleans, costs include freight insurance and ocean shipping charges Supply Chain Analyses of Exports and Imports of Agricultural Products 169 to transport the products to the seaport of Puerto Cortés on the Atlantic coast in the case of Honduras and the port of Puerto Corinto on the Pacific coast in the case of Nicaragua. Once at the port of desti- nation, several steps account for additional costs. For example, for wheat, shipping companies generally deliver the wheat free alongside ship (FAS), implying that costs are incurred in receiving the shipments and clearing customs. Next the wheat is loaded onto trucks and transported by road to the mill, where the grain is first stocked in silos and then milled. Once milled, it is packed in bags of 45 kilograms each and transported to bakeries in each country’s capital city. Similarly for rice, shipments are delivered in FAS terms and, once at the destination port, are loaded onto trucks and transported by road to rice mills, where the rice is turned into milled rice at an average conver- sion level of about 65 percent.17 The milled rice is then packed into 100-pound bags and sold to local popular market retailers or supermar- kets. In both Honduras and Nicaragua, large importers generally control most of the links of the chain taking place in-country and distribute the products directly to retailers and wholesalers, while smaller importers generally have to hire distributors or sell at the mill. Finally, corn shipments are also delivered in FAS terms and, having undergone customs procedures, are loaded onto trucks and transported by road to the feed manufacturing plant where the grain is first stocked in silos and then crushed and milled. Once the corn has been crushed or milled, it is mixed with other ingredients in bags of 45.4 kilograms each, which are then transported to animal-raising plants. Price Breakdown Bearing in mind that all of the grain supply chains end at wholesale and thus are comparable, this section provides a cross-country and product analysis. In doing so, it pays close attention to the transport and logistics costs, shedding light on the logistics challenges facing Honduras and Nicaragua in the import of grains. Tables 6.3 and 6.4 show the breakdown of costs associated with wheat imports into Nicaragua and Honduras fol- lowing the process described above.18 Figure 6.4 displays the aggregated cost components for each of the supply chains by product and country. Aside from the farm gate price,19 the largest components for the wheat, rice, and corn chains are operating costs (see table 6.5 for a definition of costs) and transportation costs. The share of transportation costs in the final price of the good is the least sig- nificant for rice and the most significant for corn. However, total logistics 170 Fernández, Gómez, de Souza, and Vega Table 6.3 Supply Chain Analysis and Cost Contributions to the Average Price of Wheat Flour Sold in San Pedro Sula, Honduras Additional cost (US$ per Cumulative Wheat as a % of % of each Supply chain elements kilogram) cost cumulative cost element Farm price in North Dakota 0.12 n.a. 100.00 13.71 Truck transport to rail terminal 0.01 0.13 91.91 1.21 Rail transport to port of New Orleans 0.06 0.19 61.05 7.54 Insurance, ocean freight, broker’s profit 0.04 0.23 51.02 4.42 Reception at port and customs clearance and fumigation 0.06 0.29 40.51 6.97 Land transportation to mill 0.05 0.34 34.58 5.81 Silage and warehousing costs at mill 0.06 0.40 29.28 7.17 Losses during extraction 0.08 0.49 24.27 9.67 Milling additives 0.05 0.54 22.01 5.81 Sacking, packing, delivery to transportation vehicles 0.02 0.56 21.22 2.32 Other fixed variable costs 0.11 0.67 17.68 12.93 Mill’s profit 0.05 0.72 16.45 5.81 Loading in trucks 0.00 0.72 16.40 0.23 Transportation to wholesalers 0.04 0.76 15.54 4.65 Wholesaler costs 0.08 0.84 14.13 8.83 Wholesaler profit 0.03 0.86 13.71 2.91 Sources: Authors’ calculations, data from U.S. Department of Agriculture, interviews with millers, U.S. International Trade Commission statistics, Honduras’s agricultural statistics, a prices survey. Note: Wheat costs include local elevator costs. New Orleans port costs are estimates. n.a. = Not applicable. costs (transportation plus other logistics costs) represent shares of the final price of the good at wholesale similar to those for wheat and rice, but almost four times larger than those for corn. The similarity in the cost structure of the corn supply chain for Honduras and for Nicaragua is noteworthy. Fieldwork conducted for this study indicates that, in both countries, large poultry and pork pro- ducers control the whole supply chain, and competition among them is high, as their production systems are similar. This is reflected in rela- tively competitive markets for yellow corn, with similar cost structures in the two countries.20 At first glance, it seems that the more competitive Supply Chain Analyses of Exports and Imports of Agricultural Products 171 Table 6.4 Supply Chain Analysis and Cost Contributions to the Average Price of Wheat Flour Sold in Managua Additional cost (US$ per Cumulative Wheat as a % of % of each Supply chain elements kilogram) cost cumulative cost element Farm price in North Dakota 0.17 n.a. 100.00 17.92 Truck transport to rail terminal 0.01 0.18 94.18 1.11 Rail transport to port of New Orleans 0.06 0.24 69.06 6.92 Insurance, ocean freight, broker’s profit 0.04 0.29 58.54 4.66 Reception at port and customs clearance and fumigation 0.06 0.34 48.77 6.13 Land transportation to mill 0.06 0.40 41.79 6.13 Silage and warehousing costs at mill 0.04 0.45 37.66 4.70 Losses during extraction 0.10 0.54 30.84 10.53 Milling additives and costs 0.07 0.62 27.14 7.92 Sacking, packing, delivery to transportation vehicles 0.01 0.63 26.76 0.94 Other fixed and variable costs 0.13 0.75 22.30 13.40 Mill’s profit 0.04 0.79 21.31 3.73 Loading in trucks 0.00 0. 79 21.28 0.11 Transport to wholesalers 0.05 0.84 20.11 4.91 Wholesaler costs 0.08 0.92 18.28 8.91 Wholesaler profit 0.02 0.94 17.92 1.96 Sources: Authors’ calculations, data from U.S. Department of Agriculture, U.S. International Trade Commission statistics, interviews with millers, Nicaragua’s agricultural statistics, prices survey. Note: Wheat costs include local elevator costs. New Orleans port costs are estimates. n.a. = Not applicable. the market is, the higher is the impact of transportation costs on the final price of the good. Concerning profit margins, the analysis concludes that profit margins are greater along the Honduran supply chain for both wheat and rice. For wheat in particular, the mill’s profit, for example, is estimated at US$50 per metric ton in Honduras and US$35 per metric ton in Nicaragua. Due to greater efficiency along the supply chain, Honduran millers seem to be enjoying larger profits. While figure 6.4 displays the aggregation of all costs involved in each of the supply chains, figure 6.5 takes a closer look at the costs incurred 172 Fernández, Gómez, de Souza, and Vega Figure 6.4 Cost Components as a Percentage of the Final Price of the Good, by Product and Country 12 12 7 17 14 18 15 % of final price of the good 24 11 23 33 41 11 5 5 9 32 10 10 33 41 12 13 37 34 33 29 27 14 18 as a as a as a gu gu gu ur ur ur ra ra ra nd nd nd ca ca ca Ho Ho Ho Ni Ni Ni wheat rice corn other logistics transportation profit margins operating costs farm gate Source: Authors’ calculations. domestically once the products arrive in Central America (from reception at port and customs clearance up to the wholesaler). This approach indi- cates that, overall, controlling for distances, domestic costs are higher in Nicaragua than in Honduras, with the exception of the rice chain. These costs represent 6 percent more as a percentage of the final price of the good for wheat imports into Nicaragua than for wheat imports into Honduras. Likewise, for corn imports, domestic costs are 4 percent higher in Nicaragua. In contrast, when looking at rice imports, domestic costs are 8 percent higher in Honduras, making this the largest difference among the three products. Based on qualitative data obtained during the inter- view process, it appears that the market for rice imports is more vertically integrated in Nicaragua than in Honduras, allowing Nicaraguan importers to control more links along the supply chain. Such vertical integration, in which the same company controls the transport to mill, the milling process, and the packaging and distribution to wholesalers, avoids the double markup that takes place when distinct companies with sufficient market power at each step of the chain earn margins. These margins thus translate into a more costly supply chain in Honduras. Supply Chain Analyses of Exports and Imports of Agricultural Products 173 Table 6.5 Breakdown of Cost Components Cost aggregates Cost components Transportation U.S. transport costs (truck transport to rail terminal or barge, rail transport or barge to Port of New Orleans) Ocean shipping costs (ocean freight, insurance, broker’s profit) Domestic transport costs (land transport to mill or feed manufacturer, land transport to wholesalers) Other logistics Reception at port and customs clearance and fumigation Silage and warehousing costs at mill Sacking, packing, delivery to transportation vehicles Loading in trucks Total logistics Transportation and other logistics Operating costs Losses during extraction Other milling losses and costs (for example, milling additives) Wholesaler costs Other fixed and variable costs (for example, administrative, marketing, and financial costs of the miller or manufacturing plant, price hedging against higher international price) Profit margins Mill’s profit Wholesaler profit Retailer profit Farm gate Average farm price (North Dakota for wheat, Arkansas for rice, and Minnesota for corn) Source: Authors’ compilation. Figure 6.5 Total In-Country Costs as a Percentage of the Final Price of the Good, by Product and Country % of final price of the good 80 69 67 63 59 56 59 60 40 20 0 as ua s a s ua ra ra gu ur g g u u ra ra ra nd nd nd ca ca ca Ho Ho Ho Ni Ni Ni wheat rice corn Source: Authors’ calculations. 174 Fernández, Gómez, de Souza, and Vega Transport and Logistics Costs Figure 6.6 illustrates the share of logistics costs in each of the six grain chains. Results from the SLS suggest that logistics costs range from 36 to 40 percent of the final price of the good for wheat imports into Nicaragua and Honduras, are similar at about 29 percent for rice imports into both countries, and range from 45 to 48 percent for corn imports. For wheat and corn, these costs are higher than the producer’s share; they are almost two to three times higher for wheat imports. In contrast, for the rice chain, the share of logistics costs is similar to the producer’s share. Within transport costs, the data show that, although the three grains are imported into Nicaragua through the Panama Canal, a journey that is four times as long as that for imports into Honduras, ocean transport costs are not significantly different between countries and products. For example, for corn, the share of ocean transport costs for both Nicaragua Figure 6.6 Logistics Costs (Transport and Other Logistics Costs) as a Percentage of the Final Price of the Good, by Product and Country 50 48 45 45 7 40 40 12 36 35 costs (% of final price) 17 29 29 30 12 25 30 14 22 20 18 10 11 15 10 4 5 9 4 5 5 5 9 4 8 3 6 6 3 3 0 as a as a as a gu gu gu ur ur ur ra ra ra nd nd nd ca ca ca Ho Ho Ho Ni Ni Ni wheat rice corn other logistics domestic transport ocean transport U.S. transport Source: Authors’ calculations. Supply Chain Analyses of Exports and Imports of Agricultural Products 175 and Honduras is 5 percent of the final price of the good at wholesale. For rice, the share of ocean transport, at 4 percent, is only 1 percent higher for Nicaragua than for Honduras. As a percentage of total trans- portation costs, however, ocean transport costs for Nicaraguan rice importers are 12 percent higher than those for Honduran importers. For wheat and corn, ocean transport costs occupy similar shares within total transportation costs. With regard to the other components of transportation—U.S. transport costs and domestic transport costs in-country—there is no significant dif- ference between Honduras and Nicaragua for any of the products. However, domestic transport costs as a percentage of the final price of the goods at wholesale are 5 percent higher (twice as high) in Honduras for rice imports, but 8 percent higher in Nicaragua for corn imports. For wheat, the share of domestic transport costs is similar for both countries. What is more important, however, is that, within transport costs, domestic costs occupy the largest share as a percentage of the final price of the goods. For the corn chain, specifically, domestic transport costs in Nicaragua are higher than the U.S. transport, ocean transport, and other logistics costs combined (30 and 18 percent, respectively). In Honduras, in contrast, domestic transport costs are similar to the other transport and logistics costs combined (22 and 23 percent, respectively). As shown in figure 6.6, the corn supply chains are the only ones for which domestic transport costs are the highest both within transport costs and within total logistics costs. For the wheat and rice chains, although domestic costs occupy the largest share of transport costs, they are second to the “other logistics” component, consisting of reception at port and customs clearance and fumigation; silage and warehousing costs at mill; sacking, packing, and delivery to transportation vehicles or drayage; and loading in trucks at the mill or feed manufacturer. Other Logistics Costs A breakdown of the “other logistics” component for all three chains shows that significant costs are related to the reception of grains at the port of entry, including those related to the carrying out of phytosanitary and san- itary inspections (see figure 6.7). In fact, the port, customs clearance, and fumigation costs represent the most important subcomponent, next to silage and warehousing costs at the mill, for wheat imports into both Honduras and Nicaragua. Likewise, for corn imports into Honduras and Nicaragua, the costs of reception at port and customs clearance and fumigation occupy the largest percentage of the final price of corn at 176 Fernández, Gómez, de Souza, and Vega Figure 6.7 Other Logistics Costs as a Percentage of the Final Price of the Good, by Product and Country 16 15% 14 logistics costs (% of final price) 12 10 9 8 7 7 6 6 5 5 4 4 4 3 2 2 2 2 1 1 1 1 1 1 1 1 0 0 0 0 as a as a as a gu gu gu ur ur ur ra ra ra nd nd nd ca ca ca Ho Ho Ho Ni Ni Ni wheat rice corn reception at port and customs clearance and fumigation silage and warehousing costs at mill sacking, packing, delivery to transportation vehicles loading in trucks Source: Authors’ calculations. wholesale. The rice chains are interesting in that, for Honduras, reception at port and customs clearance and fumigation do not seem to be as important as the sacking, packing, and delivery of the product to trans- portation vehicles; for Nicaragua, silage and warehousing costs at mill are the most important component within “other logistics.” This finding calls for a more in-depth analysis of logistics at the mill for both countries. Conclusions Logistics costs represent an important portion of the final price of delivered foods in Central America and are significant both for highly perishable goods, such as tomatoes, as well as for dry goods such as wheat, rice, and corn. These costs can range from 17 percent in the case of a small exporter of tomatoes in intraregional trade to 48 percent in the case of an importer of yellow corn from the United States to Nicaragua. Furthermore, logistics costs disproportionately affect smaller firms. In the case of fresh tomato exports from Costa Rica to Nicaragua, the weight of logistics costs for a small exporter is twice as high as for a large exporter. Supply Chain Analyses of Exports and Imports of Agricultural Products 177 Specifically, the study demonstrates that distance is not a central driver of transport costs, either for ocean transport or for domestic road transport. Even though grain imports into Nicaragua must travel longer distances in crossing the Panama Canal to Puerto Corinto in the Pacific than grain imports traveling into Honduras through Puerto Cortés, ocean transport costs as a share of the final price of the good are similar for all grain sup- ply chains. For example, ocean transport costs represent 4 percent of the final price of the rice sold at wholesale in Managua, while they represent 3 percent of the final price of the rice sold in San Pedro Sula. Similarly, road transport costs, in the case of tomato exports, equal 17 percent of the final price of a kilogram of tomatoes for a small exporter, 3.4 times as much as for a large exporter, despite the fact that the small exporter trav- els 71 kilometers less, or 0.86 times what the large exporter travels. Results from this study indicate that some of the drivers behind high logistics costs both for intraregional and for extraregional trade in DR- CAFTA countries can be addressed within four main policy areas: cus- toms integration and reform, road transport development (increasing coverage and improving quality of the road network), harmonization of sanitary and phytosanitary regulations, and port use optimization. Bottlenecks in these areas represent particular challenges for Central America, and addressing them through DR-CAFTA’s complementary agenda is crucial if the free trade agreement is to generate the most ben- efits for the region’s economies. Notes 1. Authors’ calculations based on World Integrated Trade Solution (WITS). 2. Authors’ calculations based on WITS. Excludes the United States. 3. Given the confidential nature of certain requested data, some of the figures were estimated using the results of previous supply chain analyses. See Schwartz, Guasch, and Wilmsmeir (2009). 4. Confidentiality was requested from the corporations and individuals inter- viewed for this study. 5. The exporter transfers the product to cardboard boxes, because transport- ing the product in plastic boxes would imply having to process the former as a temporary export, implying additional costs (approximately US$14.50 per trip per container and an extra US$61.80 for a carta de política). 6. Larger exporters with lower logistics costs usually sell directly to supermarket chains, while smaller and more expensive exporters sell to public markets. 178 Fernández, Gómez, de Souza, and Vega 7. This finding is interesting in itself, as one would think that the time spent at customs would be lower for the large exporter than for the small exporter. Given that the large exporter has continuous shipments and hires reliable and well-established customs agents to file customs documents that accurately reflect the volume, value, and other characteristics of their cargo, it is expected that the large exporter has developed a relationship of trust with customs authorities. For this particular study, however, the small exporters inter- viewed also expressed that they have developed good relationships with customs authorities. One small exporter did, however, emphasize that, at times, poorly skilled customs agents submit inaccurate information to customs, which, on inspection of the cargo, not only creates additional costs, but also jeopardizes the established relationship of trust. This suggests that, as long as this relationship of trust is maintained, waiting time at customs is similar for both large and small exporters. This logic also suggests that waiting time at cus- toms as a result of thorough inspections is the longest for new market entrants. 8. The requirement for shipments to be fully inspected by customs is deter- mined by a stoplight at the border. Once the customs declaration has been submitted, a stoplight determines whether the cargo is to be subjected to fur- ther inspections. If the light turns green, the cargo can be cleared, if yellow, customs documents must be reviewed thoroughly, and if red, customs must confirm that the documents submitted reflect the nature of the shipments through a detailed cargo inspection in addition to phytosanitary and narcotics inspections. If the light is red, the exporter himself must pay for approxi- mately seven people to unload the cargo. 9. See “Furgones esperan ingreso en Costa Rica y Nicaragua: Nuevo control en Peñas Blancas crea presas de varios kilómetros,” La Nación, Costa Rica, December 21, 2006, available at http://www.nacion.com/ln_ee/2006/ diciembre/21/economia937007.html. 10. Previously, samples were sent to Managua, which implied longer waiting times at the border. 11. The regulations guiding charges during unloading at the ports differ in these two countries. In Puerto Corinto, Nicaragua, importers are rewarded US$5,000 a day if they manage to unload the vessel in less than three days. If they take longer, they are fined US$20,000 a day. In Honduras, there is no fine or reward system. Importers pay a flat fee of US$7,000 a day. 12. Importers may choose to buy storage space in advance to be able to fulfill fluctuations in demand. 13. The bill of lading is issued by the carrier, acknowledging that the goods have been received on board as cargo and are being transported to a specified place and recipient. 14. Panama Canal website, available at http://www.panacanal.com/eng/maritime/ tolls.html. Supply Chain Analyses of Exports and Imports of Agricultural Products 179 15. For purposes of this study, the small exporter is defined as an independent exporter having his own transport infrastructure and traveling to Managua to sell his own product to private clients. The small exporter has a lower carry- ing capacity for tomatoes than does the large exporter, as he gathers four to five products on average, often from his own production, and consolidates the cargo. The large exporter has a year-round supply of product, purchased from well-established independent producers or producers associations, and makes at least two weekly trips from Costa Rica to Nicaragua. 16. Rough (or paddy) rice is the rice as it comes from the field. The rice kernels are still encased in an inedible, protective hull that, when separated from the kernel, can be burned as fuel for power plants and other industrial processes, be used as mulch, or become a component in abrasives and other products. 17. The conversion ratio is the percentage of final product—that is, white rice— recovered out of the total volume of paddy rice being milled. For modern mills such as the ones used in both Nicaragua and Honduras, the by-products of 100 kilograms of long-grain rough (paddy) rice are 62–68 percent milled rice, 4–5 percent rice bran, 25 percent rice husk, and 2–3 percent germ wastages. When the milled rice goes to supermarkets, the millers package the products in branded bags of various sizes (from 0.45 to 20 kilograms) at an additional cost. Retail prices are not included in the analysis performed for this study. 18. For more disaggregated information on cost regarding the supply chains for corn and rice, contact Raquel Fernández (rfernandez2@worldbank.org) or Santiago Gomez (sflorezgomez@worldbank.org). 19. Consistent with an analysis of free-on-board prices, the estimated farm price paid by Nicaraguan wheat and rice importers is 4 and 7 percent higher, respectively, as a share of the final price of the good when compared to that paid by Honduran importers. Due to larger volumes of imported wheat and exogenous factors such as the availability of better traders, Honduran importers may be better suited to import wheat at more competitive prices. 20. The authors calculate that the efficiency of Nicaragua’s feed manufacturers is about 92 percent, while that of Honduran feed manufacturers is about 95 percent. These high and comparable efficiency levels can be attributed to the level of competition, which is high overall, and pressure from the animal- raising industry in both countries. References Baier, S. L., and J. H. Bergstrand. 2001. “The Growth of World Trade: Tariffs, Transport Costs, and Income Similarity.” Journal of International Economics 53 (1): 1–27. 180 Fernández, Gómez, de Souza, and Vega Blyde, J. S., M. Moreira, and M. Volpe. 2008. “Unclogging the Arteries: The Impact of Transport Costs on Latin American and Caribbean Trade.” Inter-American Development Bank, Washington, DC; Harvard University, David Rockefeller Center for Latin American Studies, Cambridge, MA. Hummels, D. 2001. “Time as a Trade Barrier.” Unpublished manuscript, Purdue University, West Lafayette, IN. Schwartz, J., J. L. Guasch, and G. Wilmsmeeir. 2009. “Logistics, Transport, and Food Prices in LAC: Policy Guidance for Improving Efficiency and Reducing Costs.” World Bank and Inter-American Development Bank, Washington, DC. World Bank. 2006. “Costa Rica Country Economic Memorandum: The Challenges for Sustained Growth.” Report 36180-CR, World Bank, Washington, DC. ———. 2007. “Costa Rica Investment Climate Assessment.” Report 35424-CR, World Bank, Washington, DC. ———. 2009. “Uruguay Trade and Logistics: An Opportunity.” World Bank, Washington, DC. World Economic Forum. 2009. Global Enabling Trade Report. Davos, Switzerland: World Economic Forum. CHAPTER 7 Logistics Challenges in Central America José A. Barbero This chapter assesses the condition and performance of trade logistics and facilitation in Central America and the Dominican Republic to determine the region’s ability to take advantage of the potential benefits of the Dominican Republic–Central America Free Trade Agreement (DR- CAFTA). After describing the main patterns of regional trade and the core factors influencing logistics performance, the chapter presents a set of policy priorities for strengthening the system’s functioning. The Relevance of Logistics as a Factor in Trade The physical movement of goods is conditioned by numerous factors that can be organized in three major groups, which together constitute a country’s logistics system: infrastructure and transport services, logistics business management by carriers and shippers, and trade facilitation pro- cedures, including issues related to documentation, control, and provi- sion of security for trade flows. Each major group encompasses components that are interrelated in a complex network. Infrastructure— the system’s hardware—is critical to ensuring the efficient transportation and storage of goods. Business logistics include cargo owners and service providers; cargo owners are interested in improving the organization of 181 182 Barbero the supply chain, while logistics providers are interested in improving and broadening the scope of their services by investing in new equip- ment and facilities. Trade facilitation initiatives (as well as other regula- tions related to the circulation of goods) are generally under the mandate of the public sector and—together with transport regulations—could be considered the system’s “software.” These have a large impact on the sys- tem’s functioning, as they determine the processes and documents that users and service providers must understand and comply with if they wish to interact. This basic definition of the factors involved in the flow of goods gives rise to two important conclusions. First, analyses and proposals should not be constrained to infrastructure bottlenecks (as they are most times), but should also consider relevant aspects in all of the system’s components. Second, the performance of logistics systems depends on the coordinated efforts of stakeholders in both the public and the private sectors. Logistics are directly related to the organization of productive activi- ties among firms, most of them within the private sector, and have gained more relevance as the flow of goods has become more complex and pres- sures to reduce costs have increased. Until the 1980s, companies handled the transportation of inputs, distribution of products, and storage systems as separate functions, in a relatively independent manner. Subsequently, companies began to integrate them, considering logistics to cover the physical movements along the complete cycle of materials and the docu- mentation and information from the purchase of inputs to the delivery of the final product, encompassing the functions of transport, storage, inven- tory management, and packaging of goods as well as the administration and control of these flows. Thus business logistics need to be analyzed from the broader perspective of supply chains, with the aim of boosting coordination among the links to improve efficiency and reduce costs. The Impact of Logistics and Trade Facilitation on Trade Costs As trade theory develops, the positive impacts of more efficient logis- tics and trade facilitation initiatives on the costs of trade are becom- ing increasingly explicit. The benefits of trade for general welfare have been widely researched (World Bank 2009); the analytical focus on the impacts associated with reducing the costs of trade by increasing logistics efficiency was developed in recent years. This type of analy- sis is needed to build a strong case for logistics-related reforms, as the findings would help define the benefits that can accrue from these Logistics Challenges in Central America 183 initiatives. A recent World Bank document on border management presents a detailed review of the latest literature on the issue, which is summarized below: • Significance of trade costs versus tariffs. The decline in tariff levels in the past 20 years has highlighted the existence of trade costs that are not related to traditional trade policy. Research by Anderson and van Wincoop (2004) is illustrative in this sense. The authors establish a broad definition of trade costs (including transport costs, tariff and nontariff costs, legal and regulatory costs, information costs, and so forth). Based on their findings, trade costs are large, and a significant part of them stems from economic policies not directly associated with trade, such as transport policy and related regulations. “Their esti- mate of the ad valorem tax equivalent of trade cost for industrialized countries is 170 percent, of which 21 percent falls under transporta- tion costs . . ., 44 percent under border trade-related barriers, and 55 percent under retail and wholesale distribution costs. The authors assert that trade barriers in developing countries are higher than what is estimated for industrial countries” (World Bank 2009). • Time dimension of trade costs. Trade barriers bring forward costs that are linked with time delays and uncertainty in moving goods across borders. Hummels (2001) was the first to explore this issue. He argues that each additional day spent on transport reduces by 1 to 1.5 percent the chances a country has of exporting to the U.S. market. Further research by Hummels and Schaur (2009) estimates the value of cross- border delays on the basis of average depreciation and inventory carrying costs: each day of delay is equivalent to additional costs of 0.8 percent of the value of freight. • Effectiveness of trade facilitation measures. Research by Wilson, Mann, and Otsuki (2005) highlights potential gains from trade facilitation initiatives. The authors define trade facilitation in a broad manner, including port efficiency (port facilities, inland waterways, and airports), customs environment (hidden barriers to imports and bribes), regulatory environment (transparency of government policies and levels of corruption), and e-business infrastructure (speed and cost of Internet access). Based on data for 75 countries in 2000–01, the authors argue that improvements in the four areas have a positive impact on trade. More important, they highlight that, if the least-efficient countries 184 Barbero could boost efficiency halfway to the sample’s average, trade gains could amount to US$377 billion. • Infrastructure development and trade. The links between trade and infrastructure development have been explored recently by Limao and Venables (2001). Their research shows that 40 percent of trans- port costs in coastal countries and up to 60 percent in landlocked countries result from infrastructure deficiencies. The authors estimate that if landlocked countries in the lower percentiles of infrastructure development would rise to the 75 percentile, they would be able to increase their volume of trade. • Trade and institutional quality. Sadikov (2007) analyzes these issues by looking at specific measures that are under the control of customs and other regulatory authorities. His research analyzes the number of signa- tures required for exporting (considered a proxy for border barriers) and the number of procedures for registering a business (a proxy for behind-the-border barriers). Based on data for 126 countries, his find- ings show that reducing export signatures and registration procedures results in trade gains: each extra signature reduces aggregate exports by 4.2 percent. Finally, he points out that the impact of these measures is larger when products have a higher level of differentiation. • Impacts of trade costs at the firm level. A series of studies conducted recently shows that exporting firms in a country are generally larger and more efficient than firms that do not export. The process of self- selection between the two groups is determined by the existence of cross-border trade costs: only firms that are productive enough to overcome the additional costs associated with expansion to larger new markets are able to export. Falling trade costs are thus linked with important decisions at the firm level: (a) entry and exit, (b) the deci- sion to export, (c) the amount to export, (d) technology decisions, and (e) employment decisions (World Bank 2009). This strand of theory, represented by Melitz (2003), Bernard, Jensen, and Schott (2006), Yeaple (2005), and Bernard and others (2007), suggests that reduced trade costs push more firms to export and stimulate the growth of existing exporters, which, in turn, results in higher productivity. Progress in the implementation of DR-CAFTA was expected to be positive for Central America for two main reasons. The first was that the Logistics Challenges in Central America 185 agreement granted expanded access to the U.S. market, boosting trade volumes. Although the region already had certain privileges for access to the United States resulting from the Caribbean Basin Initiative, these were to deepen under DR-CAFTA. Other provisions that would foster trade included more flexible rules of origin for exports of apparel and tex- tiles from maquilas (which would help to offset tough competition from Asian countries) and technical assistance to exporters of nontraditional agricultural products (which would help them to meet the sanitary and phytosanitary standards required for entry into the United States). The latter was expected to help diversify the region’s base of exports. A sec- ond group of benefits for the region was given by the possibility of deep- ening integration among Central American countries. The fact that these agreed to make DR-CAFTA a treaty that would be applied multilaterally (instead of bilaterally between the United States and each Central American country individually) was a good starting point in this direc- tion. The ability of Central American countries to pursue a comple- mentary policy agenda in areas such as trade facilitation, regulatory reforms, transport policies, and innovation will determine to a great extent the magnitude of the benefits resulting from DR-CAFTA. The financial crisis that hit the U.S. economy in mid-2008 had strong impacts on the region; however, the prospects of DR-CAFTA remain positive. The close trade ties between the United States and Central America strengthened when DR-CAFTA came into effect: between 2005 and 2008, Central American exports to the United States grew 10 percent, and in 2008, close to 30 regional exports went to the U.S. market (EIU 2009). Thus contraction in U.S. demand in the aftermath of the financial crisis had strong effects in the region, with export earnings falling 20 percent year on year as of January–March 2009 (EIU 2009). Uncertainties about the pace of U.S. recovery outline the need to develop new export markets. On the positive side, export diversification seems to be advancing, with nontraditional exports registering growth of close to 20 percent in 2008. In addition, integration among countries in the region has been strengthened: earnings from intraregional trade have expanded by an annual average of 12 percent since 2003, according to data from the Secretariat of Economic Integration in Central America (SIECA).1 A regional policy agenda to address issues that are relevant from a trade perspective is also developing slowly, as noted below. Finally, in the short term, intraregional trade flows, together with the opening of new markets in Asia and Latin America for regional exports, will largely determine the volume of trade in the following years. 186 Barbero International Logistics Indicators The logistics performance index (LPI) is a research initiative launched by a transport and trade facilitation alliance composed of several public and private international organizations, led by the World Bank. The LPI was estimated for the first time in 2007 based on a survey of international freight operators from 150 countries, who provided feedback on their perception of several logistics attributes of the countries in which they operate and with which they trade. The responses allow for the calcula- tion of several subindexes, which together make up the LPI. The LPI is expressed through a score (from 1 to 5) and a ranking according to the position held by an individual country within the group. A second esti- mate was done in 2009, comprising 155 countries. From a global perspective, the LPI shows that the region’s logistics per- formance is relatively weak. The average regional rank (of the five Central American countries and the Dominican Republic) was 85 in 2007 and 79 in 2009, in a sample of 150 and 155 countries, respectively. The region is a clear example of the logistics gap that exists between the high-income countries (the top performers) and the middle- and low-income coun- tries. Malaysia and Thailand,2 which are relevant comparators in terms of income, ranked 29 and 35, respectively, in the 2009 survey. However, Central America’s overall performance in the LPI improved significantly between 2007 and 2009. The scores of four of the six DR-CAFTA coun- tries improved (Costa Rica, Dominican Republic, Honduras, and the Nicaragua), while those of El Salvador and Guatemala deteriorated some- what (see figure 7.1). On average, the region improved considerably the absolute value of logistics performance (from 2.48 to 2.73), which suggests that improve- ments in many logistics factors have been yielding results. If the average rank is analyzed to assess the relative—not the absolute—logistics per- formance, the region also shows some improvement, albeit more moder- ate (from 85 to 79). This suggests that other countries have also improved;3 some of the Central American countries, however, show a sig- nificant decline in their relative position, even after achieving an absolute improvement in their logistics performance. For example, Guatemala and El Salvador had small gains in absolute terms, but declined from position 75 to 90 and from 66 to 86 in the LPI country rank.4 Analysis at the subindex level shows that the region has intermediate performance in customs administration efficiency and quality of transport and information technology infrastructure. Scores on these topics are only Logistics Challenges in Central America 187 Figure 7.1 LPI Rank of Central American and Comparator Countries, 2007 and 2009 Nicaragua Guatemala El Salvador Honduras Dominican Republic Costa Rica regional average 79 85 Mexico Uruguay Chile Thailand Malaysia 0 20 40 60 80 100 120 140 LPI rank 2009 2007 Source: 2007 and 2009 LPI surveys. moderately lower than those of regional comparators such as Uruguay and Chile (the regional leader, whose performance decayed between 2007 and 2009).5 The gap in Central American scores on infrastructure quality widens significantly when compared with that of Mexico, Thailand, and Malaysia, as shown in figure 7.2. Central America receives good scores for timeliness of shipments reaching their destination, in some cases scoring higher than Chile and Uruguay. However, regional performance is weaker in the rest of the topics covered by the survey: the ease and affordability of arranging international shipments are at an intermediate level when compared with Chile and Uruguay, with El Salvador and Guatemala bringing down the regional average. Tracking and tracing capabilities are much lower than in Chile, although Central America’s average is higher than Uruguay’s score. Scores for local logis- tics competence are also low, with the exception of Costa Rica. All Central American scores are well below those of Malaysia and Thailand, with differences close to one point in most cases, except for timeliness of shipments. 188 Barbero Figure 7.2 Performance on the LPI Subindexes of Central American and Comparator Countries 4.5 4.0 3.5 3.0 LPI score 2.5 2.0 1.5 1.0 0.5 0 sia nd ile y ico e ica ic as or a a ua al gu ag bl ad ur Ch ay la aR em ex pu ug er ra nd ai lv al M av st ca at Th Re Ur Sa M Ho Co Ni Gu al El n on ica gi in re m Do international shipments logistics competence tracking and tracing timeliness Source: 2009 LPI survey. As stated, although overall performance appears to be relatively good in regional terms, the gap with high-income countries is wide. Nevertheless, good performers such as Costa Rica and El Salvador help to bring the regional average up in subindexes such as customs and infra- structure (see figure 7.3). The Global Competitiveness Report and Trade Enabling Index The results of the perception survey conducted by the World Economic Forum for the 2009 and 2010 editions of the Global Competitiveness Report (WEF 2009, 2010) provide additional information on the relative performance of Central American countries, although the results are not entirely consistent with those of the LPI. The overall condition of regional infrastructure has improved, but performance varies strongly among countries and sectors (see figure 7.4). The per- ceived quality of infrastructure is intermediate, with scores behind Logistics Challenges in Central America 189 Figure 7.3 Performance on Customs and Infrastructure of Central American and Comparator Countries, 2009 Nicaragua Guatemala El Salvador Honduras Dominican Republic Costa Rica regional average Mexico Uruguay Chile Thailand Malaysia 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 LPI score infrastructure customs Source: 2009 LPI survey. those of good performers such as Uruguay and Chile (the regional leader) and comparators in Southeast Asia. El Salvador and Guatemala are the region’s strong performers; Honduras and the Dominican Republic are at an intermediate stage, while Nicaragua and Costa Rica lag behind (although the latter shows strong improvement in the 2010 survey). Roads are the weakest sector, with regional averages well behind those of the comparator group. El Salvador is an exception, with qual- ity at levels close to those of Chile, Malaysia, and Thailand. Nicaragua and Costa Rica register low scores, bringing down the regional average for road quality. The situation is similar regarding the perceived qual- ity of ports, in which the regional average is far from the scores of comparator countries. However, most countries improved their scores on ports in the 2009 survey: Honduras is well positioned, reflecting the good performance of Puerto Cortés, while Nicaragua and Costa Rica are the weakest performers. The perception of the quality of rail- ways is remarkably low, as a result of the limited development of this 190 Barbero Figure 7.4 Quality of Overall Infrastructure of Central American and Comparator Countries, 2008–10 Malaysia 5.4 5.6 Thailand 4.8 4.8 Uruguay 4.2 3.6 Chile 5.6 5.1 regional average 3.7 3.3 Costa Rica 3.4 2.6 Dominican Republic 3.4 3.3 El Salvador 4.8 4.4 Guatemala 4.3 3.8 Honduras 3.7 3.5 Nicaragua 2.7 2.1 0 1 2 3 4 5 infrastructure quality ranking 2009–10 2008–09 Source: WEF 2009, 2010. subsector in the region. Finally, airports show a strong and balanced performance in the region, with average scores close to those of the comparator group. Central America’s performance on the global enabling trade index shows transport infrastructure and border administration at an intermedi- ate level, close to Uruguay’s scores and lagging behind those of Chile, Malaysia, and Thailand (see figure 7.5). Regional performance at the level of subindexes reflects the general scores, with some interesting points: the efficiency of import and export procedures is high in most countries, reg- istering scores close to regional best practice as well as those of Thailand and Malaysia; transparency in border administration, particularly in the Dominican Republic, Honduras, and Guatemala, is rather low, and the effi- ciency of customs administration is poor in Nicaragua and Honduras. The availability of transport infrastructure is moderate from a regional perspec- tive, but lags well behind the scores of Chile and countries in Southeast Asia. Finally, the availability and quality of transport services are quite weak in the region, except in Costa Rica and the Dominican Republic. Logistics Challenges in Central America 191 Figure 7.5 Performance on the Global Enabling Trade Index of Central American and Comparator Countries, 2009 6 5.3 5 4.74.6 4.5 4.3 trade index ranking 4.1 4.1 3.9 3.9 4.1 4 3.8 3.7 3.2 3.2 3.1 3.2 3.43.0 3.4 3.0 3.1 3.2 3 2.9 2.5 2 1 0 ica Sa ic Gu dor a ca s a e y ile ico sia nd a ua al gu ag bl ur Ch ay la ica a R em ex a pu ug er ra nd ai lv al M av st at Th Re Ur M Ho Co Ni al El n on gi in re m Do border administration transport infrastructure Source: WEF 2009. Doing Business, Trading across Borders The trading across borders indicator monitored by the World Bank’s Doing Business database is a good tool for benchmarking transport costs and lead times for import and export processes. As shown in table 7.1, Central America’s performance is at an intermediate level, well posi- tioned in regional terms, but falling behind Malaysia, Thailand, and the Organisation for Economic Co-operation and Development (OECD) countries. The number of documents required for import and export operations is reasonable from a regional perspective, although perform- ance is weaker for imports, with Central America’s regional average of seven documents lagging behind the average for OECD and Thailand, at approximately four. Time to export and import measured in days also performs well (17 days) versus the comparator group but not the OECD countries (11 days). Nicaragua experiences the longest delays for imports and exports, with a lead time of 29 days in both cases. Finally, the costs associated with external trade are poor for the region. The small differ- ence between Central American and OECD countries mostly results from the higher internal prices in OECD countries. The cost gap with other countries in the comparator group is much more significant: the cost to export or import a 20-foot container from a Central American 192 Table 7.1 Performance on Doing Business Indicators of Trading across Borders of Central American and Comparator Countries Documents to Time to export Cost to export Documents to Time to Cost to import Region or economy export (number) (days) (US$ per container) import (number) import (days) (US$ per container) Latin America and Caribbean (regional average) 7 19 1,244 7 21 1,481 OECD (regional average) 4 11 1,090 5 11 1,146 Chile 6 21 745 7 21 795 Uruguay 10 19 1,100 10 22 1,330 Mexico 5 14 1,472 5 17 2,050 Malaysia 7 18 450 7 14 450 Thailand 4 14 625 3 13 795 Central America (regional average) 7 17 1,112 8 17 1,179 Costa Rica 6 13 1,190 7 15 1,190 Dominican Republic 6 9 916 7 10 1,150 El Salvador 8 14 880 8 10 820 Guatemala 10 17 1,182 10 17 1,302 Honduras 7 20 1,163 10 23 1,190 Nicaragua 5 29 1,340 5 29 1,420 Source: World Bank, Doing Business database, 2009. Logistics Challenges in Central America 193 port is about US$1,100, more than twice the cost in Malaysia and con- siderably above the cost in Chile or Thailand. This factor is particularly relevant, as some developing countries have achieved high levels of effi- ciency in logistics and trade facilitation. Country Logistics Review In this section, we review specific logistics challenges and characteristics in the six countries that are signatories of the DR-CAFTA. Costa Rica Costa Rica’s reliance on trade as an engine of growth highlights the need to improve logistics performance, which is currently hampering compet- itiveness. Based on the conclusions of recent World Bank assessments on sectors and topics associated with logistics (World Bank 2006b, 2007a, 2007b), infrastructure and transport services appear to be the main bot- tlenecks for increased efficiency of trade flows. Costa Rica has high levels of coverage in the road sector, but years of underinvestment have led to a marked deterioration of the network. Public investment in the sector fell from a peak of 2.1 percent of GDP in 1984 to below 0.6 percent on aver- age in 1999–2005. In 2006, almost 35 percent of national roads were in poor condition. In 2006–07, the government undertook substantial investments to improve the state of roads, focusing on the links between the central plateau and the port of Limón-Moin on the Atlantic coast and Caldera on the Pacific coast, which are the high-volume corridors in the country. However, the fast-growing rates of motorization, together with increasing trade volumes, continue to put pressure on these corridors, generating congestion and deteriorating infrastructure. Ports and their associated services appear to be the weakest link. The port complex at Limón-Moin handles more than 80 percent of the volume of maritime freight, mostly in container traffic (more than 70 percent of total volume in 2007), followed by general cargo and liquid bulk. Inefficient manage- ment under the state-owned tool-port model has created serious chal- lenges for operational efficiency in Limón-Moin: the cost per ton of moving cargo from Limón-Moin is estimated to be as much as twice as high as at most other ports in Central America (EIU 2008). The port of Caldera, in contrast, was concessioned recently, and since then container movement has nearly doubled. The government has plans to increase pri- vate participation in the sector by concessioning Limón-Moin and build- ing a new mega-port in Limón through a public works concession. 194 Barbero Table 7.2 Freight Flows in Costa Rica, by Mode of Transport Value Volume % originating in a Density free trade zone Transport US$ % of Tons % of (US$ per mode (millions) value (thousands) volume ton) Value Volume Maritime 3,894 43 5,769 81 675 38 8 Air 3,720 41 52 1 71,537 91 35 Surface 1,476 16 1,300 18 1,136 13 2 Total 9,089 100 7,120 100 1,277 56 7 Source: Procomer (the agency responsible for the promotion of external trade in Costa Rica). Costa Rica has two international airports, Juan Santamaría in northwest San José and Daniel Oduber near Liberia. Management of the former was transferred to the private sector, and efforts are under way to increase its operating capacity. As shown in table 7.2, the three modes of transport participate actively in the country’s export flows. Maritime transport takes the highest share freight in terms of volume, handling mostly dry bulk cargo of low value added per ton. The main challenges faced by this subsector are the lack of modern cargo-handling equipment and difficult access to the main port facilities. The development of air freight services is closely linked to the high-tech firms, which have brought efficiency to its operation. Thus air transport handles high value added goods and is heavily used by firms based in free trade zones. Urban congestion in the outskirts of San José generates some difficulties for cargo traveling toward the airport, increas- ing transit times. Finally, surface transport handles a low level of freight in terms of both volume and value, which reflects Costa Rica’s relatively low participation in intraregional trade. Delays in the Peñas Blancas border crossing (which is the main link for trade with other Central American countries) are very significant and have a relevant economic impact, as the majority of companies using land transport for export of goods are small and medium enterprises (SMEs). Dominican Republic The provision and quality of infrastructure are regarded as relatively good in the Dominican Republic. The largest challenges for trade are mostly linked with regulatory issues and customs efficiency. Port infrastructure has improved significantly in the last five years, following the entrance of private operators. The port of Haina, located 20 kilometers west of Santo Domingo, handled the largest volume of general cargo and container traffic Logistics Challenges in Central America 195 until 2004, followed by Puerto Plata. A privately financed state-of-the-art deepwater port at Caucedo was inaugurated in 2004 and has attracted most of Haina’s existing business thanks to its large storage facilities and modern equipment. Caucedo is currently absorbing significant volumes of transshipment traffic in the Caribbean, and these are expected to grow as plans to expand existing facilities are implemented. Other important ports are Santo Domingo, which handles vehicle imports, and La Romana, which handles sugar exports. These two ports, together with Samana in the northeast of the country, are also the main destinations for cruise ships. The primary road network is in overall good condition as a result of a roads program implemented in the late 1990s, although cur- rent levels of maintenance are rather low. As the number of road conces- sions has grown, so have investments in the sector. The condition of the secondary and tertiary networks is not as good, which has a negative impact on access to ports. Airports perform well in the Dominican Republic, playing a key role in an economy that relies strongly on tourism and maquilas. Most airports were given in concession to the private sec- tor during the 1990s, and the trend toward increasing private participa- tion continues as new tourism centers develop. However, many challenges persist with regard to logistics performance in the Dominican Republic. The most important of them is customs administration, which is considered rather inefficient, mostly due to the high degree of discretion of customs officials (particularly those in charge of collecting tariffs and fees). The rates of physical inspection of cargo are high, and documentation requirements are sometimes excessive and remain in paper rather than in electronic form. The development of logis- tics operators is also intermediate; however, increasing freight volumes in Caucedo are expected to attract international logistics operators to the Dominican market, pushing service quality closer to international stan- dards. Finally, the trucking industry, which handles all of the internal flows of freight, is particularly fragmented and operates with very low lev- els of efficiency. El Salvador Throughout the 1990s, El Salvador made important improvements to enhance service provision in the transport sector: roads were rehabili- tated, and the international airport and the main freight port (Acajutla) were developed. Despite these reforms, a comprehensive infrastructure assessment carried out in 2006 found that Salvadoran firms still consider transport infrastructure as the main bottleneck in their export supply 196 Barbero chain, with associated costs representing between 15 and 22 percent of the total cost structure in the industries surveyed (the third-largest cost after labor and raw materials; World Bank 2006a).6 The country’s geogra- phy and foreign trade structure lead to intensive use of Guatemalan and Honduran ports, making the land-sea interface of paramount importance for foreign trade, emphasizing the importance of the primary network and trucking services. The road network faces important challenges: its con- dition is generally poor, and the country’s two main highways (the Carretera Litoral, which runs along the Pacific coast, and the Pan-American Highway, which runs across the country’s interior) are frequently affected by natural disasters. In addition, these major highways, which link the cen- tral region (where production and export activities are concentrated) with key trade nodes, require freight to pass through heavily congested urban areas. To address congestion around the city of El Salvador, a major ring is being built with support of a loan from the Inter-American Development Bank. Secondary and rural roads, particularly in the northern and eastern provinces, create additional challenges in terms of accessibility, as they become impassable during the rainy season. In 2006, the government created a program to address these issues, devoting additional funds for rural roads development. However, based on the findings of the Recent Economic Development in Infrastructure (REDI), trucking services are considered to be highly inefficient and unreliable. Approximately 70 percent of freight trucks in the country are more than 15 years old. The market is extremely fragmented, and the lack of professionalism in service provision generates increasing levels of exter- nalities. Congestion in the border crossings with Guatemala and Honduras also hampers the efficiency of surface transport in El Salvador. As regards waterborne trade, El Salvador’s volume of containerized cargo is close to 600,000 20-foot equivalent units and is mostly shipped via ports in Honduras and Guatemala. Acajutla has been able to capture close to 20 percent of this volume, as a result of increased efficiency and reduced port fees. Still, port performance is considered average (particu- larly for container movement) due to the lack of freight-handling equip- ment. Vessels calling in at Acajutla are, in general, small carriers whose operation is inefficient and costly. The port of La Unión is currently under reconstruction with financing from the Japan Bank for International Cooperation. Plans to concession Acajutla and La Unión have suffered delays due to congressional objections to the bidding rules. Air transport generally performs well: the El Salvador International Airport is the main hub for TACA airlines, which has helped raise the quality of service. Logistics Challenges in Central America 197 Relatively competitive airport fees contribute to an increasingly compet- itive air freight service. Guatemala The Pan-American Highway and the Pacific and Atlantic Highway are the main corridors in the country. Although the network is being expanded (a new trunk road that crosses the Petén region is currently under con- struction) and upgraded, insufficient maintenance affects the state of roads. In 2000, 10 percent of the network was in bad condition, and this grew to 30 percent in 2008. Demand for transport continues to put pres- sure on the system as trade and motorization grow. The metropolitan region is congested, and trucks are not allowed to operate in the area for six hours a day. Only half of the road network is paved, which makes it vulnerable to natural disasters and hampers accessibility, particularly in rural areas. The main ports are Santo Tomás and Puerto Barrios on the Atlantic coast and Puerto Quetzal on the Pacific coast. Ports are mostly publicly owned and operated, with some exceptions. Although the qual- ity of service is not bad, efficiency is low, and too few resources are devoted to investment. As a result, dredging is insufficient, and berths are unable to support modern cranes, so only feeder vessels are able to call in at Guatemalan ports. This has had a negative impact on maritime freight rates. Airports perform adequately with regard to air freight services. Customs administration has gone through an important modernization process; however, coordination among control agencies is still weak. The trucking industry encompasses a relatively old truck fleet, operating at low levels of efficiency. Trade is carried primarily by maritime transporta- tion, although surface transportation is becoming more common. Trucks handle approximately 30 percent of exports and 20 percent of imports. The growing importance of surface transport highlights the need to improve the efficiency of customs administration, which tends to be lower in land border crossings. Honduras The country’s two main corridors are the north-south highway, which connects Tegucigalpa with San Pedro Sula, and the Pan-American Highway, which runs alongside the Pacific coast and connects Honduras with Nicaragua and El Salvador. Although the condition of the road net- work has improved in recent years, the proportion of the national net- work in poor condition is still high, and it is even higher in the rural network, which is critical for agricultural activities. In addition, the low 198 Barbero percentage of paved roads makes the network very vulnerable to natural disasters and increases vehicle operating costs. The trucking industry reg- isters low levels of efficiency, with an extremely fragmented market and an aging fleet. Additional challenges from the logistics perspective arise from long delays and smuggling in the main border crossings (particularly those with El Salvador and Nicaragua) and from poor cargo security, particu- larly around suburban areas. Puerto Cortés, on the northern coast of the country, is Honduras’s main port and one of the most modern deepwater ports in Central America. It handles more than 50 percent of the coun- try’s exports as well as freight originated in or destined for El Salvador and Nicaragua. The levels of efficiency in its operation are considered high; however, the existing equipment is insufficient to handle increasing trade, thus generating delays. Honduras’s four main airports (Tegucigalpa- Toncontin, San Pedro Sula, La Ceiba, and Roatan) are under concession to a national private operator. However, Toncontin was relocated recently to a military base after an accident in 2008. Approximately 20 percent of Honduran trade is transported by surface through the main border cross- ings: El Amatillo (on the border with El Salvador), El Guasaule (on the border with Nicaragua), and El Corinto (on the border with Guatemala). Based on the results of the latest investment climate assessment car- ried out by the World Bank in Honduras (World Bank 2004), firms con- sider transport bottlenecks as one of the main constraints to efficient trade flows. Although overall quality of infrastructure is seen as moderate to regular, most firms consider air transport as the main bottleneck, fol- lowed by ports, particularly for large companies located in free trade zones. These companies are mostly large maquilas, which rely heavily on airports and ports to ship their products. In contrast, micro and small companies consider surface transport as the main bottleneck. Losses resulting from spoilage and breakage constitute 1.8 percent of total sales; 36 percent of firms lose merchandise while in transit (com- pared with 26 percent in Nicaragua and 40 percent in Guatemala; World Bank 2004). The survey indicates that the time to clear customs is ade- quate for exports, but high for imports, registering marked peaks of up to 27 days. These delays affect particularly small firms. Nicaragua Nicaragua’s performance in infrastructure services is among the lowest in Central America (EIU 2008). Given the limited development of ports, air transportation, and railways (which have ceased to exist), the largest Logistics Challenges in Central America 199 bottlenecks to trade lie in the road sector (World Bank 2006c). The quality of Nicaragua’s road infrastructure is the poorest in Central America: based on assessments carried out by the Ministry of Transportation and Infrastructure, more than 75 percent of the total road network was in poor condition in 2006. Nearly 50 percent of the network becomes unusable during the rainy season. In recent years, the primary network has been restored to good condition, but it represents less than 20 per- cent of the total network. Thus, the main challenge is associated with the quality of roads rather than their availability. The Pan-American Highway crosses the country from north to south and is the main road linking Nicaragua with other countries in Central America. It provides the country with access to seaports on the Atlantic coast of the isthmus, principally Puerto Cortés in Honduras and Puerto Limón in Costa Rica. Most of the country’s agricultural exports must be shipped through these ports due to the lack of paved roads linking the Atlantic and the Pacific coasts. Approximately 400,000 tons of freight a year pass through the border crossings with Honduras and Costa Rica. Poor road quality affects rural development as well as competitiveness of agricultural products, which represent more than 80 percent of total exports. Transport services also perform poorly: more than 50 percent of the trucks are more than 10 years old and operate with extremely low levels of efficiency. The country has two ports on the Pacific coast, Puerto Sandino and Puerto Corinto. The latter has recently been rehabilitated, but it still cannot handle volumes sufficient to attract large vessels. Managua International Airport handles mostly air passengers, and the volume of air freight is very low. Assessing Logistics Performance in Central America Factors affecting logistics performance can be grouped as follows: • Factors showing the most serious deficiencies in Central America are roads, ports, domestic and regional surface transportation (carried by the trucking industry), and the security of surface freight. • Factors showing considerable problems, although less severe than the previous ones, are border management and border-crossing facilities. • Factors in which problems exist, but to a lesser extent, are airports, international transport services (air and maritime), carriers’ ability to manage their supply chain efficiently, and logistics operators and inter- mediaries. 200 Barbero The Most Critical Factor: Transport Infrastructure and Services One of the weakest segments, highways have a role not just in freight logistics, but also in general mobility. Road coverage in Central America is relatively weak, with only about 15 percent of the network paved, and quality shows significant shortcomings, “failing to comply with the basic standards to ensure a smooth, safe, and effective regional traffic” (Sánchez and Wilmsmeier 2005). In addition, road infrastructure is vulnerable to frequent natural hazards (hurricanes, earthquakes, volcanic activity). Countries need to overcome those structural deficits, while con- fronting two relevant needs. The first one is the need to increase capacity, as growing trade and motorization are pushing up demand for roads in key corridors and urban areas; bottlenecks are becoming more severe in large metropolitan regions (like San José), in the access to the main gate- ways (particularly ports), and in most important intercity highways pass- ing through towns and villages. The second one is the need to ensure adequate maintenance, which requires additional resources as traffic increases. Private partnerships in highways are still rare in Central America. Poor road quality brings multiple challenges: first, it implies low levels of mobility for rural communities, hindering access to markets, health, and education among the poor. In addition, it affects the compet- itiveness of goods produced in rural areas; this is particularly relevant, as agriculture still represents more than 20 percent of regional GDP and employs close to 50 percent of the total workforce (SIECA 2009). Finally, the dynamism of intraregional trade, as well as growing flows to and from Mexico, which are largely shipped using surface transportation, outlines the need to improve the quality of roads to reduce costs. Ports are key nodes for trade, particularly general cargo and container terminals. Central American countries are relatively small, generating a limited amount of freight, and most have access to two oceans; the his- torical trend has been for each country to develop its own facilities on each shore, causing the development of myriad terminals, each with lim- ited scope.7 As regards port organization, the tendency has been to keep the traditional state-owned tool-port model, without adopting (with a few exceptions) private participation under a landlord organization scheme, as has been done in other Latin American countries, with evident success (Mexico, Uruguay, or Chile).8 Ports play a relevant role at the local level, and in many cases employees (current and retired) and munic- ipalities try to extract as much rent as possible from them and block reform, disregarding the key role of ports as gateways for trade. The lack of strong and efficient regulators has also empowered public port operators, which are highly politicized institutions that tend to reduce Logistics Challenges in Central America 201 the chances of implementing reforms. The combination of numerous small ports with small scale, organized under a low-efficiency operational model, clearly has a negative impact on the region’s competitiveness (about two-thirds of Central American trade is waterborne). Low depth, inadequate berths, and lack of modern handling equipment constrain the type of vessel that shipping lines can deploy in the region. Road freight transportation (truck transportation) is without question the most important domestic mode of transport in the region and has sig- nificant influence on logistics chains. Almost all surface transportation in Central America is carried by trucks, as railways are not operating (with a few exceptions). Trucks have increased their activity in regional trade, which is about one-fifth of total trade in Central America. Despite the importance and complexity of the sector, the lack of systematic data about road transportation is remarkable. Road transportation activities are performed almost entirely by the private sector, highlighting the impor- tance of the regulatory framework. According to SIECA data, there are currently about 20,000 trucks in the region, with more than 100 compa- nies providing international transport services. The remaining trucks are owned and operated mostly by individual owners. In accordance with the few studies available, the sector is inefficient. Typical average distances are 50,000 kilometers a year (which would be two to three times higher in an efficient operation). While truck productivity partially depends on the type of demand (type of product) and the time-space structure (direc- tional imbalance, traffic seasonality), it is also a consequence of the organ- ization of companies, which depends on the regulatory framework established by the state and the professionalism of the business sector. The impact of road transportation on a country’s logistics is greater than what initial analyses may suggest. The improvements that may be achieved in price and service quality not only can reduce the costs and transit time faced by freight providers in the short run, but also can allow companies that produce goods to develop more efficient medium-term supply strategies (Dutz 2005). Other modes of transport could potentially move the freight that is currently moved almost exclusively by truck, such as railways and mar- itime cabotage. Most railways in Central American countries have ceased to operate, and their infrastructure is antiquated. Although opportunity may exist to redevelop their activity, “investments needed to set up a new network with high technical standards would only be possible in corridors whose traffic density allows for a competitive participation” (Sánchez and Wilmsmeier 2005). Maritime cabotage—similar to the short sea shipping (SSS) developed in the European Union—is a long-standing project that 202 Barbero seeks to use ocean routes for domestic and regional trade, without the complex regulations that control ocean trade. The implementation of effi- cient SSS services requires harmonized regulations and specific port infra- structure (Sánchez and Wilmsmeier 2005). The stealing of freight from trucks is prevalent throughout the region. In Costa Rica, for example, cargo that used to be moved by truck is now being transported by air due to lack of security, at costs that are three times higher. Usually robbers target just the freight, but sometimes they also steal the truck (which is later dismantled and sold) and even kidnap the driver. Criminal activity involves products chosen by two criteria: the value of the merchandise and the ease of resale. Typical targets are food and drink, electronics, cigarettes, shoes and clothing, and medicine. This type of crime has multiple impacts: security costs and insurance premi- ums, satellite-positioning systems, guards, private surveillance posts, and other precautions that shippers and carriers need to take. It also under- mines compliance with the client’s requirements, harms the competitive position of the company responsible for the shipment, and diminishes the company’s image and its products when cargo ends up being traded through clandestine channels without quality traceability. Another neg- ative impact is felt in international traffic: in-transit freight is exempt from duties and border tariffs, but, if lost inside a territory, the customs office is obligated under the presumption of fraud to enforce the pay- ment of warranties. In accordance with Central American regulations, the vehicle is the in-transit traffic warranty. Even if it is insufficient, the warranty facilitates the expansion of intraregional trade. However, cus- toms agencies in the region have taken a negative attitude due to the fre- quency of robberies and the weakness of warranties, putting at risk a system that has facilitated trade and integration. Border Management and Border-Crossing Facilities Central American countries have gone through a process of moderniz- ing customs, such as Guatemala’s customs support for air express serv- ices, and the effort has helped to improve trade. But Central American countries, as most developing countries, need to expand the scope of intervention to facilitate trade, considering the entire border manage- ment process. Recent research conducted by the World Bank (2009) concludes that the necessary border reform is more than customs mod- ernization: While improving the performance of customs administrations remains a high priority for many countries, it is only one of the agencies involved in border Logistics Challenges in Central America 203 processing, and evidence suggests it is often responsible for no more than a third of regulatory delays. . . . This highlights the need to focus attention on reforming and modernizing border management agencies other than customs (including health, agriculture, quarantine, police, immigration, standards, etc.). There are several Central American initiatives in this regard, supported by SIECA and other multilateral institutions. DR-CAFTA includes obli- gations aimed at strengthening, improving, and modernizing the opera- tion of customs to facilitate trade among signatory parties. Provisions seek to facilitate customs procedures and reduce room for discretion. The treaty includes rules of origin designed to be easier to administer. It also requires transparency, procedural certainty, and efficiency in administer- ing customs procedures, including DR-CAFTA rules of origin. Central American countries have made a three-year commitment to accomplish goals such as the Internet publication of all norms and regulations, the automation of the clearance procedures, the electronic presentation of certificates of origin, and the implementation of management and risk evaluation systems. All signatories also have agreed to share information to combat the illegal transshipment of goods. A program of technical assistance was agreed to support Central American countries in carrying out their commitments in this area. Similar to other developing countries, Central American countries have recognized the importance of addressing these issues and are look- ing for ways to harmonize, streamline, and simplify border management systems and procedures. This has led to several initiatives: • Coordinated border management, which is based on approaches such as the co-location of facilities, close cooperation between agencies, delegation of administrative authority, cross-designation of officials, and effective information sharing • One-stop border posts, which allow neighboring countries to coordi- nate import, export, and transit processes to ensure that traders are not required to duplicate regulatory formalities on both sides of the same border • Single-window systems, which allow traders to submit all import, export, and transit information to regulators via a single electronic gateway instead of submitting the same information multiple times to different government entities. The impact of trade facilitation and border management is evident. Specific measurement initiatives show that customs procedures and 204 Barbero electronic commerce have an important impact on trade, although less so than port efficiency (Wilson, Mann, and Otsuki 2003). Their influ- ence is generally measured as the rate of trade that a country loses due to inefficient performance and, therefore, as percentage points of GDP that a country loses. The subject has been analyzed in depth by several entities of the United Nations, both multilateral and bilateral, and they have formed an alliance to coordinate their efforts (Global Facilitation Partnership for Transportation and Trade). A recent assessment carried out by the Inter-American Development Bank in the Pacific corridor concludes that border-crossing facilities— infrastructure and general layout—along this route are in very poor con- dition (IDB 2009).9 The report finds that trade facilitation initiatives have resulted in improved lead times to cross the border (particularly among Guatemala, El Salvador, Honduras, and Nicaragua, members of the CA-4 agreement). However, the situation is more complex in the borders between Guatemala and Mexico and between Costa Rica and Panama, creating bottlenecks that are particularly costly in the case of the Mexican border, where trade flows have increased steadily in recent years. According to the assessment, important works should be carried out to improve border facilities (offices and adjacent platforms) and access roads, and a preliminary group of projects could be undertaken using a private participation scheme.10 Other Factors Affecting Logistics Performance Airports have generally improved in Central American countries in the last few years, as they have in most Latin American countries. The increas- ing flow of passengers (due to growing tourism and increased economic activity) has created demand for improvements, which were made largely through public-private agreements. Latin America and the Caribbean is the region with the largest participation of the private sector in airports, both in number of facilities and in investment commitments; the pre- ferred public-private partnership agreement is the concession, in most cases including runways and terminals (World Bank, Private Participation in Infrastructure Database). The improvement has been centered on the main airports, usually one or two per country; many secondary airports still constrain the efficient movement of freight. Although freight move- ments are generally not the driver of airport modernization, freight activ- ities have benefited from it. Usually trade facilitation procedures are better organized and have faster responses in airports than in other gate- ways, particularly with regard to courier and express services. One of the Logistics Challenges in Central America 205 key elements for moving freight is the existence of freight terminals in the airport (or close by), which helps organize and expedite the loading and unloading of freight. These needs become more relevant when per- ishable products are traded: diversification of exports toward nontradi- tional “specialty” products (such as tropical fruits, flowers, or fishing products) highlights the need for coordinated efforts among the public and the private sector. Maritime transportation services basically include shipping lines with regular routes (whose freight is transported mostly in containers), bulk transportation (dry and liquid), and specialized services. Generally speak- ing, the first type of service is usually provided by regular lines, while the others are provided to satisfy specific demands. Markets for shipping serv- ices are habitually competitive, and shipping lines go where demands calls.11 The regular service available for a country depends on the deci- sions that shipping lines make on how to organize the routes, defining ports of call, frequency, and type of vessel. Therefore, although the serv- ice is provided by international commercial companies, it is dependent— to a large extent—on decisions made by governments, particularly with regard to the organization of ports. In Central America the provision of maritime service is adequate, although uneven: Guatemala and the Dominican Republic are above the regional average in the liner shipping connectivity index (prepared by UNCTAD 2008),12 while Nicaragua, Honduras, and El Salvador are below it; Costa Rica’s rating fell after 2006, reflecting a decline in the operational performance of its Atlantic ports. This is mostly the result of the port policy adopted by the countries, the market size, and the existence of alternative ports (and their landside accessibility). There is also some cartelization of refrigerated freight in the shipping lines linking Central America and the Caribbean with the U.S. Gulf of Mexico.13 Air freight transportation has few restrictions on com- mercial access and has more freedom than passenger transport services. A significant part of freight is transported in passenger planes, while the rest is transported in special freight planes. The main restrictions on the expansion of air freight transportation are centered on low demand, higher competitiveness of alternative modes (maritime and roads), and in some cases weak infrastructure at regional airports. A distinctive characteristic of air freight transportation is the need for fast processing during inspection and control procedures, to the point that customs (and other agencies) generally perform better at airports than at ports and border crossings. Large companies (national or international) have usually optimized their supply chain, following the dominant trend, to reduce total logistics 206 Barbero costs by optimizing (basically) inventory levels, transport costs, and qual- ity of service. Small and medium enterprises tend to have much higher logistics costs than large companies, as a result not only of smaller scale, but also of limited capacity to organize the flow of materials throughout the procurement, production, and distribution processes. Partial studies in Latin America show that logistics costs tend to be two to three times higher for SMEs than for large companies. SMEs are important sources of employment, and their competitiveness is of great interest to countries and subnational entities. Support for the logistics development of SMEs can be linked to territorial policies, including the deployment of logistics platforms, which offer the possibility of sharing resources. In recent years, the development of areas for conducting logistics activity has exploded, and the impact has been particularly useful to SMEs. Public cold ware- houses are a special case because Central American countries have very few public access facilities (Global Cold Chain Alliance 2009). The organization of modern logistics practice, as summarized above, has led producing and trading firms to outsource several logistics func- tions, particularly distribution. Under this scheme, a specialized firm (the logistics operator) receives the products and distributes them according to the client’s orders and the shipper’s level of service request. The logis- tics operator has different functions than traditional transport firms or cargo agents, as they provide tailored services, usually under multiyear contracts, and maintain fluid communications with cargo owners. Many transportation firms have adopted this scheme and become logistics oper- ators. This process can be viewed as one of modernization, generally led by market forces, but some agents have had difficulty adapting to the new needs, particularly small or individual trucking operators in remote areas. Logistics platforms could be implemented to enhance the efficiency of trade flows. These types of facilities play an important role in reducing the negative externalities that freight movements cause in densely populated areas. They also play a pivotal role in integrating SMEs to global supply chains by providing economies of scale in transport and inventory man- agement. Efficiency gains are larger when these facilities are in periurban areas, near ports or airports, or in areas adjacent to high-traffic border crossings. Current Initiatives Several regional initiatives are in line with the logistics issues covered in this book. They can be grouped as infrastructure, trade facilitation, trans- port services, and analytical work. Logistics Challenges in Central America 207 In the area of infrastructure, the Mesoamerican project is an agreement that was signed in 2008 between Belize, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, and Colombia. It cov- ers several areas, two of which are related to trade logistics: transport and trade facilitation. In transport, the project includes the construction of a regional highway network: the Red Internacional de Carreteras Mesoamericanas (RICAM).The first RICAM project is the Pacific corri- dor, a high-standard, 3,600-kilometer highway linking all signatory coun- tries. The initiative goes beyond the construction, operation, and maintenance of the highway and includes the modernization of facilities and equipment in border crossings along the corridor. The initiative also addresses road safety, and technical studies are currently under way to identify critical sections and define the required improvements. The Mesoamerican project includes a pillar to address challenges linked to trade facilitation, which is based on implementation of the Procedimiento Mesoamericano para el Tránsito Internacional de Mercancías (TIM), an improvement of the transit procedures aimed at reducing the time required for trucks to cross borders in transit. This proj- ect is being implemented with the support of SIECA, which is also involved in the Proyecto de Diseño y Aplicación de Políticas Comunes Centroamericanas (ADDAPCA) with support from the European Union. Initiatives under this project include regulatory harmonization of com- mercial policies and linked regulations (tariffs, customs procedures, and technical norms) among signatory countries.14 Another relevant initiative in this area is implementation of the Paso Facil among the signatories of CA-4. Paso Facil is a mechanism for expediting procedures at the border. It encompasses initiatives to improve coordination and communications among border agencies in the different countries and to adopt standard documentation. The Comité Técnico Regional Permanente de Transportes (COMI- TRANS), created under the framework of SIECA, comprises the heads of the roads directorates as well as sectoral experts. Its main objective is to reach agreement with a view to harmonizing transport policies in the region. Many relevant initiatives are under way under COMITRANS. First, the committee has recommended the creation of a regional training center for surface transportation companies, with the objective of raising the level of professionalization of firms in the sector. The center is intended to train drivers and other staff employed by transportation com- panies. COMITRANS has also undertaken the task of drafting regional manuals in sensitive areas of transport regulation. The first addresses the 208 Barbero transportation of hazardous materials and waste, which includes actions to strengthen the capacity of public agencies with oversight of the sector. The second intends to harmonize regulations for vehicle inspections, with the ultimate goal of improving the state of the existing truck fleet and road safety conditions. Finally, a regional manual on road safety intends to foster coordination of national policies in this sensitive area. The Comisión Centroamericana de Transporte Marítimo (COCA- TRAM) is leading several initiatives to strengthen harmonization of port policies in the region, such as the definition of a common port strategy and the simplification of maritime procedures. COCATRAM is also working to promote short-distance shipping services within the region. Under the framework of the Mesoamerican project, the viability of increased railway transportation in the region is currently under study. Policy Priorities to Enhance Trade Logistics The analysis in this chapter suggests two main messages: first, logistics performance is relatively weak, and there is ample room for improve- ment; second, the diversity of factors influencing logistics performance calls for a broad range of coordinated activities by the public and private sectors. Traditional trade patterns in Central America and recent trends result- ing from implementation of the DR-CAFTA highlight the need for improved logistics. First, Central America’s export base has traditionally relied heavily on agricultural commodities. Although its relevance has declined in recent years, agriculture continues to be a very relevant sector (encompassing 22 percent of regional GDP and employing 50 percent of the total workforce; SIECA 2009), particularly in Guatemala and Nicaragua. Given their low value added, transport costs (and logistics costs in general) constitute a large share of the total cost of commodities and are thus a key determinant of their competitiveness in international markets. As industry continues to develop in the region, countries have begun to reduce their reliance on agricultural commodities. However, competition from Asian countries (particularly after the expiration of the Multi-Fibre Agreement in 2005) brings to light inefficiencies in the local supply chains. Geographic proximity to the U.S. market has helped maquilas in Central America to offset some of these inefficiencies, but improvements are required if the region is to compete with Asia. The rapid growth of intraregional trade is a positive sign, as Central America’s trade is highly concentrated in the U.S. market. But it also calls for enhanced road infrastructure and harmonized customs requirements, to Logistics Challenges in Central America 209 ensure smoother flows. Finally, the expansion of nontraditional exports (such as tropical fruits or flowers) emphasizes the need to modernize logistics services to comply with the requirements of international cus- tomers, with a focus on improving air freight services and cold chain logis- tics. Based on these conclusions, the policies with the highest priorities pertain to surface transportation, border management, ports, and logistics software and institutions. Surface Transportation There is a need to increase private participation in the road sector. There are several ways to promote public-private partnerships, particularly on highways. Although most of the measures associated with increasing pri- vate participation in roads are under the oversight of national govern- ments, measures can be undertaken at the regional level to achieve stronger synergies. Recommended actions are as follows: • Implement output-based contracts for rehabilitation and maintenance. There are multiple examples in the region, like the contracts for reha- bilitation and maintenance (CREMA) contracts in Brazil, Argentina, and Uruguay or the Prestación Privada de Servicios projects in Mexico. This modality helps to increase efficiency in the allocation of funds for the road sector and could have a substantial impact given the specific challenges this sector faces in Central America (particularly those linked with natural disasters). • Analyze the possibility of concessioning specific sections of the Pacific corridor. Some sections of the corridor register sufficient levels of traffic to make concessioning attractive to private investors. Imple- menting this recommendation is especially feasible in Costa Rica and Guatemala, which have adequate frameworks for public-private partnerships. • Develop (or strengthen, depending on the country) public-private partner- ship frameworks at the national level. This step encompasses not only the legal aspects of such partnerships, but also the institutional capac- ity required to design and implement them efficiently. Efforts to har- monize the public-private partnership frameworks in Central American countries would help to avoid competition among them and thus to attract private investors. In addition, it is important to focus road investments on strategic points—urban bypasses and port access should be prioritized to improve 210 Barbero highway circulation and relieve urban congestion—and to conduct research and collect data on the road sector and security. Informed deci- sion making in infrastructure investment requires reliable data on the condition of roads. The development of a set of harmonized indicators at the regional level would be particularly helpful for managing the shared trade corridors more effectively. Additionally, a road freight security and protection program should be prepared by security forces. Firms may contribute by refining their personnel selection process, mapping the incidence of crime and sharing results with the authorities, loading trucks in ways less attractive to thieves, installing vehicle-monitoring devices, and improving channels of communication with the security authorities. Also important would be undertaking a program to modernize the trucking industry. This effort would include reviewing current regula- tions, considering the inclusion of a unified register, and establishing a cargo document to be issued by carriers (a bill of lading, as is done in most countries). In addition, a program to renew the fleet, promoting the scratching of the oldest vehicles and providing incentives to incorporate new, more efficient, and clean trucks, would improve efficiency. It also would be important to implement a professionalization program, includ- ing mandatory training for workers (linked to the issuance of a profes- sional driver’s license) and entrepreneurs. Good examples are available from the Argentine and Uruguayan experiences. Finally, it is important to harmonize regulations with key trading part- ners. As trade with Mexico increases, the lack of harmonization, forcing the loading and unloading of goods at the border, generates marked inef- ficiencies. The harmonization of standards, at least for the circulation of trailers (switching only tractor units at the border), may help to reduce logistics costs. The setup of logistics zones close to border crossings should be promoted. Ports New port legislation and institutional organization are needed. A new legal and regulatory framework would help enhance the performance of ports. Some of the key issues for the new legal framework, which should be con- sulted with the many relevant stakeholders, are creation of a centralized national policy and administrative organization, incentives to attract pri- vate investment in infrastructure, long-term planning, effective mecha- nisms for users’ participation, and an adequate regulatory entity. Logistics Challenges in Central America 211 Border Management The integration of border control functions is the main challenge to facil- itating trade. For example, conducting joint inspections (customs, agricul- ture and food, narcotics, public health) would eliminate overlapping procedures, which are costly and time consuming. To improve intrare- gional trade flows, efforts could be made to improve the layout of infra- structure and the operational flow of the busiest or most strategic border crossings. Logistics Software and Institutions Efficient business logistics need the support of regulations and promo- tional policies. Two initiatives are proposed in this regard. The first is to facilitate access to warehousing, particularly for SMEs; logistics platforms are the most efficient way to do this. The second is to support the devel- opment of cold chain facilities. Guatemala has a very small capacity of refrigerated warehouses, and they are mostly private, with only 15 percent of the total capacity of refrigerated warehouses available for public use (Global Cold Chain Alliance 2009). In this context, the development of logistics centers would have a very positive effect on trade and help reduce the negative externalities produced by freight operations in densely populated areas. Various Latin American countries have designed networks of logistics platforms (Colombia, Peru); mapping the main value chains and analyzing their performance would help to identify the most appropriate locations and the type of functions that logistics centers should perform. Since logistics centers require the coordinated partici- pation of several government entities, there would be a need to establish a coordination council—a national logistics council—at a high level of government, composed of representatives of the government areas involved (reflecting users and providers of logistics services) as well as representatives of the main private stakeholders. The council might be supported by a small technical office (a logistics observatory), responsi- ble for generating key indicators of performance and knowledge with regard to trade logistics. Notes 1. SIECA’s Regional Trade Statistics are available at http://estadisticas.sieca.int/. 2. Malaysia is classified as an upper-middle-income country, while Thailand is considered a lower-middle-income one. 212 Barbero 3. The 2009 LPI clearly shows that there has been a general improvement in logistics performance in the world. 4. The slight changes in the sample (several countries included in 2007 were no longer present in the 2009 LPI, and some new ones were incorporated) may have some effect when comparing ranks. 5. The LPI is based on perception surveys, which are subjective and, as such, highly sensitive to specific situations (positive or negative) affecting the inter- viewee. This can produce marked variations in the scores assigned. 6. Textiles, chemical products, and processed foodstuffs. 7. The region has 23 ports, 10 in the Caribbean and 13 along the Pacific coast (Sánchez and Wilmsmeier 2005). 8. The Caucedo container terminal, in the Dominican Republic, is the main exception. 9. The Pacific corridor starts in the City of Puebla, Mexico, and stretches along the Pacific coast to reach the City of Panama. It extends more than 3,000 kilo- meters and is the shorter link between Puebla and Panama, which makes it the most efficient corridor for integration in Mesoamerica. 10. The study proposes to focus on the Mexico-Guatemala border crossing (Ciudad Hidalgo-Tecun Uman) and the Costa Rica–Panama border. The large volume of trade that passes through these two crossings makes them attractive for private investors. The other two projects are a grouping of border facilities among CA-4 countries and along the border between Nicaragua and Costa Rica. 11. There is still some cartelization in the shipping services linking Caribbean ports with the U.S. Gulf of Mexico. 12. The index is calculated on the basis of five components: (1) the number of ships; (2) the container-carrying capacity of those ships; (3) the number of companies; (4) the number of services; and (5) the maximum ship size, always referring to the ships that are deployed to provide liner shipping services to a country’s port(s) (UNCTAD 2008). 13. CADA (Central America Discussion Agreement) is an organization that works to obtain common decisions for most of the shipping lines covering this route, as conferences used to do in the past. 14. Costa Rica, El Salvador, Guatemala, Honduras, and Nicaragua. Panama will be included shortly. References Anderson, J. E., and E. van Wincoop. 2004. “Trade Costs.” Working Paper on Economics 593, Boston College, Department of Economics, Boston, MA. Bernard, A., J. B. Jensen, and P. K. Schott. 2006. “Trade Costs, Firms, and Productivity.” Journal of Monetary Economics 53 (5): 917–37. Logistics Challenges in Central America 213 Bernard, A., J. B. Jensen, S. Redding, and P. K. Schott. 2007. “Firms in International Trade.” Journal of Economic Perspectives 21 (3): 105–30. Dutz, M. 2005. “Road Freight Logistics, Competition, and Innovation: Downstream Benefits and Policy Implications.” Policy Research Working Paper 3768, World Bank, Washington, DC. EIU (Economist Intelligence Unit). 2008. “Costa Rica Country Profile.” EIU, London. ———. 2009. “Latin America Economy: DR-CAFTA’s Progress amid Recession.” EIU, London, November 26. Hummels, D. L. 2001. “Time as a Trade Barrier.” Purdue University, Department of Economics, Krannert School of Management, West Lafayette, IN. Hummels, D. L., and G. Schaur. 2009. “Hedging Price Volatility Using Fast Transport.” NBER Working Paper 15154, National Bureau of Economic Research, Cambridge, MA. Global Cold Chain Alliance. 2009. Global Cold Chain Logistics Report 2008–2009. London: Global Cold Chain Alliance and Transport Intelligence. www.transportintelligence.com. IDB (Inter-American Development Bank). 2009. “Proyecto Mesoamérica / Hojas Informativas.” IDB Bank, Washington, DC. http://portal2.sre.gob.mx/mesoamer ica/dmdocuments/Hoja%20Informativa%20Transporte.pdf. Limao, N., and A. J. Venables. 2001. “Infrastructure, Geographical Disadvantage, Transport Costs, and Trade.” World Bank Economic Review 15 (3): 451–79. Melitz, M. J. 2003. “The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity.” Econometrica 71 (6): 1695–725. Private Participation in Infrastructure Database. World Bank, Washington, DC. http://ppi.worldbank.org. Sadikov, A. M. 2007. “Border and Behind-the-Border Trade Barriers and Country Exports.” IMF Working Paper wp/07/292, International Monetary Fund, Washington, DC. Sánchez, R., and G. Wilmsmeier. 2005. “Bridging Infrastructural Gaps in Central America: Prospects and Potential for Maritime Transport.” Economic Comission for Latin America and the Caribbean, Santiago. SIECA (Secretaría de Integración Económica Centroamericana). 2009. La inte- gración económica centroamericana ante los efectos de la crisis económica interna- cional. Guatemala. http://www.sieca.org.gt/site/VisorDocs.aspx?IDDOC= Cache/17990000003083/17990000003083.swf. UNCTAD (United Nations Conference on Trade and Development). 2008. Liner Connectivity Index. Geneva. WEF (World Economic Forum). 2009. Global Enabling Trade Report. Davos, Switzerland: WEF. 214 Barbero ———. 2010. Global Enabling Trade Report. Davos, Switzerland: WEF. Wilson, J. S., C. L. Mann, and T. Otsuki. 2003. “Trade Facilitation and Economic Development: Measuring the Impact.” Policy Research Working Paper 2988, World Bank, Washington, DC. ———. 2005. “Assessing the Benefits of Trade Facilitation: A Global Perspective.” World Economy 28 (6): 841–71. World Bank. 2004. “Honduras Investment Climate Assessment.” World Bank, Washington, DC. ———. 2006a. “Auditoria sobre facilitación del transporte y el comercio.” Working Paper, World Bank, Washington, DC. ———. 2006b. “Costa Rica Country Economic Memorandum.” World Bank, Washington, DC. ———. 2006c. “Project Appraisal Document: Nicaragua 4th Road Maintenance and Rehabilitation Project.” World Bank, Washington, DC. ———. 2007a. “Investment Climate Assessment.” World Bank, Washington, DC. ———. 2007b. “Public Expenditure Review.” World Bank and Inter-American Development Bank, Washington, DC. ———. 2009. “Border Management Modernization: A Practical Guide for Reformers.” World Bank, Washington, DC. Yeaple, S. R. 2005. “A Simple Model of Firm Heterogeneity, International Trade, and Wages.” Journal of International Economics 65 (1): 1–20. CHAPTER 8 Access to Credit and Productivity in Central America Inessa Love, Teresa Molina Millán, and Rashmi Shankar As we have seen elsewhere in this volume, trade liberalization is an oppor- tunity that calls for significant policy effort if countries are to realize the potential benefits. Complementary policies are examined in depth in var- ious companion pieces to this chapter, with a view to understanding how the countries of Central America can become competitive enough to take advantage of the improved market access granted by the Dominican Republic–Central America Free Trade Agreement (DR-CAFTA). Productivity improvements are an important component of this enhanced competitiveness. Here we focus on the relationship between productivity improvements, access to finance, and likelihood of exporting. A well-functioning financial system is an important component of eco- nomic growth and development. Numerous studies have found a strong and significant relationship between the level of financial development and long-run growth (Beck, Levine, and Loayza 2000; Calderón, Fajnzylber, and Loayza 2004, among others). Specifically, financial intermediaries and markets help to reallocate credit to its most productive uses and reduce transaction costs and information frictions (Levine 1997; Love 2003). The international trade literature has traditionally focused on factor endowments, technology, and scale economies as sources of comparative 215 216 Love, Molina Millán, and Shankar advantage and used these as determinants of trade flows between coun- tries. More recently, the role of finance has been named as another important influence on the pattern of international trade flows (Beck 2002). Thus, financial development may serve as a source of compara- tive advantage that might influence a country’s trade patterns. A growing literature has found evidence of the comparative advantage that financial development provides to exporters or firms entering foreign markets. In particular, some papers have argued that financially developed countries export relatively more in sectors that require more outside finance. For example, in a cross section of 56 countries and 36 industries, Beck (2002, 2003) and Svaleryd and Vlachos (2002, 2005) show that countries with better-developed financial systems have higher export shares and trade balances in industries that use more external finance. In another cross-sectional analysis for 1995, Becker and Greenberg (2005) reach a similar conclusion using different industry measures of fixed up- front costs. Similarly, Hur, Raj, and Riyanto (2004) show that a better financial environment is associated with a larger 1980–90 average share of exports in sectors with fewer internal funds and hard assets. Manova (2008) presents further evidence on the causal influence of finance on trade by exploiting shocks to the level of local financial development. She uses incidences of stock market liberalizations and shows that liberalization increases exports disproportionately more in financially vulnerable sectors that require more outside finance or employ fewer collateralizable assets. Such evidence confirms that finance has a causal influence on trade; it does not simply follow trade and growth. Another link between local financial development and trade is explored in Manova and Chor (2009), who show that tighter credit con- ditions, as measured by interbank lending rates, reduce a country’s exports to the United States. This reduction is even more pronounced in industries that are likely to face more financing constraints (such as indus- tries that require extensive external finance, have few collateralizable assets, or have limited access to trade credit). They argue that financially vulnerable industries are more sensitive to limited availability of finance and, further, that this sensitivity increased during the recent financial crises. In addition, they find that exports of countries with stronger pre- crisis fundamentals were less sensitive to a decline in trade that followed the crisis, suggesting that stronger financial markets may mitigate the oth- erwise damaging impact of a crisis. Access to Credit and Productivity in Central America 217 These studies show that development of financial markets and inter- mediaries has a significant and causal impact on trade. There are many channels through which local financial development may influence trade. One channel is direct—the provision of trade finance to exporting or importing firms. Trade finance is a critical part of the institutions that countries need to take full advantage of trade-related opportunities. As some put it, trade finance “oils the wheels of trade,” which is especially vital for countries with limited access to finance (World Bank 2003). The importance of trade finance is further underscored by the following quote (italics added) from Auboin (2007): Since more than 90 percent of trade transactions involve some form of credit, insurance, or guarantee, one can reasonably say that trade finance is the lifeline of trade. Producers and traders in developing or least-developed countries need to have access to affordable flows of trade financing and insurance to be able to import and export, and hence integrate in world trade. From that per- spective, an efficient financial system is one indispensable infrastructure to allow trade to happen. Another important channel through which financial development can affect trade is its impact on productivity, which allows firms to gain com- parative advantage in global markets. Several models provide theoretical justification for the proposition that credit affects growth through its impact on productivity. In these models, the financial sector provides real services by alleviating information and transaction costs, in particular making the longer-gestation higher-return projects more attractive (see, for example, Levine 1991; Bencivenga, Smith, and Starr 1995). However, the existing empirical evidence on this channel is still limited. At the macro level, Easterly and Levine (2001) show that total fac- tor productivity (TFP) accounts for most of the cross-country variation in economic development and growth. They go as far as to claim that factor accumulation is not important for future growth, but productiv- ity is. Levine and Zervos (1998) argue, “The major channel through which growth is linked to stock markets and banks is through produc- tivity growth.” Combining macro and micro data, Jeong and Townsend (2005) define a growth model with micro foundations and find that 73 percent of TFP growth in Thailand between 1976 and 1996 was the result of occupational shifts and financial deepening. However, rapid credit growth accompanied by resource misallocation could have an adverse impact on productivity. For example, Ghani and Suri (1999) 218 Love, Molina Millán, and Shankar argue that the rapid growth of bank credit was associated with negative growth of productivity in Malaysia because the allocation of credit was inefficient. Among related papers, Bernstein and Nadiri (1993) estimate the effect of financial structure on productivity growth in U.S. manufacturing com- panies. Their focus is on estimating the impact of the agency cost of debt and the signaling benefits of dividends on productivity growth. Nickell and Nicholitsas (1999) find that financial pressure (defined as the ratio of interest payments to cash flow) has a positive effect on productivity. Using data from the United Kingdom and Italy, Schiantarelli and Sembenelli (1999) show that firms with a larger proportion of long-term debt in their capital structure have improved subsequent performance, measured as profitability, sales growth, and total factor productivity. Similar patterns are found in Schiantarelli and Jaramillo (1999) for Ecuador and Schiantarelli and Srivastava (1999) for India. However, due to data limitations, all these studies focus on the effect of leverage on productivity. Several recent papers estimate the productivity impact of investment climate variables. Escribano and Guasch (2005) analyze the impact of dif- ferent variables from investment climate assessments in Guatemala, Honduras, and Nicaragua. Among other results, they find that, by engag- ing in an external audit of their financial statements, firms could increase their productivity. Following their methodology, Fajnzylber, Guasch, and López (2008) conduct a similar study using investment climate assess- ment data, but their pooled database includes 16 Latin American coun- tries. They also look at the productivity effect of indicators of governance, infrastructure, access to finance, and technology. However, they do not obtain consistent estimators of the productivity effect of access to credit indicators when they use their aggregate database, probably due to the heterogeneity among subregions in Latin America. Numerous studies have also established a significant positive relation- ship between export expansion and economic growth via a productivity effect. The early literature on finance and growth established the robust relationship between the level of financial development and long-run growth (King and Levine 1993; Levine 2005). Once the link was estab- lished at the macro level, subsequent literature has aimed to understand the channels through which such a link may operate. Rajan and Zingales (1998), Demirgüç-Kunt and Maksimovic (1998), and Wurgler (2000) argue that finance may operate by allowing firms to invest in potentially profitable growth opportunities and thus support efficient allocation of Access to Credit and Productivity in Central America 219 capital. Love (2003) uses the Euler-equation approach and shows that financing constraints are more severe in countries with lower levels of financial development. Productivity growth appears to be an important channel through which finance may affect overall economic development and growth. Gatti and Love (2006) use firm-level data and confirm a positive relation- ship between a firm’s access to finance and its TFP. Ayyagari, Demirgüç- Kunt, and Maksimovic (2007) find that access to finance has a significant impact on the rate of innovation, which can also be linked to TFP and growth. Two mechanisms are hypothesized. First, the export-led growth theory suggests that exporting enhances productivity growth through a learning-by-doing process. Exporters improve productivity because they enter foreign markets, which increases the competitive pressures on them, while also enabling them to exploit economies of scale. Firms that enter the international market are also more likely to acquire new tech- nology, which, in turn, contributes to productivity improvements (Almeida and Fernandes 2008). However, an alternative theory highlights a self-selection process through which only competitive firms enter foreign markets. Firms that export incur additional costs, perhaps to modify domestic prod- ucts for foreign consumption, for transportation, distribution, or mar- keting, or for skilled personnel to manage foreign networks. These costs are entry barriers that more productive firms are more likely to be able to overcome (Roberts and Tybout 1997; Bernard and Jensen 1999). Export markets are also likely to be more competitive than domestic markets, making it harder for less productive firms to enter. Firms might even be forward looking, with the desire to export leading them to improve productivity so as to become competitive in foreign mar- kets (Wagner 2007). Both mechanisms are plausible, but their importance is likely to vary across countries and industries. In fact, more evidence has found that, in the self-selection process, more efficient firms enter the export market and that this is the main reason why exporters are more productive than nonexporters (Yan, Chung, and Roberts 2000). In this chapter we analyze the relationship between productivity, access to finance, and exports using a sample of manufacturing firms in Central America. Since we run a cross-sectional regression with one year of data, we are not able to infer causality. Rather, we present a plausible argument of how higher productivity may set off virtuous circles of opportunity and growth in Central America. 220 Love, Molina Millán, and Shankar Productivity and Access to Financial Services in Central America Assessing the role of financial services in determining productivity is par- ticularly relevant in the context of regional integration in Central America. Although the region experienced high GDP growth during the years preceding the current financial crisis, productivity gains were lim- ited and should be the focus of both public policy and private sector strat- egy. Figure 8.1 explores differences in TFP growth between the median country of Central America and other regions. This evidence suggests that Central America’s TFP growth has not kept pace with that of other regions. In fact, except for the period from 1991 to 1995, Central America’s TFP growth rates have been lower than those of most other regions. However, differences with Latin America’s median country are not large. While Latin American TFP growth has improved in the past five years, both Latin America and Central America have been at the lower end of the sample since 1996. While East Asia and the Pacific, Central Asia, and South Asia experienced TFP growth of more than 1.5 percent between 2001 and 2005, TFP growth in Central America contracted slightly (–0.3 percent). From 1981 to 2005, Central America experienced average productivity growth of 0.34 percent a year, similar to Latin America and well below East and South Asia. There is, however, significant country heterogeneity within the Central American region, as shown in figure 8.2, and variation across time for individual countries. Nicaragua and Honduras are well behind the rest Figure 8.1 TFP Growth Rates, by Region, 1981–2005 4 2 growth rate (%) 0 –2 –4 e be nd ic As ral Af orth ia a ica ric m cif As nt rib a a er an ia a co Af Pa ric N Ce h Am Ca ic in ut n d d e er ra an d gh So an l an Am ra ha st Hi ia nt Sa pe Ea As tin Ce b- ro th e La st dl Su Eu Ea id M 1981–85 1986–90 1991–95 1996–2000 2001–05 Source: Auhtors’ calculations. Access to Credit and Productivity in Central America 221 Figure 8.2 TFP Growth Rates in Central America, by Country, 1981–2005 1.5 1.0 0.5 0 growth rate (%) –0.5 –1.0 –1.5 –2.0 –2.5 –3.0 ica or a as ua al ad r aR m du ag lv te n r st Sa ca a Ho Co Gu Ni El 1981–95 1996–2005 Source: Authors’ calculations. of the countries within the region. In the last decade Guatemala achieved the most improvement in productivity. The financial sector in Central America grew substantially in the last decade. The average credit-to-GDP ratio rose from 29.41 percent in 1998 to 41.35 percent in 2008, while average M2 (measure of money supply) to GDP rose from 30.5 to 37.8 percent. Although financial depth varies sig- nificantly from one country to another (see figure 8.3), financial intermedi- ation in Central America is above the average in Latin America. Another financial indicator that reflects the level of development of the region’s financial system is the cost of banking services. The interest rate margins (the accounting value of a bank’s net interest revenue as a share of its total earning assets) and overhead costs (the accounting value of a bank’s over- head costs as share of its total assets) are taken as measures of the efficiency of the financial system. Central America’s margins are similar to the aver- age of Latin American countries, but larger than those of other developing countries. Within Central America, banks in Guatemala have the highest costs. Moreover, access to credit remains limited in Central America. The composite indicator developed by Beck, Demirgüç-Kunt, and Martínez Peria (2007) measures the percentage of the adult population with access to an account with a financial intermediary. The entire region is well behind the median of Latin America, especially Nicaragua. Although international comparative data on the extent to which firms and house- holds use financial services remain limited, this index shows how far 222 Love, Molina Millán, and Shankar Figure 8.3 Credit to the Private Sector and M2 as a Percentage of GDP in Central America, by Country, 2008 60 50 40 share of GDP (%) 30 20 10 0 Costa Rica El Salvador Guatemala Honduras Nicaragua credit to the private sector M2 Source: Beck, Demirgüç-Kunt, and Levine 2000; World Bank, World Development Indicators. Central America is from other Latin American countries in terms of financial breadth. Beck, Demirgüç-Kunt, and Martínez Peria (2007) also construct indicators of geographic barriers to accessing financial services. Table 8.1 reports the geographic and demographic penetration of bank branches and automated teller machines (ATMs) in the region. Costa Rica, El Salvador, and Guatemala are around the Latin American level of pene- tration, while Honduras and Nicaragua are well behind. These indicators are only crude proxies for geographic access, however, since branches and ATMs are never distributed equally across a country but are clustered in cities and some large towns. Overall, even though the financial system has grown, access to finance remains a development issue for public policy, and productivity growth remains weak. Is there a link between lack of financial access and produc- tivity for Central American firms? We next discuss the data used to address this question. The Data We use data from the 2005 and 2006 rounds of the World Bank enter- prise survey for Central America. We construct a pooled sample with Access to Credit and Productivity in Central America 223 Table 8.1 Geographic and Demographic Penetration of Branches and ATMs in Central America, by Country Branch ATM Country Geographic Demographic Geographic Demographic and region penetration penetration penetration penetration Costa Rica 7.52 9.59 10.07 12.83 El Salvador 14.58 4.62 34.89 11.07 Guatemala 11.49 10.12 22.93 20.20 Honduras 0.46 0.73 2.22 3.56 Nicaragua 1.29 2.85 1.18 2.61 Latin America and the Caribbean 4.94 7.50 10.61 12.51 Source: World Bank 2003. Note: Geographic penetration refers to the number of branches or ATMs per 1,000 square kilometers. Demographic penetration refers to the number of branches or ATMs per 100,000 people. Table 8.2 Distribution of Firms in Central America, by Country and Size of Firm Number Size of firma Country Year of firms Micro Small Medium Large Costa Rica 2005 343 170 111 27 35 El Salvador 2006 467 122 190 53 102 Guatemala 2006 328 90 136 37 65 Honduras 2006 263 96 93 31 43 Nicaragua 2006 365 160 156 30 19 Total 1,766 638 686 178 264 Source: Authors’ calculations a. Firm size is defined as follows: micro, less than 10 employees; small, from 11 to 50 employees; medium, from 51 to 100 employees; large, more than 100 employees. information on more than 1,700 firms in the manufacturing sector from five countries in Central America. Table 8.2 reports the distribution of firms in our sample, by country and size. The sample is heavily dominated by micro and small firms, while medium and large firms constitute about a quarter of the sample. We follow the methodology in Gatti and Love (2006), who find credit to be positively and strongly associated with TFP using data from a cross section of Bulgarian firms. The survey has several indicators of usage of financial products. Firms report whether they have a credit line or an overdraft facility. As one of our indicators of financial products usage, we use a variable (line) taking the 224 Love, Molina Millán, and Shankar Table 8.3 Use of Financial Products in Central America, by Size of Firm % of firms Product Micro Small Medium Large Total Credit line 46 64 79 88 63 Checking account 74 92 95 95 86 Financial user 82 94 98 98 91 Financial obstacle 35 31 22 17 29 Source: Authors’ calculations. value of 1 if the firm has either overdraft or a credit line and 0 otherwise. We combine overdrafts and credit lines together, as both instruments rep- resent easy access to immediate liquidity and both have short-term matu- rity. In addition, in some of the surveys (such as Costa Rica) the survey instrument does not allow us to separate lines of credit from overdraft usage. About 63 percent of firms in the sample have an overdraft facility, a line of credit, or both. Credit availability increases monotonically with firm size: 88 percent of large firms and only 46 percent of microenter- prises have a credit line or overdraft facility. Table 8.3 reports the distri- bution of firms by their use of financial products and by size. We also use an indicator for firms that use checking accounts. According to the literature, especially for microenterprises, the use of checking and savings products is equally or even more important than the use of credit products. About 86 percent of firms in the sample have a checking account. Among microenterprises, only 74 percent of all firms have a checking account, but among medium and large enterprises almost all firms have a checking account (95 percent in our sample). Although the use of checking accounts increases with size, the difference among size categories is smaller than in the use of credit products. This is likely because it is easier to open a checking account than obtain a line of credit or overdraft facility, especially for smaller firms. Finally, we create an indicator variable for firms that use any financial product—that is, whether a checking account, overdraft facility, or line of credit. We call it financial user. This variable identifies firms that have any interaction with the formal financial sector and those that do not. Almost 91 percent of firms in the sample use at least one of these finan- cial instruments. This again is monotonically related to size—82 percent of microenterprises, 94 percent of small firms, and 98 percent of medium and large firms use at least one financial instrument. Thus most Access to Credit and Productivity in Central America 225 firms in our sample use at least one financial product, and only a small percentage of firms are excluded (voluntarily or involuntarily) from the formal financial sector. The survey also asks firms to rank various obstacles to doing business (rankings range from no obstacle to major obstacle). We create a dummy variable that equals 1 if the firm reports access to finance as a major or a severe obstacle and 0 otherwise. This measure of access is subjective in that it reports respondents’ perceptions of the severity of the obstacle, whereas previously discussed measures capture a more objective dimen- sion: actual use of financial products. To summarize, we use one objective and three objective measures of the use of financial services: credit line, checking account, financial user, and financial obstacle. Descriptive statistics are presented in table 8.4. Table 8.5 reports correlations among our four financial indicators. We observe a relatively low correlation between credit line and check- ing account use—about 23 percent. By construction, financial user is highly correlated with checking account or line of credit. The financial obstacle variable is not correlated with use of a checking account. This suggests again that checking account access is not perceived as a major or severe obstacle by firms in our sample. However, the financial obsta- cle variable is negatively correlated with credit line use. Thus, firms without access to a credit line are more likely to claim that financial access is a major or severe obstacle to operation of their business. When we tabulate the use of credit products against financial obstacle (not reported), we find that about 26 percent of firms with access to a credit line consider access to finance as a major or severe obstacle, compared to 35 percent of firms without access to a credit line. While the corre- lation is not large—only –10 percent—it is statistically significant at the 10 percent level. The survey asks firms to report their share of domestic and external sales. We use this information to construct two variables used in our model to assess the relationship between productivity and exports. First, we construct a dummy that equals 1 if the firm exports at least 5 percent of its sales. We find that less than 15 percent of firms in our survey are exporters. We also use another proxy for export status, constructed as the share of domestic sales in total sales. The dummy variable measure has a more straightforward interpretation, and therefore we use it as the main measure in the regressions discussed below. 226 Table 8.4 Descriptive Statistics for Survey Variables 25th 50th 75th Standard Variable Minimum percentile percentile percentile Maximum Mean deviation Number Log sales 4.52 10.30 11.69 13.41 19.82 11.92 2.26 1,667 Log value added 4.23 9.79 11.19 12.85 19.79 11.40 2.26 1,534 Log capital 3.13 8.41 9.98 11.62 18.43 10.11 2.41 1,333 Log labor 0.00 1.95 2.83 3.95 8.23 3.09 1.46 1,766 Log sales per worker 3.42 7.78 8.98 9.85 15.42 8.84 1.52 1,667 Log sales per capital –4.32 0.92 1.85 2.89 9.30 1.96 1.60 1,312 TFP –4.91 –0.63 –0.02 0.53 6.92 –0.01 1.06 1,312 TFP value added –4.41 –0.63 –0.03 0.56 7.44 0.00 1.12 1,261 Log age 0.00 2.30 2.83 3.37 4.80 2.80 0.81 1,760 LLC 0.00 0.00 0.00 1.00 1.00 0.45 0.50 1,765 Credit line 0.00 0.00 1.00 1.00 1.00 0.63 0.48 1,764 Checking account 0.00 1.00 1.00 1.00 1.00 0.86 0.34 1,762 Financial user 0.00 1.00 1.00 1.00 1.00 0.91 0.29 1,763 Financial obstacle 0.00 0.00 0.00 1.00 1.00 0.29 0.46 1,742 Source: Authors’ calculations. Access to Credit and Productivity in Central America 227 Table 8.5 Correlations among Financial Indicators Checking account Credit line Financial user Credit line 0.2371* 1 Financial user 0.7912* 0.4094* 1 Financial obstacle 0.0024 –0.1004* –0.0022 Source: Authors’ calculations. *p < .10. Estimating Productivity Firms also report the value of total sales and fixed assets as well as infor- mation on the number of employees, wages, and input costs. We use this information to obtain estimates of total factor productivity. TFP growth is often thought to be the result of product or production process inno- vations that increase the value produced using the same amount of factor inputs (that is, capital and labor). However, TFP also reflects unobserved improvements in the quality of factors of production related to human capital accumulation or the upgrading of physical infrastructure. Recent years have seen a surge of interest in productivity analysis, and an extensive literature about different methodologies for estimating pro- ductivity has been developed. In this chapter we use a production func- tion method that is based on the stochastic frontier approach and a parametric translog cost function to estimate the efficiency frontier. However, econometric issues arise when we estimate productivity as the difference between actual output and output estimated by a production function using actual quantities of inputs, as firm productivity can affect the choice of inputs. For example, firms that receive a productivity shock may alter their mix of inputs. This implies that the error and the regres- sors in our model might be correlated and that coefficient estimates obtained with ordinary least squares (OLS) might be biased. Various solutions have been proposed in the literature to overcome this problem. These include using country fixed effects that would deal with time- invariant individual effects and an instrumental variable strategy for choice of inputs. Although we are able to control for country and sector- specific effects, we do not have enough observations to build instrumen- tal variables for input values. As an alternative to fixed-effect regressions, Olley and Pakes (1996) develop a consistent semiparametric estimator, which solves the simul- taneity problem by using a firm’s investment choice to proxy unobserved productivity shocks. A strictly monotonous relationship between the proxy and output has to be met to obtain consistent estimates using this 228 Love, Molina Millán, and Shankar technique. This implies that any firm with no investment has to be dropped from the data, which reduces considerably the number of obser- vations. Following this methodology, Levinsohn and Petrin (2000) argue that using information on choice of intermediate inputs such as demand for electricity—which tracks productivity shocks quite closely and cannot be stored—allows one to control effectively for productivity shocks and thus obtain consistent and unbiased estimates of capital and labor elasticity (see the discussion in Hallward-Driemeier, Iarossi, and Sokoloff 2002). However, we cannot follow either of these procedures since they do not fit the characteristics of our data set. The principal caveats that we face in our sample are the lack of information on the previous year’s inputs and the reduced number of firms that report information on input proxies as expenditure on research and development or investment. We use two measures of total factor productivity and one measure of labor productivity. The first measure is obtained as a residual from a regression with log sales as a dependent variable. The second measure is a residual from a regression with log value added as a dependent variable, and the third measure is the ratio of sales to number of employees.1 Table 8.6 presents production function estimates obtained using pooled OLS across countries. Labor is computed as the number of employees, while capital is the stock of fixed assets at the end of the pre- vious fiscal year. We include country dummies to capture differences among countries. In an alternative specification, we add dummies for sub- sectors, but the results remain unchanged. The share of capital is esti- mated to be about 0.25, while the share of labor in output is close to 0.9. The sales regressions are estimated with an R2 above 0.7. The second col- umn reports a regression with value added as a dependent variable, which is used to obtain our main measure of TFP. The last column contains a regression of the determinants of labor productivity. We find consistent results across three regressions. Among countries, firms from Costa Rica (the omitted category in all regressions) have the highest productivity, independent of the measure of productivity used, while firms from El Salvador have the lowest. Our two measures of TFP are obtained as residuals from regressions in models 1 or 2, while our measure of labor productivity is the actual log of sales per worker (that is, the dependent variable in model 3). Table 8.7 reports correlations among our three TFP measures. We find that both TFP measures obtained as residuals are highly correlated (cor- relation of nearly 90 percent), while labor productivity has a correlation of about 68 percent with each of the TFP measures. Access to Credit and Productivity in Central America 229 Table 8.6 TFP Estimation Log value Log sales per Variable Log sales added worker Log capital 0.253*** 0.245*** (0.000) (0.000) Log labor 0.976*** 0.956*** (0.000) (0.000) Log capital per worker 0.274*** (0.000) El Salvador –1.597*** –1.446*** –1.376*** (0.000) (0.000) (0.000) Guatemala –0.228** –0.064 –0.219* (0.049) (0.671) (0.058) Honduras –0.266*** 0.011 –0.277*** (0.007) (0.918) (0.007) Nicaragua –0.559*** –0.271** –0.522*** (0.000) (0.031) (0.000) Constant 7.055*** 6.491*** 7.488*** (0.000) (0.000) (0.000) Number of observations 1,312 1,261 1,312 R2 0.723 0.67 0.327 Source: Authors’ calculations. Note: Robust p values are in parentheses. * p < .10, ** p < .05, *** p < .01. Table 8.7 Correlations among Estimated TFP and Observed Labor Productivity TFP value Labor TFP added productivity TFP 1 TFP value added 0.895*** 1 Labor productivity 0.764*** 0.679*** 1 Source: Authors’ calculations. *** p < .01. TFP and Usage of Financial Products We use estimated TFP to assess the relationship between the use of finan- cial services and firms’ productivity. Given the high correlation among the different estimates of TFP discussed in the previous section, we pres- ent the results using the TFP measure obtained from the value added regression and labor productivity. We regress estimated TFP on country and firm characteristics described in the previous section. 230 Love, Molina Millán, and Shankar To control for firms’ characteristics with regard to productivity, we employ a rich set of control variables such as size, legal status, age, and landownership. In addition, we introduce country dummies to control for country fixed effects. TFPi = α + qFIt + bXt + g EXPt + et , (8.1) where TFP is the estimated residuals from equation 8.1; FI = (Checkingd; Credit lined; Financial userd; Financial Obstacled); and Xt = {log(age), LLCd Landd, Sized Countryd}; and EXP = Exportsd. Table 8.8 reports the basic results for our OLS model. We find that firms’ age has no effect on productivity, probably because we are looking at a sample with a majority of old firms (firms’ average age in our sample is about 17 years). We do not find either a significant coefficient for the size dummy vari- ables on TFP, partially because the sample is dominated by micro and small firms. However, we find that in terms of labor productivity, micro firms are significantly below small firms, while small firms are signifi- cantly below medium and large firms (see table 8.9). This could be because of differences in education or labor skills, which are likely to be much lower in micro and small firms, due to the nature of their business. We also control for other variables that could have an effect on pro- ductivity. We find that the legal status of the firm has an impact on the level of productivity. We define variable LLC as a dummy that equals 1 if the firm has limited liability (it includes publicly listed and privately held shareholding companies, limited liability companies, and limited partnerships) and 0 if the firm is a sole partnership or a partnership with unlimited liability. Limited liability companies seem to be more pro- ductive, and the effect is robust to different specifications. By contrast, landownership or belonging to a group of firms has no significant effect on the level of productivity of the company. Productivity appears to be higher in exporting firms. We report the results with a dummy variable to account for the effect of being an exporter (dummy defined as 1 if the firm exports at least 5 percent of its sales). In an alternative specification, we use another proxy for export sta- tus (the reverse)—the share of domestic sales in total sales—and we obtain a significant and negative relationship. The OLS estimate suggests that being an exporter firm is associated with an increase in productivity Table 8.8 Access to Credit and TFP value added TFP value added Variable (1) (2) (3) (4) (5) (6) Checking account dummy 0.251* 0.192 (0.06) (0.13) Line of credit dummy 0.194** 0.153* (0.02) (0.05) Financial user dummy 0.342** (0.02) Financial obstacle –0.046 (0.57) Micro dummy 0.057 0.075 0.114 0.117 0.076 0.056 (0.65) (0.55) (0.38) (0.36) (0.55) (0.66) Small dummy 0.026 0.011 0.062 0.041 0.024 0.03 (0.81) (0.92) (0.56) (0.70) (0.82) (0.78) Medium dummy 0.078 0.081 0.077 0.081 0.068 0.076 (0.55) (0.53) (0.55) (0.53) (0.60) (0.56) Log of age 0.013 0.007 0.013 0.009 0.013 0.014 (0.79) (0.87) (0.77) (0.85) (0.77) (0.77) LLC 0.278*** 0.25** 0.275*** 0.256** 0.248** 0.267*** (0.01) (0.01) (0.00) (0.01) (0.01) (0.01) Direct exports dummy 0.184** 0.174* 0.174* 0.17* 0.172* 0.181* (0.05) (0.06) (0.05) (0.06) (0.06) (0.05) Landownership –0.056 –0.065 –0.07 –0.074 –0.063 –0.055 (0.55) (0.48) (0.45) (0.42) (0.48) (0.57) El Salvador –0.116 –0.124 –0.181* –0.173 –0.126 –0.12 (0.28) (0.24) (0.09) (0.11) (0.23) (0.27) 231 (continued next page) 232 Table 8.8 (continued) Variable (1) (2) (3) (4) (5) (6) Guatemala 0.193 0.208 0.134 0.159 0.205 0.186 (0.12) (0.10) (0.26) (0.19) (0.10) (0.13) Honduras 0.244** 0.212* 0.197 0.179 0.235* 0.228* (0.05) (0.09) (0.11) (0.15) (0.06) (0.07) Nicaragua 0.142 0.188* 0.119 0.16 0.177 0.13 (0.20) (0.09) (0.28) (0.14) (0.11) (0.25) Constant –0.277 –0.472* –0.389* –0.513** –0.582** –0.252 (0.23) (0.06) (0.09) (0.04) (0.03) (0.29) Number of observations 1,234 1,233 1,234 1,233 1,234 1,221 R2 0.03 0.04 0.04 0.04 0.04 0.03 Source: Authors’ calculations. Note: Numbers in parentheses are robust p values. * p < .10, ** p < .05, *** p < .01. Table 8.9 Access to Credit and Labor Productivity log sales per worker Variable (1) (2) (3) (4) (5) (6) Checking account dummy 0.627*** 0.515*** (0.00) (0.00) Line of credit dummy 0.426*** 0.321*** (0.00) (0.00) Financial user dummy 0.700*** (0.00) Financial obstacle –0.143* (0.07) Micro dummy –0.72*** –0.653*** –0.586*** –0.564*** –0.672*** –0.705*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) Small dummy –0.473*** –0.481*** –0.387*** –0.415*** –0.456*** –0.454*** (0.00) (0.00) (0.01) (0.00) (0.00) (0.00) Medium dummy –0.264 –0.271* –0.233 –0.246 –0.274* –0.238 (0.10) (0.08) (0.15) (0.12) (0.08) (0.17) Log of age –0.042 –0.034 –0.029 –0.025 –0.036 –0.037 (0.43) (0.53) (0.56) (0.62) (0.50) (0.50) LLC –0.1 –0.092 –0.145 –0.128 –0.116 –0.098 (0.41) (0.41) (0.18) (0.21) (0.29) (0.43) Direct exports dummy 0.578 0.519 0.578 0.53*** 0.542*** 0.564*** (0.00)*** (0.00)*** (0.00)*** (0.00) (0.00) (0.00) Landownership 0.261** 0.235** 0.249** 0.231** 0.243** 0.265** (0.01) (0.02) (0.01) (0.02) (0.01) (0.01) El Salvador –1.958*** –1.977*** –2.116*** –2.093*** –1.995*** –1.989*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) 233 (continued next page) 234 Table 8.9 (continued) Variable (1) (2) (3) (4) (5) (6) Guatemala –0.041 –0.012 –0.168 –0.113 –0.022 –0.065 (0.75) (0.92) (0.16) (0.35) (0.86) (0.61) Honduras 0.015 –0.032 –0.093 –0.107 0.008 –0.012 (0.91) (0.80) (0.49) (0.41) (0.95) (0.93) Nicaragua –0.455** –0.275* –0.531*** –0.363** –0.315** –0.505** (0.03) (0.08) (0.01) (0.01) (0.03) (0.02) Constant 9.829*** 9.241*** 9.548*** 9.135*** 9.161*** 9.869*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) Number of observations 1,630 1,628 1,630 1,628 1,629 1,608 R2 0.44 0.46 0.46 0.48 0.46 0.44 Source: Authors’ calculations. Note: Numbers in parentheses are robust p values. * p < .10, ** p < .05, *** p < .01. Access to Credit and Productivity in Central America 235 of 0.17–0.18 points. Our OLS regressions cannot establish causality, as more productive firms may choose to become exporters or, once they become exporters, their productivity may increase because of competi- tive pressures. We find that use of a credit line or an overdraft facility is positively and significantly associated with higher productivity (model 3). In addition, we find a positive and significant relationship between using a checking account and productivity (model 2). Having a checking account has a larger impact on TFP than having a line of credit. However, model 4 shows that, when included together, having a credit line remains signifi- cant, while having a checking account does not, indicating that the effect of having a checking account on value added productivity is not inde- pendent of having a credit line. The combined measure of financial user has the largest impact in terms of magnitude, while our subjective measure of financial obstacle is negative, as expected (that is, higher obstacles, lower productivity), but it is not significant. Next we assess whether our results are robust to using an alternative measure of productivity. Table 8.9 reports the results for labor productiv- ity defined as the log of total sales per worker. Log of age remains nega- tive, but not significant, while landownership, defined as a dummy variable taking the value of 1 when the company owns more than 50 per- cent of the land, is positive, but significant only in model 4. As discussed, we find that micro and small firms are less productive than large firms; the difference in productivity between medium and large firms is not significant (partially because our sample of medium and large firms is quite small). We find that credit line and checking account have the predicted and significant effect. In fact, the relationship between use of a checking account and productivity is more significant in regressions with log sales per worker than it is in regressions with a TFP measure. When both meas- ures are included together, they both have significant coefficients (in model 4), while the magnitude of the checking account coefficient remains larger than the magnitude of the credit line coefficient. In addi- tion, we find that the variable for financial obstacle now has a negative and significant relationship with labor productivity. The results we obtain are stronger with the measure of labor productivity for two reasons. First, the sample is larger because fewer data are required to estimate this regression (that is, we do not have to have values for fixed assets, which are often missing). Second, the TFP measure is an estimated variable (that 236 Love, Molina Millán, and Shankar is, a residual), while labor productivity is an actual value. Nevertheless, the fact that we obtain similar results using both measures strengthens the case that use of financial services is associated with higher productivity. An important caveat is that our results are obtained from cross-sectional OLS regression, and thus we cannot claim causality. In other words, the association between use of financial services and productivity may stem from the fact that more productive firms are more likely to be able to use financial services or from a reverse relationship—that is, firms that use financial services are able to raise their productivity. Unfortunately, without time-series data or any suitable instruments (which are absent in current surveys), we cannot establish causality. Our results should not be treated as showing a definite positive impact of financial services on productivity. Cross-Country Differences in the Relationship between TFP and Financial Products To assess the differences among countries, we run the regressions by country and look at the differences in sample composition. Table 8.10 reports proportions of financial users, by country. Costa Rica has the lowest use of a credit line by firms (43.2 percent), while El Salvador and Honduras have the highest (75.8 and 72.1 percent, respectively). These patterns are replicated when we disaggregate by size of firm (see table 8.11). We find that firms in each size category in Costa Rica report less use of credit lines than similar size firms in other countries. Firms in Costa Rica also report higher financing obstacles than firms in other countries. Differences among countries are not so large when we look at the use of checking accounts. Table 8.10 Use of Financial Instruments in Central America, by Country % of firms Country Credit line Checking Costa Rica 43.2 92.7 El Salvador 75.8 92.1 Guatemala 68.6 84.5 Honduras 72.1 92.8 Nicaragua 52.5 70.0 Total 62.8 86.3 Source: Authors’ calculations. Access to Credit and Productivity in Central America 237 Table 8.11 Use of Financial Instruments, by Country and Size of Firm % of firms Country and instrument Micro Small Medium Large Costa Rica Credit line 31 48 63 71 Checking 86 99 100 97 Finance obstacle 51 45 41 17 El Salvador Credit line 67 79 83 89 Checking 84 95 98 97 Finance obstacle 21 24 11 14 Guatemala Credit line 63 70 79 90 Checking 86 80 82 92 Finance obstacle 22 21 26 11 Honduras Credit line 51 77 100 99 Checking 84 98 99 97 Finance obstacle 32 33 6 19 Nicaragua Credit line 49 58 78 95 Checking 39 75 94 100 Finance obstacle 23 15 9 22 Source: Authors’ calculations. Tables 8.12 and 8.13 report the results by country for TFP and labor productivity, respectively, from the value added regression. In general, the results are in line with those reported for all countries estimated together, but significance is often lost because the sample size is much smaller for each country. While we find almost all positive coefficients on checking accounts and line of credit, they are not significant in all regressions. We find that having a checking account is more often sig- nificant in individual countries (three out of five in TFP regressions and four out of five in labor productivity regressions), while credit line is often not significant in individual-country regressions (only one out of five in TFP regressions is significant at the 5 percent level, while two out of five coefficients are significant in the labor productivity regres- sions). Because of the small sample size, these results should not be taken to imply that use of financial services is not important for these countries. 238 Love, Molina Millán, and Shankar Table 8.12 OLS Regression: TFP, by Country TFP value added Variable Costa Rica El Salvador Guatemala Honduras Nicaragua Checking account dummy 0.454* 0.573** 0.694*** 0.613 –0.219 (0.091) (0.025) (0.004) (0.129) (0.293) Line of credit dummy 0.241** 0.188 –0.183 0.061 0.16 (0.024) (0.298) (0.402) (0.770) (0.382) Micro dummy 0.049 0.054 –0.188 0.285 –0.366 (0.794) (0.890) (0.469) (0.279) (0.267) Small dummy 0.018 –0.159 –0.336 0.102 –0.366 (0.926) (0.319) (0.102) (0.668) (0.256) Medium dummy –0.161 0.293 0.308 0.166 –0.204 (0.559) (0.137) (0.196) (0.470) (0.578) Log age –0.002 0.25** –0.085 0.214 –0.162 (0.980) (0.045) (0.573) (0.073)* (0.172) Constant –0.524 –1.41** 0.028 –1.368** 0.789 (0.167) (0.018) (0.951) (0.031) (0.147) Number of observations 265 290 261 192 225 R2 0.035 0.079 0.081 0.068 0.025 Source: Authors’ calculations. Note: Numbers in parentheses are robust p values. * p < .10, ** p < .05, *** p < .01. Exports and Productivity In this section, we present additional evidence that exporting status of the firm is associated with higher productivity. Ideally, we would like to test whether firms become more productive after they become exporters (that is, learning by doing) or whether more productive firms self-select into exporting status. However, with only one year of data and no suitable instruments, we are not able to address this question. Here we simply present a partial correlation between exporting status and TFP, controlling for other firm characteristics associated with export status. Our dependent variable is a dummy for exporting firm status, and we use TFP as one of the regressors. As table 8.14 shows, higher productiv- ity is positively associated with higher likelihood of exporting status. As noted, our results do not provide any information about causality. The probability that a firm becomes an exporter also increases with size and differs significantly across countries. Costa Rica (which is the omitted category) and El Salvador seem to have the highest proportion of Access to Credit and Productivity in Central America 239 Table 8.13 OLS Regression: Labor Productivity, by Country log sales per worker Variable Costa Rica El Salvador Guatemala Honduras Nicaragua Checking account dummy 0.684*** 1.422*** 0.502** 0.28 0.542** (0.000) (0.000) (0.012) (0.267) (0.017) Line of credit dummy 0.089 0.569** 0.11 0.15 0.600*** (0.339) (0.019) (0.563) (0.425) (0.001) Micro dummy –1.467*** –0.506** –0.803*** –0.969*** –1.246*** (0.000) (0.030) (0.001) (0.000) (0.000) Small dummy –0.885*** –0.24 –0.311 –0.772*** –1.444*** (0.000) (0.445) (0.156) (0.000) (0.000) Medium dummy –0.577** –0.192 –0.181 –0.197 –0.367 (0.020) (0.304) (0.491) (0.308) (0.428) Log age –0.059 0.025 –0.068 0.043 –0.131 (0.392) (0.866) (0.576) (0.667) (0.275) Constant 10.214*** 5.885*** 9.522*** 9.518*** 9.776*** (0.000) (0.000) (0.000) (0.000) (0.000) Number of observations 314 439 305 235 327 R2 0.298 0.199 0.14 0.125 0.29 Source: Authors’ calculations. Note: Numbers in parentheses are robust p values. * p < .10, ** p < .05, *** p < .01. exporters. In previous specifications, we introduced controls for foreign ownership, age, and other firms’ characteristics, but they did not change our main results, which are fully consistent with those reported in chapter 3 of this book. Conclusions A growing body of empirical research strengthens the link between access to financial services and economic growth. Although the channels through which credit affects growth on the micro level are not entirely identified, we provide some evidence on the relationship between use of financial instruments and firm productivity in Central America. We also find a positive relationship between productivity and exports, even though we cannot distinguish between two hypotheses: self-selection of productive firms and learning by exporting firms. The results suggest that policy priorities should include further efforts to widen financial access and efforts to boost productivity. 240 Love, Molina Millán, and Shankar Table 8.14 Probit: TFP and Exports Y = 1 if firm exports Variable (1) (2) (3) TFP 0.118** (0.04) TFP value added 0.109* (0.08) Labor productivity 0.140** (0.01) Micro dummy –0.948*** –0.98*** –1.055*** (0.00) (0.00) (0.00) Small dummy –0.414*** –0.354** –0.531*** (0.01) (0.02) (0.01) Firm group dummy 0.001 0.055 0.002 (1.00) (0.75) (0.99) El Salvador –0.329** –0.412** –0.201 (0.05) (0.01) (0.36) Guatemala –0.43* –0.448* –0.499** (0.08) (0.07) (0.03) Honduras –0.689*** –0.756*** –0.674*** (0.00) (0.00) (0.00) Nicaragua –0.692*** –0.755*** –0.711*** (0.00) (0.00) (0.00) Constant –0.129 –0.129 –1.41** (0.33) (0.34) (0.03) Number of observations 1,284 1,235 1,633 Source: Authors’ calculations. Note: Numbers in parentheses are robust p values. * p < .10, ** p < .05, *** p < .01. Note 1. Productivity estimates can be obtained from a regression of the type Yi = Ai Kiα Liβ . Taking logs and rearranging equation 8.1, we derive a measure of labor productivity and two measures of TFP: (1) labor productivity: log(Yi/Li) = a log(Ki/Li ) + ei and (2) TFP productivity: log(Yi ) = a log(Ki) + b log(Li ) + ei. 1. TFP = ε = log ( Y ) − log Y i ˆ i i ˆ ( )i ⎛Y ⎞ ˆ 2. TFPi = ε i = log ( Yi Li ) − log ⎜ i ⎟ , ˆ ⎜L ⎟ ⎝ i⎠ where Y is firm’s output, K and L are capital and labor, a and fl are capital and labor shares. We estimate the model by OLS. Access to Credit and Productivity in Central America 241 References Almeida, Rita, and Ana Margarida Fernandes. 2008. “Openness and Technological Innovations in Developing Countries: Evidence from Firm-Level Surveys.” Journal of Development Studies 44 (5): 701–27. Auboin, Mark. 2007. “Boosting Trade Finance in Developing Countries: What Link with the WTO?” Staff Working Paper ERSD-2007-04, World Trade Organization, Geneva. Ayyagari, Meghana, Asli Demirgüç-Kunt, and Vojislav Maksimovic. 2007. “Firm Innovation in Emerging Markets.” Policy Research Working Paper 4157, World Bank, Washington, DC. Beck, Thorsten. 2002. “Financial Development and International Trade: Is There a Link?” Journal of International Economics 57 (1): 107–31. ———. 2003. “Financial Dependence and International Trade.” Review of International Economics 11 (2): 296–316. Beck, Thorsten, Asli Demirgüç-Kunt, and Ross Levine. 2000. “A New Database on Financial Development and Structure.” World Bank Economic Review 14 (3): 597–605. Beck, Thorsten, Asli Demirgüç-Kunt, and María Soledad Martínez Peria. 2007. “Reaching Out: Access to and Use of Banking Services across Countries.” Journal of Financial Economics 85 (1): 234–66. Beck, Thorsten, Ross Levine, and Norman Loayza. 2000. “Financial Intermediation and Growth: Causality and Causes.” Policy Research Working Paper 2059, World Bank, Washington, DC. Becker, Bo, and David Greenberg. 2005. “Financial Development and International Trade.” Unpublished manuscript, University of Illinois at Urbana-Champaign. Bencivenga, V. R., B. D. Smith, and R. M. Starr. 1995. “Transaction Costs, Technological Choice, and Endogenous Growth.” Journal of Economic Theory 67 (1): 52–177. Bernard, Andrew B., and Bradford Jensen. 1999. “Exporting and Productivity.” NBER Working Paper 7135, National Bureau of Economic Research, Cambridge, MA. Bernstein, J., and I. Nadiri. 1993. “Production, Financial Structure, and Productivity Growth in U.S. Manufacturing.” Working Paper, New York University, New York. Calderón, César, Norman Loayza, and Pablo Fajnzylber. 2004. “Economic Growth in Latin America and the Caribbean: Stylized Facts, Explanations, and Forecasts.” Working Paper 265, Central Bank of Chile, Santiago. Demirgüç-Kunt, Asli, and Vojislav Maksimovic. 1998. “Law, Finance, and Firm Growth.” Journal of Finance 53 (6): 2107–37. 242 Love, Molina Millán, and Shankar Easterly, W., and R. Levine. 2001. “It Is Not Factor Accumulation: Stylized Facts and Growth.” World Bank Economic Review 15 (2): 177–219. Escribano, A., and J. L. Guasch. 2005. “Assessing the Impact of the Investment Climate on Productivity Using Firm-Level Data.” Policy Research Working Paper 3621, World Bank, Washington, DC. Fajnzylber, Pablo, J. Luis Guasch, and J. Humberto López, eds. 2008. Does the Investment Climate Matter? Microeconomic Foundations of Growth in Latin America. Latin American Development Forum Series. Washington, DC: Inter- American Development Bank and World Bank; Santiago, Chile: Economic Commission on Latin America and the Caribbean, Santiago. Gatti, Roberta, and Inessa Love. 2006. “Does Access to Credit Improve Productivity? Evidence from Bulgarian Firms.” Policy Research Working Paper 3921, World Bank, Washington, DC. Ghani, E., and V. Suri. 1999. “Productivity Growth, Capital Accumulation, and the Banking Sector: Some Lessons from Malaysia.” Policy Research Working Paper 2252, World Bank, Washington, DC. Hallward-Driemeir, M., G. Iarossi, and K. Sokoloff. 2002. “Exports and Manufacturing Productivity in East Asia: A Comparative Analysis with Firm- Level Data.” NBER Working Paper 8894, National Bureau of Economic Research, Washington, DC. Hur, Jung, Manoj Raj, and Yohanes E. Riyanto. 2004. “The Impact of Financial Development and Asset Tangibility on Export.” Unpublished manuscript, National University of Singapore. Jeong, Hyeok, and Robert M. Townsend. 2005. “Sources of TFP Growth: Occupational Choice and Financial Deepening.” IEPR Working Paper 05.28, Stanford University, Institute of Economic Policy Research, Palo Alto, CA. King, Robert G., and Ross Levine. 1993. “Finance and Growth: Schumpeter Might Be Right.” Quarterly Journal of Economics 108 (3): 717–37. Levine, Ross. 1991. “Stock Markets, Growth, and Tax Policy.” Journal of Finance 46 (4): 1445–65. ———. 1997. “Financial Development and Economic Growth: Views and Agenda.” Journal of Economic Literature 35 (2): 688–726. ———. 2005. “Law, Endowments, and Property Rights.” Journal of Economic Perspectives, American Economic Association 19 (3): 61–88. Levine, Ross, and Sara Zervos. 1998. “Stock Markets, Banks, and Economic Growth.” American Economic Review 88 (3): 537–58. Levinsohn, J., and A. Petrin. 2000. “Estimating Production Functions Using Inputs to Control for Unobservables.” NBER Working Paper 7819, National Bureau of Economic Research, Cambridge, MA. Love, Inessa. 2003. “Financial Development and Financing Constraints: International Evidence from the Structural Investment Model.” Review of Financial Studies 16 (3): 765–91. Access to Credit and Productivity in Central America 243 Manova, Kalina. 2008. “Credit Constraints, Equity Market Liberalizations, and International Trade.” Journal of International Economics 76 (1): 33–47. Manova, Kalina, and Davir Chor. 2009. “Off the Cliff and Back: Credit Conditions and International Trade during the Global Financial Crisis.” Working Paper, Stanford University, Palo Alto, CA. Nickell, S., and D. Nicholitsas. 1999. “How Does Financial Pressure Affect Firms?” European Economic Review 43 (8): 1435–56. Olley, S., and A. Pakes. 1996. “The Dynamics of Productivity in the Telecommunications Equipment Industry.” Econometrica 64 (6): 1263–97. Rajan, Raghuram G., and Luigi Zingales. 1998. “Financial Dependence and Growth.” American Economic Review 88 (3): 559–86. Roberts, Mark, and James R. Tybout. 1997. “The Decision to Export in Colombia: An Empirical Model of Entry with Sunk Costs.” American Economic Review 87 (4): 545–64. Schiantarelli, Fabio, and Fidel Jaramillo. 1999. “Access to Long-Term Debt and Effects on Firms’ Performance: Lessons from Ecuador.” Policy Research Working Paper 1725, World Bank, Washington, DC. Schiantarelli, Fabio, and Alessandro Sembenelli. 1999. “The Maturity Structure of Debt: Determinants and Effects on Firms’ Performance: Evidence from the United Kingdom and Italy.” Policy Research Working Paper 1699, World Bank, Washington, DC. Schiantarelli, Fabio, and Vivek Srivastava. 1999. “Debt Maturity and Firm Performance: A Panel Study of Indian Companies.” Policy Research Working Paper 1724, World Bank, Washington, DC. Svaleryd, H., and J. Vlachos. 2002. “Markets for Risk and Openness to Trade: How Are They Related?” Journal of International Economics 57 (2): 369–95. ———. 2005. “Financial Markets, the Pattern of Specialization, and Comparative Advantage: Evidence from OECD Countries.” European Economic Review 49 (1): 113–44. Wagner, Joachim. 2007. “Exports and Productivity: A Survey of the Evidence from Firm-level Data.” World Economy 30 (1): 60–82. World Bank. 2003. “Trade Finance for Emerging Markets.” International Finance Briefing Note 29, World Bank, Washington, DC. Wurgler, Jeffrey. 2000. “Financial Markets and the Allocation of Capital.” Journal of Financial Economics 58 (1–2): 187–214. Yan, Aw B., Sukkyun Chung, and Mark J. Roberts. 2000. “Productivity and Turnover in the Export Market: Micro-level Evidence from the Republic of Korea and Taiwan (China).” World Bank Economic Review 14 (1): 65–90. CHAPTER 9 Are Food Markets in Central America Integrated with International Markets? An Analysis of Food Price Transmission in Honduras and Nicaragua Mario A. De Franco and Diego Arias Central American countries have recently signed a free trade agreement with the United States—the Dominican Republic–Central America Free Trade Agreement (DR-CAFTA)—and are negotiating another one with the European Union and others. Food products, for the most part, are included in such agreements. This study seeks to shed light on whether the domestic markets of these food products are integrated with interna- tional markets by assessing the transmission of international prices to domestic prices of key agricultural commodities in Nicaragua and Honduras. In other words, we analyze to what degree (if any) a change in the international price of a given food product changes the domestic price The authors would like to acknowledge the comments received from John Nash, Miguel Robles, Miguel Gómez, Dante Mossi, Daniel Lederman, J. Humberto López, Nabil Chaherli, Rashmi Shankar, Hector Peña, Martin Gurría, and Raquel Fernández. 245 246 De Franco and Arias of that same good, at the level of the consumer and producer as well as in different regions within each country. This analysis provides important evidence of the price dynamics that guide recommendations for a complementary public policy agenda of agricultural trade liberalization in the region. Price transmission is a key indicator of the extent to which domestic food markets in Honduras and Nicaragua are integrated into international markets; however, the impli- cations of this evidence for trade liberalization are limited because the price transmission indicator only measures the degree to which short- term movements in international food prices are translated into domestic food markets, not price levels. Trade liberalization may indeed lower food prices in the region, even if short-term price transmission continues to be low or sometimes nonexistent. There are several ways to analyze the relationship between interna- tional and domestic prices. The first one (price wedge analysis) analyzes the difference between international and domestic prices. This involves analyzing data on prices, but also on transport costs and border protection (tariffs, fees, and others). This price wedge analysis was done for most of the same products and the same countries by Peña and Arias (2010), and the results are presented later in this chapter for comparison. The second methodological approach (price transmission analysis) analyzes the varia- tion in percentages (growth) of international versus domestic prices. This is the approach taken in this chapter. The analysis presented here does not assess welfare, only price behavior. Price behavior is important in a context of trade liberalization, as the reduction in border protection is supposed to translate into a reduction in prices for local consumers. With less border protection, farmers are expected to receive clearer market signals from international prices, enabling them to take advantage of export opportunities and higher inter- national prices, as occurred during the 2007–08 global food crisis. Several related studies on the welfare effect of the DR-CAFTA (Bussolo and oth- ers 2010) assume a high elasticity of international price transmission to domestic prices. However, for these assumptions to hold, there needs to be a price transmission process that is perfect (or almost perfect) and timely (without much delay). If such transmission from international to domestic prices is not perfect or timely, it is difficult to imagine how food consumers and farmers could benefit from trade liberalization. Evidence from previous analysis worldwide is mixed. Some coun- tries and some products present high price transmission, while others present very little. A summary of the literature on price transmission Are Food Markets in Central America Integrated with International Markets? 247 of agricultural products in Latin America was prepared by the Food and Agriculture Organization (FAO 2006), and the different studies reach a consensus that analyses of price transmission processes must consider the entire supply chain (upstream and downstream) to under- stand the results. Imperfect price transmission can be explained by several factors, but the two most basic explanations are (a) the existence of noncompetitive market structures in which one agent has sufficient market power to establish or influence a market price above the marginal cost and (b) the fact that, even in competitive market structures, price transmission is imperfect as long as there are costs to price adjustments at some point within the supply chain (Vavra and Goodwin 2005). These are important conclusions, as they point out that when trade liberalization is not accompanied by a review of domestic market structures or adjustment costs within a supply chain, it is unlikely to have the expected benefits for consumers and farmers. However, even in a situation of complete free trade (no tariffs) in which the import market for a specific good is managed by a single importing firm (monopoly) or a group of colluded firms (cartel), the degree of price transmission will be determined by the degree to which consumers are able to substitute that good with another one. In other words, the degree to which the monopoly or cartel can arbitrarily set domestic prices will depend on the elasticity of substitution of the demand curve for that good. The same applies to a domestic food processor or exporter. The elasticity of substitution of crops being produced by farmers will determine the degree to which the buyer (monopoly) or colluded buyers (monopsony) are able to fix prices arbitrarily at the farm gate. Adjustment costs can also play an important role even with strong domestic competitiveness and free trade. Domestic food prices can be “sticky” due to several market characteristics such as menu costs (the rela- beling and reprinting of price lists); marketing costs (the negative impact on consumer demand of shifting prices); logistics costs (the larger volatil- ity in inventory, storage, or transport costs from unexpected changes in demand generated by continuous price changes); and corporate image costs (the impact of staple food price volatility on the company’s reputa- tion). Furthermore, fluctuations in domestic food prices can undermine contract farming or result in the nonfulfillment of current contracts between farmers and buyers. The food products studied here were selected based on their impor- tance in the agriculture sector and rural economy of the country as well 248 De Franco and Arias as their weight in the consumer basket, in particular of low-income households. A mix of import and export products was chosen to allow comparison and yield public policy recommendations. The seven prod- ucts selected are highly tradable, in terms of exports and/or imports. Table 9.1 presents the weight of exports in overall production and the weight of imports in overall supply for Nicaragua. These seven food prod- ucts represent 47.4 percent of the consumption basket of the lower income quintiles in Nicaragua and 74.7 percent of the consumption bas- ket of households in extreme poverty in Honduras. Therefore, studying these products is to study a large part of the economy of these countries, in particular the rural economy. To assess the price behavior of these products, the following questions were used to guide the econometrics analysis in both countries: • Are changes in international prices transmitted throughout the domestic supply chain? • What is the time lag of price transmission? • Is price transmission asymmetric? In other words, do prices behave the same when they increase as when they decrease? • Is price transmission homogeneous between geographic regions within the country? The chapter begins by presenting a model and the results of the esti- mates of the price transmission process for the agricultural commodities Table 9.1 Weight of Tradables of Select Food Products in Nicaragua, 2005 % of total % of total Product production supply For export Sugar 45.7 — Coffee 89.1 — Meat (beef ) 58.0 — Beans 45.7 — For import Maize (all) n.a. 13.3 Rice (paddy) n.a. 33.6 Vegetable oil n.a. 33.9 Source: Central Bank of Nicaragua. Note: — = Not available; n.a. = not applicable. Are Food Markets in Central America Integrated with International Markets? 249 selected in Honduras and Nicaragua. It describes the industries of these products to characterize the functioning of these markets at a domestic and regional level and presents a simple model in which conditions are derived for an imperfect transmission of variations in international prices to domestic prices. The model is developed both for import and for export goods. The model assumes, as per the evidence presented, the exis- tence of adjustment costs and a noncompetitive domestic market struc- ture in which the elasticity of substitution of demand and supply plays a key role (following Dixit and Stiglitz 1977). This is followed by a section analyzing the results. A final section recommends some areas for public policy to complement agricultural trade liberalization in Honduras and Nicaragua. Price Transmission of International to Domestic Prices of Food Products This section explains the methodology used in this chapter and summa- rizes the results of the price transmission model. The first step in determining the most convenient econometric process to follow is to determine the degree of integration of the study variables (as with all time-series analysis) to see whether they are stationary or non- stationary. To determine the degree of integration of the study variables, we use the unit root test Dickey-Fuller–generalized least squares (DF-GLS) with critical values of Eliott-Rothenberg-Stock, which is more powerful than tests such as augmented Dickey-Fuller (ADF), Phillips-Perron, and Kwiatkowski-Phillips-Schmidt-Shin (KPSS) tests according to Maddala and Kim (2004) and Dutoit, Hernández, and Urrutia (2009). This unit root test is applied to the natural log of prices through a general to partic- ular method, as proposed by Enders (1995), which consists of proving ini- tially the existence of unit root through a more complete model that includes the trend, constant, and lags.1 Thus when the variables are non- stationary, we use cointegration techniques to avoid the spurious correla- tion problem. To determine the cointegration equation, we use the Johansen procedure after proving that the price variables have an order of integration larger than 0 but equal between them (Johansen and Juselius 1990). This cointegration technique consists of the formulation of an autoregressive vector of the variables in levels until the errors in each of the equations for each vector make up white noise. Subsequently, from this autoregressive vector a system is formed by the variables in first differences and a set of stationary combinations of the variables in levels. 250 De Franco and Arias To generate such combinations (known as cointegration vectors or long- term relations), we apply Johansen’s maximum likelihood to the vector autoregression (VAR) model proposed. Finally, to determine whether the cointegration relations are significant, the trace and maximum value test is used. The cointegration relations are estimated in pairs, between the international and the domestic price, at the producer level, at the con- sumer level, or by region. In other words, we do not estimate a VAR for the entire set because of the difficulties of controlling the large amount of information within a VAR and of generating a reasonable long-term relationship. Nevertheless, although the VAR allows us to generate individual models for domestic prices (for consumers, pro- ducers, and regions) with their respective cointegration relationships (when they exist), such VARs present three important limitations: (a) they are estimated using the econometric program E-views,2 which does not present the option of considering international prices as super-exogenous variables, so such prices are considered in general terms as endogenous,3 which is not correct; (b) they do not consider the existence of causality between domestic prices (the fact that con- sumer prices can affect producer prices and vice versa); and (c) they do not consider the possibility of different corrections when domestic prices are above the level suggested by the cointegration equation compared to the situation when they are below the level suggested by the long-term relationship. Therefore, VAR is used only to determine the cointegration equation (or long-term relationship) between a domestic and an international price. The effective transmission of international prices to domestic prices is analyzed through a system of equations of the first logarithmic differences of domestic prices characterized by (a) the elimination of causality of domestic to international prices (considering the latter as strongly exogenous); (b) the statement of a possible causality between domestic prices (between consumer and producer prices and between regions); and (c) the possibility of asymmetries in the error correction of the cointegration equation. In this system, a convenient number of lags is introduced in the price variables, as in other variables such as salaries, exchange rates, input prices, and others. The final system is not necessarily estimated with an equal number of lags or of explanatory variables for the individual equations; thus the seemingly unrelated regressions method, as suggested by Enders (1995), is followed in search of efficiency in the estimation of parameters. Are Food Markets in Central America Integrated with International Markets? 251 Consequently, to determine the existence of asymmetries in the error correction of the cointegration equation, a binary variable is included that is equal to 1 when the remainder of the cointegration relation is positive and equal to 0 when negative. Then the statistical significance of the parameter of this binary variable is tested through the t statistic that appears in the system of equations. When the order of integration of the price variables does not match, the variable for which the order of inte- gration does not match the rest is excluded. When no cointegration rela- tion is found between variables, we proceed in the same form, estimating the system of equations, but excluding the remainder of the long-term relationship, given that this does not exist. Finally, when the domestic and international prices are determined to be stationary variables, a system of equations similar to the nonstationary case is estimated, but with three differences: (a) the variables are included not in first differences but in levels; (b) the lag remainder of one period of the cointegration equation is not excluded, given that this does not exist; and (c) the asymmetry of international price transmission is tested with dummies that take the value of 1 when international prices increase and 0 when they decrease. Table 9.2 presents a summary of the results of the price transmission found in Nicaragua following the econometric methodology just Table 9.2 Growth in Domestic Prices in Nicaragua Given a Permanent Increase of 10 Percent in the International Price percent Rest of the Average Average Managua country domestic domestic consumer consumer producer price consumer price price growth price growth Steady Steady Steady Steady Product 1 month statea 1 month statea 1 month statea 1 month statea Vegetable oil 1.4 9.7 0.8 7.9 — — — — Rice 0.0 3.4 0.0 3.7 0.0 5.7 0.0 4.3 Sugar 0.0 0.1 0.0 0.2 0.0 0.3 0.0 0.1 Coffee 0.4 1.0 0.0 2.3 0.0 0.2 0.0 0.5 Meat (beef ) 0.0 0.0 0.0 0.0 0.0 1.8 0.0 1.7 Maize 0.0 1.8 0.0 2.2 — — — — Beans 3.2 4.6 1.7 3.0 0.0 6.1 0.0 6.7 Source: Authors’ calculations. Note: — = Not available. a. Increase after 45 months. 252 De Franco and Arias described. The first four columns show the growth of average domestic producer and consumer prices, and the last four columns show the growth in consumer prices in Managua and the rest of the country after one month and in the steady state as a result of a permanent variation of 10 percent in the international price of each product.4 In the short run (one month), price transmission from international to domestic prices is low or nonexistent in all the food products studied in Nicaragua. Furthermore, not even in the long run (after 45 months) is price trans- mission complete,5 with the exception of vegetable oil, where price transmission is almost 100 percent. In the case of beans, high price transmission is also observed in the long run. In the case of Honduras, consumer prices in different regions of the country were obtained for the period January 2000 to June 2009. Table 9.3 presents a summary of the estimated changes in domestic consumer prices after one month and at the steady state (45 months) given a 10 percent change in international prices. These results are backed by simple statistical tests of correlations between both variables. At first sight, these simple statistical tests also show the low correlation between international and domestic prices for the period studied. To clarify the dynamics of price transmission, 12-month trajectories for each product are calculated given a 10 percent change in the inter- national price of each product (see figures 9.1 and 9.2). For Nicaragua, both consumer and producer prices are available for all products, but regional price data are available only for rice, sugar, maize, and meat (the other products are not available), provided by the Ministry of Agriculture and Forestry; the same econometric methods are used as explained in the previous section. For Honduras, only consumer prices are available for the different regions of the country, with the exception of beans. As shown in figures 9.1 and 9.2, in both countries the dynamics for coffee and rice are quite different. While rice presents very small price transmission estimates in both countries, even in the long term, coffee prices to farmers adjust fully after three months in Nicaragua. Table 9.4 presents the findings with respect to the transmission of consumer prices by region for both countries and with respect to consumer versus producer prices for Nicaragua. For all cases, there is no evidence of price transmission asymmetries. The system of equations does not yield different results between a sit- uation where the international price increases versus one where it Table 9.3 Change of Consumer Prices in Honduras Given a Permanent Increase of 10 Percent in International Prices, by Region percent Center North (metropolitan) Center (rest) (metropolitan) North (rest) West East South Steady Steady Steady Steady Steady Steady Steady Product 1 month state 1 month state 1 month state 1 month state 1 month state 1 month state 1 month state Rice 0.5 3.4 0.0 3.3 0.2 3.6 0.0 4.1 0.6 3.9 0.4 3.4 0.0 2.6 Sugar 0.1 –0.1 0.2 0.0 0.3 0.0 0.4 0.1 0.2 –0.1 0.2 –0.2 0.2 0.0 Coffee 0.5 3.0 0.7 3.6 0.3 2.1 0.4 2.2 0.3 1.7 0.5 2.6 0.4 2.6 Meat (beef ) 0.0 5.4 0.0 6.0 0.0 5.6 0.0 6.9 0.0 5.1 0.0 6.8 0.0 5.0 Maize 0.0 4.8 0.0 3.8 0.0 4.3 0.0 3.1 0.0 3.6 0.0 4.5 0.0 4.8 Vegetable oil 0.0 7.3 0.0 6.9 0.0 5.3 0.0 6.0 0.0 4.4 0.0 6.8 0.0 5.3 Source: Authors’ calculations. Note: a. Increase after 45 months. 253 254 De Franco and Arias Figure 9.1 Transmission of Rice and Coffee Prices in Hondurasa a. Rice prices 4.5 4.0 price transmission (%) 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 0 1 2 3 4 5 6 7 8 9 10 11 12 months b. Coffee prices 2.5 price transmission (%) 2.0 1.5 1.0 0.5 0 0 1 2 3 4 5 6 7 8 9 10 11 12 months central, metropolitan central, other north, metropolitan north, other west east south Source: Authors’ estimations. a. Increase in local rice prices given a 10 percent increase in international prices. decreases. In conclusion, for Nicaragua, coffee is the only product that presents high price transmission when looking at producer and consumer prices as well as at regional differences, although transmission is delayed a few months. For Honduras, consumer prices also show little reaction to international price changes within the first month. Even in the long term, most of the products present a price transmission that is less than Are Food Markets in Central America Integrated with International Markets? 255 Figure 9.2 Transmission of Rice and Coffee Prices in Nicaraguaa a. Rice prices 4.5 4.0 3.5 price transmission (%) 3.0 2.5 2.0 1.5 1.0 0.5 0 0 1 2 3 4 5 6 7 8 9 10 11 12 months Managua rest of the country b. Coffee prices 16 14 price transmission (%) 12 10 8 6 4 2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 months consumer processor producer Source: Authors’ estimations. a. Increase in local rice prices given a 10 percent increase in international prices. 50 percent. The only products with a price transmission more than 50 percent are meat and vegetable oil. Understanding the Price Transmission Results To understand the results, we present some basic characteristics of the agribusiness sectors involved in the production and commercialization of these food products and then propose a simple, general model for 256 De Franco and Arias Table 9.4 Findings of Estimates of Price Transmission Analysis of Select Food Products in Honduras and Nicaragua Product Honduras Nicaragua Vegetable oil No or very little change is After three months, consumer observed during the initial prices change only 2 percent months, but after 12 months, and producer prices change prices change between 0.5 3 percent. After 12 months, percent in the western region prices change 6.0 and 5.1 percent, and 4.5 percent in the northern respectively. (nonmetropolitan) region. Rice (paddy and After the first month, there is After one month, producer prices processed) almost no change, and after do not change and consumer 12 months, consumer prices prices change only 0.5 percent. change between 2.5 and 4.0 After six months, producer and percent, with the northern consumer prices change less than region presenting the highest 3 percent. However, after three months, prices seem to change transmission and the southern more in Managua than region the lowest. in the rest of the country. Sugar No or very little change No significant change is is observed. found through the entire first 12 months. Regional price differences are minimal. Coffee During the first month, a change Consumer prices (for ground between 0.25 and 0.75 percent is coffee) do not change significantly observed; after 12 months during the first 12 months, but consumer prices change producer prices for green bean between 1.50 and 2.25 percent, coffee change more than 6 with the northern region percent during the first three months and 10 percent during the presenting the lowest first 12 months; producer prices for transmission and the southern ground coffee change 10 percent region the highest. after the first three months. Consumer prices for ground coffee change more in Managua than in the rest of the country. Maize (white) No change is observed during the During the first eight months, first month, but in the second there are no significant changes month, price changes range in domestic consumer or from 2.5 percent in the southern producer prices or by regions. region to 0 percent in the eastern region. After 12 months, changes vary from 2.5 percent in the northern (nonmetropolitan) region to 6.0 percent in the southern and eastern regions. (continued next page) Are Food Markets in Central America Integrated with International Markets? 257 Table 9.4 (continued) Product Honduras Nicaragua Meat (beef ) No or very little change is observed There is no transmission of during the first month; after 12 international prices into domestic months, consumer prices change prices. 1.50 percent in the northern region and –0.25 percent in the southern and central (nonmetropolitan) regions. Beans Data are not available. During the first month, producer and consumer prices change 3.2 and 1.7 percent, respectively. After 12 months, they change 3.1 and 3.9 percent, respectively. Source: Authors’ calculations. imperfect international price transmission. At the end of the section, we compare the findings with the results on price wedges presented by Peña and Arias (2010). Agribusiness Industry Structure Information on the number, market share, and type of agribusinesses in Nicaragua and Honduras has been difficult to obtain; however, some evi- dence is presented in tables 9.5 and 9.6 for Nicaragua. With the exception of beans and to some extent dairy products, the domestic market structure concentrates market power (market share) in a small number of trading and processing companies. This situation indi- cates a potential for reduced competition in these agribusiness markets, which could explain some of the low price transmission estimates pre- sented in the previous section. Modeling Imperfect International Price Transmission The model produced in this study has a simple, general structure. Its pur- pose is to derive, in a consistent manner and with reasonable assumptions, the conditions in which international price transmissions are reflected in a perfect or imperfect fashion in domestic markets. By finding these con- ditions, we can then derive public policy and program recommendations to complement trade liberalization and improve consumer and producer welfare. The model assumes that demand for a product composed of import and local inputs, Q, with respect to the total final goods of the economy, 258 De Franco and Arias Table 9.5 Number and Market Share of Large Agribusiness Companies in Nicaragua, 2005 Product Number of large companies Market share (% of total supply) Sugar 2 90 Meat (beef ) 3 70 Vegetable oil 3 90 Poultry 2 75 Wheat (flour) 3 100 Rice 1 70 Source: Central Bank of Nicaragua. Table 9.6 Number and Export Share of Large Exporters in Nicaragua, 2005 Number of Number of large export Export (import) market Product producers (import) companies share (% of total exports) Sugar — 2 90 Meat (beef ) > 35,000 4 90 Coffee 35,000 2 60 Dairy products — 1 40 Beans — 3 30 Rice (imports) — 1 84 and 45a Source: Customs Division, Managua, Nicaragua. Note: — = Not available. a. 84 percent within quota and 45 percent over quota imports. QG, is simply a function of its own price, P, with respect to the general price index of the economy, PG, elevated to the elasticity of substitution, s, defined as −σ Q ⎛ P ⎞ =⎜ ⎟ . (9.1) QG ⎝ PG ⎠ Assuming that, in line with the evidence presented previously, there is only one seller (or a group of colluded sellers) of the good; to supply this good, the seller(s) acts as the only buyer for the local producers (farmers), who, in turn, sell the product at price Pp and quantity Qp. Moreover, because of storage, package, and financial market imperfections and difficulties, the seller(s) is the only importer of the good, Qm, of similar quality at price Pm. These assumptions, although extreme, are not far from reality in the countries and products being studied. Ignoring other costs for simplification purposes, the seller(s) of the good maximizes its profits, Π, in a traditional fashion: Π = PQ – PmQm – PpQp. (9.2) Are Food Markets in Central America Integrated with International Markets? 259 Nevertheless, the supply of farmers with respect to the set of final goods responds to producer prices with respect to the general price index of the economy, given a certain (low) elasticity of substitution of produc- tion, g, by other goods: γ Qp ⎛ Pp ⎞ =⎜ ⎟ . (9.3) QG ⎝ PG ⎠ Given that the total imports of this good are not greater than the total quantity minus the purchases made to local farmers, Qm = Q – Qp. (9.4) Substituting equations 9.3 and 9.4 in the profit equation, we obtain γ 1+ γ PQG P1−σ PmQG P −σ PmQG Pp QGPp Π= − + − . (9.5) PG σ − PG σ − γ PG γ PG For the monopolist or cartel described above, the chosen variables are the price to be paid to local farmers (Pp) and the sale price (P). Deriving equation 9.5 with respect to P, equaling 0 and solving for P, we obtain the results of the model based on the elasticity of substitution assumption of Dixit and Stiglitz (1977): σ P= Pm . (9.6) σ −1 Equation 9.6 says that the price at which the monopolist (or cartel) can sell is determined by the elasticity of substitution of the international price. Following an equivalent process with respect to producers, we arrive at a purchasing price fixed by the monopolist that is a function of the elasticity of supply given international prices: γ Pp = Pm . (9.7) 1+ γ If s or g tend toward large values—in other words, if the elasticities are high—consumer and producer prices will tend to be equal to international prices; when the elasticities are low, the opposite occurs. The results in equations 9.6 and 9.7 correspond to situations where adjustment costs do not exist; their role in this theoretical model is for reference, given the fact that they are used later on as the base scenario to compare the evolution of domestic prices with respect to a change in international prices. 260 De Franco and Arias In determining the variation in domestic prices with respect to the variation in international prices (elasticity of domestic prices with respect to international prices), one needs to include not only the elasticity of substitution and elasticity of production of that good, but also the adjust- ment costs that occur with a price change. To complete the model, we express it in growth rates. Noting each initial price with the subscript 0, growth rates are expressed in the following fashion: Pm0 → Pm0 (1+ m), (9.8) P0 → P0 (1+ ε c ), and (9.9) Pp 0 → Pp 0 (1+ ε p), (9.10) where m, ec, and ep, are the growth rates of import, consumer, and pro- ducer prices, respectively. Using these three definitions of price, we can rewrite the profit function expressed in equation 9.5 in terms of growth rates of consumer, producer, and import prices: QG 1−σ Q −σ Π(ε c ,ε p , m) = ⎡P (1+ ε c )⎤ − Pm0 (1+ m) −G ⎡ o (1+ ε c )⎤ −σ ⎣ o ⎦ P σ ⎣ ⎦ PG PG QG γ QG 1+ γ + Pm0 (1+ m ) ⎡ P (1+ ε p )⎤ − γ ⎡Po (1+ ε p )⎤ γ ⎣ o ⎦ P ⎣ ⎦ (9.11) PG G − A1ε c2− A2ε p . 2 This structure allows us to include the two new terms on the right- hand side of equation 9.11, A1 and A2, which represent the adjustment costs to the firm from changing consumer and producer prices, respec- tively. As in similar models, the quadratic form is used to derive the model, rather than just to represent the empirical behavior of these costs, given that the simplified form includes the existence of adjustment costs, because prices both increase and decrease with respect to the baseline. Deriving equation 9.11 with respect to ec and ep, equaling to 0, and sim- plifying, we obtain the following first-order conditions to maximize profits: ⎡ QG −σ ⎤ −σ −σ −1 ⎢ −σ P0 ⎥ ⎡(1− σ )(1+ ε c ) P0 + σ Pm0 (1+ m )(1+ ε c ) ⎣ ⎤ = 2 A1ε c . (9.12) ⎦ ⎣ PG ⎦ ⎡QG γ ⎤ ⎡ γ −1 γ ⎤ ⎢ γ Pp 0 ⎥ ⎢γ Pm0 (1+ m)(1+ ε p ) − (1+ γ )Pp 0 (1+ ε p) ⎥ = 2 A2 εp . ⎣ ⎦ (9.13) ⎣ PG ⎦ Are Food Markets in Central America Integrated with International Markets? 261 Given the above, consumer price elasticity with respect to interna- tional prices, hc, and producer price elasticity with respect to interna- tional prices, hp, can be calculated by taking the total difference of equations 9.12 and 9.13 with respect to ec, ep, and m and assuming that we start from an initial situation where growth rates are 0, so that the results established in equations 9.6 and 9.7 hold true.6 In this context, the price transmission elasticities are defined as follows: hc = ∂ec/∂m, hp = ∂ep/∂m. After several manipulations we obtain equations 9.14 and 9.15, which express the key variables that determine the transmission of international prices to consumer and producer prices, respectively: QG −σ σ Pm0 P0 ∂ε c PG σ − = . (9.14) ∂m Q 2 A1+ σ P 0 −G P0−σ PG σ m QG γ γ Pm0 P0 ∂ε p PGγ = . (9.15) ∂m Q 2 A2+ γ P 0 G P0γ m γ PG If in equation 9.14 the elasticity of substitution of consumption goods is low, we conclude that, if A1 > 0, the international price transmission into domestic (consumer) prices is less than 1—in other words, the trans- mission is imperfect. However, independent of the elasticity of substitu- tion of demand, if A1 = 0, we conclude that the transmission elasticity equals 1—the transmission is perfect. Finally, if in equation 9.15 the elasticity of supply is low, we conclude that, if A2 > 0, the transmission of a change in international prices into domestic (producer) prices is less than 1—in other words, the transmis- sion is imperfect. However, independent of the elasticity of substitution of supply, if A2 = 0, the transmission elasticity of international prices is equal to 1—the transmission is perfect. Similar conclusions are obtained when considering exportables (goods that are produced in the domestic market and exported) and when the firm has domestic market power given that it is the only one that sells in the domestic market. In fact, if we assume that one firm buys an amount of inputs QP domestically at price Pp and then transforms them to produce a good to be sold in the domestic market (quantity Q) at 262 De Franco and Arias price P and exported (quantity QX) at price Px, then the profit function is as follows: Π = PxQx + PQ – PpQp. (9.16) However, assuming that the supply of domestic inputs follows equa- tion 9.17, where QG is a quantum measure of the rest of the goods in the economy and PG is the producer price index for the rest of the goods in the economy, we then have the following: γ Q p ⎛ Pp ⎞ =⎜ ⎟ , (9.17) QG ⎝ PG ⎠ where g is the elasticity of transformation of input Qp with respect to the rest of the goods in the economy. There also exists demand for the good sold by the monopoly in the domestic market, which presents an elastic- ity of substitution –s : −σ Q ⎛ P ⎞ =⎜ ⎟ . (9.18) QG ⎝ PG ⎠ Given equations 9.16, 9.17, and 9.18 and assuming in the meantime the absence of adjustment costs, then the producer (farmer) establishes the sale price of the good in the domestic market and the purchase price of inputs in the domestic market according to the following relationships: γ Pp = Px 1+ γ (9.19) σ P= Px . σ −1 The prices in equation 9.19 are used as reference to analyze the varia- tions in domestic prices when export prices change in light of the adjust- ment costs of price changes. Using a procedure similar to the case of imports, we obtain a profit equation that includes adjustment costs and looks as follows: Q γ Q −σ Π = Px (1+ x ) G ⎡Pp (1+ ε p )⎤ − Px (1+ x) −G ⎡P (1+ ε d )⎤ γ ⎣ ⎦ σ ⎣ ⎦ PG PG , (9.20) 1− σ QG 1+ γ QG + ⎡P (1+ ε d )⎤ ⎣ ⎦ −σ − ⎡ Pp (1+ ε p )⎤ ⎣ ⎦ γ − Aεd − Bε 2 , 2 p PG PG where A and B are the adjustment cost coefficients of price variations, ed and ep represent the percentage change in domestic prices (consumers Are Food Markets in Central America Integrated with International Markets? 263 and input suppliers, respectively) with respect to the baseline, and x rep- resents the change in export prices with respect to the baseline. Finally, if the same procedure is used for the case of imports, we obtain the follow- ing sensitivities of domestic prices to export prices: QG −σ σ Px0 Pd 0 ∂ε d PG σ − = ∂x Q σ 2 A + σ P 0 −G Pd−0 PG σ x (9.21) QG γ γ Px0 Pp 0 ∂ε p PGγ = . ∂x Q γ 2B + γ Px0 G Pp 0 γ PG Similar to the case of importables, the existence of small adjustment costs of domestic prices (even if A and B are slightly above 0) is enough for the elasticity of substitution and transformation of equations 9.17 and 9.18 to play a significant role in the degree of international price trans- mission to domestic prices. The model shows that the transmission elasticity between interna- tional and domestic prices is a function of the elasticity of substitution of demand and supply and the adjustment costs. Therefore, the low price transmission found in the analysis must necessarily be explained by one or several of the following factors: (a) low elasticity of substitu- tion of supply, (b) low elasticity of substitution of demand, or (c) high price adjustment costs. It is no surprise that the elasticity of substitution of supply of food products is low. Changes in agricultural production in general are costly and have time lags of several months, and sometimes more than a year, to switch from one crop to another. The main factor that could enable (constrain) faster changes in agricultural production in response to price changes is technology (or the inefficiency in the adoption of improved technologies). Technological advantage could produce plant varieties with shorter life cycles, or better soil conditioning for switch- ing crops (or livestock) could increase the elasticity of supply for food products by allowing shorter time periods and reducing the costs of substituting the production of one product by the production of another. Other factors that influence the elasticity of supply of agricul- tural products could be the amount of spare capacity, either in the form of human or physical capital (idle land) and the amount of stock 264 De Franco and Arias of the good in question. Unfortunately, we were unable to find infor- mation to estimate the elasticity of supply for these products in Nicaragua and Honduras. With respect to the elasticity of the substitution of demand, we were also unable to find time-series information that could capture the signif- icance of the elasticity of substitution of demand for determining the low price transmission for each food product. However, cross-sectional infor- mation is available through the Living Standards Measurement Survey (LSMS),7 so an analysis is undertaken to assess the importance of the elas- ticity of substitution of demand in determining domestic prices across regions. The analysis is based on a model assuming that the elasticity of substitution of demand is an endogenous variable that depends on other factors such as income per capita, education, demographic factors, and spatial factors.8 The results for Nicaragua show the following: 1. The baseline category (or reference category), which is a household in the fifth income quintile, with a woman head of household over 54 years of age, with no education, and residing in urban Managua, has a significant impact on the elasticity of substitution of demand for all food products, except powdered milk. This category also presents higher values for elasticity of substitution of demand for vegetable oil and meat (beef) than for beans, eggs, maize, and rice. 2. The education of the head of household increases significantly the elasticity of substitution of demand for rice, plantains, beans, and eggs. 3. The elasticity of substitution of demand for coffee, chicken meat, beef, and powdered milk does not change with geographic location. 4. The presence of a male head of household increases the elasticity of substitution of demand for vegetable oil, sugar, beans, maize, bread, and tortillas, but has no effect on other products. 5. The presence of younger members in the household with respect to older (above 54 years old) has a statistical significance in reducing the elasticity of substitution of demand for eggs and tortillas, but has the opposite effect for powdered milk. For the rest of the food products, age composition of the household does not seem to play a role. 6. Households in the central rural region have elasticities of substitution of demand that are higher for eight food products (sugar, coffee, chicken, meat, powdered milk, milk, maize, and tortillas); however, those in urban Managua have the lowest in the country, with the exception of vegetable oil, beef, and bread. Are Food Markets in Central America Integrated with International Markets? 265 7.Following table 9.7, consumers in the lower income quintiles have in general higher elasticities of substitution than those in higher income quintiles. This could be explained in part by the fact that the food products selected make up more than 80 percent of the consumption basket of the lowest income quintile, so that a change in food prices immediately forces households in this income segment to adjust to their budget restriction. However, in some special cases the changes in elasticity could be explained by the greater practical difficulty that consumers in different quintiles have in substituting these products. For Honduras, the estimates show similar results: 1.When the head of household is a man, the elasticity of substitution of demand decreases for rice, sugar, beans, eggs, milk, maize, plantains, and tortillas. For vegetable oil and beef, the opposite occurs. 2. With respect to regional differences, there is no systemic finding, with the exception of rice, sugar, and plantains, for which the elasticity of substitution of demand increases when going from the central urban region to the eastern rural region. However, for coffee it is the oppo- site (elasticity of substitution of demand decreases). 3. The regions with the highest elasticity of substitution of demand are the eastern rural region for rice, sugar, and beans and the central urban Table 9.7 Elasticity of Substitution of Demand for Select Food Products in Nicaragua, by Quintile Product 1 2 3 4 5 Vegetable oil –0.20 –0.38 –0.57 –0.84 –1.34 Rice (grains) –1.45 –0.95 –0.75 –0.35 0.31 Sugar –1.40 –0.89 –0.63 –0.29 0.20 Coffee (ground and beans) –0.99 –1.03 –0.94 –0.90 –0.77 Chicken –0.96 –0.87 –0.78 –0.79 –0.96 Beef –1.00 –0.94 –1.02 –1.04 –1.27 Beans (grains) –1.37 –0.92 –0.52 0.05 0.82 Eggs –1.19 –1.10 –1.01 –0.93 –0.75 Powdered –1.04 –1.03 –0.99 –0.96 –0.92 Milk –0.96 –0.85 –0.69 –0.48 –0.10 Maize (grains) –0.74 –0.42 –0.15 0.12 0.47 Bread –0.79 –0.81 –0.82 –0.79 –0.62 Green bananas, plantains, or banana squares –1.02 –0.93 –0.86 –0.79 –0.58 Tortillas –1.00 –0.90 –0.79 –0.67 –0.47 Source: Authors’ calculations based on regression analysis. 266 De Franco and Arias region for coffee and beef. The lowest elasticity of substitution of demand is in the central urban region for rice, sugar, and beans and in the eastern rural region for coffee and beef (inverse results for these two regions for the same food products). 4. A higher percentage of children in the household between birth and four years old with respect to members 54 years of age and older produces higher elasticity of substitution of demand for eggs, milk, plantains, and tortillas. However, for chicken, it is the opposite. In contrast, the higher percentage of 15–54 year olds with respect to members 54 years of age and older produces lower elasticity of substitution of demand for rice, sugar, beans, and maize. However, for chicken, it is the opposite. 5. When the level of poverty of the household increases, the elasticity of substitution of demand for rice, sugar, beans, maize, plantains, and tor- tillas increases (see table 9.8). This is because these products consti- tute a relatively larger portion of the consumption basket of low- income households. 6. For the other food products, the change in the elasticity of substitu- tion of demand moves in the opposite direction; however, these other food products make up a smaller portion of the lower-income house- hold’s consumption basket. To grasp the extent to which the elasticity of substitution of demand for the food products studied here explains domestic price Table 9.8 Elasticity of Substitution of Demand for Select Food Products in Honduras, by Level of Poverty Product Extreme poverty No extreme poverty No poverty Vegetable oil 0.22 –0.78 –0.70 Rice –1.62 –0.87 –0.11 Sugar –1.58 –0.89 –0.13 Coffee –0.26 –0.62 –1.12 Chicken 0.76 –0.12 –0.93 Beef –0.59 –1.21 –1.74 Beans –0.99 –0.90 –0.06 Eggs –0.09 –0.57 –0.63 Milk –0.64 –0.71 –0.81 Maize –1.32 –0.87 –0.37 Bread 0.38 –0.38 –0.67 Plantains –1.32 –1.06 –0.71 Tortillas –1.85 –1.60 –0.67 Source: Authors’ calculations based on regression analysis results. Are Food Markets in Central America Integrated with International Markets? 267 differences, we construct an indicator of consumer price deviations with respect to their average and analyze, through regressions, several sup- ply and demand factors for their impact.9 The main results show that for Nicaragua a decrease in the elasticity of substitution of demand for rice, sugar, coffee, beans, eggs, milk, bread, and plantains is accompa- nied by an increase in the price of the food product with respect to the average price in the economy. This could be evidence that, in regions or income segments of the population where the elasticity of substitu- tion of demand is low, firms with significant market share (monopolists or cartels) can increase market prices, which reduces the degree of international price transmission. To assess the importance of demand factors in explaining domestic price variations, we compare the R2 of regressions with and without demand factors. As table 9.9 shows, demand factors in Nicaragua explain more than 20 percent of the deviations of household prices for vegetable oil, coffee, rice, chicken, milk, and plantains, while in Honduras the list includes vegetable oil, coffee, eggs, maize, bread, and plantains. Finally, the third factor that has an influence over international price transmission is the price adjustment cost. As mentioned in the introduc- tion, such costs can be due to several factors, but for agricultural and food products they are likely due mainly to logistics costs, which refer to larger volatility in inventory, storage, or transport costs due to unexpected Table 9.9 Importance of Demand Factors in Explaining Domestic Food Price Deviations in Nicaragua and Honduras % of impact Product Honduras Nicaragua Vegetable oil 39.4 80.5 Rice 9.3 34.7 Sugar 1.9 13.5 Coffee (ground and beans) 27.7 23.1 Chicken 9.6 24.0 Beef 11.8 16.2 Beans (grains) 4.6 9.8 Eggs 28.4 8.1 Milk 8.8 32.6 Maize 49.2 15.6 Bread 43.0 8.6 Green bananas, plantains, or banana squares 38.8 59.9 Tortillas 2.5 9.8 Global average 22.9 25.9 Source: Authors’ calculations based on regression analysis results. 268 De Franco and Arias changes in demand generated by continuous price changes. Other costs such as menu costs, marketing costs, or corporate image costs are likely to play a smaller role in international price transmission, especially for basic staple foods. Unfortunately, no systemic time series for the logistics costs of these products exists for Nicaragua and Honduras, but a recent logistics costs study (Fernández and others 2010) presents interesting results that point to the importance of such costs in the supply chain for wheat and rice in these countries. The results show that producer prices are less than 25 percent of the price paid by the consumer of that same product, while most of these margins (often more than 75 percent) are related to the logistics costs of taking the product from the farm gate to the consumer’s hands. It is no surprise that adjustment costs play an impor- tant role in the degree to which international prices are transmitted into domestic prices. Price Transmission versus Price Wedges The Organisation for Economic Co-operation and Development (OECD) methodology for measuring the market price support (MPS) of agricultural products is useful for understanding the extent to which domestic food prices in Honduras and Nicaragua are composed of price supports.10 The methodology takes the international price of a given agri- cultural product and, by adding known transport costs, calculates the farm gate international price, which is the price that the producer or con- sumer would pay if there were no border protection or market distortions (law of one price). The difference between the international price (at farm gate) and the producer (or consumer) price is known as the price wedge (or MPS in the OECD methodology nomenclature). Tables 9.10 and 9.11 pres- ent the results obtained by Peña and Arias (2010) for select food products in 2006 and 2007 for Nicaragua and Honduras. In this analysis, price wedges are explained by policy-based price interventions (border meas- ures, such as tariffs and export taxes) and “other” factors. These “other” fac- tors may be the same factors that produce low elasticities of international price transmission, and this is shown by the fact that beans seem to pres- ent the most competitive market structure, with a relatively larger portion of the price wedge (MPS) than other products, explained by border pro- tection. This is the same result obtained in the price transmission analysis for Nicaragua, where beans present a relatively higher international price transmission estimate than the rest of the products. Are Food Markets in Central America Integrated with International Markets? 269 Table 9.10 Price Wedges for Select Agricultural Products for Honduras, 2006 and 2007 L per ton, unless otherwise noted International Domestic price Year and price at farm at producer % attributed % attributed to product gate level level Difference to tariff rates other factors 2006 Rice 3,694.40 7,887.00 4,192.60 17.0 83.0 Beans 9,382.00 11,089.80 1,707.80 77.9 22.1 Maize 3,239.40 6,116.00 2,876.60 7.2 92.8 Pork 27,937.80 22,355.10 –5,582.70 — 100.0 2007 Rice 4,136.50 9,372.00 5,235.50 0.0 100.0 Beans 11,343.80 24,805.00 13,461.20 12.0 88.0 Maize 4,240.30 9,174.00 4,933.70 9.8 90.2 Pork 41,736.70 22,556.50 –19,180.20 — 100.0 Source: Peña and Arias 2010. Note: The Honduran currency is the lempira. The applied rate corresponds to a weighted customs rate that takes into account the actual imports in and out quota with corresponding tariff rates. — = Not available. Table 9.11 Price Wedges for Select Agricultural Products in Nicaragua, 2006 and 2007 C$ per ton, unless otherwise noted International Domestic price Year and price at farm at producer % attributed % attributed to product gate level level Difference to tariff rates other factors 2006 Maize 2,832.60 2,970.80 138.20 0.0 100.0 Rice 4,609.80 4,328.70 –281.10 — 100.0 Beans 6,957.40 9,170.40 2,213.00 91.6 8.4 Soy 4,424.40 4,918.60 494.20 0.0 100.0 Beef 32,304.40 54,491.10 22,186.70 0.0 100.0 Pork 38,561.20 41,377.60 2,816.40 0.0 100.0 2007 Maize 3,827.70 4,887.20 1,059.50 0.0 100.0 Rice 5,215.90 4,483.20 –732.70 — 100.0 Beans 10,605.70 16,476.40 5,870.70 53.2 46.8 Soy 6,185.10 4,766.70 –1,418.40 0.0 100.0 Beef 33,835.10 54,037.90 20,202.80 0.0 100.0 Pork 42,001.10 53,028.80 11,027.70 0.0 100.0 Source: Peña and Arias 2010. Note: The Nicaraguan currency is the córdoba. The applied rate corresponds to a weighted customs rate that takes into account the actual imports in and out quota with corresponding tariff rates. — = Not available. 270 De Franco and Arias Main Conclusions and Policy Implications This study’s findings indicate that a complementary agenda should accom- pany agricultural trade liberalization so that consumers and farmers in Honduras and Nicaragua may fully benefit from international price signals and market opportunities. Given that the transmission of international to domestic prices for the agricultural products studied here is largely imper- fect, a careful assessment of welfare should be done in the context of future trade liberalization, since food consumers and farmers are receiving very low and delayed transmission of price signals. These results are con- firmed by testing for regional differences, for consumers, and for farmers in an aggregated and disaggregated fashion. They also are confirmed by assessing estimates of a recent agriculture price wedge analysis for these countries. When fluctuations in international prices are transmitted only slowly and imperfectly to domestic markets, consumers and producers at any point in time may be making decisions based on prices that do not rep- resent the real social costs and benefits of their actions, implying losses in the economic welfare of society. While logistical costs play an important role in the price wedge differ- ential, they may not be an important determinant of price transmission. But the cost of price adjustment may affect the latter. It is important to distinguish between price wedges and price transmission when discussing the welfare effects of trade liberalization. Price wedge analysis is a static assessment of the average difference at the farm gate between domestic and international prices, while price transmission is a dynamic analysis of changes in the relation between international and domestic prices, which is one indicator of the degree of market integration. The level of transac- tion costs may not necessarily be the only culprit in the low or incomplete transmission of price signals between international and domestic agricul- ture markets. Evidence from Nicaragua suggests that, for most of the agriculture sup- ply chains studied here (except for beans), there is little competition in the country’s domestic market structure. A few Nicaraguan companies own the majority share of the market, both to purchase and export agricultural products and to import and sell food domestically. Further assessment is required of other countries and food products to identify whether a lack of competition is the main cause of imperfect transmission of international agricultural prices into domestic markets. Obtaining information about the structure of domestic agriculture and food markets could shed light on country-specific impediments to Are Food Markets in Central America Integrated with International Markets? 271 increasing agricultural growth, reducing poverty, and improving rural com- petitiveness. Information on domestic market structure was difficult to obtain for this study, particularly for Honduras. But, as this chapter points out, even in a context where the domestic market structure concentrates purchasing and selling power in a few agribusiness companies, price transmission could be high. The drivers of imperfect price transmission in a context of a monopsony or cartel are a low elasticity of substitution of demand, low elasticity of substitution of supply, or high costs of price adjust- ment. Thus, even with a concentrated market structure in the agribusiness sector, public policy could increase the degree to which market signals reach domestic farmers by improving the capacity for and lowering the costs of switching from producing one food product to producing another (elastic- ity of substitution of supply), as well as lowering the costs for agribusinesses (traders and processors) to adjust the prices they pay farmers. Price transmission dynamics vary among products and countries. Honduras and Nicaragua display similarities and differences in commodi- ties, economies, and cultures. This suggests that distinct public policies should be developed for the Central American countries, to support their transition to free trade. For example, public policies respecting highly internationally traded agricultural commodities such as coffee and veg- etable oil should address international price volatility as a key source of risk for farmer and consumer income. For other products with low price transmission estimates, such as meat and sugar, public policy could focus more on managing domestic market risks, such as market power, produc- tivity, product quality, and local food availability. That said, coffee con- sumers in Honduras seem more responsive to international prices than consumers in Nicaragua, given the higher price transmission coefficients. Public policies to expand domestic demand for coffee in Honduras should address international price volatility, while in Nicaragua they could focus solely on domestic constraints. These are but a few examples of the differences among countries and products. Without complementary and differentiated public policies to improve international price transmission, domestic consumers and farm- ers in Honduras and Nicaragua will continue to lag in their response to international market signals in agricultural trade. Notes 1. The number of lags is determined by the information criteria of Schwarz (1978), with a maximum of 36 lags. 272 De Franco and Arias 2. All econometric estimations are done with this program. 3. The most one can do is to consider that international prices are weakly exoge- nous; in other words, that these prices do not suffer corrections when devia- tions from the cointegration relation exist. 4. No significant differences are found in the behavior of domestic prices as a result of positive or negative changes in international prices. 5. The value of prices after 45 months is irrelevant from a public policy point of view, but not from a theoretical one. 6. This is done to simplify the results. 7. Encuesta de Medición Nacional de Vida in Spanish. 8. The equation is as follows: Q/QG = (P/PG)-σ, where Q is the quantity of a given product composed of imported and local inputs, QG is the total quan- tity of final goods in the economy, P is the price of the same product, PG is the general price index of the economy, and σ is the elasticity of substitution of demand for that product. 9. For each food product, an average price paid by all households in the LSMS is estimated. Then the price that each household has to pay for each food product is divided by the average of each product for the entire set. This coef- ficient is a measure of the deviation of the price that the household must pay for a given food product with respect to the average price of that same prod- uct in the economy. 10. See the OECD methodology for details: http://www.oecd.org/dataoecd/ 33/48/32361345.pdf. References Bussolo, Maurizio, Samuel Frieje-Rodríguez, Calvin Djiofack, and Melissa Rodríguez. 2010. “Trade Shocks and Welfare Distribution in DR-CAFTA Countries.” World Bank, Washington, DC. Dixit, Avinash K., and Joseph E. Stiglitz. 1977. “Monopolistic Competition and Optimum Product Diversity.” American Economic Review 67 (3, June): 297–308. Dutoit, Laure C., Karla Hernández, and Cristóbal Urrutia. 2009. “Transmisión de precios para los mercados del maíz y arroz en América Latina.” The Selected Works of Laure C. Dutoit. http://works.bepress.com/laure_dutoit/6/. Enders, W. 1995. Applied Econometric Time Series. New York: Wiley. FAO (Food and Agriculture Organization). 2006. Agriculture Price Transmission in Latin America and the Caribbean in the Context of Trade Liberalization. Santiago, Chile: FAO. www.rlc.fao.org/prior/desrural/fao-bid/. Are Food Markets in Central America Integrated with International Markets? 273 Fernández, Raquel, Henry Vega, Santiago Flores, and Francisco Estrazulas. 2010. “Logistics Analysis of Selected Food Products in Nicaragua, Honduras, and Costa Rica.” World Bank, Washington, DC. Johansen, Soren, and Katarina Juselius. 1990. “Maximum Likelihood Estimation and Inference on Cointegration.” Oxford Bulletin of Economics and Statistics 52 (2): 169–210. Maddala, G. S., and In-Moo Kim. 2004. Unit Roots, Cointegration, and Structural Change. Cambridge, U.K.: Cambridge University Press. Peña, Hector, and Diego Arias. 2010. “An Assessment of Price Wedges in Central America.” World Bank, Washington, DC. Schwarz, G. 1978. “Estimating the Dimension of a Model.” Annals of Statistics 6 (2): 461–64. Vavra, Pavel, and Barry Goodwin. 2005. “Analysis of Price Transmission along the Food Chain.” Food, Agriculture, and Fisheries Working Paper 3, OECD, Paris. CHAPTER 10 Intellectual Property Rights and Foreign Direct Investment: Lessons for Central America Walter G. Park This chapter discusses the effects of intellectual property rights (IPRs) on foreign direct investment (FDI) and the role that regional economic inte- gration may play in determining those effects. The discussion is applied to the Dominican Republic–Central America Free Trade Agreement (DR-CAFTA) region. In terms of GDP and population, the DR-CAFTA region is a relatively small market with geographic advantages related to its proximity to the U.S. market. The region accounts for a small share of U.S. outward FDI. Most U.S. FDI in the region is concentrated in the wholesale trade and manufacturing industries, such as textiles. Other FDI from Asia occurs in agriculture and fishing. Given the characteristics of the DR-CAFTA market and its potential growth, the aim of this chapter is to analyze how strengthening IPRs in the context of economic integra- tion will influence the incentives of U.S. and other foreign firms to acquire or establish subsidiaries in this region. Chapter 15 of the DR-CAFTA lays out a comprehensive set of provi- sions to raise intellectual property standards and enforcement mecha- nisms in the region. The agreement calls for the ratification of or accession to the Copyright Treaty of the World Intellectual Property Office, Patent 275 276 Park Cooperation Treaty, Trademark Law Treaty, Madrid Agreement Concerning the Registration of Trademarks, and other global treaties. The agreement calls for national treatment, strengthens protection for digital products, and contains provisions for technological protection measures (such as prohibitions on circumvention devices). Enforcement levels and resources for IPRs are to be commensurate with the enforcement of laws in general. The agreement also protects pharmaceutical and agricultural chemical data that are submitted to regulators for purposes of evaluating safety and efficacy, the public disclosure of which may enable unfair com- mercial use of the data. The agreement has provisions to extend the terms of pharmaceutical patents if delays in marketing approval result in an unreasonable curtailment of the effective patent term. The question is, how influential are these and other intellectual property provisions for FDI into the region? Regional economic integration (via a free trade agreement) can affect foreign direct investment, as can intellectual property rights in general. But intellectual property reforms induced by a free trade agreement may have particular characteristics and effects on FDI. For example, a strengthening of IPRs will influence FDI in combination with a change in market size and market access. Furthermore, IPRs may matter differently depending on the nature of the FDI—that is, whether it is for production, research, sales, or distribution. This chapter is organized as follows. It begins by reviewing some descriptive statistics on intellectual property regimes and foreign direct investment within the DR-CAFTA region. Three measures of IPRs are examined: an index of patent protection based on statutory and case laws, an index of IPRs based on surveys of business executives, and rates of soft- ware piracy. Two sources of FDI data are examined: United Nations Conference on Trade and Development (UNCTAD) data and U.S. Bureau of Economic Analysis (BEA) data on U.S. outward FDI in the region. It then reviews some theoretical and empirical studies on the rela- tionship between FDI and regional economic integration, followed by a review of some theoretical and empirical studies on the relationship between FDI and IPRs, of which there are two types of studies. One type focuses just on FDI, and the other examines FDI alongside other modes of technology transfer. The chapter then builds on the literature reviewed to analyze the effects of IPRs on FDI within the context of an economic region such as DR-CAFTA. A final section provides some concluding remarks. Intellectual Property Rights and Foreign Direct Investment 277 Trends in IPRs and FDI in the DR-CAFTA Region It would be useful to start with a review of some trends in intellectual property rights as well as some trends in inward and outward foreign direct investment in the DR-CAFTA region. For perspective, these trends are compared to a reference group of countries in Latin America. These descriptive statistics are provided in tables 10.1–10.6. First, since the early 1990s, intellectual property laws have evolved in the DR- CAFTA region. In particular, patent rights have expanded. Table 10.1 shows an index of patent rights (from Park 2008). Although intellectual property protection encompasses many kinds of rights—patents, copy- rights, trademarks, geographical indications, industrial designs, and so forth—patent rights are likely to be the most relevant type of IPR for businesses that engage in inventive activity and for technological trans- fers that involve new inventions. The index of patent rights ranges from 0 (weakest) to 5 (strongest). The value of the index is obtained by aggre- gating the following five components: extent of coverage, membership in international treaties, duration of protection, absence of restrictions on rights, and statutory enforcement provisions.1 As table 10.1 shows, El Salvador has the strongest patent system in Central America. The Dominican Republic has the weakest. All six DR- CAFTA countries have adopted stronger patent law provisions since the Table 10.1 Strength of Patent Protection in DR-CAFTA Countries and Comparison Groups, 1990–2005 Country or region 1990 1995 2000 2005 Argentina 1.71 2.73 3.98 3.98 Brazil 1.28 1.48 3.59 3.59 Chile 2.26 3.91 4.28 4.28 Colombia 1.13 2.74 3.59 3.72 Mexico 1.36 3.14 3.68 3.88 Costa Rica 1.16 1.56 2.89 2.89 Dominican Republic 2.12 2.32 2.45 2.82 El Salvador 1.71 3.23 3.36 3.48 Guatemala 0.88 1.08 1.28 3.15 Honduras 1.25 1.9 2.86 2.98 Nicaragua 0.92 1.12 2.16 2.97 Latin America Mean 1.35 2.28 3.18 3.42 Standard deviation 0.44 0.77 0.73 0.42 Source: Park 2008. 278 Park Trade-Related Aspects of Intellectual Property Rights (TRIPS) agreement came into force in 1995. Five of the six DR-CAFTA countries are below the average strength of patent protection in Latin America. Only El Salvador is above the mean during 2000–05. As of 2005, the patent pro- tection levels of the DR-CAFTA countries are all below those of the five largest Latin American economies: Argentina, Brazil, Chile, Colombia, and Mexico. Table 10.2 provides an idea of the sources of the recent strengthening of IPRs in the DR-CAFTA region. All six member states are signatories to the TRIPS agreement, to the Paris Convention for the Protection of Industrial Property, and to the Berne Convention for the Protection of Literary and Artistic Works. However, none is a member of the Madrid agreement. Only Nicaragua thus far is a member of the International Union for the Protection of New Varieties of Plants. However, patent pro- tection for pharmaceuticals and chemicals in Nicaragua remains an issue. Otherwise, the other five countries have expanded the subject matter of Table 10.2 Intellectual Property Provisions in DR-CAFTA Countries, 2005–07 Costa Dominican Provision Rica Republic El Salvador Guatemala Honduras Nicaragua TRIPS agreement 1 1 1 1 1 1 Paris convention 1 1 1 1 1 1 Berne convention 1 1 1 1 1 1 Patent Cooperation Treaty 1 0 0 0 0 1 Madrid agreement 0 0 0 0 0 0 Patentability of chemicals 1 1 1 1 1 0 Patentability of pharmaceuticals 1 1 1 1 1 0 Patentability of software 0 1 0 0 0 0 Utility model protection 1 0 1 1 1 0 Plant and variety protection 0 0 0 0 0 1 Pretrial injunctions 1 1 1 1 1 1 Compulsory licensing for not working 0 1 0 0 0 1 Sources: Sinnott, Sinnott, and Cotreau 2008; Park 2008. Note: 1 = Signatory or available; 0 = Not a signatory or not available. Intellectual Property Rights and Foreign Direct Investment 279 patenting to include chemicals and drugs. The patentability of software remains an issue, except in the Dominican Republic. Four countries allow for utility model protection (that is, for adaptive and minor inventions, such as tools). All six countries allow for preliminary injunctions against an accused infringer while a patent case is pending. This is a useful mech- anism for enforcing patent rights. Still, four of these countries issue com- pulsory licensing for patented inventions that are, from the authorities’ perspective, inadequately exploited (either by local production or by importation). An alternative perspective on IPRs is provided by a survey of business executives conducted by the World Economic Forum (WEF). One of the survey questions in the WEF’s Global Competitiveness Report asks respon- dents in each country to rate, on a scale from 1 (lowest) to 7 (highest), whether intellectual property rights are well protected. The responses in each country are then averaged.2 A shortcoming of the IPR part of the survey is that a single question (or response to it) lacks nuance. The ques- tion the survey poses is very broad, since IPRs include patents, copyrights, trademarks, geographical indications, trade secrets, industrial designs, and so forth. Other drawbacks are that the survey responses are subjective or based on perceptions, not on actual rulings or prevailing legal conditions. The overall rating for a country may also not be fully comparable to the ratings of other countries since a different sample of respondents rates each country. For example, it is hard to compare a score of 3.5 for Costa Rica and a score of 3.0 for the Dominican Republic. Had the same group of people scored both countries, at least the scores could be comparable in an ordinal sense. Notwithstanding these limitations, the surveys pro- vide useful information on the actual experiences of firms with IPR pro- tection in their countries. The statutes may, for example, provide for preliminary injunctions, but in practice obtaining one may be a time- consuming and bureaucratic process. Furthermore, what may drive business behavior is the firm’s perception of the adequacy of IPRs rather than the stated provisions in the legal statutes. Business perceptions of IPR adequacy fell in 2005 but rose thereafter to reach a peak in 2006. The signing of the United States and DR-CAFTA free trade legislation by the White House in August 2005 may have con- tributed to the spike in perception (but a more formal statistical test would better explain the temporal patterns). After 2006, perceptions appear to fall toward levels that may be more consistent with the levels of IPRs prevailing in these countries. In terms of the relative perception of the adequacy of IPR protection across countries, Costa Rica, the 280 Park Dominican Republic, and Honduras are above the Latin American mean, while El Salvador, Nicaragua, and Guatemala are below. Another perspective on the IPR regimes in DR-CAFTA can be gleaned from statistics on software piracy rates in the region. The Business Software Alliance estimates the rate of piracy as the ratio of the level of piracy to total sales (that is, the sum of legitimate sales and illegitimate sales). The level of piracy is the difference between total installations of software and legal shipments of software. Since the TRIPS agreement was enacted, piracy rates have fallen significantly in the region. In 1995, rates were in the 90–98 percent range. In 2008 they were between 60 and 80 percent. However, those rates are still relatively very high. Only Costa Rica’s rate of piracy is below the mean rate of piracy in Latin America. The other five DR-CAFTA countries have piracy rates in the high 70 per- cent range. This may be why business perceptions of IPR adequacy remain low. To the extent that the software industry is representative, IPR enforcement remains an issue. More effective deterrents and resources for intellectual property agencies are needed. Table 10.3 shows that the three measures of the state of IPRs are cor- related in the Latin American region. Business perceptions of IPR ade- quacy are generally high in countries that have strong patent systems. Piracy rates have an inverse correlation with patent strength and business perceptions of IPRs; that is, they are lower if patent rights are stronger and if IPRs are more adequately enforced. As these are simply correlations, causality cannot be established without a more formal statistical analysis. For example, IPR issues and reforms may take on greater importance in regions that have high levels of piracy, imitation, and infringement. Thus IPR laws may respond to piracy and perceptions, as well as vice versa. Table 10.3 Correlations among Intellectual Property Measures Patent law index Piracy rate Intellectual property perception Patent law index 1 Piracy rate –0.46 1 Intellectual property perception 0.42 –0.71 1 Sources: The patent law index is from Park (2008), the piracy rate is from the Business Software Alliance’s estimates of software piracy, and the intellectual property perception is from the World Economic Forum’s survey of business managers. Note: Sample size is 18 Latin American countries (for 2005): Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, the Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Paraguay, Peru, Trinidad and Tobago, Uruguay, and República Bolivariana de Venezuela. Intellectual Property Rights and Foreign Direct Investment 281 To reveal some trends in FDI in the DR-CAFTA region, table 10.4 presents statistics on inward and outward FDI flows in the region, while table 10.5 shows the same for stocks. The flows and stocks are in nominal U.S. dollars. As such, the growth rates of FDI activity are downplayed, since the figures are not in real terms. Some important and interesting cross-sectional observations can nonetheless be made. First, there are far more inward FDI flows into DR-CAFTA countries than outward FDI flows from them. DR-CAFTA countries are not a major source of global capital. The inward flows of capital are important to DR-CAFTA coun- tries insofar as they represent a fairly significant percentage of gross fixed capital formation. In 2007, foreign capital inflows equaled almost half of domestic investment in El Salvador (see table 10.4) and about a third in Costa Rica. However, in Guatemala, FDI inflows are just 10 percent of domestic fixed investment. For Latin America as a whole, FDI inflows account for about a fifth of gross fixed capital formation. Over time, for all six countries, the ratio of FDI inflows to gross fixed capital formation has increased, indicating greater exposure to global supplies of capital and some trend expansion in inward FDI flows. FDI flows into and out of DR-CAFTA pale in comparison to those of the world or even of Latin America as a whole. The inward and outward stocks of FDI in DR-CAFTA tell a similar story (see table 10.5). As a percentage of GDP, the inward stock of FDI in 2007 was 55 percent for Nicaragua, about 20 percent for the Dominican Republic, almost 14 percent for Guatemala, and about 30 percent for the other countries in the group. Again, the stock of FDI in DR-CAFTA is a small percentage of the stock of FDI capital in Latin America and the world as a whole. Again, while the stock of FDI is in nominal rather than real dollars, the ratio of FDI stock to GDP suggests that it has expanded relative to mar- ket size, measured by GDP. Table 10.6 presents data on U.S. outward FDI to the DR-CAFTA region. The United States is, of course, an important player in the DR- CAFTA and has been a significant source of inward FDI for the region. In 2008, the United States accounted for about a fifth of the stock of inward FDI in Costa Rica and about half of the stock in El Salvador. Altogether, about 47 percent of the stock of inward FDI in Latin America (and other Western Hemisphere) countries is due to the United States.3 Table 10.7 shows the industry composition of U.S. FDI in DR- CAFTA, along with the composition in some reference groups. The fig- ures here are an average of 2004 and 2008. Most of U.S. FDI occurs in 282 Table 10.4 Flows of Foreign Direct Investment in DR-CAFTA Countries and Comparison Groups, 1980–2008 Country or group and mode 1980 1990 1995 2000 2005 2006 2007 2008 Costa Rica Inward flow 52.7 162.4 336.9 408.6 861.0 1,469.1 1,896.1 2,021.0 Inward flow as % of GrossCap 4.0 11.1 15.1 14.4 22.7 32.6 33.4 Outward flow 4.5 2.1 5.5 8.5 –43.0 98.1 262.5 5.9 Outward flow as % of GrossCap 0.3 0.1 0.2 0.3 –1.1 2.2 4.6 Dominican Republic Inward flow 92.7 132.8 414.3 952.9 1,122.7 1,528.3 1,578.9 2,884.7 Inward flow as % of GrossCap 5.5 8.0 16.4 19.7 20.5 23.5 20.5 Outward flow 0.0 0.0 14.6 61.0 20.8 –61.3 –16.8 –19.1 Outward flow as % of GrossCap 0.0 0.0 0.6 1.3 0.4 –0.9 –0.2 El Salvador Inward flow 5.9 1.9 38.0 173.4 511.1 241.1 1,508.5 784.2 Inward flow as % of GrossCap 1.1 0.3 2.1 7.8 19.6 8.0 46.0 Outward flow 0.0 0.0 –2.3 –5.0 112.9 –26.4 100.3 65.4 Outward flow as % of GrossCap 0.0 0.0 –0.1 –0.2 4.3 –0.9 3.1 Guatemala Inward flow 110.7 59.3 75.3 229.6 508.3 591.6 745.1 837.8 Inward flow as % of GrossCap 8.1 5.7 3.4 7.0 10.2 9.7 10.8 Outward flow 2.0 0.0 –19.1 40.1 38.2 40.0 25.4 16.3 Outward flow as % of GrossCap 0.1 0.0 –0.8 1.2 0.8 0.7 0.4 Honduras Inward flow 5.8 43.5 69.4 381.7 599.8 669.1 815.9 877.0 Inward flow as % of GrossCap 0.8 6.0 6.2 20.6 24.7 22.3 21.8 Outward flow 1.0 –1.0 –2.0 6.5 1.0 0.6 1.0 1.8 Outward flow as % of GrossCap 0.1 –0.1 –0.2 0.4 0.0 0.0 0.0 Nicaragua Inward flow 12.5 0.7 75.4 266.5 241.1 286.8 381.7 626.1 Inward flow as % of GrossCap 4.6 0.1 11.7 23.6 17.4 19.3 22.5 Outward flow 0.0 0.0 0.4 8.0 18.1 21.0 9.2 16.1 Outward flow as % of GrossCap 0.0 0.0 0.1 0.7 1.3 1.4 0.5 Latin America and the Caribbean Inward flow 6,415.8 8,926.1 29,513.0 98,354.6 77,069.7 93,303.2 127,491.4 144,377.1 Inward flow as % of GrossCap 3.3 4.0 8.8 25.8 16.2 15.9 17.8 Outward flow 898.8 299.7 7,459.2 49,579.0 35,967.2 63,619.4 51,741.1 63,207.0 Outward flow as % of GrossCap 0.5 0.1 2.2 13.0 7.6 10.9 7.2 World Inward flow 54,076.4 207,273.3 341,144.3 1,381,675.2 973,329.1 1,461,074.1 1,978,837.9 1,697,353.2 Inward flow as % of GrossCap 2.1 4.1 5.3 20.0 10.0 13.6 16.2 12.3 Outward flow 51,549.8 239,111.1 361,679.3 1,213,794.8 878,987.7 1,396,915.5 2,146,521.6 1,857,734.0 Outward flow as % of GrossCap 2.1 5.0 5.6 17.6 9.0 13.0 17.5 13.5 Source: UNCTAD 2009. Note: Figures are in millions of current U.S. dollars. The % of GrossCap denotes percentage of gross fixed capital formation. 283 284 Table 10.5 Stocks of Foreign Direct Investment in DR-CAFTA Countries and Comparison Groups, 1980–2008 current US$, millions, unless otherwise noted Country or group and mode 1980 1990 1995 2000 2005 2006 2007 2008 Costa Rica Inward stock 497.1 1,323.7 409.1 2,709.1 5,416.9 6,780.5 8,802.8 10,818.0 Inward stock as % of GDP 8.1 18.2 3.5 17.0 27.1 30.5 34.0 36.8 Outward stock 7.2 44.1 66.3 86.1 153.6 262.9 525.8 531.6 Outward stock as % of GDP 0.1 0.6 0.6 0.5 0.8 1.2 2.0 1.8 Dominican Republic Inward stock 238.7 571.5 –1,835.0 1,672.8 5,276.0 6,960.6 8,523.3 11,408.0 Inward stock as % of GDP 2.9 6.1 7.1 15.7 19.5 20.8 24.8 Outward stock 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Outward stock as % of GDP 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 El Salvador Inward stock 154.3 212.1 293.0 1,973.1 4,166.5 4,407.8 5,916.3 6,701.4 Inward stock as % of GDP 4.1 4.4 3.1 15.0 24.4 23.6 29.0 30.3 Outward stock 56.1 53.3 74.0 310.1 283.7 384.0 449.4 Outward stock as % of GDP 1.2 0.6 0.6 1.8 1.5 1.9 2.0 Guatemala Inward stock 701.0 1,734.2 2201.6 3,419.9 3,319.2 3,897.8 4,617.6 5,455.4 Inward stock as % of GDP 10.0 25.4 16.9 19.9 12.2 12.9 13.8 14.3 Outward stock 23.7 92.8 250.3 293.4 315.6 331.9 Outward stock as % of GDP 0.2 0.5 0.9 1.0 0.9 0.9 Honduras Inward stock 5.8 292.9 555.6 1,391.6 2,708.3 3,333.9 4,223.8 5,112.2 Inward stock as % of GDP 0.2 8.1 11.8 19.4 27.8 30.8 34.3 36.3 Outward stock 0.0 0.0 0.0 0.0 23.5 24.6 26.1 24.6 Outward stock as % of GDP 0.0 0.0 0.0 0.0 0.2 0.2 0.2 0.2 Nicaragua Inward stock 121.4 144.8 384.2 1,414.5 2,461.0 2,747.8 3,129.5 3,755.6 Inward stock as % of GDP 5.8 5.3 12.1 35.9 50.7 51.8 55.1 59.3 Outward stock 0.4 22.4 93.9 114.9 124.1 140.2 Outward stock as % of GDP 0.0 0.6 1.9 2.2 2.2 2.2 Latin America and the Caribbean Inward stock 40,959.5 110,546.8 185,122.7 502,487.2 817,560.1 933,610.3 1,125,109.4 1,181,615.7 Inward stock as % of GDP 5.3 9.5 10.3 24.4 31.5 30.9 32.0 25.7 Outward stock 47,518.1 57,642.9 87,892.1 204,387.9 335,424.2 430,344.6 500,548.1 561,432.9 Outward stock as % of GDP 6.2 5.0 4.9 9.9 12.9 14.2 14.2 8.3 World Inward stock 705,211.4 1,942,207.2 2,915,311.4 5,757,359.9 10,050,885.0 12,404,439.0 15,660,498.0 14,909,289.0 Inward stock as % of GDP 6.5 8.8 9.8 18.0 22.3 25.4 28.7 24.4 Outward stock 548,932.5 1,785,583.9 2,941,724.2 6,069,881.8 10,603,662.0 12,953,546.0 16,226,586.0 16,205,663.0 Outward stock as % of GDP 5.3 8.4 9.9 19.0 23.5 26.6 29.7 26.4 Source: UNCTAD 2009. 285 286 Park Table 10.6 Amount of U.S. Foreign Direct Investment in DR-CAFTA Countries and Comparison Groups, Historical Cost Basis, 2004–08 current US$, millions Country or group 2004 2005 2006 2007 2008 All countries 2,160,844 2,241,656 2,477,268 2,916,930 3,162,021 Latin America and other Western Hemisphere countries 351,709 379,582 418,429 508,711 563,809 Costa Rica 2,687 1,598 2,105 2,265 2,525 Dominican Republic 1,028 815 789 766 960 El Salvador 851 934 626 1,559 3,215 Guatemala 410 386 436 614 915 Honduras 755 821 864 640 700 Nicaragua 131 163 237 162 Source: U.S. Bureau of Economic Analysis, http://www.bea.gov/international/di1fdibal.htm. the manufacturing sector of DR-CAFTA. About half of U.S. FDI is in the manufacturing sector of the Dominican Republic. The exception is Nicaragua, where about 43 percent of U.S. FDI is in wholesale trade. Within manufacturing, the food and beverages sector in Guatemala is a major recipient of U.S. FDI. The food and electrical equipment sectors receive a significant share of U.S. FDI in Costa Rica. The sectoral distri- bution of U.S. FDI is somewhat more distinct in DR-CAFTA than in Latin America as a whole, where about a third of U.S. FDI is in finance and insurance and almost 40 percent is in holding companies. This reflects U.S. FDI across the world as a whole. Less than a fifth of U.S. global outward FDI is in manufacturing, a little more than a fifth is in finance, and a little more than a third is in holding companies. Wholesale trade accounts for just over 5 percent of U.S. global outward FDI. Thus, in comparison to these reference groups—that is, Latin America and the world as a whole—U.S. FDI in DR-CAFTA exhibits a difference in spe- cialization or motivation. For example, manufacturing production (due to lower labor costs) and wholesale trade (due to geographic location between North and South America) appear to be the key areas of focus in the DR-CAFTA region. Regional Integration and FDI This section briefly reviews previous studies on the relationship between regional integration and FDI. There is a large literature on this, so this chap- ter defers to studies cited in this section, which provide a more thorough Table 10.7 U.S. Foreign Direct Investment in DR-CAFTA Countries and Comparison Groups, by Industry, 2004–08 Average share of total industries (%) Other Western Hemisphere Dominican Industry All countries countries Costa Rica Republic El Salvador Guatemala Honduras Nicaragua Mining 4.8 5.4 0.0 0.0 0.5 0.0 0.0 –18.0 Manufacturing 17.7 11.7 37.8 52.0 14.3 33.9 22.2 0.0 Food 1.3 1.2 5.9 2.7 0.0 13.5 0.4 0.0 Chemicals 4.1 2.6 3.8 5.9 0.0 0.0 1.1 0.0 Metals 1.0 1.0 1.5 2.6 –1.0 0.1 0.5 0.0 Machinery 1.1 0.6 0.0 0.0 0.0 0.2 0.0 0.0 Computers and electronics 2.4 0.1 0.0 –1.0 0.0 0.0 0.0 0.0 Electrical equipment and related 0.7 0.3 4.6 0.0 0.0 0.0 0.0 0.0 Transportation 2.1 1.5 0.0 0.0 0.0 0.0 –0.6 –1.5 Other manufacturing 5.1 4.4 0.0 39.5 0.0 0.0 0.0 0.0 Wholesale trade 5.7 4.0 4.8 13.5 0.0 11.6 9.2 43.3 Information 3.2 1.7 0.3 0.0 0.3 0.6 0.0 –0.7 Depository institutions 3.7 0.3 0.0 0.0 0.0 0.0 0.0 13.3 Finance and insurance 20.1 33.3 0.0 0.1 1.6 19.2 7.7 0.0 Professional, scientific, technical services 2.5 0.4 3.3 0.1 0.0 0.2 0.0 3.5 Holding companies 35.5 38.5 0.0 –0.1 0.0 –1.1 –0.1 0.0 Other industries 6.8 4.8 –4.7 5.8 31.1 19.3 0.0 –1.2 Source: U.S. Bureau of Economic Analysis, http://www.bea.gov/international/di1fdibal.htm. 287 288 Park background. As of yet, there are limited, if any, empirical economic stud- ies on DR-CAFTA, since this is a new agreement. Most of the evidence is based, therefore, on other experiences with regional integration; for example, the European Community (EC), the Southern Cone Common Market (Mercosur), the North American Free Trade Agreement (NAFTA), the Association of South East Asian Nations (ASEAN), and the United States–Canada Free Trade Agreement. In a review of previous empirical studies, it would, of course, be useful to examine which prior cases best approximate DR-CAFTA. It would be especially useful to understand the basic principles or mechanisms by which regional integra- tion influences FDI. It then becomes an empirical issue as to which mech- anisms are most applicable to or observed in the DR-CAFTA region. Regional economic integration typically leads to a reduction in within- region trade barriers and investment restrictions. Studies on the relation- ship between integration and foreign direct investment focus mostly on the impact on inward, rather than outward, FDI. Theoretically, on the one hand, the easing of investment restrictions should enhance inward FDI. On the other hand, the easing of trade barriers may reduce FDI to the extent that the main motivation for FDI is to evade tariffs (that is, tariff- jumping motivation) or defuse tariffs by setting up a subsidiary that employs and produces locally. Another important motivation for FDI is to exploit intangible assets, such as a firm’s intellectual property assets (trademarks, copyrights, patents, or trade secrets) or marketing expertise. Another channel by which regional economic integration should affect FDI is through market size. In addition to providing a source of greater demand for a multinational firm’s products (which may be served more efficiently through local production rather than through exports), the larger common market enables a firm to spread the fixed costs of affiliate investments. To the extent that freer trade and investment stimulate eco- nomic growth, regional economic integration also produces dynamic effects: increased growth enhances the future profitability of the common market, thereby attracting more FDI. The larger and faster-growing mar- ket may, in turn, feed the incentives of multinational firms to innovate— to create new products or improve the quality of existing products. This should stimulate the research and development (R&D) of parent firms and augment the stock of intangible assets.4 Empirical studies on the effects of regional economic integration on foreign direct investment are based either on descriptive statistical analy- ses or on formal econometric modeling. Examples of the former include Mirus and Scholnick (1998), a study that focuses on U.S. FDI into Canada Intellectual Property Rights and Foreign Direct Investment 289 after the bilateral trade agreement between the United States and Canada. The data analysis here suggests a positive response of U.S. FDI to the agreement. The authors argue that this evidence dispels the notion that U.S. FDI was motivated by tariff jumping (since FDI continued and intensified after Canadian tariffs were lowered). They also dispel the notion that U.S. FDI occurred to take advantage of Canada’s natural resources (such as oil and timber). The evidence indicates that U.S. FDI increased in the manufacturing sectors, not in resource extraction. Lastly, the study finds evidence of agglomeration effects—that is, economies enjoyed by firms from clustering. One source of these effects may be knowledge spillovers and improved opportunities to learn about new technologies and markets; another may be the availability of more sup- porting industries (producers of components and services) that would not otherwise be available in less dense markets. Mirus and Scholnick (1998) find that FDI is greater in those sectors where U.S. firms were already present in Canada. Blomstrom, Globerman, and Kokko (2000), however, do not detect any significant cross-border affiliate activity between Canada and the United States after the free trade agreement. They argue that the United States–Canada Free Trade Agreement constituted a minor environmental change in the business climate. The two countries had already engaged in much cross-border investment such that increased regional economic integration had a marginal effect. Instead they find that regional eco- nomic integration has larger effects on FDI if the integration involves a northern country and a southern country (as in the case of NAFTA) or a southern country and another southern country (as in the case of Mercosur), rather than a northern country and another northern country (as in the case of the United States–Canada Free Trade Agreement). Blomstrom, Globerman, and Kokko (2000) identify two critical factors that determine the extent to which regional economic integration boosts FDI. The first is the existence of sufficient trade and investment liberal- ization. The second is the presence of good locational advantages in the regions concerned. For example, post-NAFTA, Mexico received a larger influx of FDI, not so much from U.S. and Canadian firms, but from firms outside NAFTA. Mexico has a locational advantage: proximity to the United States. Along with cheaper Mexican labor, foreign firms would find easier supply routes into the United States and thus have an incentive to invest in Mexican subsidiaries. Moreover, NAFTA occurred alongside other reforms in Mexico, such as investment and regulatory reforms, that may have been the more significant drivers of inward flows of FDI. 290 Park FDI in the Mercosur region also increased substantially, but was not evenly distributed within the region. Inward FDI increased especially in the larger markets of Argentina and Brazil, but not significantly in the smaller markets of Paraguay and Uruguay. Of course, other factors were involved (in addition to regional economic integration); namely, macro- economic stability in the larger member countries, which helped to reduce investor risk and uncertainty. Where the southern region can serve as an export platform for the products of foreign multinational firms, regional economic integration can attract vertical FDI in particular. In this case, different regions can contribute to the different stages of a product’s value added. A free trade and liberalized investment region can allow dif- ferent multinational affiliates to specialize more efficiently according to their location-specific advantages, whether they are resources or local know-how and skills. Balasubramanyam and Greenaway (1993) also study a regional eco- nomic integration case involving a northern country and a group of north- ern countries, namely, Japan and the European Community. In this case, the European common market had a significant influence on Japanese FDI into the EC. For Japan, the EC represented an important environ- mental shift, since prior Japanese affiliate activity in the EC was limited. However, Balasubramanyam and Greenaway (1993) argue that Japanese FDI was driven more by protection-defusing motives. There was concern that a “Fortress Europe” would block imports from Japan—hence the motivation for establishing Japanese branches within Europe. However, because the EC common market had tremendous growth potential, Japanese FDI continued to expand into the EC. Ultimately, the longer-run driver of Japanese FDI was the desire to s