Report No. 31458 - HN Honduras Investment Climate Assessment (In Two Volumes) Volume II November 27, 2004 Finance, Private Sector and Infrastructure Latin America and the Caribbean Region Document of the World Bank CURRENCY EQUIVALENTS (March 2003) FISCAL YEAR January 1 ­ December 31 Vice President: David de Ferranti Country Director: Jane Armitage Sector Director: Danny Leipziger Sector Leader: Manuel Sevilla Sector Manager: Susan Goldmark Investment Climate Unit Manager: Axel Peuker Task Manager: Marialisa Motta ii MAIN ABBREVIATIONS AND ACRONYMS ADR Alternative Dispute Resolution CABEI Central American Bank for Economic Integration CADERH Advisory Committee for Human Resources Development CAFTA Central American Free Trade Agreements CAS Country Assistance Strategy CEPAL Comisión Económica para América Latina y el Caribe CIN Comisión Interinstitucional de Normalización CITE Centro de Innovaciones Tecnológicas CNBS Comisión Nacional de Banca y Seguros COHCIT Consejo Hondureño de Ciencia y Tecnología COHEP Consejo Hondureño de la Empresa Privada CONATEL Comisión Nacional de Telecomunicaciones CUPROFOR Centro de Utilización y Promoción de Productos Forestales DGIC Dirección General de Investigación Criminal DGPI Dirección General de Propiedad Intelectual DICTA Dirección de Ciencia y Tecnología Agropecuniaria ECLAC Economic Comisión for Latin America and the Caribbean ENEE Empresa Nacional de Energía Eléctrica ESNACIFOR Escuela Nacional de Ciencias Forestales FIAS Foreign Investment Advisory Service FDT1 Fondo de Desarrollo de Telecomunicaciones (Primero) FDT2 Fondo de Desarrollo de Telecomunicaciones (Segundo) FDI Foreign Direct Investment FIDE Fundación de Inversión y Desarrollo de Exportaciones FHIA Fundación Hondureña de Investigación Agrícola GDP Gross Domestic Product GNI Gross National Income GMP Good Manufacturing Practice HACCP Hazard Analysis Critical Control Point Systems IC Investment Climate ICA Investment Climate Assessment ICE Instituto Costarricense de Electricidad ICS Investment Climate Survey ICT Information and Communication Technologies IDB Interamerican Development Bank IMF International Monetary Fund IV Instrumental Variables IHCAFE Instituto Hondureño del Café INCAE Instituto Centroamericano de Administración de Empresas INFOP Instituto Nacional de Formación Profesional IHNFA Instituto Hondureño de la Niñez y de la Familia ITU International Telecommunication Union ISO International Organization for Standardization MSME Micro, Small and Medium Enterprises NAC National Anti Corruption Council NAFTA North American Free Trade Agreements NGO Non Governmental Organization NQC National Quality Council OECD Organization for Economic Cooperation and Development OLS Ordinary Least Squares P Productivity PPIAF Public-Private Infrastructure Advisory Facility iii PRSC Poverty Reduction Support Credit RE Random Effects R&D Research and Development SERNA Secretaría de Recursos Naturales y Ambiente SUBTEL Subsecretaría de Telecomunicaciones TELESP Telecomunicações de São Paulo TFP Total Factor Productivity TSC Tribunal Superior de Cuentas UNAH Universidad Autónoma de Honduras WBI World Bank Institute WEF World Economic Forum WRANP Worldwide Responsible Apparel Production iv Honduras Investment Climate Assessment Table of Contents ACKNOWLEDGEMENTS .......................................................................................................................................ix INTRODUCTION.......................................................................................................................................................x CHAPTER 1. THE HONDURAN ECONOMY, CAFTA AND THE INVESTMENT CLIMATE: AN OVERVIEW .......................................................................................................................................................12 CHAPTER 2. RED TAPE, CORRUPTION AND CRIME ...................................................................................31 CHAPTER 3. INFRASTRUCTURE........................................................................................................................52 CHAPTER 4. QUALITY, INNOVATION AND LABOR SKILLS......................................................................72 CHAPTER 5. ACCESS TO AND COST OF CREDIT..........................................................................................88 CHAPTER 6. CONCLUSIONS: MOVING FORWARD......................................................................................99 BIBLIOGRAPHY....................................................................................................................................................101 ANNEX A : SAMPLE DESCRIPTION.................................................................................................................106 ANNEX B : STANDARD TABLES .......................................................................................................................109 ANNEX C : INVESTMENT CLIMATE DETERMINANTS OF PRODUCTIVITY IN HONDURAS: ECONOMETERIC METHODOLOGY ........................................................................................................119 ANNEX D : QUALITY SYSTEMS AND R&D IN HONDURAS.......................................................................151 v List of Boxes Box 1-1: The CAFTA Agreement ...............................................................................................................................18 Box 1-2: What is an Investment Climate Assessment? ...............................................................................................19 Box 1-3: Productivity Analysis: some background information to help interpret the results of Table 1-3 .................25 Box 1-4: Productivity Analysis: A Summary of the Methodology..............................................................................29 Box D-1: Main Components of a Typical National Quality System .........................................................................151 Box D-2: Institutions investing in R&D and supporting technology diffusion among Honduran firms ...................153 List of Tables Table 1-1: Determinants of GDP Growth....................................................................................................................14 Table 1-2: Changes in average tariffs, Central and Latin America..............................................................................17 Table 1-3: Elasticities and Semi-elasticities of Investment Climate (IC) Variables on Productivity, after Controlling for Other Firm, Industry and Country Characteristics................................................................24 Table 1-4: Average (log) Productivity Gains and Losses due to Investment Climate Variables.................................26 Table 1-5: Investment Climate at a Glance..................................................................................................................28 Table 2-1: Criminal activities affecting firms..............................................................................................................46 Table 3-1: Incidence, length and impact of power outages, international comparison................................................53 Table 3-2: Supply and demand in the electricity sector, 1985-2002............................................................................54 Table 3-3: Incidence, length and impact of telephone cuts, international comparison (2002).....................................57 Table 3-4: Incidence, length and impact of transport interruptions, international comparison (2002)........................63 Table 3-5: Merchandise losses while in transport, international comparison (2002) ..................................................63 Table 3-6: Productivity indicators for Puerto Cortés compared with international standards (1999)..........................69 Table 5-1: Key credit indicators, international comparison.........................................................................................95 Table C-1: General Plant Information and Production Function Variables...............................................................132 Table C-2: Investment Climate (IC) Variables..........................................................................................................133 Table C-3: IC and Plant Characteristics (C) Variables..............................................................................................134 Table C-4: Investment Climate -- Perception (PE) Variables ..................................................................................135 Table C-5: Number of firms that enter in the IC regressions, by Industry and by Country.......................................135 Table C-6: Number of firms that enter in the IC regressions, by Industry and by Year............................................136 Table C-7: List of Significant ICA Variables and their Units of Measurement ........................................................137 Table C-8: Correlation between Solow residuals in levels and estimated productivity.............................................138 Table C-9: Production function parameters from the restricted estimation...............................................................139 Table C-10: Production function parameters from the unrestricted estimation by industry: Cobb Douglas Specification................................................................................................................................................140 Table C-11: Production function parameters from the unrestricted by industry estimation: translog specification. ...............................................................................................................................................141 Table C-12: Two step unrestricted by industry estimation for young and old firms..................................................142 Table C-13: Two step unrestricted by industry estimation for small and large firms. ..............................................143 Table C-14: Elasticities or semi-elasticities and percentage of R-square Productivity Contribution of Each explanatory variable, after controlling for the other IC and Plant Control variables...................................144 Table C-15: Guatemala -- percent average (log) productivity gains and losses due to Investment Climate (IC)..............................................................................................................................................................145 Table C-16: Honduras -- percent average (log) productivity gains and losses due to Investment Climate (IC)..............................................................................................................................................................146 Table C-17: Nicaragua -- percent average (log) productivity gains and losses due to Investment Climate (IC)..............................................................................................................................................................147 vi List of Figures Figure 1-1: GDP Growth (Annual percentage)............................................................................................................12 Figure 1-2: Determinants of Growth in Honduras (1970-2000) -- A Growth Accounting Analysis..........................13 Figure 1-3: Structure of the export economy (2001). Honduras..................................................................................14 Figure 1-4: Composition of agricultural exports. Honduras........................................................................................15 Figure 1-5: Manufacturing. Composition of value added to the economy -- excluding maquila. Honduras. ............16 Figure 1-6: Maquila. Value added to the Honduran economy.....................................................................................16 Figure 1-7: Change in export to GDP between the 1980s and the 1990s. ...................................................................16 Figure 1-8: The perspective of Honduran firms -- main constraints on growth mentioned by more than forty percent of surveyed firms .....................................................................................................................21 Figure 1-9: Percentage of sales lost due to quantifiable constraints ............................................................................21 Figure 2-1: Number of days needed to acquire each license required to register a firm..............................................33 Figure 2-2: Number of days of inspections, international comparison........................................................................35 Figure 2-3: Types of inspections. Honduras, Nicaragua and Guatemala. (Percentage values)....................................35 Figure 2-4: Days of inspections, by size. Honduras. ...................................................................................................35 Figure 2-5: Percentage of management time spent with regulators, by size. Honduras. .............................................35 Figure 2-6: Bribery needed to get things done (percentage of sales), international comparison. ................................36 Figure 2-7: Percentage of firms that pay a bribe for each service, Honduras..............................................................36 Figure 2-8: Informal economy (percentage of GNI)....................................................................................................37 Figure 2-9: Percentage of contract paid to win a public contract, international comparison.......................................40 Figure 2-10: Percentage of contract paid to win a public contract, by size. Honduras................................................41 Figure 2-11: Percentage of contract paid to win a public contract, by industry. Honduras.........................................41 Figure 2-12: Percentage of firms that do NOT believe that the courts will uphold their contracts and property rights in a dispute, international comparison. .................................................................................42 Figure 2-13: Percentage of firms that had delays in payments and went to court and average time needed to settle a case, by size. Honduras. ....................................................................................................................43 Figure 2-14: Percentage of firms that had delays in payments and went to court and average time needed to settle a case, by Region. Honduras................................................................................................................43 Figure 2-15: Crimes reported to and solved by the Police. International Comparison. (Percentage values)...............46 Figure 2-16: Frequency of crime and number of cases, by geographical area. Honduras. ..........................................47 Figure 2-17: Crime losses and security expenses, by geographical area. Honduras. /a...............................................47 Figure 2-18: Frequency of crime and number of cases, by size. Honduras. ................................................................49 Figure 2-19: Crime losses and security expenses, by size. Honduras. (Percentage values).........................................49 Figure 3-1: Number of hours and percentage of sales lost because of power cuts -- by firm size. Honduras. /a....................................................................................................................................................................53 Figure 3-2: Distribution and Transmission Losses as Percentage of Output, Central American Countries ................55 Figure 3-3: Average waiting time for telephone line installation (Number of days)...................................................57 Figure 3-4: Firms that use e-mail and websites to communicate with clients and suppliers .......................................59 Figure 3-5: Firms that use e-mail to communicate with clients and suppliers, by size--Central American countries........................................................................................................................................................59 Figure 3-6: Percentage of firms that lost merchandise while in transport and loss suffered. By Department. Honduras. ......................................................................................................................................................65 Figure 3-7: Percentage of firms that lost merchandise while in transport and loss suffered. By Size. Honduras. ......................................................................................................................................................65 Figure 3-8: Map of Honduras ......................................................................................................................................66 Figure 3-9: Transport cost perceived as a major constraint to CAFTA.......................................................................66 Figure 3-10: Losses while merchandise is transferred (percent of consignment value) -- for firms that experienced a loss..........................................................................................................................................67 Figure 3-11: Customs clearance time for imports (days, mean values).......................................................................67 Figure 4-1: ISO certified firms -- international comparison. Percentage values. .......................................................73 Figure 4-2: ISO Certified firms in Honduras by size. Percentage values. ...................................................................73 vii Figure 4-3: ISO Certified firms in Honduras by sector. Percentage values.................................................................73 Figure 4-4: Firms using quality standards other than ISO -- international comparison. Percentage values...............74 Figure 4-5: Honduran firms using quality standards other than ISO, by size. Percentage values. ..............................74 Figure 4-6: Honduran firms using quality standards other than ISO, by sector. Percentage values............................74 Figure 4-7: Innovating firms -- international comparison..........................................................................................77 Figure 4-8: Technology adoption, international comparison.......................................................................................78 Figure 4-9: Firms with computerized equipment.........................................................................................................78 Figure 4-10: Honduran firms licensing foreign technologies, by size. Percentage values...........................................80 Figure 4-11: Honduran firms licensing foreign technologies, by sector. Percentage values. ......................................80 Figure 4-12: The limited role of linkages in technology adoption--Central American countries...............................80 Figure 4-13: Percentage of firms providing formal training, international comparison ..............................................84 Figure 4-14: Percentage of Honduran firms providing formal training, by firm size..................................................84 Figure 4-15: Reasons cited by Honduran firms for not training their employees........................................................84 Figure 4-16: Percentage of Honduran firms using private training schools, by firm size............................................86 Figure 4-17: Percentage of Honduran firms using private training schools, by sector................................................86 Figure 5-1: Share of firms with loans, international comparison, percentage values. .................................................89 Figure 5-2: Share of firms demanding and obtaining loans, by size. Honduras, percentage values............................89 Figure 5-3: Share of firms demanding loans, maquila versus non-maquila. Honduras. Percentage values. ...............90 Figure 5-4: Main Sources of finance for working capital needs, international comparison, percentage values.............................................................................................................................................................91 Figure 5-5: Sources of finance for working capital needs, by size, Honduras, percentage values. .............................91 Figure 5-6: Constraints preventing firms from applying to a loan...............................................................................92 Figure 5-7: Interest rates, percentage values................................................................................................................93 Figure 5-8: Interest Rates in USD for Large Firms ....................................................................................................93 Figure 5-9: Average value of collateral (as percent of loan), international comparison.............................................94 Figure 5-10: Sources of collateral (as percentage of loan), Honduras........................................................................94 Figure 5-11: Administrative expenses over assets, international comparison .............................................................96 Figure A-1: Size Distribution of the Sample ............................................................................................................106 Figure A-2: Exporting Profile....................................................................................................................................106 Figure A-3: Industrial Profile ....................................................................................................................................107 Figure A-4: Exporting Industries...............................................................................................................................107 Figure A-5: Geographical Distribution of the Sample...............................................................................................107 Figure A-6: Geographical Distribution of Selected Industries ..................................................................................107 Figure C-1: Olley-Pakes decomposition of Solow residual by country.....................................................................148 Figure C-2: Olley-Pakes decomposition of Solow residual by firm size...................................................................149 Figure C-3: Olley-Pakes decomposition of industry Solow residual.........................................................................150 Figure D-1: Institutional Setting of a Typical National Quality System....................................................................152 Figure D-2: Institutional Setting of the Honduran National Quality System.............................................................152 viii ACKNOWLEDGEMENTS This report was prepared by a team lead by Marialisa Motta and composed of Aquiles Almansi, Geeta Batra, José Barbero, Bernice K. Van Bronkhorst, Modibo Khane Camara, Ximena Clark, Paulo Correa, Juan Miguel Crivelli, Lydie Ehouman, Alvaro Escribano, Ana Margarita Fernandes, Marc Forni, Isabel Sanchez Garcia, Michael Goldberg, Jose Luis Guasch, Carlos Gomez, Luke Haggarty, Lucio Monari, Elena Morachiello, André Pizarro, Francesca Recanatini, Heisnam Singh, Stefka Slavova, Rajeev Swami, Clemencia Torres, Eloy Eduardo Vidal and Eduardo Zolezzi. At various stages, the team received precious guidance from Jane Armitage, Marianne Fay, Susan Goldmark, Danny Leipziger, Guillermo Perry, Axel Peuker, and Manuel Sevilla. The investment climate survey was designed by a DEC team composed by Ximena Clark, Pablo Fajnzylber and Luke Haggarty and was implemented by CID Gallupp. The econometric analysis was performed by a team composed by Alvaro Escribano, Jose Luis Guasch, Ana Margarita Fernandez and Heisnam Singh. Stefka Slavova reviewed and summarized the findings of the econometric analysis in the Report. The report benefited from background papers prepared by Araceli de Leon, Julio Fuster and Diane Thompson. Special thanks go to Aarre Laakso, who provided invaluable editorial support and substantial and timely help with the Executive Summary. The report was prepared under the general supervision of Jane Armitage -- Country Director for Central American countries, and Danny Leipziger -- Director of the Finance and Infrastructure Department of the Latin America Region. The Honduras team collaborated closely with the teams working on the Nicaragua and the Guatemala Investment Climate Assessments. Lily Franchini, Nelida Mattos and Eric Palladini provided precious document, production and editing support. The team received thoughtful contributions from distinguished representatives of the Honduran private and public sectors during two seminars held in Tegucigalpa at the end of March, 2004. The team would like to thank all the members of the Honduran Competitive Commission and of the Consejo Hondureno de la Empresa Privada (COHEP) for their valuable insights and for contributing to improve the overall quality of this work. Special thanks go to the Vice President of the Republic of Honduras, Vicente Williams, for his precious comments and advice; the Executive Director of FIDE, Vilma Sierra, the coordinator of the World Bank Competitiveness Project, Santiago Herrera and Juan Pablo Carias and Maribel Banegas for providing useful inputs and information at various stages of development of this report, and for their invariable help with all World Bank initiatives related to productivity, trade and the investment climate in Honduras. Peer reviewers are Theresa Bradley (CICIC, World Bank), Mary Hallward-Driemeier (DECRG, World Bank), and Joseph Manoharan Owen (LCCHN, World Bank). The team would like to thank the peer reviewers and all the other Bank colleagues that provided useful comments on the first version of this report. ix INTRODUCTION 1. Objectives, rationale and scope of the study. The objective of this study is to evaluate private sector constraints on growth in Honduras. The study does this by presenting the results of a survey of 450 Honduran manufacturing firms conducted during the year 2003 (the Investment Climate Survey, or ICS) to assess key bottlenecks that detract from productivity and growth. The study argues that ongoing trade liberalization initiatives will not, per se, lead to sustainable increased economic growth in Honduras, and that reducing bottlenecks to firms' operation is necessary for Honduras to exploit the opportunities offered by the trade agreements and start on a path of sustainable growth and reduced poverty. The study addresses constraints on growth by looking at governance, access to infrastructure and credit, technology adoption and labor skills. These four variables stood out on the basis of quantitative evidence from the ICS, including both descriptive statistics and the results of an econometric analysis based on the ICS data. 2. Target audience. This study targets both the Honduran public and the private sectors. Its main aim is to stimulate and inform a dialogue between the Honduran Government and representatives of local firms in all areas related to the investment climate. Its dissemination objectives are two fold. One the one hand, we hope that the report will help the Government devise a microeconomic reform agenda that could help increase private sector's productivity and growth. On the other hand, we would like the study to help the private sector identify new potential for growth, become more dynamic, identify and exploit linkages with local and foreign firms and public institutions that could increase its innovation and productivity levels. The study does not attempt to stipulate a comprehensive and final set of policy reforms. Rather, it should be considered as a first input for discussion and for design of a Government Action Plan to improve the country's investment climate. 3. Organization of the study. Key findings of the Study are summarized in Volume 1 -- Executive Summary and Policy Recommendations. The detailed Investment Climate Assessment is included in Volume II, which is divided into five main sections. The first section presents a brief assessment of the performance of the Honduran economy over the past few years, an overview of the key elements of the investment climate and a brief description of the Honduras Investment Climate Survey. The rest of the report analyzes the results of the survey in the four key areas of the investment climate:1 (1) business-government relationships -- with specific reference to red tape, procurement, the judiciary system and crime; (2) infrastructure -- energy, information and communication technologies (ICTs), and transport and logistics; (3) the capacity of Honduran firms to move up the value chain by introducing innovations, adapting new technologies, certifying the quality of their processes and products, and improving the skills of their labor force; and (4) access to and cost of finance. For each variable, the report examines the results of the survey, presents complementary evidence to better assess the country's progress in each area, and suggests key policy recommendations. 4. Comparator countries. To the extent possible, Honduran firms' performance and views are compared to those of firms located in other Central American countries and other Regions. Comparator countries have been chosen on the basis of three criteria: (1) similar income countries (e.g., Ecuador, Nicaragua, Guatemala), (2) countries that are competing with Honduras (e.g., China and Bangladesh -- both operating in the apparel manufacturing sector) and (3) countries where the World Bank has already conducted investment climate surveys and for which the results of the surveys are available and comparable with those of Honduras. 1This report does not address the country's macroeconomic framework. Macroeconomic policies are covered by the IMF's PRGF Program. See CHAPTER 1 in this volume for further justification for focusing on microeconomic constraints. x 5. Sources of information for diagnostic and policy recommendations. The report describes the main findings of the survey and provides policy recommendations for each area covered by the survey, including a brief summary of the Government's ongoing and planned initiatives and some additional suggestions. The ICS is a diagnostic tool that helps identify key bottlenecks for firms' development and growth, but it is not designed to devise policy recommendations. The policy recommendations included in the study draw from background papers that were prepared to complement the findings of the ICS and to design the World Bank Honduras Trade Facilitation and Productivity Enhancement Project2. Other essential documents used as references for policy recommendations include the World Bank Country Assistance Strategy (2003) and the Honduras World Bank PRSC Matrix (2004) -- providing short- and medium-term recommendations that the Government and the World Bank has discussed during the past few months. Each Chapter of the Study will include references to documents used for additional recommendations related to specific topics. 2See a description of the Project at: http://web.worldbank.org/external/projects/main?pagePK=104231&piPK=73230&theSitePK=40941&menuPK=228 424&Projectid=P070038 xi CHAPTER 1. THE HONDURAN ECONOMY, CAFTA AND THE INVESTMENT CLIMATE: AN OVERVIEW3 1.1 The overarching challenge for the Honduran economy is to address its low and fluctuating economic growth, which has been driven by natural shocks, macroeconomic imbalances and microeconomic constraints on private investment and productivity. GDP growth in Honduras has been volatile over the past decade, ending with a downturn in 2000-2002. Although GDP growth averaged 3 percent between 1990 and 2002, growth rates varied considerably -- ranging from a peak of over 6 percent in 1993 to a low of ­2 percent in 1999. Moreover, GDP growth has fallen by over 2 percentage points since early 2000. Both the volatility in annual rates of growth and the sharpness of the recent downturn have been greater in Honduras than in lower-middle income countries worldwide or in Latin America on average (see Figure 1-1). Important factors contributing to Honduras' poor performance include: (a) natural shocks, especially the devastation caused by natural disasters in 1974 and 1998 (hurricane Mitch), and the precipitous drop in coffee prices, down by over two thirds since 1998; (b) cyclical downturns caused by international crises (e.g., the Asian, Argentinean and Brazilian crises, as well as the recent recession in the US) that resulted in terms-of-trade loss (down 24 percent from 1997 to 2002)4; (c) recurrent macroeconomic imbalances -- recently (1995-2000), the appreciation of the local currency coupled with an increase in real wages greater than productivity gains inflated labor costs and reduced the cost advantage of Honduran exports; and (d) regulatory, institutional and structural factors -- microeconomic constraints that deter private investment and productivity growth. Figure 1-1: GDP Growth (Annual percentage) 7 6 5 4 3 2 1 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 -1 -2 -3 Honduras LAC lower-middle income Lower middle income Source: World Bank, WDI 2003. 1.2 Although cyclical trends, macroeconomic factors and natural shocks have played an important role in shaping Honduras' poor growth pattern, this study focuses on the future potential for reducing microeconomic constraints on productivity and growth. While acknowledging the irrefutable importance of macroeconomic soundness and of cyclical trends, as well as the potentially destructive economic effect of natural shocks, this study focuses on assessing microeconomic constraints on productivity and growth. The analysis of microeconomic determinants of growth and productivity has gained interest in recent years, both among researchers and among Governments worldwide. There is a growing body of evidence -- and a growing consensus -- on the need for Governments to go beyond 3 This Chapter was prepared by Marialisa Motta and Stefka Slavova -- summary of econometric analysis and methodology. Juan Miguel Crivelli and Elena Morachiello provided valuable inputs. 4World Bank, Latin America Regional Database. 12 sound macroeconomic policies to help remove hurdles to private sector development and devise policies that can increase firms' productivity and growth. A first step in this direction is the definition of new instruments to better identify and quantify the exact barriers, costs and risks that firms face in various countries. 1.3 The importance of microeconomic factors -- and structural reforms to improve them -- in Honduras becomes more clear when looking at the composition of the country's GDP growth, which has been slowed by a sharp, consistent decrease in Total Factor Productivity (TFP). During the 20th century, investments in physical capital and infrastructure were thought to be the force driving economic growth. However, the failure of many developing countries to grow despite increased levels of investment indicated that increasing labor and capital was not, per se, sufficient.5 Recent research shows that TFP -- i.e., the way in which capital and labor are used, rather than their simple accumulation -- may be the key to growth. In fact, TFP explains as much as 50 percent of GDP increase in OECD countries and China.6 Conversely, Honduras' GDP growth over the last thirty years has been driven by a (substantial) increase in the contribution of labor growth to GDP and a (limited) increase in capital growth. Growth in Honduras has been significantly slowed by a sharp, consistent decrease in TFP -- see Figure 1-2.7 This indicates that the ability of the Honduran economy to increase the value added per unit of output, or to increase output per unit of input has been -- and continues to be -- very limited. Figure 1-2: Determinants of Growth in Honduras (1970-2000) -- A Growth Accounting Analysis Labor Growth Capital Growth TFP Growth GDP Growth 6.00% Labor growth 4.00% GDP growth 2.00% Capital growth 0.00% 1971-1980 1981-1990 1991-2000 -2.00% -4.00% TFP growth -6.00% 5See Easterly (1999). 6 "It turns out that it is higher productivity that has performed the new economic miracle in Asia. Chinese productivity increased at an annual rate of 3.9 percent during 1979-94, compared with 1.1 percent during 1953-78. By the early 1990s, productivity's share of output growth exceeded 50 percent, while the share contributed by capital formation fell below 33 percent". Hu, Z. and M. Khan Why is China Growing So Fast? (International Monetary Fund -- Economic Issues, 2004). 7H. Juan-Ramon, in a study of the determinants of growth in Honduras, confirms that lack of growth in the country can mainly be attributed to the offsetting negative influence of low labor and capital productivity (International Monetary Fund -- Policy Discussion Paper, 1999). 13 Table 1-1: Determinants of GDP Growth GDP growth Labor growth Capital growth TFP growth (percent) (percent) (percent) (percent) 1971-1980 5.39 2.05 2.12 1.22 1981-1990 2.43 2.98 1.14 -1.69 1991-2000 3.21 5.30 1.87 -3.95 Source: Loayza, Fajnzylber, Calderón, 2002. Note: TFP is the residual. Labor and Capital Growth contribution are calculated as follows: Percent Labor Growth = percent Growth Employment * percent Growth Worked Hours * percent Growth Human Capital Stock Percent Human Capital Growth = Weighted average of educational attainments by shares of adult population with different educational levels. Percent Capital Growth = percent Growth Physical Capital * (1 ­ percent Growth Unemployment Rate) 1.4 Low productivity levels mirror the under-development of the Honduran economy and its dependence on the export of a few agricultural commodities, which -- despite recent diversification -- is still significant. While the importance of agriculture as a component of GDP has been decreasing (from 23 percent in 1997 to about 14 percent in 2001, in part due to Hurricane Mitch) -- see Figure 1-3 -- agriculture remains the largest employer in the Honduran economy, accounting for about 60 percent of jobs and around one-third of all merchandise export earnings. Historically, the Honduran agricultural sector has been dependent on few commodities. To the current day, coffee and bananas account for 48 percent of the country's total agricultural exports -- see Figure 1-4. Higher value agricultural products -- such as cultivated shrimp, melon, pineapple and palm oil -- contribute about 52 percent of the agricultural exports. Increasing production and exports of these products, as well as of other value added agricultural crops (e.g., tropical fruits, fresh vegetables), would free the economy from dependence on the price fluctuations of traditional commodities and ensure higher returns. Developing an advanced agro processing industry would also help exploit the country's rich endowment of agricultural resources. Recent private sector investments in higher value agricultural activities -- approximately US$100 million has been invested during the last two years, mainly in palm oil -- confirm that Honduras' agricultural sector is diversifying. Several Government initiatives are catalyzing this change. Together with private and non profit local institutions, the Government has recently supported export diversification programs and facilitated the organization of groups of producers of valued added crops to help them devise successful export strategies. More efforts along these lines are needed in the future to help the economy diversify its production and export base and start on a more stable, sustainable growth path. Figure 1-3: Structure of the export economy (2001). Honduras. Honduras - GDP Agriculture 14% Other Industry 11% Services 55% Manufacturing 20% Source: World Development Indicators. 14 Figure 1-4: Composition of agricultural exports. Honduras. 100.0% 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 48 percent Bananas 20.0% and Coffee 10.0% 0.0% 1993 1994 1995 1996 1997 1998 1999 2000 2001 BANANO CAFE MADERA AZUCAR CAMARON CULTIVADO CAMARON DE EXTRACCION LANGOSTAS TABACO MELONES PIÑAS ACEITE DE PALMA Source: Banco Central de Honduras. 2004 1.5 Maquilas are playing an increasingly important role in the manufacturing sector, but face increasing competition from Asian firms. The Honduran manufacturing industry may be divided into four main sub sectors -- see Figure 1-5 and Figure 1-6. First are the relatively long-established food processing and drink industries, as well as textiles and clothing for the domestic market, which make up about 67 percent of total value added. Second are the processing industries related to the principal agro- exports, such as sugar milling, meat packing, seafood and paper and pulp manufacturing (other manufacturing in the charts). These sectors are dominated by the two US fruit companies, Chiquita and Dole.8 Third are chemical, metal and machinery, and transport equipment industries (8-9 percent of the total value added). Finally is maquila (offshore assembly for re-export), which operates mainly in the apparel sector. Maquilas emerged after the 1990s and grew very rapidly. Honduras is now the largest maquila manufacturer in Central America, and maquila is the country's main export industry (US$560 million in net foreign-exchange earnings in 2002). While the maquila sector has grown rapidly -- especially after export tax promotion incentives were eliminated -- little diversification outside of the sector's main activity has taken place. Increasing competition from Chinese and other Asian firms -- thanks to the low cost of the labor force, as well as their increasing ability to produce valued-added products at low prices -- is forcing Honduran maquilas to rethink their strategies and better choose the niches in which they want to operate. 8The Economist Intelligence Unit, Honduras Country Profile 2003 (Economist Intelligence Unit, 2004). 15 Figure 1-5: Manufacturing. Composition of value added Figure 1-6: Maquila. Value added to the to the economy -- excluding maquila. Honduras. Honduran economy. 50.0% 12.0% Honduras - Industrial Structure 45.0% Value Added Maquilas over Good Exports Value Added Maquilas over GDP 10.0% Chemicals 40.0% Textiles and 3% clothing 35.0% 8% 8.0% 30.0% Other 25.0% 6.0% manufacturing 30% 20.0% 4.0% 15.0% Food, beverages and tobacco 10.0% Machinery and 59% 2.0% transport 5.0% equipment 0% 0.0% 0.0% 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 p/ 2001 p/ 2002 p/ Source: World Development Indicators Source: Banco Central de Honduras. 2004 Note: Maquila figures are reported separately from the rest of the manufacturing industry because the Central Bank of Honduras keeps separate statistics for the maquila industry. For maquilas, we use value added, rather than exports, because these firms import most of their supplies, so using export values would over-rate maquilas' contribution to the economy. 1.6 All in all, export increase and export diversification into higher value-added activities are still in their infancy in Honduras. The (already) low level of tariffs points to the need to undertake complementary measures to increase the country's participation in the international economy. Figure 1-7 shows an international comparison of the change in export to GDP ratio between the 1980s and the 1990s. Honduras' export ratio has increased by only 13 percent, versus 74 percent in China and more than 100 percent in Bangladesh, for example. This is despite the fact that, by the end of the 1990s, Honduras had the lowest tariffs among all comparator countries included in Figure 1-7 and most Latin American countries -- see Table 1-2.9 In fact, average tariffs in Honduras have decreased from 42 percent in 1989 to 8 percent in 1999 (below all Latin American countries). This evidence -- together with low TFP and high dependency on a few low value-added sectors -- points to the need to go beyond simple tariff reduction and trade liberalization to increase and diversify the country's exports. Figure 1-7: Change in export to GDP between the 1980s and the 1990s. Peru -28.1% Brazil -19.1% Guatemala -3.5% Honduras 13.3% Pakistan 37.1% China 74.0% India 75.5% Bangladesh 108.5% -60.0% -20.0% 20.0% 60.0% 100.0% 140.0% Source: World Development Indicators 9See also D. Dollar, M. Hallward-Driemeier and T. Mengistae (2003). 16 Table 1-2: Changes in average tariffs, Central and Latin America 1989 1995 1999 Honduras 41.9 9.7 8.1 Nicaragua N/A 10.7 10.9 Guatemala 16.0 12.0 7.6 El Salvador 16.0 10.2 5.7 Costa Rica 16.4 11.2 3.3 México 10.6 12.6 10.1 Argentina 43.7 10.5 11.0 Brazil 42.2 12.0 13.3 Chile 15.1 11.0 10.0 Peru 68.1 16.3 13.0 Source: Perry G., Lederman D. and Suescún R. (2002) based on IDB figures. STRENGTHENING THE IMPACT OF CAFTA ON THE HONDURAN ECONOMY BY IMPROVING THE INVESTMENT CLIMATE 1.7 The Central America Free Trade Agreements (CAFTA) will give Honduras a unique opportunity to increase its participation in the global economy. Thanks to the ongoing CAFTA negotiations, Honduras will have the opportunity to accelerate its integration into the global economy. Box 1-1 describes the objectives, scope and status of the CAFTA agreements. Several studies have tried to assess the potential impact of CAFTA on Central American economies. Some key findings are based on lessons from the impact of trade agreements in other countries -- e.g., the experience of Mexico with NAFTA. A recent study on NAFTA10 found mixed results. On the one hand, trade increase in Mexico has been impressive: exports have grown from US$50 million in 1994 to US$134 million in 2002. In addition, exports diversified and technology transfer and FDI increased. On the other hand, economic growth and employment have not improved significantly. Furthermore, the least advanced states in the South of the country have not benefited from the increased openness of the economy. In a paper on CAFTA, Jaramillo shows that mixed results are also likely to occur in Central American economies.11 10G. Perry, D. Lederman, W. Maloney, and L. Serven Lessons from NAFTA for Latin America and the Caribbean (World Bank, 2003). 11C. F. Jaramillo. Que Sabemos de los efectos del CAFTA? (World Bank, 2004). 17 Box 1-1: The CAFTA Agreement Free trade agreement negotiations between the United States and five Central American countries (Guatemala, El Salvador, Costa Rica, Honduras and Nicaragua) began formally in January 2003 and were completed eleven months later, on December 17, 2003. Nine rounds of negotiations were completed over the course of the year, culminating in the signing of the agreement by four of the five nations -- Guatemala, El Salvador, Honduras and Nicaragua. Costa Rica signed onto the agreement a month later, on January 26, 2004. The Central American Free Trade Agreement (CAFTA) is a second-generation trade agreement, similar to NAFTA and the U.S.-Chile FTA. In other words, it does not cover free trade of goods only, but stipulates free trade for all sectors, including services. It opens market access by locking in low tariffs with clearly defined rules of origin. Relevant market sectors include agriculture, manufacturing and services. In addition to free trade of goods and services, the agreement contains provisions regarding non- market access issues, such as labor and environmental regulations, intellectual property rights, and dispute settlement. In view of this agenda, negotiations proceeded in five working groups: a) market access (including agriculture); b) investment and services; c) government procurement and intellectual property; d) labor and the environment; and e) dispute settlement and other institutional issues. The adoption of CAFTA generates significant benefits for Central American countries, such as making access to the U.S. market more permanent, creating an irrevocable anchor to free trade.12 CAFTA also grants exporters and investors a fuller set of protection and dispute resolution rules. Before CAFTA can be implemented, it faces a final daunting obstacle -- ratification by the U.S. congress. In fact, it seems unlikely in an election year swirling with protectionist rhetoric that the proposal of ratification will even be voted on by the legislature. Source: World Bank -- CAFTA study (ongoing) and Honduras Trade Facilitation and Productivity Enhancement Project 1.8 Existing studies show that trade openness can translate into higher foreign trade, sustainable productivity, and growth only if complemented by policies aimed at improving the country's overall investment climate. In the NAFTA study, Perry et al. show that to take full advantage of the trade agreement and ensure that it translated into higher economic growth, Mexico should have made specific efforts to improve the quality of its overall business environment. Cross country studies confirm this evidence. In a study conducted in 2003, Bolaky and Freund demonstrate that the effect of increased trade on growth is reduced by more than half in countries with a poor business environment. The analysis shows that countries' poor investment climate restricts growth because, even after liberalization, resources cannot move to the most productive sectors and the most efficient firms. 1.9 This report considers four key areas of the investment climate. The investment climate is usually defined as the host of institutional factors and policies that influence private sector development. The World Bank is currently preparing Investment Climate Assessments for various countries worldwide. The standard methodology is summarized in Box 1-2. This report groups Honduras' investment climate factors and policies into four main areas:13 a. Firm-government relationships -- with specific reference to red tape, corruption, procurement, the judiciary system and crime; b. The quality and quantity of the physical and financial infrastructure, such as energy, transport and telecommunications; 12Under the current trade regime -- the Caribbean Basin Initiative (CBI) -- these nations enjoy nearly equal access to U.S. markets, but due to the unilateral nature of the agreement, long term investors cannot be confident that preferential trade status will continue. 13 While a sound macroeconomic framework is a necessary precondition for improving a country's investment climate, macroeconomic issues are not covered in this report. Refer to paragraph 1.2 and footnote 1 for the reasons. 18 c. The economy's ability to move up the value chain by introducing innovations, adopting new technologies, improving quality of processes and products, and improving the skills of its labor force; and d. Access to financial services. 1.10 The impact of specific investment climate factors on enhancing the effects of trade agreements and improving economic growth is being evaluated by a growing number of studies. Several studies have shown that improving access to infrastructure -- especially ports, airports and customs -- has a direct impact on increasing the returns on trade agreements.14 Access to finance has also been proven to enhance firms' opportunities to take advantage of economic openness. Research also demonstrates that broader efforts to improve governance -- including reducing excessive regulations, corruption and crime -- are essential to fully reap the benefits of trade treaties.15 Finally, a growing body of evidence confirms that active Government policies aimed at increasing innovation, technology transfer and labor skills facilitate development of backward and forward linkages between local and foreign firms and between larger and smaller companies, resulting in higher growth rates.16 Box 1-2: What is an Investment Climate Assessment? Investment climate assessments systematically analyze the conditions for private investment and enterprise growth in a country, drawing on the experience of local firms to pinpoint the areas where reform is most needed to improve the private sector's productivity and competitiveness. By providing a practical foundation for policy recommendations and involving local partners throughout the process, the assessments are designed to give greater impetus to policy reforms that can speed the private sector's growth. Produced by the World Bank Group in close partnership with a public or private institution in each country, the investment climate assessments are based on a survey of private enterprises designed to find out what difficulties they encounter in starting and running a business. The survey captures firms' experience in a range of areas -- financing, governance, regulation, tax policy, labor relations, infrastructure services, technology, and training, among others. All these are areas where difficulties can add substantially to the costs of doing business. The survey attempts to quantify firms' costs related to the investment climate bottlenecks. Using a standard methodology, the assessment then compares the survey findings with those in similar countries to evaluate how the country's private sector is competing. The findings of the survey, combined with relevant information from other sources, provide a practical basis for identifying the most important areas for reform aimed at improving the investment climate. The findings and policy recommendations emerging from the assessments are discussed extensively with the private sector and other stakeholders in the country. This broad dissemination of the findings is aimed at engaging not only policymakers but also business leaders, investors, nongovernmental organizations, and the donor community in shaping the national private sector development strategy, forging consensus on the priorities for reform of the investment climate, and laying the groundwork for concrete responses to the problems identified. Source: Batra, G. Guatemala Investment Climate Assessment 14See, among others, J. Wilson, C. Mann, and T. Otsuki (2004). For evidence of the specific links between these factors and economic growth, see D. Dollar, M. Hallward-Driemeier. and T. Mengistae (February 2003) 15D. Dollar, M. Hallward-Driemeier and T. Mengistae (February 2003); J. Wilson, C. Mann, and T. Otsuki (2004). 16Closing the Gap in Technology and Education in Latin America (World Bank, 2003). G. Perry, D. Lederman, W. Maloney, and L. Serven. Lessons from NAFTA for Latin America and the Caribbean (World Bank, 2003). 19 THE HONDURAS INVESTMENT CLIMATE SURVEY: SUMMARY FINDINGS AND OVERALL IMPLICATIONS FOR THE PUBLIC AND PRIVATE SECTORS 1.11 To assess the level of development of each Investment Climate Variable in Honduras -- and the extent to which each constrains firms' operation -- this study presents the results of a survey of 450 Honduran manufacturing companies. While a number of studies have assessed the links between investment climate variables and economic growth, most available investment climate indicators don't help Governments identify what specifically needs to be done to help firms increase their productivity and growth. For this reason, this study goes down to the level of the firms and asks them what key factors are constraining their development. 1.12 Definition of the sample for the Honduras Investment Climate Assessment. The Honduras survey was carried out in 2003. The survey was designed by the World Bank and a private firm, who administered it under World Bank supervision. The survey covers a stratified sample of 450 manufacturing firms17 operating in 11 industries,18 across 14 departments.19 The firm population distribution and sample were drawn from two lists of firms provided by the Central Bank of Honduras, for maquila and non-maquila firms. Sample size was determined at the sector level. In each sector, the sample was stratified by employment size (using the Neyman allocation) and then proportionally allocated to provinces. For employment size, three ranges were used: 5 to 9 employees, 10 to 49 employees, and 50 or more employees. This exercise was performed after restricting the sample frame to the main manufacturing sectors in Honduras, responsible for 82 percent of total manufacturing employment in the country. The departments that were covered in the sample are responsible for more than 99 percent of employment in those sectors. In order to deal with the problem of non-response -- caused either by the inability to find firms that changed address or exited the market, or by the unwillingness of some firms to participate in the survey -- samples of substitute firms were drawn from both the original sample frame and from frames obtained from various chambers of commerce and industry. The substitute firms were drawn, to the extent possible, from the same strata as the firms in the main sample that failed to respond to the survey. Additional information on the sample and its composition is provided in Annex A. SUMMARY FINDINGS OF THE SURVEY AND OVERALL IMPLICATIONS FOR THE PUBLIC AND THE PRIVATE SECTORS 1.13 Summary findings of the Honduras Investment Climate Survey show that the four pillars of the investment climate have a clear impact on the performance of Honduran firms -- firms rate them as critical constraints on growth and point out that they significantly increase their cost of doing business. Figure 1-8 and Figure 1-9 summarize key results of the ICS. More detailed information 17Hotels were also included, but their responses are not included in the survey because the data were received too late. 18These are grouped in eight main categories: food and tobacco (24 percent of the sample); apparel; furniture and wood (24 percent of the sample); non metallic minerals (cements and ceramics) (11 percent of the sample), chemical and rubber (7 percent of the sample), metal products (7 percent of the sample), beverages (5 percent of the sample) and textiles (4 percent of the sample) -- see Annex 1. 19 Departments in the Center South Region: Choluteca, Comayagua, El Paraiso, Fco. Morazan, Valle and Yoro. Departments in the North Coast Region: Atlantida, Colon, Cortes and Islas de la Bahia. Departments in the West Region: Santa Barbara and La Paz. Note that departments have been aggregated differently in different chapters according to the aim of the analysis. 20 about the results of the survey may be found in Table 1-5: Investment Climate at a Glance on page 14 at the end of this Chapter and ANNEX B on page 109 near the end of the Report. Figure 1-8 below lists the factors that at least 40 percent of surveyed companies rate as key constraints on growth. While the chart represents firms' subjective rankings -- and thus should be taken cautiously -- it does indicate the need to improve governance by reducing corruption and crime, and to increase access to credit. The rankings also highlight the importance of two factors that are outside the scope of this study -- macroeconomic stability and economic and regulatory uncertainty. Figure 1-9 shows an objective ranking of Honduran firms' losses due to shortcomings of the business environment. This analysis relates only to quantifiable factors in the survey and should not be taken as the only (or most important) way to identify weaknesses in Honduras' investment climate. Nevertheless, the chart is quite revealing. Total losses due to quantifiable shortcomings of the business environment are equivalent to 12.9 percent of firms' sales, a considerable amount. The costs of poor governance (losses due to corruption, crime and red tape) consume 7.6 percent of total sales, and losses associated with infrastructure bottlenecks (due to inefficiencies in transport, energy and telephone communication) eat up 5.3 percent of total sales. Figure 1-8: The perspective of Honduran firms -- main constraints on growth mentioned by more than forty percent of surveyed firms Corruption 63 Crime and violence 61 Cost of financing 60 Macroeconomic instability 52 Access to financing 51 Availability of financing 49 Economic and regulatory uncertainty 47 0 10 20 30 40 50 60 70 Governance factors Finance factors Source: Honduras ICS Figure 1-9: Percentage of sales lost due to quantifiable constraints Water interruptions 0.2 Phone interruptions 0.3 s Time spent with regulators 0.6 Total losses osL Crime 0.9 as percentage of e of sales: usaC Transport losses 1.8 12.9 Payment to get things done 2.5 Electricity interruptions 3.0 Security costs 3.6 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Average Percentage of Sales Loss for all Firms Losses related to infrastructure Losses related to governance Source: Honduras ICS 1.14 Systematic econometric analysis conducted as part of this report for Honduras, Nicaragua and Guatemala and utilizing the ICS data for all three countries also points to all four IC areas as being significant for firm productivity. For the purposes of the present report, total factor productivity (TFP), or more generally productivity, is considered to be that part of the production of goods (sales) that 21 is not explained by the main production inputs (labor, intermediate materials and capital).20 Inference on the impact of key IC variables related to the four investment climate areas -- governance, infrastructure, innovation and technology, and finance -- was done by the use of panel data regression techniques whose results are displayed in Table 1-3. Box 1-3 (on page 25) and Box 1-4 (on page 29) in this Chapter and ANNEX C (starting on page 119) near the end of the Report provide more information on the methodology used to conduct the econometric analysis. The results of the analysis indicate that most IC variables belonging to each of these areas are highly statistically significant and that the direction of the impact, for all the productivity measures used, is as expected. Governance, corruption and crime-related variables have in general a negative impact upon productivity. Thus, the regulatory burden -- as measured by the number of days spent on government inspections and regulation-related activities -- has a sizeable impact on productivity. An extra day spent in such activities reduces productivity between 5.8 percent and 10.7 percent. Similarly, a lack of security -- measured by the prevalence of crime -- affects productivity negatively and significantly. For every new criminal attempt suffered productivity is reduced between 1.8 percent and 3.2 percent. Bureaucratic procedures should be made more transparent and efficient, so that there are is no room for productivity gains (between 1.3 percent and 3.3 percent) by those firms which make payments to deal with bureaucracy faster. Infrastructure variables -- such as power outages, losses during transportation, access to the Internet and number of days needed to clear customs -- are also relevant for productivity. Among these variables, internet access has a particularly strong positive effect -- having Internet access raises productivity between 11 percent and 15 percent. The rest of the infrastructure variables are indicative of a poor investment climate and therefore create significant losses in productivity. There is clear evidence that labor training and innovation -- as measured by external training provided for firm employees and the share of staff engaged in R&D -- increase productivity and -- in the case of labor training -- with a strong positive impact. The effects of quality certification and other forms of innovation -- such as the share of computer-controlled equipment -- as well as of education of the labor force are positive but less pronounced21. Among the quality, innovation and labor training variables, we find that external training of their workers raises firms' productivity between 8.9 percent and 11.7 percent. A one-percent increase in the fraction of firm staff engaged in research and development (R&D) activities increases productivity between 0.6 percent and 0.7 percent. Investing in ISO quality certification, and increasing computer-operated plant equipment also raise productivity but their significance varies with the definitions of productivity (see Box 1-3). Finally, financial and corporate governance variables also exhibit positive large and statistically significant effects on productivity. In fact, the highest positive effects on productivity are estimated to stem from having externally audited financial statements, which raise productivity between 11.6 percent and 17 percent. Therefore, the productivity analysis results are very much in line with the perception and costs data -- exhibited in Figure 1-8 and Figure 1-9 -- on governance, finance and infrastructure. 1.15 Further econometric analysis performed on the individual country data allows us to see the impact of the average country-wide investment-climate variables on country-wide average productivity. As part of the productivity analysis, individual country-by-country evaluation was also done to assess the impact of investment climate (IC) variables on average productivity in Honduras. The results for Honduras are presented in Table 1-4. The net effect of the four IC variables is to decrease average (log) productivity in Honduras between 4.3 percent and 11.3 percent, depending on the productivity measure used. In absolute terms, regardless of the sign of the effects, IC variables related to Red Tape, Corruption and Crime affect average Honduran (log) productivity from 5.3 percent to 14.8 percent. The net effect of the four infrastructure variables on average (log) productivity ranges from ­2.8 20 We employ several productivity measures depending on the functional form of the production function considered, the level of aggregation and the estimation methodology. Box 1 and the Annex provide detailed information about this econometric approach and the empirical results obtained. 21Those variables were significant, and with similar elasticity values, if estimated by random effects (RE). 22 percent to ­5.9 percent. The aggregate effect of the infrastructure variables -- regardless of the signs -- is of the order of 4.9 percent to 9.3 percent. The IC variables related to Quality, Innovation and Labor Skills displayed in Table 1-2 exert only positive effects on average (log) productivity. Their cumulative effect on average (log) productivity in Honduras ranges from 1.5 percent to 2.5 percent. The finance and corporate-governance variables display increases in average (log) productivity gains of similar magnitude -- from 1.2 percent to 2.7 percent. Finally, the additional control variables (firm age and share of imported firm inputs) produce additional average (log) productivity gains, from 1.7 percent to 3.8 percent. 23 Table 1-3: Elasticities and Semi-elasticities of Investment Climate (IC) Variables on Productivity, after Controlling for Other Firm, Industry and Country Characteristics Restricted OLS Estimation Unrestricted OLS Estimation Solow Res. 1 step Solow Res. 1 step Cobb-Douglas Translog Cobb-Douglas Translog Red Tape, Corruption and Crime Number of days spent in inspection-and -0.097** -0.099** -0.058 -0.101** -0.107** -0.068* Regulation-related work [0.043] [0.043] [0.041] [0.043] [0.043] [0.041] Fraction of sales undeclared to the tax authority -0.601** -0.612** -0.416* -0.676** -0.767*** -0.593** for tax purposes [0.263] [0.260] [0.243] [0.267] [0.255] [0.250] Payments to deal with bureaucracy 0.031** 0.033*** 0.015 0.031** 0.030** 0.013 "faster", percent of sales [0.012] [0.012] [0.012] [0.012] [0.012] [0.012] Number of criminal attempts suffered -0.029** -0.031*** -0.018 -0.032*** -0.029** -0.018 [0.012] [0.012] [0.011] [0.012] [0.011] [0.012] Infrastructure Average duration of power outages (log) -0.095* -0.085* -0.072 -0.088* -0.075 -0.024 [0.052] [0.050] [0.047] [0.052] [0.050] [0.049] Days to clear customs for imports (log) -0.097** -0.106** -0.125*** -0.105*** -0.105** -0.119*** [0.041] [0.042] [0.039] [0.041] [0.043] [0.041] Shipment losses (fraction of sales lost) -1.860** -2.119** -1.229 -1.948** -2.530*** -2.063** [0.850] [0.867] [0.811] [0.852] [0.871] [0.825] Dummy for Internet access 0.147*** 0.144*** 0.119*** 0.139*** 0.128*** 0.111*** [0.038] [0.041] [0.039] [0.038] [0.039] [0.037] Quality, Innovation and Labor Skills Fraction of computer-controlled 0.119 0.13 0.132* 0.117 0.131 0.084 Machinery [0.082] [0.082] [0.079] [0.083] [0.085] [0.086] Fraction of total staff engaged in R&D 0.594** 0.667** 0.581** 0.589** 0.607** 0.580** [0.287] [0.294] [0.270] [0.285] [0.280] [0.268] Dummy for ISO quality certification 0.154 0.167 0.024 0.142 0.176* 0.105 [0.102 [0.103] [0.093] [0.101] [0.105] [0.103] Fraction of total staff with 0.036 0.048 0.03 0.033 0.054 0.056 Secondary education or higher [0.058] [0.059] [0.059] [0.057] [0.060] [0.062] Dummy for training provided beyond 0.117*** 0.105*** 0.089*** 0.116*** 0.110*** 0.098*** "on-the-job" training [0.036] [0.036] [0.033] [0.036] [0.036] [0.034] Finance and Corporate Governance Dummy for incorporated firm 0.150*** 0.140*** 0.115*** 0.146*** 0.132*** 0.117*** (Sociedad Anonima) [0.039] [0.042] [0.038] [0.039] [0.040] [0.038] Dummy for external audit of financial 0.168*** 0.159*** 0.121*** 0.170*** 0.150*** 0.116*** Statements [0.042] [0.042] [0.038] [0.042] [0.042] [0.038] Other Control Variables Age of the firm (log) 0.050** 0.049** 0.046** 0.051*** 0.043** 0.032* [0.020] [0.020] [0.018] [0.20] [0.019] [0.018] Share of imported inputs (fraction) 0.113** 0.101** 0.096** 0.110** 0.074 0.101** [0.049] [0.051] [0.047] [0.049 [0.051] [0.048] R-Squared 0.20 0.20 0.36 0.26 0.51 0.94 Notes: (1) Robust standard errors shown in parentheses under coefficient estimates. (2) Significance is given by robust standard errors. * Significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. (3) The regressions include a constant, industry dummies, country dummies and year dummies. 24 Box 1-3: Productivity Analysis: some background information to help interpret the results of Table 1-3 In order to evaluate the economic impact of investment climate (IC) variables on productivity, data were collected through the Investment Climate Surveys (ICS) of Honduras, Nicaragua and Guatemala. In the econometric analysis, all the observations from the firms of these three countries were pooled, generating a combined sample of 1,300 observations.22 This pooling is justified on grounds that if we were to conduct productivity analysis for each country separately, we would not be able to make reliable inferences for industries due to the insufficient number of observations.23 Table 1-3 presents the results of the pool regression estimated by ordinary least squares (OLS) to get the impact of investment-climate (IC) variables on six different measures of productivity (see Box 1-4 for more on the six productivity measures). Each row of the table shows the estimated regression coefficients of each explanatory IC variables, obtained by running each (log) productivity measure on all the explanatory IC variables listed in the table, as well as controlling for industry, country and year dummies. In addition, the estimated regressions also control for the age of the firm and the share of its imported inputs. The table shows elasticity and semi-elasticity estimates for both restricted and unrestricted OLS regressions, depending on whether the input coefficients of the production function are allowed to vary by industry (unrestricted), or assumed the same for all industries (restricted).24 For purposes of reading the table, the IC variables are divided into several categories related to: Red Tape, Corruption and Crime; Infrastructure; Quality, Innovation and Labor Skills; and Finance. 22The sample covers 461 firms in Guatemala, 370 -- in Honduras, and 469 -- in Nicaragua. 23For example, in such a country-by-country analysis we would end up with an industry having only 5 observations in Honduras. However, pooling the three countries together gives at least 33 observations in the industry with the minimum number of observations, thus allowing for more reliable statistical results. 24For more on the restricted and unrestricted estimation and the functional forms of the production function see Box 1. 25 Table 1-4: Average (log) Productivity Gains25 and Losses due to Investment Climate Variables Restricted OLS Estimation Unrestricted OLS Estimation Solow Res. 1 step Solow Res. 1 step Cobb-Douglas Translog Cobb-Douglas Translog Red Tape, Corruption and Crime Number of days spent in inspection-and -6.34** -6.03** -2.15 -6.66** -6.74** -2.9* Regulation-related work Fraction of sales undeclared to the tax authority -4.48** -4.28** -2.37* -5.13** -5.51*** -2.91** for tax purposes Payments to deal with bureaucracy 1.84** 1.81*** 0.51 1.85** 1.72** 0.50 "faster", percent of sales Number of criminal attempts suffered -0.82** -0.83*** -0.25 -0.91*** -0.78** -0.34 Cumulative contribution -9.80 -9.33 -4.26 -10.85 -11.31 -5.65 Cumulative absolute contribution 13.48 12.95 5.28 14.55 14.75 6.65 Infrastructure Average duration of power outages (log) -2.89* -2.45* -1.19 -2.73* -2.21 -0.47 Days to clear customs for imports (log) -3.73** -3.8** -3.09** -4.1*** -3.89** -2.99*** Shipment losses (fraction of sales lost) -0.74** -0.79** -0.41 -0.79** -0.97*** -0.54** Dummy for Internet access 1.74*** 1.6*** 1.85*** 1.7*** 1.45*** 0.87*** Cumulative contribution -5.62 -5.44 -2.84 -5.92 -5.62 -3.13 Cumulative absolute contribution 9.10 8.64 6.54 9.32 8.52 4.87 Quality, Innovation and Labor Skills Fraction of computer-controlled 0.20 0.21 0.22* 0.20 0.22 0.10 Machinery Fraction of total staff engaged in R&D 0.40** 0.42** 0.40** 0.40** 0.39** 0.25** Dummy for ISO quality certification 0.18 0.18 0.05 0.18 0.2* 0.08 Fraction of total staff with 0.29 0.37 0.12 0.27 0.42 0.30 Secondary education or higher Dummy for training provided beyond 1.34*** 1.13*** 1.16*** 1.35*** 1.22*** 0.73*** "on-the-job" training Cumulative contribution 2.41 2.31 1.95 2.40 2.45 1.46 Finance and Corporate Governance Dummy for Incorporated firm (Sociedad Anonima) 0.91*** 0.79*** 1*** 0.9*** 0.77*** 0.47*** Dummy for external audit of financial 1.71*** 1.52*** 1.49*** 1.76*** 1.47*** 0.77*** statements Cumulative contribution 2.62 2.31 2.49 2.66 2.24 1.24 Other Control Variables Age of the firm (log) 2.95** 2.69** 2.37** 3.06*** 2.42** 1.23* Share of imported inputs (fraction) 0.78** 0.66** 0.92** 0.77** 0.49** 0.46** Cumulative contribution 3.73 3.35 3.29 3.83 2.91 1.69 Grand Total Contribution -6.66 -6.80 0.63 -7.88 -9.33 -4.39 Grand Total Absolute Contribution 31.34 29.56 19.55 32.76 30.87 15.91 Note: (1) The asterisks correspond to the significance level of the variables in the corresponding OLS regression: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. 25Those percentage productivity gains and losses are measured as the percentage average productivity contribution of each IC variable of Honduras, relative to average (log) productivity of the firms in Honduras. 26 A policy agenda to improve the investment climate requires broad-based action and close collaboration between the public and private sectors, as well as among different Government agencies. Summary indicators related to all four investment climate variables -- and the relative performance of Honduras as compared to other countries26 in these areas -- are included in Table 1-5: Investment Climate at a Glance below. Improving the investment climate in Honduras will require cross sector strategies and actions geared towards aligning economic policy, public management, infrastructure, technology, FDI, innovation, training and finance policies. This can be done only by establishing close collaboration between the private and the public sectors and among various Ministries in the public sector to push forward coordinated initiatives that can unleash the private sector's potential for growth. Although the Honduran Government has already made significant progress in this respect -- the Competitive Commission is an excellent example of effective collaboration between the private and the public sectors -- more effort in this direction will be needed in the future. Ensuring that the Competitive Commission has sufficient capacity to implement reform initiatives agreed upon by its members and creating Technical Committees that can take responsibility for specific reform initiatives are essential. The recent creation of a Technical Committee overseeing Competition Policy initiatives and the plans for a Committee for Administrative Simplification prove that the Government is moving in the right direction -- and that improving the Honduran investment climate is feasible. 26See paragraph 4 in the Introduction on criteria used to select comparator countries for this study. 27 Table 1-5: Investment Climate at a Glance Topic Honduras Nicaragua Guatemala Brazil Ecuador 1995 2001 1995 2001 1995 2001 1995 2001 1995 2001 Macro Indicators GNI per capita, PPP (current international $)a 2,260 2,450 1,780 1,880 3,430 3,880 6,410 7,180 na na Population (Millions)a 5.6 6.6 4.4 5.2 10.0 11.7 159.5 172.4 11.5 12.9 Exports and Import of goods (% of GDP) 72.3 66.3 82.7 99.9 37.2 39.4 14.2 23.2 47.2 54.5 GDP growth (annual %) (1990-1995 vs 1996-2001)a 3.6 3.0 1.5 4.6 4.3 3.6 3.2 2.1 2.7 1.7 Gross capital formation (% of GDP) a 31.6 32.4 24.9 33.8 15.1 15.2 22.3 21.4 21.6 25.7 Foreign direct investment, (% of GDP) a 1.3 3.0 4.1 8.9 0.5 2.2 0.7 4.4 0.2 6.3 Average Years of Education b 4.1 4.4 3.1 4.6 6.5 Governance Days to register a business c 100.0 32.0 50.0 na na Number of Days spent dealing with Inspections c 12.1 21.3 9.9 13.7 15.5 Senior Management Time Spent dealing with 8.5 7.8 10.3 7.8 15.6 Regulators(%) c Payments necessary "To get things done" (% Sales) c 2.5 1.7 3.7 na 2.6 Informal Economy (% GNI) a 49.6 45.2 51.5 39.8 34.4 Bribes on Government Contracts (% of Contract) c 4.6 4.3 3.7 12.2 9.0 Percentage of firms who don't trust the judicial systemc 38.6 41.0 47.7 22.5 56.4 Share of firms experiencing a crime c 30.0 27.0 5.2 23.1 37.0 Losses due to crime (% of Sales) 3.2 3.3 3.8 0.6 1.3 Security Cost (as a percentage of total costs) c 3.6 2.9 5.5 1.9 5.2 Infrastructure Indicators Share of Firms that experienced a power cut c 92.1 88.7 74.1 64.1 71.1 Sales lost due to power outages (percent) c 3.0 3.9 2.3 1.3 2.5 Days to obtain an electricity connectionc 32.5 31.4 62.3 25.6 41.2 Telephone mainlines (per 1,000 people) a 27.0 46.1 22.1 31.2 28.6 64.7 85.1 182.1 60.9 103.7 Days to obtain a telephone connection c 175.0 118.4 47.7 18.2 129.7 Losses due to telephone interruption (% sales) c 0.3 0.7 0.2 0.3 1.1 Personal computers (per 1,000 people) a 10.8 8.9 3.0 12.8 17.3 50.1 13.1 23.3 Share of Firms that use email for business purposes c 50.2 38.5 66.4 92.0 83.2 Share of Firms that use a web site for business purposes c 21.8 16.6 29.2 73.1 55.4 Roads, paved (% of total roads) a 20.4 20.4 10.0 11.0 26.0 34.5 8.9 5.5 12.7 18.9 Losses while in transport c 1.6 1.6 1.7 1.0 1.5 Losses due to transport interruptions c 0.1 0.6 0.4 0.5 0.5 Days to Clear Customs - Importsc 5.1 5.2 9.4 13.8 17.1 Days to Clear Customs - Exports c 1.9 2.0 2.3 8.4 7.1 Innovation Share of Firms with ISO Certification c 5.4 3.3 3.3 14.1 16.8 Share of Firms with Other Quality Certification c 17.0 20.3 11.9 9.1 21.0 Share of Firms Introducing Technological Innovations (Last two years)c 67.3 67.9 80.9 na na Share of Firms with a Foreign License c 15.2 9.1 19.6 7.5 23.4 Share of Firms with Computerized equipment c 21.4 11.1 23.5 na na Share of Firms providing formal trainingc 49.3 37.4 55.0 67.1 68.9 Share of Firms using private training providersc 43.5 37.2 42.2 74.5 na Finance Interest Rate in USD for Large Firms c 11.0 13.3 10.2 na na Share of Firms with a Loan c 51.2 43.7 43.5 34.7 50.3 Share of Firms with Audited Financial Statements c 43.2 30.2 34.2 19.1 47.1 Percentage of Revenues Reported for tax purposes c 68.0 66.4 77.0 67.4 79.8 aWDI -- b "Closing the Gap in Education & Technology" World Bank 2002 -- c World Bank Investment Climate Dataset Source: Investment Climate Surveys. 28 Box 1-4: Productivity Analysis: A Summary of the Methodology This Box is based on the analysis of Escribano and Guasch (2004) "Assessing the Impact of the Investment Climate on Productivity using Firm-Level Data: Methodology and the Case of Guatemala, Honduras and Nicaragua". ANNEX C on page 119 below includes a detailed explanation of the methodology. The present box provides a summary of the main methodological issued considered and the key steps and procedures undertaken. Escribano and Guasch (2004) argue that TFP, or productivity in general, includes the effect of any variable affecting the production process which is different from the main inputs of production -- labor (L), capital (K) and intermediate inputs (M). In particular, TFP is assumed to include the effects of IC variables such as infrastructure, governance, quality of labor, innovation, research and development (R&D), technological and regulatory changes, competition and any other firm- specific characteristics or control variables (C). Prior to measuring the impact of each of these IC variables on productivity, we have to face a challenging identification problem due to common causes simultaneously affecting both the inputs of production (K, L and M) and TFP, i.e. the usually unobserved IC and control variables mentioned above. Therefore, all of the inputs of production (K, L and M) are endogenous and will be correlated with productivity (TFP), which invalidates least squares estimation techniques run in two steps, where one first estimates TFP and then runs a regression of TFP on IC variables. This endogeneity problem of the three inputs also invalidates recent alternative estimation techniques based on Olley and Pakes (1996) and Levinsohn and Petrin (2003) procedures, when applied to annual observations of IC variables. To overcome this endogeneity problem of the inputs, Escribano and Guasch (2004) propose a structural estimation approach which allows for several sources of simultaneity associated with the IC variables. They also discuss the conditions required to estimate the rate of growth of TFP, based on the rates of growth of the production function variables, and the conditions required to estimate TFP based on its levels (logs). The conclusion is that due to the characteristics of this dataset -- based on the 2003 Investment Climate Surveys (ICS) of Guatemala, Honduras and Nicaragua -- it is more appropriate to use the level (logs) of TFP and the levels (logs) of the other variables which enter the production function, rather than their corresponding rates of growth. There is no universally accepted measure of productivity or total factor productivity (TFP). The analysis followed in this report seeks to measure the impact of different investment climate (IC) variables on (log) productivity, controlling for other (firm, industry, country, etc.) characteristics. Inevitably, any empirical evaluation of the impact of the IC variables will be contingent upon the productivity measure used. This methodology aims to obtain robust and consistent conclusions from all productivity measures, providing a consistent range of productivity gains and losses due to IC variables. The empirical analysis of this report relies on twelve different productivity measures but we only report in Table 1-3 and Table 1-4 the results based on six productivity measures, since the results based on the other six are very similar (see ANNEX C). 29 Escribano and Guasch (2004) make clear that the 2-step approach should only be used when we either have good instrumental variables (IV) for the inputs (L, M and K), or when we obtain the input-output elasticities by "growth accounting techniques" (using the cost shares of the inputs) to get the Solow residuals (TFP), Solow (1957). The Solow approach is one of the procedures that we apply in this report. In the second general approach (1-step approach) that we apply we estimate the parameters of the inputs (L, M and K) in an extended production function where we simultaneously estimate the impact of the IC variables while controlling for other firm, industry and country characteristics. This 1-step estimation approach is applied to a Cobb-Douglas and a Translog production function under different aggregation conditions. We consider a restricted estimation case, where we assumed that the input coefficients are the same for all industries; and an unrestricted estimation case, where we allow the input coefficients to vary across different industries. Under the restricted estimations, the average cost shares are calculated as the average share of each input (K, L and M) across the whole sample of manufacturing firms over the years 2001 and 2002; under the unrestricted estimation the cost shares are calculated as the average industry cost share of each input across all firms of the three countries belonging to the same industry over the last two years (i.e. a separate cost share of each of the three factor inputs (K, L and M) is calculated for each of the nine industries covered by the ICS). To reduce the least squares simultaneity bias associated with the fact that the IC variables may be correlated with the error of the extended production function, we use averages of the ICit variables taken at the region-industry cells ( IC jt ) for each country. The coefficients on the investment climate variables ( ICit ) and the firm-specific characteristics (Cit ) are usually estimated at the same level of aggregation, i.e. one coefficient per ICit and per Cit for the whole pool sample while, as we said before, the input coefficients (input-output elasticities) of the production functions are estimated at two different levels of aggregation. In Table 1-3, we presented the results of six (out of the twelve) different measures of productivity obtained: two measures based on Solow residuals, using ordinary least squares (OLS) in the second step, and four productivity measures based on OLS estimation in a 1-step estimation. Furthermore, we also applied random effects (RE) estimation, instead of OLS, getting another set of six productivity results, (see ANNEX C). All the different regression specifications include industry dummies, country dummies and a year dummy to control for differences in firm productivity due to industry- specific, country-specific and time-specific effects. For a more detailed technical discussion of the main estimation techniques, see ANNEX C. 30 CHAPTER 2. RED TAPE, CORRUPTION AND CRIME27 2.1 Governments can facilitate -- or obstruct -- doing business. Sound institutional regimes -- characterized by high political stability, respect for the rule of law, and a low level of corruption -- and the efficient provision of public services by the government are essential for private sector development and growth. While the ultimate responsibility for starting businesses and increasing their productivity lies with firms themselves, governments have important roles to play. First, they are responsible for maintaining a country's overall macroeconomic stability. Although analyzing the macroeconomic environment in Honduras is beyond the scope of this study, macroeconomic stability is necessary for businesses to function effectively. Second, governments can simplify business processes by reducing excessively cumbersome interactions between public agencies and private firms, together with making the exchanges that are necessary as efficient as possible. When starting a firm, entrepreneurs are likely to be defeated by long and expensive start up procedures and may decide to keep their firm informal. Furthermore, when firms first start operating, interactions with the government can be disruptive and costly. As a result, firms may resort to bribing the public actors with whom they interact, especially if it seems to be the best -- or even the only -- way to get things done. Efficient governments, by contrast, provide firms with useful and timely services, increasing their potential to become more competitive in local and international markets. Governments can also be good business partners for firms, facilitating procurement of goods and services in a competitive manner. Efficient and effective judicial systems can enforce firms' rights when settling a dispute. Finally, governments have a responsibility to protect firms, ensuring that they operate in a safe environment, free of crime and violence. 2.2 Extensive research shows that better governance translates into higher levels of investment and economic growth and reduces informality. Honduran firms consider business- government relationships to be a serious constraint on growth. When asked to rank various investment climate bottlenecks, sixty three and sixty one percent of surveyed firms respectively indicate that corruption and crime are major impediments for growth -- more than for all other investment climate factors (see Figure 1-8 on page 21 above). Several cross country studies show that a country that significantly improves governance dimensions such as the rule of law, corruption, the regulatory regime, and voice and democratic accountability can expect, in the long run, a dramatic increase in its per capita income and other social indicators. In fact, Kaufmann (2004) shows that an improvement in governance by only one standard deviation can ultimately result in up to a fourfold increase in per-capita income.28 Burdensome regulations and increased levels of corruption are also associated with higher levels of informality, which in turn contribute to reduced investment and growth. 2.3 This section presents the results of the Honduras investment climate survey with regard to the relationship between businesses and the Government. This study focuses on aspects of governance that were thoroughly covered in the ICS and emerged as most critical in 27 This Chapter was prepared by Marialisa Motta with inputs from Stefka Slavova ­ summary of the econometric analysis and Judicial System. Valuable inputs and advice were also provided by Luke Haggarty, Bernice K. Van Bronkhorst, Francesca Recanatini and Rajeev Swami. Miguel Crivelli and Marc Forni provided excellent research assistance. 28 D. Kaufmann, The Global Competitiveness Report, (World Economic Forum, 2004) page 154. For additional references on topics related to governance and growth, see Knack and Keefer (1997), Mauro (1995), Easterly and Levine (2002), Rodrik, Subramanian and Trebbi (2002), Sachs and Warner (1997) and Diamond (1999). 31 the results of the survey.29 It is organized in four parts: (1) red tape and corruption in the process of business registration and operation; (2) business-government interactions in procurement processes -- doing business with the government; (3) the effectiveness and efficiency of the judiciary system; and (4) crime and violence. While this section does not consider all variables related to governance, it aims to provide an overview of the elements of governance that most directly influence business operation in Honduras. 2.4 Sources of information for policy recommendations. Besides the CAS and the PRSC, key reference documents for the policy recommendations provided in this chapter include: background papers on Business Registration and Operations prepared for this Study; the Honduras Country Financial Accountability Assessment (2003), background documents on Civil Service Reform prepared as inputs for the PRSC (2004), the Doing Business Report (2004), and the IDB Peace and Citizen Coexistence Project for the Municipalities of the Sula Valley (2004). REGISTERING AND OPERATING A BUSINESS 2.5 Registering a firm in Honduras is a very lengthy process. According to the results of the ICS, it takes 90 days on average to register a firm in Honduras,30 about twice as long as in Guatemala and three times as long as in Nicaragua. All procedures are lengthier in Honduras than in Nicaragua and Guatemala (although acquiring an environmental permit is very time consuming in Guatemala). The constitution act, environmental and health permits, and licenses to operate -- some of which are issued by the municipalities -- as well as "other licenses" (e.g., garbage, export licenses, migration, and INFOP) all require more than one month each -- see Figure 2-1. 29One important element of governance that is not discussed here, for example -- because it was only marginally addressed by the ICS questionnaire -- is "grand corruption" (e.g., influencing the content and application of laws and regulations, purchasing public sector positions, and political contributions). For a comprehensive overview of issues related to corruption in Honduras, see the report by T. Gurgur and F. Recanatini, Governance and Anti-Corruption in Honduras: An Input for Action Planning (WBI, 2002), which is based on the results of a survey conducted by WBI and MERCAPLAN in Honduras during the year 2000. The survey targeted 2000 Honduran citizens, 200 private firms and 1,403 public officials, and assesses all the main determinants of governance and corruption in the country. 30Based on responses from 72 firms that registered after 2000. 32 Figure 2-1: Number of days needed to acquire each license required to register a firm 180 Total number of days to start a firm 165** 160 (average) * 140 90 32 56 120 100 80 60 51 48 45 40 40 27 31 30 20 16 21 23 25 17 1719 20 10 12 12 13 5 5 0 Honduras Nicaragua Guatemala Constitución of the firm Registro Mercantil Registro de la Dirección de Ingresos Licencia de Operaciones* Health Ministry MAGFOR (for NC and Guate) Environment Ministry Other Source: Honduras Investment Climate Survey. *Total time to register is lower than the sum of days of each procedure because some procedures can be conducted in parallel **Only 8 firms require an environmental permit in Guatemala, and several of them did not answer the question on the overall time needed to register a firm, so the number should be taken cautiously. 2.6 The registration process is longer for medium and small businesses, and for firms operating in industries that require special permits (e.g., health and environment). Procedures are longer for medium and small firms (it takes 125 days to register a medium firm and 124 days to register a small firm). Registration times are also longer for firms operating in the food industry (which need health permits), the plastics and chemical sectors (for which environmental permits are the main bottleneck), and the wood and metal industries (where registration is delayed by operational permits). Interestingly, the survey results show that when hiring an intermediary (and 77 percent of the firms that started a company after the year 2000 did), the inscription process takes longer. This finding, which is consistent in all three countries, may be due to the fact that firms only contract the intermediaries after having tried to attain the permits themselves and becoming frustrated with the process, or to the fact that some firms do not even attempt to register by themselves because they believe that the permit gathering exercise will be too long and costly. 2.7 In the country's most important cities -- Tegucigalpa and San Pedro Sula -- the length of a few specific procedures delays the registration process substantially. Registration times in Francisco Morazan and Cortes (where the two main cities, Tegucigalpa and San Pedro, are located) are close to the average (79 and 91 days respectively).31 However, getting an operational permit takes a long time in both Tegucigalpa and Cortes, indicating the need to reform the municipalities' procedures. Getting a health permit in Tegucigalpa is also a lengthy process, while other permits (including garbage, export licenses, migration, and INFOP) are particularly cumbersome in San Pedro. The ICS data are by and large consistent with the findings of a recent World Bank Doing Business study that assessed the time and cost of business registration worldwide.32 According to the Doing Business report, the time needed to register a "standard firm"33 in Tegucigalpa is 80 days -- similar to the ICS finding of 79 days.34 The Doing 31Excluding outliers. 32World Bank, Doing Business in 2004: Understanding Regulation (Washington, DC: World Bank and Oxford University Press, 2004), hereafter cited as Doing Business. 33A limited liability company located in the capital, with initial capital of about US$9,200. Background 33 Business study also found that the cost of registering a firm in the capital is US$670. The report points out that, although the time required to register a business in the capital has decreased during the last year (from 146 to 80 days), the costs have increased (from US$567 to US$670). The change in time was driven largely by elimination of judicial permits (Calificación Judicial), reduction of time needed for the registration at the Mercantile Registry and shortened time of operational permits. The downside of reduction in time to request operational permits has been an increase in the municipal taxes to obtain them (from US$23 to US$92), which has driven the overall increase in the costs of business registration. 2.8 Once a firm is registered, interacting with the Government seems to be less of a hurdle, although firm-government interactions become more frequent and require more managerial time as firms grow. Inspections of firms in Honduras do not appear overly onerous. Honduran firms report about 12 days of inspection per year, versus 21 for Nicaraguan firms and 10 for Guatemalan companies -- see Figure 2-2.35 Honduran municipalities, as well as health (sanitation) and environmental agencies inspect firms proportionally more than the same institutions in the neighboring countries, while tax and labor inspections are less frequent in Honduras then in Nicaragua and Guatemala -- see Figure 2-3. The distribution of the number of inspections across firm sizes in Honduras is also similar to the distribution in comparator countries, with large firms receiving more inspections (22 days per year) than Micro Small and Medium Enterprises (MSMEs). There appear to be marked jumps in regulatory costs associated with increases in size of the firm, particularly when transitioning from micro to small (from 7 to 11 days) and -- especially -- from medium to large (from 11 to 22 days). These jumps in days of inspections closely match the differences in management time spent on regulation by firm size -- see Figure 2-4 and Figure 2-5. Firms operating in the food and beverage industry and textile companies receive more inspections than the average -- as expected, given these firms' need to meet sanitary and environmental requirements. Firms in Tegucigalpa are also (slightly) more likely to be inspected than firms located on the East Coast and in the Sula Valley. Firms located in La Paz receive the highest number of visits from the government. papers for the Doing Business report (2004). Draft. 34 The figures are remarkably close considering the differences between the Doing Business study and the ICS surveys. First, the "standard" firm of the Doing Business study is a limited liability company, while the firms surveyed by the ICS are of various kinds (including, for example, sole proprietorships, partnerships and cooperatives). Second, rather than reviewing the registration process of an "ideal" firm, the ICS collects data from real companies that have gone through the registration process during the last two years. Third, the ICS was conducted in the spring of 2003--and asked companies to report on registration processes starting in the year 2000--while the last Doing Business survey was conducted in November 2003. 35 Note that the objective should not be to reduce the inspections to zero. While excessive inspections are likely to be costly and associated with corruption, too few inspections are also problematic. Governments should strive to reach a balance between protecting the public interest and unduly burdening firms. 34 Figure 2-2: Number of days of inspections, Figure 2-3: Types of inspections. Honduras, Nicaragua international comparison and Guatemala. (Percentage values) Tax Inspector Labor and Social Security Fire and Building Safety Sanitation/ Epidemiology Police Environmental Ecuador 15.5 Officials from the municipality Guatemala 9.9 Nicaragua Pakistan 10.5 Guatemala Honduras 12.1 Nicaragua 21.3 Honduras 0.0 5.0 10.0 15.0 20.0 25.0 Total days spent with officials from: all agencies 0% 20% 40% 60% 80% 100% Figure 2-4: Days of inspections, by size. Honduras. Figure 2-5: Percentage of management time spent with regulators, by size. Honduras. Total 12.1 Total 8.6% Large 21.7 Large 13.4% Medium 10.0 Medium 9.5% Small 10.9 Small 8.6% Micro 4.1% Micro 6.5 0.0% 5.0% 10.0% 15.0% 0.0 5.0 10.0 15.0 20.0 25.0 Management Time Spent with regulator (%) Numer of Inspection Days with zeros Source: Investment Climate Surveys (All figures above) Note: Average number of inspections is across all firms, not just those that received an inspection. 2.9 Despite the (relatively) low number of interactions that firms have with government agencies in Honduras, bribes are a common practice to ensure that the Government gets things done. Bribery is relatively more prevalent in Honduras than in most comparator countries except Guatemala and Ecuador -- see Figure 2-6. On average, Honduran firms pay 2.5 percent of their sales to "persuade" the Government to get things done for them.36 Bribery rates are highest for infrastructure services such as the state-owned telephone company (more than 14 percent of the surveyed firms reported that a bribe is necessary to get a connection) and the privatized electricity distribution companies (9 percent of firms) -- see Figure 2-7. Between 3 percent and 8 percent of Honduran firms also pay bribes for import licenses and health, water connection and construction permits. Micro and small firms dedicate a higher proportion of their sales to bribes (3.5 percent and 3.2 percent respectively, versus 0.4 percent for large firms). Firms operating in the chemical, furniture and wood, and food and tobacco industries also pay more than the average. (See paragraph 2.17 in the procurement section below for an explanation of possible reasons for differences in bribes by firm size and industry). 36To assess this, the questionnaire asks the interviewed firm whether other firms in its industry have to pay a bribe to get things done. 35 Figure 2-6: Bribery needed to get things done Figure 2-7: Percentage of firms that pay a bribe for each (percentage of sales), international comparison. service, Honduras. 16.0% Guatemala 3.7 14.4% Ecuador 2.6 14.0% Honduras 2.5 12.0% Bangladesh 2.5 10.0% 9.0% Pakistan 2.0 8.0% 7.5% 6.7% China 1.9 6.0% Nicaragua 1.7 4.0% 3.4% 3.1% Tanzania 1.5 2.0% 1 2 3 4 0.0% Telephone Electrical Import Permit Health Registry Water connection Construction Permit Unofficial payments to get things done (% ann. sales) connection connection Source: Investment Climate Surveys. 2.10 The econometric analysis shows that firms' productivity is affected by the amount of time they spend on Government inspections and how much they bribe the Government to get things done. The econometric analysis (see Table 1-3 and Table 1-4 in CHAPTER 1, as well as ANNEX C) indicates that Honduran firms' productivity is significantly affected by the number of days they spend on inspection- and other regulation-related activities. An increase in the time spent by firms on these activities by one day results in a decrease in average productivity ranging from 5.8 percent to 10.7 percent, controlling for industry, country and year characteristics as well as a set of other firm-level and investment-climate variables (based on the regional results in Table 1-5). This result is mostly relevant for old firms. Such a pronounced negative effect on productivity of the time businesses spend in dealing with Government officials and inspectors may explain the needs for firms to extend "extra" payments in exchange for a reduction in the time needed for completion of inspections and bureaucratic processes. The econometric analysis also indicates that firms that make payments to "speed up and grease" bureaucratic processes enjoy a productivity increase -- an additional payment of 1 percent of total firm annual sales increases firms' productivity between 1.3 percent and 3.3 percent (see Table 1-3). The interpretation of this finding is that firms that make such unofficial payments to deal with bureaucracy faster gain a competitive advantage over firms that choose not to make such payments. The contribution of the average percentage of firm sales spent on such payments to average productivity in Honduras is of a similar magnitude to that in Nicaragua, but lower than in Guatemala (see ANNEX C for the analysis of average productivity in each country). 2.11 In addition to increasing costs for existing firms, the level of red tape and corruption faced by formal firms, together with the obvious tax advantages, may help to explain the fact that about half of the Honduran economy is informal. Avoiding taxes is the main reason that firms remain informal. However, several studies37 have shown that high levels of regulation, particularly if accompanied with substantial discretion and corruption, contribute to increased informality. Informality, in turn, has negative consequences on both public finance and economic growth.38 In Honduras, the combination of inefficient registration processes, the need to spend time with regulators and submit to various inspections once registered, and the requirement to pay additional bribes to get things done when interacting with the government may make staying informal an attractive option for many firms. When the costs of tax and labor requirements are added, it is a simple decision for many micro and small firms to eschew formality. Evidence of informality abounds in the country. Firms in the sample report that informal firms account for 20 37See, for example, Johnson et al. (1998), Schneider and Enste (2000). 38Kaufmann, Kraay and Zoido-Lobaton (1998). 36 percent of the market and estimate that firms in their sectors declare between 61 percent (micro) and 73 percent (large) of their sales to tax authorities. 2.12 The productivity analysis indicates that the proportion of annual firm sales that are not declared to the tax authorities exerts a statistically significant negative effect on firm productivity. A one-percent increase in the fraction of undeclared sales (out of total annual sales) is estimated to reduce firms' productivity between 0.42 percent and 0.77 percent (see Table 1-3 in CHAPTER 1 on page 24 above). This negative effect of the average share of undeclared sales on average country-wide productivity is stronger in Honduras than in Guatemala, and of a slightly lower magnitude than in Nicaragua (see ANNEX C). The fact that undeclared sales reduce productivity -- controlling for a multitude of additional factors -- should not come as a surprise. It can be surmised that the cost savings made due to paying a lower tax bill can be more than offset by losses of Government services to firms due to their not paying their full tax bill. Or else, ceteris paribus, firms which tend to underreport their taxable incomes are also less likely to allocate resources efficiently and will thus have lower productivity. 2.13 The impact on the Government of losses due informality can be estimated at about 6 percent of GDP. Independent estimates place the informal economy in Honduras at 50 percent of GDP, about the same as in Guatemala and more than in Nicaragua (45 percent), the Dominican Republic (36 percent) and Costa Rica (26 percent)39 -- see Figure 2-8. As the average value- added tax in Honduras is 12 percent, a rough estimate of the Government's losses due to informality would be 6 percent of GDP. (A precise computation would require estimating the profits of the informal private sector and considering losses related to both the value added tax and other taxes, such as income and property taxes, that informal firms don't pay.) Reducing informality to the levels of Costa Rica, for example, would increase Honduras' GDP by about 3 percent. Figure 2-8: Informal economy (percentage of GNI) Bolivia Guatemala Honduras Nicaragua AVERAGE Ecuador Costa Rica Chile 0 20 40 60 80 Source: Doing Business Database. 2.14 To accelerate business registration and operating procedures, the Government should focus on implementing the reforms it recently passed. Streamlining processes in various public agencies (including municipalities) is the Government's main challenge in terms of business administration. Efforts to make Honduras more responsive to private sector 39Figures from Schneider (2002) 37 requests for administrative streamlining are a highlight of the current administration. The Ley de Simplificación Administrativa (June 2002) framed this effort by directing every office of government to design and implement its own administrative reform program. There has already been some notable progress (e.g., time reduction in several key steps of business registration). Recently, the Government and the World Bank discussed additional specific measures to reduce the time and cost of business administration and operation. The implementation of these initiatives is expected to start in the next few months.40 First, the Government will sponsor the creation of a Business Simplification Committee modeled after a similar organization in Chile, which will be composed of representatives of the public and private sectors and will be responsible for all initiatives related to simplifying regulation of business administration. The Committee will take the lead in identifying areas in need of administrative reform, solicit the interest of the government offices responsible for these areas (dueños de tramites), and assist with the re-design and implementation of new procedures in the selected government offices. Second, simplification pilots will be conducted with a few selected municipalities where administrative simplification is urgently needed (e.g., Tegucigalpa and San Pedro Sula) and a few public agencies that have lengthy procedures (e.g., health and environment). Finally, the Government and the Bank plan to analyze in more detail the specific administrative bottlenecks faced by firms in industries that have the lengthiest procedures: e.g., food and maquila. 2.15 Corruption could be reduced and transparency in public management could be improved by strengthening Government efforts to improve public sector financial management, reform the civil service, and promote decentralization and accountability at the local level.41 Shortly after taking office, the Maduro administration renewed the mandate of the National Anti Corruption Council (NAC). The NAC is comprised of both government and civil society representatives and is now implementing a plan to reduce corruption and increase transparency in the country. Key causes of corruption that the Government is discussing with the World Bank and other organizations include: (a) the Government's weak financial management system and lack of expenditure controls; (b) politicization of civil service appointments, substitution and promotion; and (c) insufficient decentralization of responsibility (coupled with a lack of accountability) at the local level. Recommended actions in these three areas are: e. Improvements to the Government's financial management system could help reduce corruption by making financial information available for public scrutiny, thereby not only increasing the transparency of Government expenditures but also enhancing the utility of the financial information it provides. To achieve these goals, the following measures -- which the Government is currently discussing with donors -- should be taken: (i) upgrading human resources for financial management throughout the public sector; (ii) improving the public sector's budgeting execution and accounting and financial reporting; and (iv) providing technical assistance to the treasury, to ensure that it can carry out its supervisory functions and to improve the Government's ability to exercise greater fiscal and cash management discipline. 40 These initiatives will be financed under the World Bank Trade Facilitation and Productivity Enhancement Project. See the description of the administrative barriers component in the Project Appraisal Document (World Bank, 2003), pages 11 and 60-63. The Project was approved by the World Bank Board in October 2003 and is expected to become effective in April 2004. 41This section draws from the Honduras Country Assistance Strategy (World Bank, 2003), pages 18 and 19; the Draft Honduras PRSC Matrix (World Bank internal document, 2004), specifically the section on public resource management; and the Draft Honduras Financial Accountability Assessment (World Bank internal document, 2003). 38 f. To improve quality and reduce corruption in its civil service,42 the Government and donors are discussing the following possible measures: (i) restoring control over wage management; (ii) introducing competitive merit-based recruitment and promotion processes; (iii) limiting the number of political appointments in the civil service; and (iv) creating an independent body with authority to regulate the implementation of the civil service reform. g. To improve decentralization and promote accountability at the local level, the Government and donors are discussing the possibility of: (i) creating a single agency to serve as the normative entity for the coordination and regulation of the municipal sector and to support development of the municipalities and the local communities; (ii) reforming the law of municipalities to transfer more responsibility to the local level, with appropriate safeguards and accountability to ensure that overall fiscal discipline is not compromised; (iii) providing technical assistance to improve planning and monitoring capacity at the local level,43 including training as well as standard tools to facilitate the decentralization process (e.g., strategic plans, and budget and accounting toolkits); and (iv) strengthening overall monitoring capacity at the central level, as well as local mechanisms of social control necessary to hold local officials accountable to their communities and the state,44 in order to ensure that implementation at the local level is conducted in an effective and transparent manner. 2.16 The Government is in the process of approving a reform of the existing Property Rights System, which -- among other positive effects -- will help reduce informality. The insecurity of property rights in Honduras has long been a source of tension and has contributed to the weaknesses of the country's investment climate and its high poverty.45 The new Government administration has designed a reform aimed at providing better registration and protection of property rights. This entails passing the Ley de Propriedad -- which is currently under approval -- and creating a new Property Rights Institute -- the Instituto de Propriedad (IP) -- which will integrate the existing Cadastre and the Real Estate Registries. This reform is expected to create confidence in the registry system and increase respect for property rights, in turn reducing informality, improving the overall investment climate and facilitating access to credit (see Chapter 5). PROCUREMENT: HOW MUCH ARE FIRMS WILLING TO PAY TO DO BUSINESS WITH THE GOVERNMENT? 2.17 Like firms in other countries, Honduran firms have to pay for procurement contracts. Honduran MSMEs and firms operating in the chemical and wood and furniture 42 In addition to the sources quoted in footnote 41, see also the World Bank background document Honduras Civil Service Reform, prepared for the Honduras PRSC. 43 Experience in some Latin American countries has demonstrated that, not only may overall fiscal discipline be compromised by a lack of sanctions applied to municipalities that are out of compliance with financial reporting requirements, but even basic tenets of fiscal management and execution may not be understood or managed properly at the sub-national level. Technical assistance, especially with respect to execution of local investment plans, is necessary in order to mitigate precisely these kinds of risks. 44Again, experience in some Latin American countries has demonstrated that this is a necessary safeguard. 45B. Trackman, W. Fisher, and L. Salas. The Reform of Property Registration Systems in Honduras: A Status Report. (July, 1999). Available at: http://cyber.law.harvard.edu/prs/Hondu.html 39 industries pay more. Honduran firms report that, to win a contract,46 they have to pay about 4.6 percent of the value of the contract -- about one third of what Brazilian firms pay, but more than firms in all other comparator countries (including Bangladesh) and more than twice the amount paid by Chinese companies (see Figure 2-9). MSMEs pay significantly more than large firms -- medium firms, for example, pay 5.6 percent of the value of the contract, versus 1 percent for large firms (see Figure 2-10). Also, chemical companies and companies operating in the wood and furniture industries pay high prices to win the bids -- up to 8.6 percent for chemical firms and about 8.2 percent for wood -- while textile, garments, food and beverage firms pay significantly less than the average (see Figure 2-11). Smaller (one time) payments may be associated with smaller (but more frequent) procurement contracts. This could be the case for MSMEs, or for firms operating in industries such as food and drinks, for example. The higher bribes paid in certain industries may also be symptoms of more engrained corruption in public-private relationships in some sectors of the economy. Differences in frequency and amounts of procurement bribes may also be due to the level of sophistication of the industry. For example, exporters pay significantly less than non-exporters, and apparel firms -- mostly large maquilas -- pay less then firms in the wood and furniture industries. Finally, smaller and less advanced firms may see "giving gifts" as normal and acceptable behavior. This would imply that lowering tolerance for corruption and increasing the ability of firms to say "no" or to report solicitations of illegal payments without fear of reprisals would help to reduce the incidence of corruption. Figure 2-9: Percentage of contract paid to win a public contract, international comparison. China 2.1 Tanzania 3.9 Nicaragua 4.3 Guatemala 4.3 Bangladesh 4.3 Honduras 4.6 Brazil 12.2 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 Gifts expected as % value government contracts Source: Investment Climate Surveys. 46To assess this, the questionnaire asks the interviewed firm whether other firms in its industry have to pay a percentage of the contract to be awarded a procurement bid. 40 Figure 2-10: Percentage of contract paid to win a Figure 2-11: Percentage of contract paid to win a public contract, by size. Honduras. public contract, by industry. Honduras. Total 4.6 Total 4.6 Textile 0.0 Non Metalic Mine 4.6 Micro 4.5 Metal Products 4.2 Small 4.6 Furniture and Wo 8.2 Food and Tobacco 2.1 Med 5.6 Chemical and Rub 8.6 Beverages 1.1 Large 1.0 Apparel 2.1 0 1 2 3 4 5 6 0 2 4 6 8 10 Source: Honduras Investment Climate Survey. 2.18 To reduce corruption and improve procurement efficiency, the Government should strengthen the newly created Office of the Auditor General. In this initial transition period, it should rely more on external auditors.47 The Government has recently taken clear steps to increase public sector accountability and transparency of procurement by establishing a strong Office of the Auditor General through the merging of the office of the Controller General and the Ethics Office. It has also secured approval of a new State Procurement Law that brings all Government agencies under a single set of rules, provides for the establishment of a new Procurement Regulatory Office, mandates open competitive procedures for procurement, and includes some important disclosure provisions. Furthermore, a set of regulations for an Audit Court -- the Tribunal Superior de Cuentas (TSC) -- has been promulgated. However, considerable effort is still needed to effectively implement the provisions of the new law and to ensure that procurement is conducted in a transparent way. Most importantly, the Audit Court needs to develop the capacity to assume its role as the supreme oversight agency responsible for ensuring accountability of government procurement processes and contracts. Specific challenges include (i) improving the audit skills of the Audit Court's staff, who need to have sufficient knowledge of internationally accepted standards and audit methodologies; (ii) reviewing and approving the eligibility of private auditing firms hired by public sector entities, which could be used while the Audit Court trains its staff; and (iii) establishing a new national procurement office and communicating its role to the various government agencies, the private sector and the general public. 2.19 Further measures that the Government should take include re-engineering procurement processes -- at various levels -- and developing online procurement services. The Government needs to standardize its procedural manuals, sample contracts and other procurement documents. New, standardized modules should be developed for both national and local agencies, including municipalities. In addition, the Government should create an electronic public information system for procurement, and introduce an electronic system for transactions. The experiences of Chile and other Latin American countries (e.g., Brazil) with online public procurement can provide a useful reference in this respect. Chilescompra, for example, has succeeded in making business opportunities with the Government more transparent, in turn reducing transaction costs for both firms and the Government, as well as sharply reducing 47See World Bank, Draft Honduras Financial Accountability Assessment (World Bank internal document, 2003), pages 33-36; the Honduras Country Assistance Strategy (2003), page 18; and the Draft PRSC Matrix (2003). 41 opportunities for corruption.48 An analysis of the procurement processes currently followed by MSMEs and by firms operating in the wood and chemical industries would be particularly useful for clarifying why they kick back such a high percentage of their procurement contracts. The introduction of simplified online procedures for MSMEs, as well as the provision of access to computers in public agencies (see Chapter Two) would be particularly useful for facilitating MSMEs' use of electronic procurement systems. THE JUDICIAL SYSTEM 2.20 Honduran firms do not trust the judicial system for settling their disputes and prefer to resort to alternative mechanisms. Thirty nine percent of Honduran firms believe that the judicial system will not uphold contracts or property rights in a dispute, a similar proportion as in Nicaragua and less than in Guatemala, but significantly more than in Peru and Brazil -- see Figure 2-12. In fact, only 24 percent of all firms with payment disputes went to court in 2002 (56 firms out of 233), and firms who went to court waited an average of 38 weeks to get their cases settled -- see Figure 2-13. The proportion of firms that resort to the judicial system in Honduras is the same as in Nicaragua, but Nicaragua's courts are faster (it takes them 18 weeks on average to resolve a case). As a result of the distrust in -- and the inefficiency of -- the local courts, most Honduran firms with payment disputes resort to alternative resolution mechanisms, such as direct negotiation or a lawyer. Figure 2-12: Percentage of firms that do NOT believe that the courts will uphold their contracts and property rights in a dispute, international comparison. 60 56 50 48 41 39 40 30 23 23 20 10 0 Ecuador Guatemala Nicaragua Honduras Peru Brazil Source: Investment Climate Surveys. 48 "Under the Chilean government procurement e-system, companies that wish to do business with the public sector do not need to search through newspapers or the Web for information about bidding opportunities. Instead, they need only to register a single time in the areas in which they do business (e.g., office furniture, construction services, IT consulting, etc.). Whenever a public agency needs to purchase goods or contract a service, it will fill out a request in the electronic system, specifying the kind of operation and including all the documentation and information associated with the request. Automatically, the system sends an e-mail to all the private companies registered in that selected area, minimizing response time and providing an equal opportunity for all firms (that have access to the Internet). The system also provides, on-line, all the information related to procurement operations, including the public organization's name, address, phone, e-mail, fax and position of the public officer in charge of the operation. Finally, at the conclusion of the bidding process, the e-system provides the results: who participated, the proposals, the economic and technical scores, and, lastly, who won the bid or obtained the contract. Historical information about the public organization's purchases and contracts is also made available". Chile's Government Procurement e-system. Available at: http://www1.worldbank.org/publicsector/egov/eprocurement_chile.htm 42 2.21 Trust in the judicial system varies with firm size and location. Micro and small firms trust the judiciary less than medium and large firms. Only 6 percent of micro and 19 percent of small firms went to court in 2002, and their cases took longer to get settled -- 92 weeks for micro firms and 50 weeks for small firms (versus 7 weeks for large firms). Trust in the courts and court efficiency also varies across regions. For example, in the North West (San Pedro Valley), the courts are relatively efficient (29 weeks on average to resolve a case), and firms rely more on the judiciary (31 percent of firms located here go to court). By contrast, only 23 percent of firms go to court in Tegucigalpa, and they have to wait 70 weeks to get their cases resolved. (See Figure 2-14.) Overall distrust in the Honduran courts is confirmed by the results of the survey conducted by WBI in Honduras in the year 2000.49 The WBI survey shows that firms, citizens and public officials consider the judicial branch of the Government to be the most corrupt institution in Honduras. The majority of respondents also specify that the system is too complicated, time consuming, inefficient and slow, and point out that paying a bribe is a pre-condition for a speedy judicial process. Fifty-seven percent of the firms surveyed by WBI consider the judicial system unfair and the judges not credible. The WBI survey also shows that users in the North and the West are more satisfied with the courts than users in the rest of the country, consistently with the findings of the ICA survey. Figure 2-13: Percentage of firms that had delays in Figure 2-14: Percentage of firms that had delays in payments and went to court and average time payments and went to court and average time needed needed to settle a case, by size. Honduras. to settle a case, by Region. Honduras. Number of weeks to solve Number of weeks to solve the case the case large 43% 7 South 0NS NS medium 36% 48 Tegucigalpa 23% 70 small 19% 50 North West 31% 29 micro 6% 92 East Coast 10% 8 total 24% 38 Total 24% 38 0% 10% 20% 30% 40% 50% 0% 10% 20% 30% 40% Firms that had delays in payments and went to court (%) Firms that had delays in payments and went to court (%) Source: Honduras Investment Climate Survey. 2.22 The Government should continue its effort to promote judicial independence, reduce corruption and improve the overall effectiveness and efficiency of the judicial system. Recognizing the shortcomings of the judicial system, the Maduro administration has embarked upon a program to strengthen the rule of law, promote judicial independence and improve performance.50 Reforms have recently been passed and implementation work has commenced. The first step was the enactment of a constitutional amendment (in 2002) to enable the selection and appointment of a new Supreme Court through a transparent process involving civil society participation and open competition. The number of judges was increased by 8 to 15, and the term of office was increased to seven years, greatly improving judicial independence. Since its appointment last year, the new Supreme Court has taken the lead in improving the performance of the Judicial Branch and has already implemented a proactive program to fight 49WBI (2002), see footnote 29. 50This section draws from the Honduras Country Assistance Strategy (2003). Note that the Government's focus to date has mainly been on legal reforms with respect to the criminal procedure code. Other areas -- including commercial and labor matters -- have received very limited attention. In addition, the overall process has been strong on studies, but incomplete in implementation. Honduras Judicial Branch Modernization Project, Project Concept Note (World Bank internal document, 2003). 43 corruption within the judiciary, introduce oral procedures, develop training plans for staff and judges and establish justice of the peace courts. The Government is now considering several initiatives to build judicial branch governance capacity, win the confidence and trust of citizens, and improve access to justice for high priority target populations.51 Possible new initiatives in these areas should be carefully coordinated to avoid inefficiencies and duplications. Efforts to gather judicial statistics to better understand where the key bottlenecks of the judicial system are would also be useful. Similar pilot diagnostic studies have already been completed in several Latin American countries -- e.g., Mexico, Brazil, the Dominican Republic and Argentina -- and they have proven very useful. The Mexico study, for example, found that 60 percent of debt collection cases did not advance beyond the stage of admission of the complaint. It also found that, because of the large number of cases abandoned early on in the proceedings, judicial workloads are lighter than commonly thought.52 2.23 Specific effort is needed to improve the judicial system's capacity to resolve business disputes. While significant progress has been made in the last few years to improve the criminal procedure codes, the civil code still needs to be revised -- especially with reference to commercial and labor matters. As the ICS and WBI surveys demonstrate, a general lack of transparency and effectiveness in the judicial system is hurting Honduran firms. The business confidence of foreign investors is also likely to be affected. While firms' distrust of the judicial system is based on general impressions of corruption and overall court inefficiencies, it is also influenced by judges' limited knowledge of specific issues related to business resolutions. Training on corporate settlements, including case management to make business resolution more effective, would increase judges' knowledge and experience in these areas, in turn helping to restore firms' confidence in the system. Introducing specialized courts or specialized commercial judges in the regular civil courts may also be a suitable way to improve firms' access to justice. This system has proven effective in several Latin American countries, such as Colombia, which reintroduced specialized courts in 1996. 51 The World Bank is currently preparing a project aimed at: (i) building Judicial Branch governance capacity, (ii) building the confidence and trust of citizens by implementing first stage anti-corruption, transparency and accountability initiatives, and (iii) improving access to justice for high priority target populations. See Honduras Judicial Branch Modernization Project, Project Concept Note (World Bank internal document, 2003). 52See Linn Hammergren, Reforming Courts: The Role of Empirical Research, PREM Note No. 65 (World Bank, March 2002). The Mexico study was conducted by a World Bank team with local researchers in the Federal District, and gathered information about 464 debt collection cases brought under a special procedure that allows for expedited dispute resolution. Resolution times at first instance for Mexico's debt collection cases ranged from 29 to 977 days, with a median of 223 days (between filing and judgment, not including duration of execution of judgment). However, 81 percent of cases filed were abandoned -- although the case files were technically open, the parties were not pursuing their claims. The study found that only 30 percent of first-instance debt collection judgments in Mexico were appealed -- far fewer than is commonly thought. In addition, it was found that the duration of appeal proceedings was relatively short compared to duration at first instance, contravening the popular notion that appeals cause great delays. A cause for concern was the fact that few court records were found indicating execution of the judgments for these debt collection cases. Since judicially supervised auctions constitute the primary mechanism for sale of debtors' assets in Mexico, the latter finding possibly reflects defendants refusing to make payments even after a judgment has been issued against them. Furthermore, the study established that executing a first- instance judgment required twice as much time as reaching the judgment (15 months on average until the first bid for those cases which required execution of a judgment). For more findings on the Mexico study as well as the Argentina study, see Regional Research Project Report No. 26966, An Analysis of Court Users and Uses in Two Latin American Countries, (World Bank, October 2003). 44 2.24 Outside-the-court resolution mechanisms such as direct negotiations can be very useful and should be scaled up. Alternative Dispute Resolution (ADR) mechanisms -- such as mediation and arbitration -- which require the involvement of a third party in guiding the contenders toward resolution, can also be very useful, particularly in countries with ineffective courts.53. A few years ago, the Honduran Chamber of Commerce set up an ADR system (Centro de Conciliación y de Arbitraje). Firms report it to be efficient, yet it is not used very frequently. An evaluation of this system -- and an assessment of the possibility of scaling it up or replicating it -- is warranted. Strengthening ADR mechanisms and ensuring access to them for a larger number of firms (especially MSMEs) is needed to ensure speedy resolution of contract disputes, especially in the interim period during which the judiciary is trying to improve its overall efficiency and transparency. CRIME 2.25 Crime against firms increases the cost of doing business. One in three firms in Honduras suffered a criminal attack in 2002. Firms try to avoid becoming the targets of crimes -- which generate direct losses -- by spending on security. The number of firms subject to criminal attacks and the frequency of attacks (victimization rate) among Honduran surveyed firms were very similar to neighboring Nicaragua, and lower than in Guatemala.54 Thirty percent of Honduran firms were victims of crime (from theft, robbery, vandalism or arson) during 2002 -- see Table 2-1. They suffered 4 attacks on average during the year, losing 3.2 percent of their sales. The average loss for all firms -- not just the ones that suffered an attack -- was 0.9 percent. To prevent crime losses, firms incur further costs by investing in security systems. These outlays were equivalent to 3.6 percent of total costs (for all firms), 53 percent less than in Guatemala but 24 percent more than in Nicaragua. Firms pay for private security guards (vigilantes), security walls or doors, alarms, and insurance against robbery of goods and vehicles. Besides increasing firms' total costs by adding to security expenses, the frequency of crime in Honduras induces behavioral changes in entrepreneurs and workers, some of whom change their work schedule, or have stopped serving some areas of the country because they are deemed unsafe.55 53Experience from the region recommends the use of out-of-court settlements. Argentina's Ministry of Justice implemented a pilot project during 1996-1997 in 20 civil courts, requiring mandatory mediation in commercial cases. This required training of mediators, which was accomplished by the Ministry of Justice and a local NGO -- Fundación Libra. The results of the pilot led to two-thirds of filed cases being settled in mediation. Mediation times were also short compared to litigation. See: World Bank Legal Vice Presidency, Argentina: Legal and Judicial Sector Assessment (World Bank, 2001). 54To put these figures into a broader perspective, consider that Honduras has relatively high victimization rates compared to other Latin American countries. Statistics compiled by Latinobarometro in 1999 show that about 37 percent of Honduran households had at least one member victimized in 1999, close to the rates in Colombia and Nicaragua, while Guatemala had the highest rate, 55 percent. While the Latinobarometro survey does not specify the specific type of crime, we can assume that these figures refer mainly to property crimes. Latinobarometro, as reported by Pages, C. and A. Gaviria. (1999). Figures from Leyva (2000) on homicide rates are even more worrisome. According to these statistics, Honduras' homicide rate in 2000 was 46.3 per 100,000 inhabitants -- second only to El Salvador's and Colombia's and twice as high as most other Latin American countries (even Guatemala fares better than Honduras in this ranking, with 38 homicides per 100,000 people). 55Rubio (2002), for Inter-American Development Bank (IDB). The study includes a survey of households and firms aimed at assessing the incidence and causes of violence in the San Pedro Sula Valley. The survey served as a base for the recently approved IDB Peace and Citizen Coexistence Project for the Municipalities of the Sula Valley, aimed at reducing violence in this area. (See the discussion of the 45 Table 2-1: Criminal activities affecting firms Percentage Number of Losses to crimes Losses to crimes Cost of security to of firms that criminal cases in (as percentage of (as percentage of avoid crimes suffered 2002, average sales) (For sales) (For all (percentage of crimes (For victims only) victims only) firms) total costs) /a (For all firms) Honduras 30 4.03 3.2 0.9 3.6 Nicaragua 27 3.76 3.3 0.9 2.9 Guatemala 52 5.41 3.8 1.9 5.5 Source: Investment Climate Surveys. /a: Note that the average number of criminal cases for all surveyed firms is 1.2 in Honduras. 2.26 The econometric analysis demonstrates that criminal attempts affect firms' productivity negatively and significantly. With every new criminal attempt, firms' productivity is reduced between 1.8 percent and 3.2 percent depending on the productivity measure used and controlling for other firm characteristics, investment-climate variables and country, industry and year effects (see Table 1-3 in CHAPTER 1 on page 24 above). Therefore, reducing the average number of criminal attempts for Honduran firms-victims from 4 (its 2002 value, see Table 2-1) to 1 per year, for example, would raise productivity between 5.4 percent and 9.6 percent. The regional analysis also points to crime mostly reducing productivity in old and large firms. The country-by-country analysis shows that the average number of criminal attempts suffered has a larger (in absolute terms) effect on average productivity in Guatemala than in Honduras and Nicaragua (as shown in ANNEX C). 2.27 The police solve very few cases. Interestingly, only 55 percent of crimes are reported to the police.56 One of the reasons that firms do not call the police is their poor ability to solve crimes. Surveyed firms indicated that only 10 percent of reported crimes were solved in 2002. This figure is extremely low, compared to Nicaragua, for example, where police solve 29 percent of crimes -- see Figure 2-15. Figure 2-15: Crimes reported to and solved by the Police. International Comparison. (Percentage values) Percentages of 29 cases solved 6 10 53 Percentages of cases reported 70 55 0 10 20 30 40 50 60 70 80 Honduras Guatemala Nicaragua Source: Investment Climate Surveys. project's initiatives in paragraph 5.) 56 These results are confirmed by the IDB Survey, which shows that only 27 percent of surveyed households that have been subject to a criminal event reported it. Most of the people that did not report the crime (38 percent) said that they did so because the police would not do much (la policía no haría nada), followed by lack of evidence (18 percent) and fear of retaliation (17 percent) (Rubio, 2002). 46 2.28 The incidence and costs of crime vary by Region. Companies situated in the West and the North (the Sula Valley) suffer more attacks than firms in the rest of the country, suggesting the need to support specific interventions in this area. Thirty-eight percent of firms located in the North West suffered crimes, with an average of 4.7 criminal attacks per firm each year, versus 30 percent of firms and 4 attacks on average in the whole country -- see Figure 2-16. Losses experienced by the firms that are attacked in the North West are similar to losses experienced by firms in other regions, but companies in the North West spend more than the average on security (4.4 percent of their costs, versus 3.6 percent on average). See Figure 2-17. High criminal rates in the Sula Valley are confirmed by statistics of homicides, which rank Cortes (the center of the Valley) as the most dangerous place in Honduras, with 107 homicides per 100,000 people57 (versus 52 on average for the country). Since Cortes and the Sula Valley are home to most Honduran maquilas and the target of major foreign investment, crime prevention and reduction in this area is essential to continue attracting FDI. According to the survey, the safest places in Honduras are El Paraiso in the South (zero attacks on firms in 2002) and some areas of the East and the Coast (e.g., Yoro and Olancho), while the broader Tegucigalpa area is in the middle. Confidence in the police (as indicated by whether or not crimes are reported) and the ability of the police to solve cases also vary by location. For example, while firms in the North West and the East Coast report about 63 percent of crimes, those in Francisco Morazán (broader Tegucigalpa) report just 37 percent of them. The percentage of cases solved is similar in the North West and Tegucigalpa (14 percent and 10 percent respectively), but low (only 1 percent) on the East Coast. Figure 2-16: Frequency of crime and number of cases, by Figure 2-17: Crime losses and security expenses, by geographical area.58 Honduras. geographical area.59 Honduras. /a Average number of attacks (for victims) Total 3.2% 3.6% Total 30% 4.0 3.0% North West (incl. Cortes) 4.4% North West (incl. Cortes) 38% 4.7 3.0% Tegucigalpa 28% 3.7 Tegucigalpa 3.4% East and Coast 23% 2.7 East and Coast 3.1% 2.7% South 17% NS 0.0% NS South 1.5% 0% 5% 10% 15% 20% 25% 30% 35% 40% Cost of security (% of all costs), for all firms Losses (% of sales), for victims Percentage of firms that experienced an attack Source: Honduras Investment Climate Survey Source: Honduras Investment Climate Survey *NS = Not Significant. Only 4 of the 29 firms located in the South *NS = Not Significant. Only 4 of the 29 firms located in the South (the smallest sample) suffered a crime, and for this reason we do not (the smallest sample) suffered a crime, and for this reason we do not report average number of attacks. report percentage of losses. /a: Security Costs (as percentage of all costs) data are for all firms, while losses due to crimes (as percentage of total sales) are only for 57 Leyva (2001). As a reference, consider that London has a homicide rate of 2.4 per 100,000 people, Moscow of 18.2, Washington DC of 45.8 (averages per year from 1998 to 2000), Barclay, G. and C. Tavares (2002). 58The Regional aggregation -- based on vicinity and number of observations available -- is as follows: North West: Cortes, Santa Barbara, Copan, Valle and Choluteca Tegucigalpa (broader): Francisco Morazan and Comayagua South: El Paraíso and La Paz East Coast: Atlántida, Olancho, Colon, Islas de la Bahia and Yoro. 59See previous note. 47 firms that experienced a criminal attack. 2.29 Incidence and cost of crime also vary with firms' size. For large firms, crime is more frequent, but less costly than for MSMEs. Police intervention is also less effective for small firms. Forty four percent of large firms were victims of crimes in 2002, with 6.2 attacks on average, versus 30 percent of firms and 4 attacks on average for all firms. Only 20 percent of micros experienced crimes (2 on average for the victims). (See Figure 2-18.) The proportion of losses suffered by firms, conversely, increases with size, with victimized small firms reporting that they lose up to 5.9 percent of their sales due to robbery and crime (versus 3.2 percent on average and 1.3 percent for large firms). (See Figure 2-19.) This is predictable, because the limited total revenues of smaller firms make robberies more likely to have a greater impact. Investments in security are also inversely related to size: small firms invest more (as a proportion of their total costs) than medium and large firms (but micros are an exception here, investing less than firms in all other groups). Our analysis does not yet indicate whether investment in security is effective in fending off crime,60 but evidence from Rubio (2002) shows that there is a positive association between criminal acts and investment in security -- i.e., the more spent on security, the higher the probability of being the victim of crimes.61 The propensity to report a crime is similar across firms of various sizes, with medium firms being slightly more inclined to call the police and micros being more reluctant -- medium firms report 67 percent of criminal cases they suffer, versus 55 percent on average and 40 percent for micros. Once in action, the police solve 10 percent of the cases, and their effectiveness seems to be related to the size of the victimized firms -- 15 percent of large firms' cases are resolved, versus only 5 percent and 7 percent of small and micros' respectively. The variation by firm size may be related to the levels of effort that the police expend for different classes of victims, but it may also have to do with the amount and quality of evidence that the firms provide -- large firms are more likely to have sophisticated security equipment such as video cameras. 60This analysis will be conducted for the next version of the report. 61Rubio (2002), page 31. This could be because the expenses incurred by small firms are simply not effective in protecting them from violence, or it could be because the expenditures on security equipment actually increase the likelihood of becoming the target of an attack by attracting the criminals' attention. It could also be that firms who have already been the victims of crime are thereby motivated to buy security equipment. Finally, it might be that firms that are located in high-crime areas (and that are therefore more likely to be victims of an attack) are motivated to buy security equipment because they know they are at risk. If, as we might expect, the equipment is "effective" (i.e., it reduces the chances of a crime) but not "perfect" (i.e., it doesn't completely prevent any crime), firms that have it could still tend to have higher victimization rates, simply because of where they are located. 48 Figure 2-18: Frequency of crime and number of Figure 2-19: Crime losses and security expenses, by cases, by size. Honduras. size. Honduras. (Percentage values). Average number of attacks (for victims) Total 3.2% Total 30% 4.0 3.6% Large 1.3% Large 44% 6.1 3.6% 1.5% Medium 32% 3.8 Medium 4.6% Small 5.9% 4.8% Small 29% 2.9 Micro 3.5% 1.9% Micro 20% 2.1 0% 1% 2% 3% 4% 5% 6% 7% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Cost of security (% of all costs), for all firms Losses (% of sales), for victims Percentage of firms that experienced an attack Source: Investment Climate Survey Source: Investment Climate Survey /a: Security Costs (as percentage of all costs) data are for all firms, while losses due to crimes (as percentage of total sales) are only for firms that experienced a criminal attack. 2.30 Whose fault is it? Gangs -- panderillos and maras, usually street children and youth -- are responsible for the crimes in only a limited number of cases. Contrary to common belief, which tends to attribute most of the crime in Honduras to organized groups (the maras),62 Honduran firms report that gangs are responsible for only about 25 percent of all crimes, with slightly higher figures for MSMEs than for large firms. Organized crime seems to be much less active in Honduras than in Guatemala, for example -- where more than 65 percent of firms reported suffering criminal attacks by gangs. These findings are confirmed by the results of the IDB survey.63 Based on interviews with the police, the IDB finds that of about 9,000 young members of the maras in the Sula Valley, only 10 percent could be responsible for serious crimes and little more than 1 percent for homicides. A review of about 40,000 reports at the Dirección General de Investigación Criminal (DGIC) between January 1996 and May 1999 shows that the main crimes committed by youth belonging to the maras were robberies (55 percent), which are likely to affect individuals, households and perhaps micro and small firms. Only 7 percent of criminal acts committed by the maras were attacks on properties, which are more likely to affect larger firms.64 2.31 The Government has already taken clear steps to reduce violence at the National and municipal levels. The citizen security and defense policy, included in the President's document "My Commitment to You," stresses crime reduction as one of the Government's main objectives to improve the country's investment climate. In August 2003, the Government passed a controversial law prohibiting participation in the maras, which are now considered criminal 62Some of the most infamous groups are Mara Salvatrucha and Mara 18. 63See footnote 55. 64See also: Amnesty International Honduras, Zero Tolerance...For Impunity. Extra judicial Executions of Children and Youths since 1998 (Amnesty International, 2003) -- "Some of the Honduran media frequently blame the maras or street children and youths for the growing public insecurity. In fact, in real terms, statistics indicate that crimes committed by minors are minimal. The Honduran government has recognized that juvenile delinquency constitutes 5 percent of the offences and crimes recorded in the country. In 1999, a study undertaken by the Honduran Institute for Children and the Family (Instituto Hondureño de la Niñez y la Familia -- IHNFA) noted that only 0.02 percent of all murders committed in Honduras were attributable to minors. But both officials and the media continue to place responsibility for the majority of crime on street children and members of maras." 49 organizations. The Government has also decided to decentralize part of the police force to enable it to become more knowledgeable about the regions and more effective in its actions. Other aspects of the Government's program include design and implementation of crime prevention plans, involvement of community police in the municipalities, amendments to the police personnel law, increases in police salaries, and the Social Policy and Coexistence Law (Decree 226 of May 2002). In March 2003, the Government and the IDB signed a loan aimed at reducing criminality and improving the levels of peace in the Sula Valley. The loan specifically targets younger people (between 15 and 25) and is aimed at crime prevention and youth protection. The project initiatives include: (a) institutional support to the seventeen municipalities of the Sula Valley and key agencies (police and public prosecutors), with the aim of designing and implementing municipal action plans to reduce crime; (b) prevention activities to reduce juvenile crime, including development of a program for violence prevention in schools, and the construction of sports and recreation centers for promoting positive use of free time; (c) a pilot community police program in the northwestern division and organization of a comprehensive system for police oversight and accountability; and (d) a communication campaign to facilitate better citizen coexistence. 2.32 The Government should support initiatives similar to the Sula Valley's crime prevention program in other violent regions of the country (e.g., the broader Tegucigalpa area). Experience from Latin America and elsewhere suggests that one of the most effective entry-points for crime and violence prevention is at the local or municipal level -- as featured in the Sula Valley initiative. Other Honduran municipalities with high crime rates -- such as Tegucigalpa and Colon -- could benefit from similar programs. In fact, the Government and the World Bank are currently discussing development of crime prevention pilots in several urban areas. An effective crime prevention and reduction strategy must be based on comprehensive `Safety Audits' (i.e., crime and violence diagnostics). A successful implementation of the strategy requires a three-tiered approach: (1) judicial and policing reform -- ensuring that order, fairness, and access to due process is maintained in the day-to-day activities of the community, reducing the fear of crime; (2) social prevention -- targeted multi-agency programs that address the causes and risk factors associated with crime and violence; and (3) situational prevention -- measures that reduce the opportunities for particular crimes and violence problems through urban spatial interventions such as the Crime Prevention Through Environmental Design (CPTED) methodology. This involves integrating crime prevention with principles of design of public spaces such as public buildings, housing, lighting, public transport, recreational spaces, and parks. 2.33 Harmonizing national statistics on crime and conducting specific studies on the causes of crime would help reduce crime. Current crime statistics -- collected by Dirección General de Investigación Criminal, the Policía Preventiva, the Departamento Medico Forense del Ministerio Publico and the Secretaria de Salud Publica -- are fragmented and in some cases still manually managed.65 An effort should be made to rationalize and unify them. This would help in creating a national database, which would make it easier to conduct more in depth studies on the causes of crime and shed light on possible solutions (e.g., further analysis on the types of crimes committed by people that do not belong to the maras -- affecting 75 percent of firms that suffer a crime -- and their causes could help reduce crime against the corporate sector, especially medium and large companies). 2.34 Firms can also help to reduce the probability of crime and the impact that it may have. On the one hand, firms can try to reduce the likelihood of becoming victims of crime by further increasing their security measures. While increasing security costs may not reduce 65H. Leyva (2001) and IDB (2003). 50 crime,66 it may be the only option available in the short term -- and it may prove effective if the security equipment installed is adequate. Thus, improved lighting and granting safe transportation for employees, for example (particularly after regular hours) could be helpful. On the other hand, firms, as well as chambers of commerce and other private organizations could consider participating in municipalities' and government's programs to reduce violence and crime. The Chamber of Commerce of Bogotá, for example, has been an active and important partner in the city's local crime and violence reduction program, which has succeeded in reducing homicide rates by 50 percent in seven years. While this may represent an additional cost for firms in the short term, it is likely to help reduce incidence of crime in the long term, making it unnecessary for firms to lose their income to robberies and investment in security equipment. Firms could, for example, provide conflict resolution and mediation services to their employees, crime victim support services and domestic violence reduction programs. These programs will raise awareness and help employees who are victims of domestic violence and those that commit it. Larger firms might also consider Early Childhood Education programs and vacation programs for children of employees -- both proven to reduce crime and violence in the community. Finally, firms could contribute funds for local diagnostic studies, support for community-level recreational activities, staff-community mentoring programs and participation in youth training, internships, and job placement programs for local at-risk youth. 66See the discussion in paragraph 2.29 and footnote 61 on page 48. 51 CHAPTER 3. INFRASTRUCTURE67 3.1 Efficient infrastructure -- energy, transport and telecommunication services -- increases firms' opportunities to grow by giving them access to new markets (locally and internationally) and reducing their production and distribution costs. An increasing number of studies provide converging evidence of the links between infrastructure and growth in Latin America.68 Baffes and Shah (1998) indicate that the elasticity of output to infrastructure is between 0.14 and 0.16 in Bolivia, Colombia, Mexico and Venezuela. In a recent study, Easterly and Serven (2003) show that the differential evolution of infrastructure assets in Latin America and East Asia widened the cross regional gap in GDP per worker by some 30 percent over 1980- 97. Guasch and Kogan (2001) also show that improvements in infrastructure (roads, ports and telecommunication) can help to significantly reduce firms' inventory levels, and thus the cost of doing business, especially when accompanied by effective regulation and private participation in associated markets. A preliminary study conducted by Dollar, Hallward-Driemeier and Mengistae (2003) based on the results of investment climate assessments conducted in various countries also shows a positive effect of infrastructure (and finance) on firms' productivity.69 This section of the report presents the findings of the ICS, and complementary evidence and policy recommendations in three infrastructure areas critical to Honduran firms: (1) access to and use of electricity, (2) access to telephony and the Internet, and (3) access to transport and logistics systems. Policy recommendations in the infrastructure chapter draw from the CAS (2003), the PRSC (2004), inputs to the Honduras Development Policy Review (2004) and the PPIAF Report "Private Solutions for Infrastructure in Honduras" (2003). ENERGY 3.2 Honduran firms experience long, frequent and costly power outages. The industrial electricity price in Honduras compares well with the prices in other Central American countries: US$0.089/kWh, versus 0.087 average in Central America at the end of 200270, but quality of services is less satisfactory. Ninety-two percent of Honduran firms have suffered at least one power cut in 2002, more than firms in any other comparator country except Bangladesh -- see Table 3-1. The average number of power cuts experienced by surveyed firms in 2002 was 29. On average, Honduran firms were in the dark for 127 hours (about five days) in 2002, more than firms in any other comparator country except Nicaragua -- a high measure by any standard. Honduran firms lost about 3 percent of their total revenues on average as a result of power cuts. Firms that experienced a cut lost 5.1 percent of total sales. Complementary analysis shows that Honduran electric circuits, especially in Tegucigalpa and San Pedro Sula, suffer from overloads and low voltages, which further reduces the quality of power supplied and increases the risk of additional costs to consumers (from motor burnouts, for example). 67 This Chapter was prepared by Marialisa Motta and Lucio Monari, Clemencia Torres and Eduardo Zolezzi (Energy); Carlos Gonzalez and Eloy Vidal (ICT) and José Barbero and Ximena Clark (Transport and Logistics). Stefka Slavova summarized the results of the econometric analysis. Susan Goldmark and Luke Haggarty provided valuable suggestions. Miguel Crivelli provided excellent research assistance. 68On the other hand, the debate on the impact of infrastructure on productivity and growth has not yet been empirically settled-- see A. Estache, V. Foster, and Q. Wodon (World Bank, 2002), page 9, for an overview of several studies providing contrary evidence on this matter. 69 Dollar, Hallward-Driemeier and Mengistae, Investment Climate, Infrastructure and Trade: A Comparison of Latin America and Asia. (World Bank, 2003): Draft. 70 CEPAL, from official data provided by countries. This is the latest figure available that allows for consistent international comparisons. 52 Table 3-1: Incidence, length and impact of power outages, international comparison Percentage of firms Number of times the Total hours of Percentage of sales that experienced firms experienced power electricity cut lost due to power power outages outages /a /a outages /a Honduras 92 29 127 3.0 Nicaragua 89 34 129 3.9 Guatemala 74 14 60 2.3 Peru 63 6 67 1.9 Ecuador 71 12 32 2.6 Brazil 64 5 28 1.5 Bangladesh 94 25 N/A 3.0 China 75 5 N/A 2.0 Source: Investment Climate Surveys. /a : these data are for all firms, not just the ones that experienced interruptions. Note that losses for Honduran firms that did experience a cut (92 percent of the sample) were 5.1 percent of total sales. 3.3 Power cuts affect MSMEs most. Surveyed Honduran large firms reported being in the dark for about 80 hours in 2002 (see Figure 3-1), less than the national average. MSMEs, on the other hand, experience longer (and more frequent) power outages. Medium firms were particularly affected: 42 cuts in 2002, resulting in 156 hours of electricity disconnection. Possible explanations for these differences are the higher reliability of electricity connections for larger firms and the fact that 65 percent of large firms reported having self-generators, versus 20 percent of small and 36 percent of medium firms. As a result of different length and frequency of power outages, small firms lost 3.5 percent of their sales, versus 2 percent for large firms. Figure 3-1: Number of hours and percentage of sales lost because of power cuts -- by firm size. Honduras. /a Percentage of Sales Lost Due to Power Cuts (all firms) Total 127 3.0 Large 80 2.0 Medium 156 3.0 Small 145 3.5 Micro 133 3.3 0 50 100 150 200 Source: Honduras Investment Climate Survey. 3.4 The econometric analysis shows that power outages have a limited, but significant negative effect on firm productivity. A one-percent increase in the average duration of power outages (measured in hours per day) results in 0.02 percent to 0.1 percent decline in productivity, depending on the particular productivity measure used -- see Table 1-3 in CHAPTER 1 on page 24 above. The variable is significant in three of the six estimated regressions, albeit only at the 10 percent level of significance. The pooled regional estimates also indicate that the negative impact of outages on productivity generally affects old firms more than young ones. The country-by- country comparison establishes that among the three countries the negative impact of the average duration of power outages on average productivity is more pronounced in Honduras and Nicaragua than in Guatemala -- see Table 1-4 in CHAPTER 1 on page 26 above. This finding is in line with the results of the previous paragraph, whereby Honduras and Nicaragua have a larger 53 share of firms experiencing power outages, the highest duration of outages and incur the largest losses due to electric power outages. 3.5 The country's electricity capacity is highly dependent on the hydrological conditions of El Cajón. Dry seasons reduce its generation capacity significantly, which could further increase the likelihood of firms' losses due to power cuts. In the last fifteen years, increases in demand for electricity in Honduras have been matched by increases in plant capacity -- see Table 3-2 -- with private generation gaining a considerable share (61 percent of the total supply in 2002). However, adequate reserve capacity in the system depends on good hydrological conditions of the Francisco Morazán hydroelectric plant (El Cajon), the largest plant in the country (300MW, 29 percent of total installed capacity in 2003). The generating limitations of El Cajon are intrinsic to hydrological variability --very dry years may reduce its generating capacity by half. In 2002, for example, El Cajon generated 41 percent less energy than in 2000 (production decreased from 1,553 GWh to 911 GWh). Dependence on hydrological factors imply that the country may need electricity imports (which have in fact occurred during the past four years) or, in their absence, energy rationing to maintain adequate supply. The system will also require in the medium term further increases in generating capacity to meet continuing raises in demand (maximum demand reached 852 MW in 2003, 6.77 percent more than in 2002, and there has been an average yearly increase in demand of 8.4 percent between 1997 and 2001).71 ENEE has recently contracted two new Purchasing Power Agreements (PPAs) for a total of 410MW, but of these, only 235 MW represent a net addition, while the rest will replace existing diesel plants. This additional generation could come from new projects in Honduras, or from the Central American integrated electricity market, which would result from the implementation of the SIEPAC project in 2006-2007. Table 3-2: Supply and demand in the electricity sector, 1985-2002 Installed Maximum Net Generation (GWh) Net Available Capacity Demand Imports Energy Sales Losses ( Year (MW) (MW) Public Private Total (GWh) (GWh) (GWh) percent) 1980 207.8 156 854 1 855 9.3 864 759 12.1 1985 560.3 220 1,346 6 1,352 (127.9) 1,224 1,065 13.0 1990 532.6 351 2,274 0 2,274 (334.2) 1,939 1,490 23.2 1995 755.9 504 1,915 883 2,798 (18.3) 2,779 2,028 27.0 1999 911.3 661 2,175 1,269 3,444 135.7 3,580 2,832 21.0 2000 919.8 702 2,262 1,477 3,739 275.2 4,014 3,289 18.1 2001 921.5 759 1,914 2,045 3,959 308.0 4,267 3,421 19.8 2002 1,072.5 798 1,617 2,545 4,162 415.1 4,577 3,634 20.6 Source: CEPAL (2003) 3.6 Reliability problems are compounded by a significant level of losses, particularly in distribution and transmission. International comparisons show that Honduras has above average levels of distribution and transmission losses -- 21 percent of total production, versus 13 percent in El Salvador, 8 percent in Costa Rica and 5 percent in Chile (see Figure 3-2). Distribution losses, which account for the bulk of the total losses, are mainly explained by illegal connections in poor neighborhoods in the urban periphery and the underestimation of bills to commercial clients of the Empresa Nacional de Energía Eléctrica (ENEE), which is the vertically integrated public monopoly.72 71See demand projections for the Honduran electricity sector in PPIAF (2002). 72ENEE owns the hydroelectric generation plants and the transmission and distribution facilities, operates 54 3.7 Poor system reliability and high losses reflect inadequate investment to upgrade and maintain the system, especially in transmission, but also operational inefficiencies at all levels, especially in the distribution segment. Despite its effort, ENEE has not been able to reduce losses in a sustainable manner. Several areas are in urgent need of efficiency improvements. For example, ENEE has currently 476 employees per 100,000 customers, while the ratio among private power distribution utilities in Latin America normally ranges between 3 and 12 employees per 100,000 customers.73 Figure 3-2: Distribution and Transmission Losses as Percentage of Output, Central American Countries 5 Costa Rica 8 12 El Salvador 13 17 Panama 19 20 Honduras 21 23 Nicaragua 26 0 5 10 15 20 25 30 Source: World Development Indicators and CEPAL (2003) 3.8 To improve the quality and reliability of electricity provision in the country, the Government should undertake a restructuring of ENEE to achieve greater efficiency in operation, investment and maintenance. The Government should also promote active participation of the private sector through competitive mechanisms. While the generation segment has been opened to private participation, transmission and distribution remain under ENEE's monopoly. The public company also owns 49.7 percent of the total installed capacity in the country.74 To bring more transparency to the operation of ENEE and better understand where the main bottlenecks are, ENEE is considering creating separate business units -- and separate accounts -- for its various areas (generation, transmission, dispatch and distribution). Note that, while ENEE has already outsourced meter reading, billing and collection to a private contractor, to date this has not produced sufficient gains because of its limited scope. To improve distribution efficiency in the future, the Government should consider introducing mechanisms to involve the private sector. 3.9 Expansion of electricity generation is being undertaken through the award of contracts to private firms at competitive prices. Additional investment in new hydroelectric and thermal generation is required in the medium term to ensure meeting increasing demand -- advance planning is needed because bringing new plants on stream requires time. The 2001 PER estimated that the power sector's investment requirement would average US$170 million per year over the next five years, with 70 percent dedicated to expanding generation and 20 percent to expansion of the transmission grid. In view of the high opportunity the Dispatch Center, and purchases power from private generators through Power Purchase Agreements (PPAs). 73Honduras DPR, Draft (World Bank, 2004). 74CEPAL, 2002. 55 costs of public funds, this expansion should be financed with private sector capital to the maximum possible extent. How much of the necessary new capacity will come from projects inside Honduras or from investments in other Central American countries will depend on the progress in the implementation of the Central America electricity integrated market and on the overall conditions for private investments. In view of the large domestic natural resources -- especially hydro -- alternatives for developing new generation projects in the country will need to be carefully evaluated. To facilitate private investment and increase generation capacity, the roles and responsibilities of the different government departments and agencies, including the regulator and ENEE, need to be clarified. This would allow streamlining approval processes and clearance procedures, as well as improving accountability. 3.10 The Government will need to modernize the legal framework of the electricity sector, and strengthen the institutional capacity of policy agencies and the regulator in order to create an enabling environment to attract long term investors and bring tangible benefits to end users. Several draft laws have proposed to reform the overall structure of the Honduras power sector, but none have been approved.75 In addition, the policy making capacity in the energy sector is inadequate given the complex issues. While the Ministry of Environment (SERNA) is formally responsible for energy policies, the Comisión Nacional de Modernización, under the Presidencia is de facto working on all energy-related policy issues, while the Comité Energético meets only occasionally. Important policy and institutional reforms to be pursued over the next few years -- besides transforming ENEE to improve its efficiency, promoting more active private participation in the sector and increasing generation capacity -- include: (a) reinforcing the functions of sector policymaking and regulation; (b) establishing a sub-secretariat of energy under the Ministry of the Presidency to provide technical support and continuity in energy sector policies and strategies; (c) strengthening the electricity regulator, ensuring its independence and making a clear provision for its financing, either through government budget or through fees paid by the end users; and (d) defining transparent procedures for appointing the commissioners, to ensure insulation of the regulator from political interference. 75The first reform proposal was submitted to the Government in 1994 -- see the detailed description of the proposal in PPIAS (2002). In 1999, the Government proposed a new sector law, which would introduce a competitive wholesale electricity market with agents including generators, distributors, transmission companies, traders and large consumers. In the process, sector governance would be radically reformed, with a sharp reduction in the state's role in service production and a major increase in the importance of its regulatory function. ENEE would be "unbundled" both vertically and horizontally. Distribution would be privatized. ENEE's generating units would be broken up into separate operating companies, and the transmission network would be assigned to a separate company. The development of thermal generation would be undertaken privately, licensed by SERNA. Tax breaks on imported fuel would be extended to all generation (at present only power sold to ENEE has them). The development of renewable energy sources would be concessioned. To guarantee a competitive structure, no generator who had more than 25 percent of firm capacity could participate in new public offers to sell energy. In mid 2001, the Government decided to abandon the reform effort. In January 2002, soon after the new Maduro administration took office, the Government expressed a renewed interest in reforming the electricity sector. In November of 2002, the administration circulated, for a limited audience, a document with key aspects of the proposal (Comisión Presidencial de Modernización del Estado -- Propuesta de Diseño de Reforma del Subsector Eléctrico). At the end of 2003, the Government expressed renewed interest in the matter one more time, although it has not yet declared its position on the November 2002 proposal. 56 INFORMATION AND COMMUNICATION TECHNOLOGIES (ICT) 3.11 Honduran firms wait six months on average to get a telephone connection and, once they have a phone, are likely to experience service interruptions.76 Honduran firms' waiting times are the highest among all comparators -- see Figure 3-3. The problem is most severe for firms located in the western region, which can wait more than a year for a connection. Fraud is seen as a possible way to speed up the process -- 14 percent of the firms included in the survey claimed that they made "extra" informal payments to obtain a telephone line. The quality of the service is also unsatisfactory. Twenty-four percent of surveyed firms that have a telephone connection experienced telephone interruptions in 2002 -- see Table 3-3. The average number of telephone cuts in 2002 was 6.5--more than firms in any other comparator country except Nicaragua and Ecuador--with small firms and those located in the western region reporting (respectively) 11 and 36 service interruptions per year.77 On average, Honduran firms were disconnected for 56 hours (7 working days) in 2002. Large firms experience the longest cuts--88 hours, or about 10 working days. Honduran firms stated that they lost on average about 0.3 percent of their total revenues because of telephone cuts; firms that actually experienced a telephone cut (25 percent of the sample) lost about 3.1 percent of their total revenues. Figure 3-3: Average waiting time for telephone line installation (Number of days) Honduras 175 Ecuador 130 Nicaragua 118 Guatemala 48 Brazil 18 Peru 10 0 50 100 150 200 Average waiting time for phone installation Source: Investment Climate Surveys. Table 3-3: Incidence, length and impact of telephone cuts, international comparison (2002) Percentage of firms Number of times the that experienced firms experienced Total hours of Percentage of sales lost Country telephone cuts telephone cuts /a telephone cuts /a due to telephone cuts /a Honduras 24.8 6.5 56 0.3 Nicaragua 26.2 7.0 79.5 0.7 Guatemala 19.6 2.6 18 0.2 Peru 16.0 0.6 50 N/A Ecuador 36.6 12.4 203 1.1 Brazil 27.1 2.1 27 0.3 Source: Honduras Investment Climate Survey. 76Econometric evidence shows that access to telephony has a significant role in increasing firms' earning opportunities. Information and Communication Technology Strategy Paper (World Bank, 2003). 77Complementary analysis shows that Honduras electric circuits, especially in Tegucigalpa and San Pedro Sula, suffer from overloads and low voltage, creating conditions conducive to power cuts. 57 /a : these data are for all firms, not just the ones that experienced interruptions. Note that losses for Honduran firms that did experience a cut (25 percent of the sample) were 3.1 percent of total sales. 3.12 Long distance tariffs are higher in Honduras than in any comparator country. While making a local call is very cheap in Honduras -- the cost of a 3-minute local call in 2002 was US$0.06, lower than Guatemala and Nicaragua at US$0.08 -- international tariffs are high. This is due to the fact that Hondutel, the state-owned telecom company, cross-subsidizes local calls with artificially high international tariffs.78 The tariff imbalance affects firms more than individual customers, because the former generate most of the international traffic. It is estimated that only one percent of subscribers (mostly firms) had phone bills higher than US$85 per month in 2000. However, they generated 23 percent of Hondutel's revenues.79 Neither the Government nor Hondutel have many incentives to change the tariff regime, because they are dependant on the lucrative international long distance market. 3.13 Cellular tariffs are also high -- although they have been decreasing rapidly since competition was introduced in the cellular segment. Tariffs for mobile phones are also higher in Honduras than in other countries. Until last year, the cellular market was a monopoly dominated by Celtel. As soon as Megatel, a second operator, entered the market in 2003, the price per minute of a prepaid cellular call dropped from US$0.50 to US$0.25.80 Despite recent progress, tariffs need to fall much farther in order for cellular to act as a substitute and absorb the unsatisfied demand for fixed telephone lines -- consider, by comparison, that in 2002, prepaid cellular tariffs per minute were US$0.12 in El Salvador and US$0.11 in Guatemala.81 3.14 The Internet is not a common means of communication for Honduran firms -- especially MSMEs -- and it is expensive. Fifty percent of Honduran firms use e-mail regularly when communicating with their clients and suppliers, and 22 percent use web sites to get information. The rates are lower for firms located in Olancho and the West (only 13.5 percent and 33 percent respectively use e-mail to communicate with clients and suppliers). These rates are lower than those of firms located in any other comparator country except Nicaragua--see Figure 3-4. In Honduras, as in all countries, the use of the Internet varies with firm size--the larger the firm, the more likely it is to communicate electronically (see Figure 3-5). Disparities in Internet use across firms of different sizes are deeper in Honduras than in Guatemala--11 percent of micro firms and 40 percent of small firms use the Internet in Honduras, versus 33 percent and 57 percent in Guatemala. The main causes for low Internet use in Honduras are likely to be the low penetration rates of computers,82 a lack of high technology skills among people working in MSMEs, and an underestimation on the part of entrepreneurs and managers with regard to the potential of the Internet as an effective means of communication and a conduit for process innovation. Another possible constraint on adoption of the Internet--especially among MSMEs-- is the high cost of dial-up connections and the deficient coverage of the fixed network. Hondutel, which holds a monopoly on backbone cables, has not provided the required network capacity to private companies and Internet service providers (ISPs). As a result, the Comisión Nacional de 78US$0.85 per minute to the US (CONATEL, 2003). 79ITU (2003). 80According to local estimates, mobile prices have dropped further in recent months. However, official and internationally comparable data are not available yet. 81 Prepaid cellular tariff per minute from Telemovil, El Salvador and PCS Digital, Guatemala (February, 2004). 82In 2002, there were 10.8 PCs per 1,000 inhabitants in Honduras, the same as in Nicaragua but fewer than in all other Central American countries and one fourth of the Latin American average (ITU, 2004). 58 Telecomunicaciones, CONATEL, has authorized private companies to establish their own backbone channels with the condition that they do not resell their capacity to third parties. That regulation has spawned many under-utilized private networks. Firms (especially MSMEs)83 pay for this inefficient transmission infrastructure through the high tariffs on dial-up Internet connections. The retail cost of 20 hours of dial-up Internet access at the end of 2002 was estimated to be US$40, which is the equivalent of 52 percent of the monthly GNI per capita.84 The cost for the equivalent Internet usage in Nicaragua and Guatemala was estimated at US$51 and US$31 respectively.85 Figure 3-4: Firms that use e-mail and websites to Figure 3-5: Firms that use e-mail to communicate with communicate with clients and suppliers clients and suppliers, by size--Central American countries Figure 3 ­ Uses Internet regularly to communicate with client/supplier Total Percentage of firms email website Large Nicaragua 16.6 38.5 Medium Honduras 21.8 50.2 Small Peru* 57.8 Micro Guatemala 29.2 66.4 0 20 40 60 80 100 Ecuador 55.4 83.2 Brazil 73.1 Nicaragua Honduras Guatemala 92 0 20 40 60 80 100 Source: Investment Climate Surveys, 2003 * Asked firms if they had access to the Internet. 3.15 Internet access can generate large benefits for those firms that have it. The regional regressions of the productivity analysis shows that firms with Internet access have from 11 percent to 15 percent higher productivity than firms without Internet access -- see Table 1-3 in CHAPTER 1 on page 24 above. This effect is particularly strong for old firms and is highly significant statistically in all estimated specifications. The country-by-country analysis shown in Table 1-4 in CHAPTER 1 on page 26 above indicates that the average dummy for Internet access (which is equal to the probability that a firm has Internet access) has an impact on average productivity in Honduras, which is somewhat higher than that in Nicaragua but lower than in Guatemala. This suggests that the probability of having Internet access is higher in Guatemala than in Honduras and Nicaragua -- something that is corroborated by Figure 3-4 and Figure 3-5 above. 3.16 To reduce waiting times, improve reliability, and realign the costs of telephony with international standards, the Government should reform the telecom sector by privatizing Hondutel, opening fixed telephony to competition, and rebalancing tariffs. The unreliability of telephone service in Honduras appears to be the result of Hondutel's low productivity and 83Dial up connections are used mainly by small and medium firms. 84ITU (2003). 85ITU (2003). According to local estimates, Internet costs have decreased during the last year. However, official and internationally comparable data are not available yet. 59 insufficient investments in network improvements. Honduras is one of the few countries in Latin America to still hold a state-owned monopoly to provide basic telecommunications services.86 Hondutel's productivity is one of the lowest among Latin American operators, as confirmed by the fact that there are approximately 126 fixed lines per employee. Other regional state-owned operators such as Andinatel from Ecuador and ICE from Costa Rica have approximately 195 and 210 lines per employee respectively87. Efficient private phone companies in Latin America (e.g., CTC in Chile, Telesp in Brazil and CTE in El Salvador) usually have between 250 and 300 fixed lines per employee. Hondutel has not invested in modernizing and expanding its telephone network in recent years. The network digitalization rate remains at 85 percent, whereas it is close to 100 percent in comparable countries.88 In 2002, Hondutel added only 20,000 lines and had the lowest investment per telephone subscriber in Latin America.89 3.17 The Government has already taken clear steps to initiate reform of the telecom sector. In 2003, the Government approved a sector policy whose objectives were to issue additional cellular licenses, establish a franchising scheme to provide fixed line services in partnership with Hondutel, reform sector legislation, and restart privatization of Hondutel. A second cellular license was awarded in mid-2003 to Megatel, eliminating Celtel's monopoly in the mobile market. Hondutel is also expected to receive a cellular license by 2004. Hondutel's exclusivity is due to expire in December 2005, but the Government has decided to jumpstart the process and let private parties begin providing services in 2004. In September 2003, it created a franchise program (Telefonía para Todos) by which private investors sign commercial agreements with Hondutel to provide services in areas that remain without telephone access.90 Those franchisers will be able to apply for a license once the sector is liberalized in 2005. According to Hondutel, more than 250,000 new lines are expected to be installed before the end of 2005. Given that the sector requires a new legal and regulatory framework providing clarity and investment certainty to private operators, the Government also plans to amend the current legislation. Finally, the Government expects to complete the privatization of Hondutel by 2004. 3.18 The Government's effort should be further strengthened in the coming years. Key measures to improve the effectiveness and efficiency of the Telecom Sector, which would improve firms' use of ICT and have positive implications on the country's overall investment climate, include:91 (a) terminating Hondutel's monopoly in the international calls segment and granting new licenses for new entrants; (b) opening the fixed-line segment to competition, to increase the variety of services such as voice-over-IP and broadband, and allow for competitive tariffs -- to enable development of an effective competitive framework, the Government should focus on allowing unrestricted entry to new competitors and adopting a technology-neutral service approach; (c) rebalancing tariffs to reflect the true cost of phone calls, which will reduce the costs of international calls, benefiting exporters and MSMEs; and (d) strengthening 86The ICS data suggests that firms are more likely to perceive access to telecommunications infrastructure as a constraint in countries that have not liberalized their telecommunications sector. The lower ratings for access to telecommunications infrastructure in Honduras, Ecuador, and Nicaragua coincide with state- owned monopolies. 87Data from ICE, Andinatel and Hondutel. 88ITU (2003) 89ITU (2003) 90Apertura del Sector de Telecomunicaciones: El Proyecto Telefonía Para Todos (CONATEL, 2004). Mimeo. 91See also the Honduras PRSC Matrix and the Public Infrastructure and Growth Chapter of the DPR (World Bank, 2004). Draft. 60 CONATEL, the regulatory agency, and exploring long term options for institutionalizing sector policy making. 3.19 The Government is considering promoting universal telephony schemes to increase connection rates. This is particularly important for MSMEs, especially in areas that are currently disconnected (Olancho and the West). The low telephony92 rates among MSMEs and in remote areas suggest that privatization of Hondutel and the expected sector reform are likely to be insufficient to reduce existing gaps in access to telephony -- both between urban and rural areas and between large firms and MSMEs. The Government is currently considering launching universal access initiatives funded through a development fund. This will help spread the use of telephony, especially in remote areas.93 The fund could be designed following the minimum- subsidy scheme of the Chilean Fondo de Desarrollo de Telecomunicaciones for public telephony (FDT1).94 The universal telephony program is likely to benefit firms (especially micro and small firms) that do not currently have a connection, as well as people living in the disconnected areas. 3.20 The Government could help ensure that the conditions are right for MSMEs to make productive use of computers and the Internet. These conditions are as important as (if not more important than) having an Internet connection. First, the Government could address some policy issues that would help firms take advantage of the opportunities offered by the Internet. Passing an e-commerce law, protecting intellectual property rights on the Internet, increasing online security, and introducing electronic invoicing would encourage Honduran firms to use the Internet as a business tool (e.g., to sell online) rather than just as a means of communication. Second, the Government should drive diffusion of ICT among MSMEs through direct use of ICT. OECD evidence has shown that direct use of ICT by the Government is one of the most effective ways of facilitating ICT adoption in the private sector. The Government could, for example, introduce online procurement, taxation and simplification of its agencies' internal business procedures. Third, the Government could support development of training and advisory programs focusing on integrating digital tools into MSMBs' business processes, coupled with development of online information and simplified hardware and software tools. ICT training should emphasize value added, rather than simple information search, and should help entrepreneurs integrate digital tools into their business processes.95 Training should be sector-, location- and (as much as possible) firm-specific, because general ICT training has been found to be ineffective. Private-public partnerships could also be established to promote creation of online production chains linking MSMEs to larger firms, to support MSMEs engaging in e-commerce, and to develop cluster- and industry-focused portals and market places. Finally, the Government should aim at increasing the overall ICT skills of the population, which are likely to be the most important complementary asset for the implementation of any ICT program. 3.21 The Government should assess the potential impact and sustainability of the "Ciber- Honduras" initiative before implementing it. Publicly co-financed Internet Centers should be developed cautiously and their results thoroughly evaluated before scaling up. The Honduran Council on Science and Technology (COHCIT) has preliminary plans for a "Ciber- Honduras" initiative. It consists of various projects such as developing a Government intranet, 92 While this is not assessed by the Investment Climate Survey, telephony penetration rates are low in Honduras. See ITU indicators (2004). 93 The Government is considering launching its first universal access scheme for telephony in 2005 -- World Bank, Draft Honduras PRSC Matrix (World Bank, 2004), hereafter cited as Draft PRSC Matrix. 94See B. Wellenius (2003) for a complete description and evaluation of FDT1. 95See J. Hanna (World Bank, 2003). 61 multiple official portals, and increasing access to computers. The initiative has not been actively pursued.96 The specific objectives and contents of this initiative should be reviewed and priorities assessed before implementing the plan. The plan should be evaluated in light of the likely sustainability for each proposed program, and its potential impact. If supported, provision of universal access to the Internet should be designed on the basis of successful international experience. The design and evaluation of the Internet Centers should be based on four key criteria.97 The first is private ownership and management of the Centers -- small for-profit local companies with previous experience in the IT field are ideal operators to ensure sustainability of the Centers. The Government could encourage bidding to open MSME Internet Centers by existing organizations such as Chambers of Commerce, municipalities, and broader Technology Centers offering MSMEs access to training, advisory services, technology equipment and ICT -- see the overall description of these Centers in paragraph 4.19 on page 81 below. Second, Centers should be opened in areas where there is strong demand from the local business community. Ensuring business participation in the design and implementation of the project is critical for assessing and stimulating demand. Third, ICT complementary assets should be in place before opening the Centers. Complementary assets include availability of ICT training, online local content and easy-to-use software tools matching the specific needs and means of MSMEs. Finally, the Government should ensure coordination with local business institutions (i.e., Chambers of Commerce, Chambers of MSMEs, and industrial organizations). As mentioned above, it is advisable to start these initiatives on a small scale and to evaluate their impact and sustainability before scaling them up. While they can be successful if designed and managed well,98 they can also turn into unsustainable subsidized ventures. TRANSPORT AND LOGISTICS 3.22 Logistics costs -- costs related to receiving inputs and distributing goods from the production location to the final customer -- are an important component of firms' operational expenses. The effectiveness of logistics systems depends on the conditions of all the infrastructure services upon which firms rely when transferring their goods--roads, carriage (by land, sea and air), ports, airports, border crossing facilities, and entry point controls (e.g., customs). The effectiveness of the systems also depends on how firms organize their internal supply chains--i.e., on the firms' production, distribution and inventory management. Bottlenecks at each step of the logistics system generate delays and losses to firms. The section below presents the findings of the ICS together with complementary figures on logistics systems, and preliminary policy recommendations for improving performance in these areas. 3.23 Few Honduran manufacturing firms experience transport interruptions, but when they do, they incur significant losses. Only 9 percent of Honduran firms experienced a transport interruption in 2002, significantly fewer than in comparator countries (for example, about one third of the proportion in Guatemala) -- see Table 3-4. Firms that experienced an interruption declared that they had 4 stops per year on average. Transport interruptions are defined as all shortfalls associated with the transport infrastructure (e.g., late deliveries of supplies or late shipment of goods for sale because of delays in ports, airports or the road system). The interruptions lasted about eleven hours (for a total of about 4 working days in a year) and 96Policy Note on the E-readiness for Competitiveness for Nicaragua and Honduras, World Bank, 2003 97C. Kenny and M. Motta (World Bank, 2003). 98Countries with positive experiences include Canada, UK, Denmark, Sweden, Finland and Spain. In Chile, the pilot centers opened before launching FDT2 have also been successful (Subtel, 2000). 62 generated losses of 2.9 percent of total revenues for the firms that experienced an interruption (similar to the effects in Guatemala and significantly less than in Nicaragua). Table 3-4: Incidence, length and impact of transport interruptions, international comparison (2002) Percentage of Number of times Average Losses due to Losses due to firms that the firms duration of transport transport experienced experienced transport interruption interruption transport transport interruptions (as percent of (as percent of Country interruptions interruptions /a (hrs) /a sales) /a sales) /b Honduras 9.3 4.1 10.6 2.9 0.1 Nicaragua 15.5 7.4 18.2 6.8 0.6 Guatemala 29.9 7.0 8.4 2.8 0.4 Source: Investment Climate Surveys /a : For firms that experienced an interruption. /b : For all firms. 3.24 In addition to transport interruptions, many surveyed firms experience losses of merchandise while their goods are in transport. Thirty-six percent of Honduran firms lose some of their merchandise while in transit--versus 26 percent in Nicaragua and 40 percent in Guatemala -- see Table 3-5. These losses are due to breakage and spoilage,99 which in turn are related to poor conditions of the roads and lack of an efficient logistic supply chain. Losses occurring while the merchandise was in transit amounted to 4.5 percent of total consignment value for firms that experienced a loss -- higher than in Nicaragua and in other Latin American countries. Table 3-5: Merchandise losses while in transport, international comparison (2002) Percentage of firms that Losses while merchandise Losses while merchandise experienced merchandise losses is transferred is transferred Country while in transport (as percent of sales) /a (as percent of sales) /b Honduras 36.9 4.5 1.6 Nicaragua 26.3 6.2 1.6 Guatemala 40.0 4.2 1.7 Source: Investment Climate Surveys. /a : For firms that experienced an interruption. /b : For all firms. 3.25 Econometric evidence suggests that shipment losses reduce firm productivity. Shipment losses -- defined as the fraction of sales lost during transportation due to spoilage, theft, breakage, and other deficiencies in the transport means used -- were found to exert a statistically significant negative impact upon firm productivity (see Table 1-3 in CHAPTER 1 on page 24 above). The magnitude of this negative impact ranges from 1.2 percent to 2.5 percent, i.e. a one-percent increase in the fraction of shipment losses (out of the total value of the shipment consignment) is estimated to reduce firm productivity by 1.2 percent to 2.5 percent. These effects are most relevant to small and old firms. Further individual country-by-country analysis reveals that there is no major difference in the impact of average shipment losses on average productivity across the three countries. 3.26 In sum, total losses due to inefficiencies in the transport and logistic systems are high and affect the cost of doing business. Transport is also seen by Honduran firms as a critical constraint on taking advantage of the opportunities offered by the CAFTA agreements. Total losses due to transport constraints are 1.8 percent of sales for Honduran firms, 99 A simple correlation allowed us to determine that these losses are not related to crime, but to inefficiencies in infrastructure and the supply chain. 63 about one third of total quantifiable losses due to infrastructure (see Figure 1-9 on page 21 above). Honduran firms are well aware of the fact that transport is a bottleneck for their operation. When asked what they see as the key constraints on taking advantage of CAFTA, between 16 percent and 28 percent of surveyed companies say that transport costs are a critical problem. This indicates that, especially when projected into a more competitive environment, firms see transport as a key conduit to -- or bottleneck for -- increased exports.100 3.27 A preliminary analysis of the pattern of transport losses points to possible inefficiencies in the ports and airports of the Sula Valley, and in the land transport system of the East (Olancho) and North Regions.101 The two most important factors that seem to determine transport losses are regional location and size. On the basis of these elements, we can group Honduran firms into several categories, as follows: a. The first category contains large firms located in Cortes;102 these are maquilas mainly in the apparel industry, relying heavily on ports and airports for their exports. Figure 3-6 shows that a high percentage of firms located in these areas suffer merchandise losses (45 percent of the total, versus 37 percent on average). While losses experienced by these firms are lower than the average, this is likely to be due to the high concentration of large firms, for which the incidence of losses on total revenues tends to be relatively lower (see Figure 3-7). The high percentage of firms that suffer merchandise loss in Cortes suggests possible inefficiencies in the Region's port (Puerto Cortés) and airport (San Pedro Sula), although the potential impact of frequent strikes on delaying the road traffic should also be taken into account. b. The second group includes firms located in Tegucigalpa. In the capital, firms experiencing merchandise losses, and the amount of losses, are respectively close to, and slightly higher than, the average. An analysis of transport interruptions also reveals that, in the capital, interruptions occur to more firms than average, and are of slightly longer duration, but have a limited impact on firms. These somewhat mixed results reflect the mixed composition of firms in the capital (various sizes and various sectors) as well as the mixed composition of transport systems used (roads, airports, trucks) and indicate the need for a more detailed analysis of the city's transport system to identify which elements are most constraining. c. The third category of firms is located in the East Region (the Olancho and Yoro departments) --mainly micros and small businesses that rely heavily on the road system because they don't have access to the sea or to any major airport. In this area, about one in three firms experience merchandise losses and the total value of losses is 5.5 percent of total sales. Firms located in these areas also experience a high number of interruptions (of 100Although this question relates specifically to costs, rather than transport functions in general, it is a good proxy for assessing firms' perception of the adequacy of the overall transport system in view of CAFTA. 101The Regional aggregation -- based on vicinity, type of transport systems and number of observations available -- is as follows: North Coast: Santa Barbara, Copan, Atlantida, Colon, Islas de la Bahia Cortes: Cortes Tegucigalpa (broader): Francisco Morazan and Comayagua South: El Paraíso, La Paz, Valle and Choluteca East: Olancho and Yoro 102Seventy-one percent of the 157 maquilas in Honduras are located in the San Pedro Sula Valley (including Choloma and Villanueva); the rest are distributed among Tegucigalpa, Puerto Cortés, La Ceiba, Comayaga and surroundings (Asociación Hondureña de Maquiladores, 2004). 64 longer duration than the average). The findings of this preliminary diagnostic point to possible inefficiencies in surface transportation (the road network and track freight), while the high incidence of losses is likely to be due to the small size of most firms operating in these Regions (and included in the sample) -- for these firms, transport interruptions are likely to have a proportionally higher impact on revenues. d. The fourth group contains firms located in the South. Here, firms are mixed (various sizes and sectors). They rely mainly on the road network and (in part) on the San Lorenzo Port. In this area, 30 percent of firms experience losses while merchandise is in transport, and losses are close to the average. The combination of incidence and amount of transport and losses is better than in the rest of the country. It is also interesting to note that surveyed firms located in these areas did not report any transport interruptions. This would point to the fact that the land transport system, the San Lorenzo Port, the border crossing facilities with Nicaragua, and other elements of the supply chain are relatively more developed in the South than in the rest of the country. e. Finally, the fifth group includes firms located in the North and the Coast areas (excluding Cortes).103 These are mainly micro and small businesses, which are likely to rely heavily on the road networks -- and possibly on ports. The percentage of firms that suffer merchandise losses in these areas is relatively low (one in four firms), but the impact on sales is very high (9.1 percent of total sales). Even more than in the East Region, this is likely to be directly related to the (very small) size of the firms. As shown in Figure 3-7, only a small percentage of micro firms are affected by merchandise losses (23 percent) -- this is related to the fact that micros tend to use transport systems less, because they often buy and sell in the areas where they are located. However, when micros do incur losses, their amount is high (6.2 percent). Unreliability of the surface transportation system, inefficiencies at the border crossing with Guatemala and possible port inefficiencies (some firms located in the North Coast rely on Puerto Cortes) are likely to be the main sources of infrastructure inefficiencies. Figure 3-6: Percentage of firms that lost merchandise Figure 3-7: Percentage of firms that lost while in transport and loss suffered. By Department. merchandise while in transport and loss suffered. Honduras. By Size. Honduras. Losses (% of sales) Losses (% of sales) For firms that had a loss For firms that had a loss Total 4.5 Total 4.5 North Coast 9.1 Large 2.5 South 4.9 Medium 3.7 East 5.5 Small 5.7 Teguc 5.3 Cortez 2.8 Micro 6.2 0% 10% 20% 30% 40% 50% 0% 10% 20% 30% 40% 50% Source: Investment Climate Survey 103See footnote 101. 65 Figure 3-8: Map of Honduras 3.28 The findings of this preliminary analysis are confirmed by the results of a more direct ICS question on key transport bottlenecks to take advantage of the opportunities offered by CAFTA. Figure 3-9 shows that, while all firms rate airports as the main potential constraint in view of CAFTA (clearly related to increased expectations for export opportunities), firms located in the free zone and in Cortes -- large maquilas -- rate both airports and ports as more important than do other firms. On the other hand, for firms located in the East and in the North, as well as for micros, surface transportation is a key constraint. For firms located in the South, on the Coast and in Tegucigalpa, transport costs do not appear to be a main constraint for CAFTA, with the exception of transport by air in the capital, which does appear as a bottleneck. Figure 3-9: Transport cost perceived as a major constraint to CAFTA tniart 45% onsc 40% orjam 35% 30% ngiralc 25% 20% desmrif 15% 10% 5% of 0% % Total free zonefree zone ast th cro all Co Cortez EastNorth Sou Teguc Mi Sm diumLarge Me no Air transport costs Land transport costs Maritime transport costs Source: Honduras Investment Climate Survey. 3.29 Firms operating in the perishable industries are the ones that suffer most from losses while merchandise is in transport. Firms in the food and beverage industries suffered merchandise losses of 6.9 percent and 5.8 percent respectively -- see Figure 3-10. This suggests 66 the need to better assess the impact of each infrastructure bottleneck along the supply chain for these specific industries -- including warehousing, distribution centers, and conditions of land, air and sea fleet -- and to devise specific measures to reduce them. Development of an effective "cold supply chain" would be particularly important to ensure better preservation and increased exports of these goods. Figure 3-10: Losses while merchandise is transferred (percent of consignment value) -- for firms that experienced a loss. Total 4.5 Food and tobacco 6.9 * Beverage 5.8 Non metalic minerals 4.8 Metal Products 4.6 Furniture and Wood 3.6 Chemical and Plastic 3.5 Apparel 2.1 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 * Forty one firms included in the sample, of which only 3 are tobacco firms. Source: Honduras Investment Climate Survey. 3.30 Customs clearance time is adequate for exports, but high for imports, especially for small firms. Average customs clearance time is adequate for exports (2 days), with relatively low extreme values (3 days); nonetheless, these values are more than double for small firms. Clearance time is somewhat higher for imports (5 days) with pronounced peaks (up to 27 days), severely affecting small companies, particularly in Tegucigalpa and the Pacific coast, as well as non-exporting firms outside the free zone. Average time is reasonable when compared with other countries (8 days in China, 7 in Korea and Thailand); however, the longest clearance time is too large for small firms (27 days in Honduras, versus 9 days in China's main ports, and 21 days in India). The longer time for imports is due to the greater severity of incoming product controls, as well as the imbalance in flow volumes, imports being twice the volume of exports at the national level. Figure 3-11: Customs clearance time for imports (days, mean values) 30 25 20 15 10 5 0 TotalMicro SmMediumLarge all ast tez DomestiForeign free fre c r Co Co East NorthSouthTeguc non-exportexport ers ers zonee zone no avg period (mean) longest period (mean) Source: Honduras Investment Climate Survey 67 3.31 The econometric analysis shows that more days needed to clear customs reduce productivity. The regional analysis of firm productivity across Honduras, Nicaragua and Guatemala -- shown in Table 1-3 in CHAPTER 1 on page 24 above -- establishes that the number of days necessary for firm imports to clear customs exerts a negative and statistically significant impact on firm productivity, controlling for firm characteristics, investment-climate- related variables and industry, country and year effects. A one-percent increase in the average number of days necessary to clear customs is estimated to decrease productivity by 0.1 percent. These results are particularly relevant for young and small firms. The country-by-country analyses of average productivity -- shown in Table 1-4 in CHAPTER 1 on page 26 above -- suggests that the average number of days to clear customs in Honduras affects average productivity with an impact of a similar magnitude (in absolute terms) to the same effect for Nicaragua, but lower than in Guatemala. Therefore, days to clear customs appear longer on average in Guatemala than in Honduras and Nicaragua. 3.32 Increasing private sector participation in Puerto Cortés is necessary to increase the capacity and efficiency of the Port and the overall maritime transport conditions of the Sula Valley. Puerto Cortés is Honduras' dominant national port, moving 6 million tons and 350,000 TEUs per year.104 It is located in the Sula Valley, serving mainly the maquila, agriculture and forestry sectors. It is critical to trade with the US105 and is becoming a Regional hub -- already second only to Puerto Limon in Costa Rica.106 The high percentage of firms experiencing merchandise losses in this area (see Figure 3-6) may be a reflection of Puerto Cortés' inefficiencies. In fact, the performance of the Port is below international standards. It has long waiting times and low productivity in transferring cargo (especially dry bulk cargo). Total costs incurred by users are also significantly higher than international norms because of the high cost to the users of the long waiting times--see Table 3-6. The port is currently under a concessioning process aimed at increasing efficiency and expanding capacity.107 The Government plan includes certification of the Port in accordance with international standards and approval of a Ley General de Puertos, which, among other things, will enable private sector participation in the management of port.108 The minimal requirements for satisfactory implementation of the plan include expanding the Port through private financing, outsourcing functional services, and transitioning one terminal to private cargo management.109 To develop the port as a regional hub, discrimination of freight coming from (or going to) neighbor countries should be eliminated. A Center for Logistic Activities should also be developed in the Port area,110 including logistics value added services such as warehousing, packaging and quality control, container repairs, a truck center, customs, and financial agencies. This will promote customer fidelity and attract 104One TEU is equivalent to a 20 foot container. 105Puerto Cortés handles sixty-one percent of Honduras' exports and 54 percent of its imports, mostly to and from the East Coast and the Gulf. 106Demand is growing at between six and seven percent per year (World Bank, 2004). 107Puerto Cortés is expected to be awarded to a private concessionaire by February, 2005. A law should be passed by mid 2004. 108It is estimated that investments of US$80 million are needed in Puerto Cortés in the next five years for a new container terminal and to improve the handling of dry bulk cargo. Draft PRSC Matrix (see note 93) and PPIAF (2002). 109See a more detailed description of recommendations for reform of Puerto Cortés in PPIAF (2002), pp. 76-79. 110The Logistic Center is expected to be included within the port concession. 68 more volume, reducing unit costs and inducing the consolidation of production and distribution hubs. Private participation should also be encouraged in other Honduran ports.111 Table 3-6: Productivity indicators for Puerto Cortés compared with international standards (1999) Attribute and Category International Standard Puerto Cortés Waiting time of ships ( percent of time in the port) Container ships <5 percent 22 percent Dry bulk vessels <20 percent 26 percent Cargo Volume handled (metric tons per hour at dock) General cargo 89 24-55 Dry Cargo in Bulk 1000 89 TEUs per hour /a 14 25 Tons per dock meter per year General cargo 800-1000 1658 TEUs /a 8000-10000 1383 Unit cost for cargo management (TEUs /a) $120-$180 $135 Margin over sales >8 percent 26.6 percent Source: World Bank, 2002, PIAF Study: Honduras report on infrastructure policy. /a: One TEU is equivalent to a 20-foot container 3.33 The shortcomings of the concession of the San Pedro Sula airport must be overcome to ensure that it can support the expected increase in air traffic in the Sula Valley following CAFTA. An in-depth study on the overall efficiency of Honduran airports,112 especially San Pedro Sula and Tegucigalpa, should be carried out to identify adjustments needed in view of CAFTA. As shown in paragraph 3.27(a), maquilas located in the Sula Valley have a high incidence of merchandise losses and transport interruptions, and they rate the cost of air transport as one of the main constraints on exploiting CAFTA opportunities. Honduras' most important airport for freight movement is San Pedro Sula. The airport is mainly linked to the maquilas, which operate primarily with freight airplanes. While the evidence from the diagnostic is less clear on the effects of possible inefficiencies of the Toncontín airport (in Tegucigalpa), firms located in the capital also rate improvement in air transport as key. Studies show that the principal problems of Toncontín are its short landing strip, lack of electronic instruments for landing control, limited passenger terminal capacity, high fares caused by low volumes, lack of competition (the market is dominated by TACA), and lack of effective mechanisms for consumer protection.113 The San Pedro Sula, Tegucigalpa and other international airports have been given in concession during 2000. The concessionaire shows some delays in expanding capacity and charges high fees at fiscal warehouses. The shortcomings of the current concessions should be 111The other Honduran ports are Castilla, Tela, La Ceiba, Roatán (Pacific Coast) and San Lorenzo (Atlantic Coast). The total cargo for all Ports in Honduras in 2002 was 7,803 tons, of which 82 percent was handled by Puerto Cortés, 10 percent by San Lorenzo, and 7 percent by Castilla. La Ceiba and Roatán mainly support tourism (Comisión Centroamericana de Transporte Marítimo, 2003). 112 Honduras has four international airports: Tegucigalpa, San Pedro Sula, La Ceiba and Roatán, which have experienced a 9 percent increase in passengers per year between 1995 and 2000. The San Pedro Sula airport is experiencing increases in cargo of 10 percent per year due to the growth of the maquila industry. 113PPIAF (2002). 69 overcome by increasing control over the private operators for the completion of their investment programs and the regulation of their fares.114 3.34 The land transportation system -- including highways, the trucking industry, border crossings and customs -- could be improved, especially in areas where firms seem to suffer most from transport interruptions (the East, the North and in the capital). Road security should also be strengthened. The (relative) unreliability of surface transportation in Honduras is due to four main elements: (1) deficiencies in the road infrastructure (often related to lack of maintenance), (2) inefficiencies related to the trucking fleets and their operational organization, (3) inefficiencies at border crossings and (4) insufficient road security. First, the condition of the Honduras highways is relatively weak (40-40-20 good, fair and bad), with up to 65 percent in bad condition in tertiary gravel roads, particularly relevant for agricultural activity. Road conditions have improved in recent years, thanks to increased funding and better maintenance. However, the damaging impact of the Mitch hurricane in 1998 and the deterioration of urban roads (maintained by local municipalities) are still serious issues. Also, studies show that highways need maintenance and that development of the road network in the country has been unbalanced. In the future, the Government should keep improving maintenance and rehabilitation of existing roads,115 ensuring the transfer of established funds to the Fondo Vial. Concessioning some main highways might also help reduce the burden on public resources and improve efficiency. It is important to note that the Government is currently making an effort to improve the road that connects the south of the country with Puerto Cortes, which will be essential to ensure increases in traffic to the Port both from Honduras and from the neighboring countries. 3.35 A second cause of the (relatively) poor reliability of the Honduran surface transportation system may be its old trucking fleet. About 65,000 trucks use the roads in the country, and their industry organization is weak, resulting in inadequate service and high prices. It is an aged truck fleet, with inefficient operations, and extended waiting times in terminals. There are no regulations--at least none that are enforced--on axle loads, driver working times and conditions, or vehicle conditions. The lack of regulatory oversight generates significant negative externalities, such as accidents and road destruction. Trucking industry regulations should be reviewed, looking for long-term efficiency and sustainability, encouraging professionalism and the cultivation of modern logistics operators. 3.36 A third set of problems in land transportation are the significant delays and smuggling occurring at border crossings. Twenty-one percent of Honduras' imports and 17 percent of its imports are regional trade,116 moved through the country's border crossings--one with El Salvador (El Amatillo), three with Nicaragua (Guasaule being the most important) and one with Guatemala (close to San Pedro Sula). Delays should be specifically assessed for each border crossing and measures taken to reduce the problems. Finally, improving the land transport system requires better security, to reduce thefts and smuggling. Truck robbery is a serious problem, particularly close to cities, which may be affecting small firms. Port security should be improved to ensure compliance with regulations adopted by US Customs after 9/11/2001, by the maritime operators, and by the WCO, leading to greater control of exports to the USA. Finally, to facilitate border crossing and ensure that customs continues to operate in an efficient manner, automatic, paperless, non-discretionary processes should be implemented at the border crossings, 114The concession contracts have recently been renegotiated, rescheduling investments. 115The Draft PRSC Matrix (see note 93) includes adequate maintenance of at least 60 percent of the total road network by March 2005, and 70 percent of the total road network by March 2006, as key objectives. 116According to some analysts, these estimates might be low. 70 and the border agencies should adopt statistical systems harmonized with the neighboring countries. 3.37 A more in depth study of the bottlenecks faced by Honduran firms along the supply chain would help identify the exact logistical bottlenecks firms face -- in turn supporting more targeted policy recommendations. While the ICA survey provides some first indications on problems experienced by Honduran firms in relation to the transport system, more in-depth studies, preferably targeting specific industries in selected regions, would help identify bottlenecks along the supply chain. Logistical constraints for companies operating in the food industry, for example, are different from those faced by maquilas or firms operating in the wood and forestry sector. Literature on logistical constraints on firm growth is very limited in general and practically non existent in Honduras. To better assess the specific impact of each bottleneck along the chain, the World Bank is currently financing a supply chain study in Honduras (with a focus on the apparel, sea food, coffee and tourism industries). 71 CHAPTER 4. QUALITY, INNOVATION AND LABOR SKILLS117 4.1 Adopting international quality standards, encouraging technological innovation, and improving labor force skills are interrelated measures that are critical to increasing firms' productivity and growth. As the report Closing the Gap in Education and Technology has shown, productivity differences between countries and between firms within countries are profoundly affected by differences in levels of innovation, access to -- and capacity to adapt -- foreign technology, and the availability of a highly skilled labor force.118 In light of the now widely acknowledged relevance of these factors, the Honduras investment climate survey dedicated a module to assessing how Honduran companies perform in these areas. The section is divided in three parts: (1) use of quality systems -- which is considered a first step to help companies improve their processes and better integrate with the local and international economy; (2) innovation and technology adoption; and (3) training systems and labor skills. For each section, the report presents the results of the ICS, together with complementary evidence from existing studies and some policy recommendations. The policy recommendations draw from various background papers prepared for the Investment Climate Assessment and the World Bank Honduras Trade Facilitation and Productivity Enhancement Project (2004), from the FIAS Report on Reform of Labor Skills Training (2004) and from inputs to the Honduras Development Policy Review (2003). THE USE OF QUALITY SYSTEMS 4.2 Very few Honduran firms, and almost no MSMEs, have ISO 9000 or ISO 14000 certifications, and there are no ISO certifiers in the country. ISO norms are management system standards that state what an organization must do to manage processes influencing quality (ISO 9000) and processes influencing the impact of the firm's activities on the environment (ISO 14000).119 Adopting ISO standards allows firms to standardize and improve the quality and efficiency of their production and distribution procedures. The ICA survey found that only about 5 percent of Honduran firms surveyed are ISO-certified, a similar percentage as in Nicaragua and Guatemala, but ten times lower than in China--see Figure 4-1. In Honduras, large firms and firms operating in the apparel industry do somewhat better--about 1 in 8 large firms and 1 in 10 firms operating in the apparel industry has an ISO certificate -- see Figure 4-2. However, ISO certification is practically unknown among Honduran MSMEs and firms operating in some key sectors of the Honduran economy (e.g., furniture and wood, ceramics, glass, and metal 117 This Chapter was prepared by Marialisa Motta. Stefka Slavova summarized the results of the econometric analysis. Valuable inputs and advice were provided by Paulo Correa and Isabel Sanchez Garcia, Diane Thompson and Julio Fuster (background papers) and Miguel Crivelli (research assistance). The labor section draws heavily from a FIAS paper prepared by Geeta Batra, Honduras: Reform of Labor Skills Training to Strengthen the Competitiveness of the Honduran Private Sector -- Internal Draft. (FIAS, International Finance Corporation and World Bank, 2004). 118For an in-depth discussion of the role of technology, innovation and skills on countries' and firms' productivity, refer to the above-mentioned report (World Bank, 2004), which addresses in detail not only the role that each factor plays but also the interactions among the various factors. Further evidence for the impact of these factors on the productivity of Honduran firms will be provided in the next version of this study -- when the results of the productivity analysis will be included. 119See a detailed description of ISO management system standards at: http://www.iso.ch/iso/en/iso9000-14000/basics/general/basics_1.html 72 products)120 -- less than 6 percent of MSMEs and firms operating in these sectors have an ISO certificate. See Figure 4-2 and Figure 4-3. Currently, there are no ISO certifiers in Honduras. Thus, to be ISO-certified, Honduran firms must rely on foreign institutions, which increases their certification costs and may prevent them from pursuing certification. Figure 4-1: ISO certified firms -- international comparison. Percentage values. China 50 Malaysia 31 Brazil 19 Pakistan 17 Ecuador 17 Poland 9 Honduras 5 Nicaragua 4 Guatemala 3 0 10 20 30 40 50 60 Source: Investment Climate Surveys Figure 4-2: ISO Certified firms in Honduras by size. Figure 4-3: ISO Certified firms in Honduras by sector. Percentage values. Percentage values. Total 5 Total 5 Apparel 10 Large 13 Chemical and Rub 6 Med 6 Food and Tobacco 6 Metal Products 3 Small 3 Non Metalic Minerals 2 Micro 2 Furniture and Wood 2 0 2 0 2 4 6 8 10 12 14 4 6 8 10 12 Source: Investment Climate Surveys 4.3 Adoption of other quality standards is more common in Honduras, but still rarer than in comparator countries. Quality standards other than ISO (including product- and industry-specific standards) are more common in Honduras than ISO certification.121 About 18 percent of Honduran firms hold a quality standard certificate, but even these are still less widespread in Honduras than in Nicaragua or in other comparator countries -- see Figure 4-4. As with ISO certificates, other quality certifications are used mostly by large firms and firms operating in the apparel industry -- 42 percent of large firms and 36 percent of firms in the 120Complementary studies (ISO Survey of ISO 9000 and ISO 14000 Certificates) show that, out of more than two thousand registered local firms in Honduras, only twenty have an ISO 9000 certification, and the majority of them are local branches of multinational corporations that have mandated ISO 9000 certification. 121The most common certificates are those received by clients (21 percent of the total), followed by WRANP -- Worldwide Responsible Apparel Production (13 percent of the total), Hazard Analysis Critical Control Point Systems (HACCP, 7 percent of the total) and Good Manufacturing Practice (GMP, 6 percent of the total). 73 apparel sector have a quality certificate -- see Figure 4-5 and Figure 4-6. This shows a duality in the Honduras manufacturing sector between the apparel industry, which is more advanced than in other Central American countries and mostly composed of large maquila firms that are relatively better integrated with the global economy, and all other industries, which appear to be significantly less aware of the importance of quality certification. Figure 4-4: Firms using quality standards other than ISO -- international comparison. Percentage values. China 29 Ecuador 21 Nicaragua 20 Honduras 17 Guatemala 12 0 5 10 15 20 25 30 35 Source: Investment Climate Surveys Figure 4-5: Honduran firms using quality standards Figure 4-6: Honduran firms using quality standards other than ISO, by size. Percentage values. other than ISO, by sector. Percentage values. Total 17 Total 17 Apparel 36 Large 42 Chemical and Rub 18 Med 18 Food and Tobacco 18 Metal Products 13 Small 10 Furniture and Wood 4 Micro 2 Non Metallic Minerals 4 0 10 20 30 40 50 0 5 10 15 20 25 30 35 40 Source: Investment Climate Surveys 4.4 Quality certification is likely to facilitate exports and improve the efficiency of business processes. Nevertheless, most Honduran firms seem to perceive quality as a cost rather than an investment that would increase their export potential. By obtaining a quality standard certificate (ISO or other), Honduran firms can be recognized by international companies as reliable partners. Such firms can then more easily sell their products abroad, either as inputs to local or foreign firms' production processes, or directly to final customers. The relatively small number of certified firms in Honduras indicates that local entrepreneurs are either not aware of the potential returns of quality certification or consider them too costly. This suggests the need to provide specific incentives to facilitate adoption of quality systems, especially among MSMEs. 4.5 There is some econometric evidence that ISO certification is associated with higher productivity, but the results are sensitive to the productivity measure used. The regression analysis of the impact of investment-climate variables on productivity indicates that in the regional regressions a dummy variable for having ISO certification is associated with higher levels of productivity, however the result is significant in only one out of six alternative definitions of productivity (see Table 1-3 in CHAPTER 1 on page 24 above). The effect is as 74 follows: obtaining an ISO certification is estimated to raise firm productivity from 2.4 percent to 17.6 percent. This result is particularly relevant for large firms. However, the result is merely suggestive due to the lack of significance in five out of six estimations. Expanding the definition of the quality certification variable to include not only ISO, but also other types of certification, does not change the results. There are no major differences between the individual country results -- the contribution of the probability that a Honduran firm is ISO-certified to Honduras' average firm productivity is of a similar magnitude to those in Guatemala and Nicaragua (see Table 1-4 in CHAPTER 1 on page 26 above). 4.6 Limited laboratory accreditation increases firms' costs and delivery times. Because Honduras does not have a local Accreditation Authority for accrediting laboratories to international standards, local labs must be accredited by an internationally-recognized global authority. It is estimated that only three of the forty existing government and private labs that provide testing services in the country -- mainly for food products for the domestic and export markets -- are accredited by the US Department of Agriculture. Because all three accredited laboratories are in Tegucigalpa, producers must ship their samples to the capital for testing and wait for results. This increases producers' costs and delivery times, and creates disincentives to sending samples to the labs, affecting firms' export potentials. 4.7 To increase demand for quality certification and accreditation, the Government is considering financing technical assistance and matching grant programs for local firms, especially MSMEs, and laboratories. Preliminary assessments show that Honduran firms need training on business management, operations, and quality procedures. Laboratories need capacity building on calibration, management, and current analytical methods, as well as specific training to meet the requirements of ISO/IEC 17025, the international standard that is the basis for laboratory accreditation. The Government, with support from the World Bank, has recently approved a program to provide technical assistance and matching grants to MSMEs and non- accredited laboratories -- especially in areas with a high concentration of firms, to reduce companies' testing costs. Technical assistance and matching grant programs will be complemented by quality promotional campaigns targeting the management of MSMEs and labs, as well as trade associations and key local industry leaders. 4.8 To improve the supply of Quality Systems,122 Honduras should develop a well- functioning Quality Infrastructure, including a National Quality Council, an internationally recognized Accreditation Authority, institutes that can certify firms according to ISO and other technical standards, and an efficient National Quality Information Service. These initiatives are currently in the initial stages of development.123 Standards development in Honduras is coordinated by Consejo Hondureño de Ciencia y Tecnología (COHCIT), which is supported by the Comisión Interinstitucional de Normalización (CIN).124 COHCIT's lack of resources and diverse responsibilities prevent it from fully performing its quality systems functions. The development of a separate Quality Council is an important prerequisite for spreading the use of Quality Systems among local firms and labs. The Council should be lean, i.e., composed of a few experts that can define quality standards and coordinate the Quality System Infrastructure -- Figure D-1 on page 152 below shows the framework of the typical Quality Councils of more advanced countries as a reference. The Council's main function would 122See Box D-1 on page 151 for a more detailed description of various components of a typical National Quality System. 123These initiatives will be sponsored through the Trade Facilitation and Productivity Enhancement Credit. 124This institution reports directly to the President of Honduras. 75 be to oversee the development and implementation of the National Quality System and to act as the focal point for all standards and conformity-assessment activities, including the creation of sector-specific quality norms, especially in industries in which Honduras has comparative advantages (e.g., apparel and tourism). 4.9 Creation of an internationally recognized Accreditation Authority125 and development of ISO certifiers will increase the efficiency of accreditation and certification services in Honduras. The creation of a Honduran Accreditation Authority will ensure that local laboratories and firms are not obliged to turn to foreign authorities to get accredited and certified. To build the credibility needed to become internationally recognized, the newly created Honduran Authority should be granted an interim period during which it would review the performance of Honduran laboratories and firms in conjunction with an internationally recognized foreign organization. To create a network of Honduran ISO certifiers -- as mentioned in paragraph 4.2 above, there are none in Honduras -- the Government should consider strengthening local organizations that are providing various non-ISO quality certifications, facilitating their transition into ISO-certified institutions. Finally, the Government could consider providing incentives to create new certifying institutions (for ISO and other technical standards). 4.10 The National Quality Information System should be rationalized and strengthened. Information on standards, technical regulations, and conformity assessment programs in Honduras is currently not readily available because it is scattered across many government departments. This threatens the country's ability to comply with the requirements of its trading partners and makes it more difficult and costly for local firms to collect information on export requirements, as well as for foreign firms to understand the requirements for selling their goods and services in Honduras. To improve the quality of local information services, the Council should consider carrying out a thorough assessment of existing sources of information on quality -- e.g., the existing Technical Barriers to Trade Enquiry Points at the Ministries of Trade and Health, and the information center of Consejo Nacional de Ciencia y Tecnología. In addition, it should ensure that the staffs of local information service centers are properly trained and that the centers subscribe to international databases providing information on quality. TECHNOLOGY ADOPTION AND INNOVATION 4.11 While Honduras is far from the technological frontier, the ICS demonstrates a certain level of innovation among Honduran firms, which upgrade and renovate their product offerings in order to remain competitive, at least in their local markets. The fact that Honduras is far from the technological frontier is confirmed by a variety of measures.126 Honduras' level of research and development, for example, is one of the lowest in Latin and Central American (R&D was 0.05 percent of GDP in Honduras in 2000, versus 0.1 percent in Nicaragua and 0.58 percent on average in Latin America).127 The very small number of patent 125Accreditation Authorities are responsible for accrediting institutions that in turn review and guarantee the quality of laboratories (accreditation) and firms (certification). 126The World Economic Forum ranks Honduras 77th among 102 countries on Innovation. Honduras ranks slightly higher than Guatemala (79th) and Nicaragua (81st), but lower than most other comparator countries -- Ecuador is ranked 72nd, Costa Rica 61st and Bolivia 52nd. Key indicators used by the WEF to assess innovation include: R&D spending, protection of intellectual property rights, quality of scientific institutions, collaboration between firms and scientific institutions, number of patents per capita, average years of patent renewal, and number of countries in which the patents are applied for. The data are collected based on a mix of qualitative and quantitative questions (World Economic Forum, 2004). 127 Red Iberoamericana de Indicadores de Ciencia y Tecnología (2004). Government-financed R&D 76 applications filed in the US (the ratio of patents filed per 100,000 inhabitants in the last ten years was 0.17 in Honduras, slightly higher than the ratio in Guatemala, but lower than in other more developed Latin American countries -- Mexico had 0.61 patents per 100,000 inhabitants and Chile 0.68 patents per 100,000 inhabitants, for example)128 also indicates that Honduras' ability to innovate is very limited. The fact that Honduras does not invest in R&D and does not develop cutting-edge research and technologies is expected, to a certain extent, given its low income level. In that light, the ICS finding that more than 45 percent of Honduran firms introduce some sort of innovation -- by changing processes, upgrading existing product lines and introducing new products (see Figure 4-7) -- is positively surprising. Only Brazilian companies and, to a lesser extent, Ecuadorian companies, appear more dynamic than Honduran companies across all dimensions of innovation. Although these figures display a certain amount of innovation among Honduran firms, the scope of the product or process innovations introduced does not seem to be broad enough to significantly improve the firms' competitiveness, nor to ensure that they successfully participate in international markets.129 Figure 4-7: Innovating firms -- international comparison Brazil Ecuador Nicaragua Guatemala Honduras Bolivia 0 20 40 60 80 100 Process change New product line Upgrade an existing product line Source: Investment Climate Surveys 4.12 The econometric analysis demonstrates that technology adoption -- measured by the share of firm employees dedicated to R&D -- affects firm productivity positively and significantly. The fraction of total firm employees engaged in R&D raises firm productivity and is statistically significant in all six regression specifications. The magnitude of the impact is as follows: increasing the fraction of total staff engaged in R&D activities by 1 percent is estimated to raise firm productivity between 0.6 percent and 0.7 percent. This impact is especially strong for old firms and for small firms. The country-by-country analysis indicates that the positive projects are also widely dispersed and mostly financing basic R&D. Private sector financed R&D projects are a small proportion of the total (5 percent in 2000). Finally, more than one third of Government financed projects are also executed by Government organizations, leaving little role for the private sector. COHCIT (2004). 128World Development Indicators (2004). COHCIT indicates that 101 patents were filed in Honduras in 2000, of which 70 were granted (versus 96 granted in Guatemala in the same year). Of the patents filed in Honduras, 66 percent (67) were filed by foreigners, and most (66) were in industrial design. J. Fuster (2004). 129Exports in Honduras are low compared to other Central American countries. In the manufacturing sector, exports amount to 5.7 percent of GDP, versus 11.4 percent in El Salvador and 19 percent in Costa Rica, for example. In Central America, only Nicaragua and Guatemala perform worse than Honduras (their exports are 4.5 percent and 3 percent of GDP respectively). 77 contribution to productivity of the average fraction of total staff engaged in R&D is similar across Honduras, Guatemala and Nicaragua. 4.13 Sixty-seven percent of surveyed Honduran firms adopted a technological innovation in the past three years.130 Acquisition of new machinery or equipment is the most important channel for technological introduction. As in many low-income countries similar to Honduras, technology transfer and innovation mainly take place by incorporating new capital equipment into the production process -- see Figure 4-8. These results are consistent with the high level of capital good imports as a percentage of GDP observed in Honduras in the period 1997-2000.131 Other relevant channels for innovation include employment of key personnel and adaptation of existing technology within the firm -- the latter being less relevant in Honduras (and other Latin American countries) than in the East Asian economies. This pattern of innovation is common across all firm sizes and all sectors. Figure 4-8: Technology adoption, international Figure 4-9: Firms with computerized equipment132 comparison 90 80 Nicaragua 70 60 50 40 Guatemala 30 20 10 0 uador Peru Brazil Honduras h China as ala em Ec Banglades ndur Ho NicaraguaGuat 0% 5% 10% 15% 20% 25% 30% New equipment Hiring Developed or adapted within the firm Licensing/turnkey ops Source: Investment Climate Surveys 4.14 Even though Honduran firms claim to innovate by buying new equipment, the technological content of such equipment appears low. While approximately 50 percent of all firms surveyed133 claim to innovate by acquiring new equipment, a substantial number of 130 Honduras' (relatively) high levels of technology transfer are confirmed by the findings of the World Economic Forum, which ranks Honduras 53rd among 102 countries on Technology Transfer, significantly better than on Innovation (77th rank -- see footnote 126). Honduras' Technology Transfer rank is also better than several comparator countries: Guatemala (54th), Nicaragua (57th), Ecuador (62nd) and Bolivia (68th). Key elements used by the WEF to assess technology transfer include relevance of FDI as a source of technology transfer and licensing of foreign technology as a common way to acquire technologies, but these rankings are based on qualitative questions only (World Economic Forum, 2004). On Honduras' levels of technology adoption, see also: P. Correa A Preliminary Assessment of the Honduras National Innovation System (World Bank, 2004). 131 In 2001, imports of capital goods in Honduras were 15 percent of GDP, higher than most Central American countries except Nicaragua. 132Data for international comparators are not available. 133 This is obtained multiplying 67.3 percent (Honduran firms that introduce innovation) by 80 percent (firms that innovate by introducing new equipment). Note that Figure 4-7 shows simple percentages only (i.e., 67 percent for Honduras), because the question "Did you introduce an innovation or not in the last three years?" was not asked in all countries. 78 Honduran firms (79 percent) do not own computerized machines (see Figure 4-9). This proportion is higher for firms operating in the wood and the non-metallic mineral134 industries (91 percent and 90 percent respectively). Chemical companies and exporting firms (in any sector) perform better -- only 64 percent and 62 percent respectively do not have computerized equipment. While the type of equipment used--and its level of technological sophistication -- is highly dependent on the characteristics of each industry, Honduran firms' overall low use of advanced machines may account for low productivity levels and further hinder innovative behavior. 4.15 Indeed -- according to the productivity analysis -- the share of firm machinery which is computer-controlled affects firm productivity positively, albeit not always significantly. Increasing the fraction of computer-controlled production equipment (out of total equipment) by 1 percent is expected to raise firm productivity by 0.1 percent, but the significance of this result is not robust to different definitions of productivity --see Table 1-3 in CHAPTER 1 on page 24 above. Out of six estimations shown in Table 1-3 the estimated coefficient is found significant only once -- using the restricted OLS estimations with a Translog form of the production function. The result is particularly relevant for old and large firms. There are no differences in the estimated impact of the average fraction of computer-controlled plant equipment on average productivity across the three countries (see Table 1-4 in CHAPTER 1 on page 26 above). Furthermore, the individual contributions toward average productivity of the fraction of computer-controlled equipment are small in all three countries. 4.16 Technology licensing is not a common channel for acquiring new technologies. Technology licensing, which is a common means of innovation in East Asian economies, is not widespread among Honduran firms. Only 3.3 percent of Honduran companies include technology licensing as one of the three most important channels of innovation, versus 4 percent in Guatemala, 7.8 percent in Ecuador and 24 percent in China -- see Figure 4-8.135 In fact, in response to a more direct question, only 15 percent of Honduran firms reported licensing foreign technologies, versus 19.6 percent in Guatemala.136 Most of the firms that license foreign technologies are large maquilas operating in the apparel sector (29 percent) and large firms operating in the chemical sector (24 percent) -- see Figure 4-10 and Figure 4-11. Licensing or turnkey operations from domestic sources are also uncommon. 134This includes ceramics, glass and cement. 135Note that here we report simple percentages only to allow for international comparison -- see note 14. 136Figures for other countries are not available. 79 Figure 4-10: Honduran firms licensing foreign Figure 4-11: Honduran firms licensing foreign technologies, by size. Percentage values. technologies, by sector. Percentage values. Total 15 Total 15 Non Metalic Mine 0 Large 34 Metal Products 7 Med 17 Furniture and Wo 7 Small 12 Food and Tobacco 17 Micro 2 Chemical and Rub 24 Apparel 0 10 20 30 40 29 Source: Investment Climate Surveys 4.17 Firms' linkages -- with clients or suppliers, industry associations or R&D institutions -- play a marginal role for technology transfer and innovation in Honduras. Figure 4-12 shows that collaboration with parent companies, universities, clients and suppliers, trade associations or consultants are not key sources of innovation in Honduras (similar to Nicaragua and Guatemala). In fact, the sum of all these possible sources of innovation for Honduran firms is less than innovation acquired by buying new equipment. Honduras would benefit from policies strengthening linkages. In fact, as existing literature shows, linkages have played a key role in development of applied research and innovation in more advanced Regions, such as East Asia and Europe (particularly in the north).137 Figure 4-12: The limited role of linkages in technology adoption--Central American countries 90 80 70 60 50 40 30 20 10 0 Honduras Nicaragua Guatemala New equipment Equipment suppliers Trade Fairs/Study Tours In cooperation with clients Transferred from parent company Consultants Business/industryassoc Universities, public institutions Source: Investment Climate Surveys 4.18 In sum, there seems to be a mismatch between Honduran firms' apparent dynamism in introducing innovation and buying new equipment, on the one hand, and their low levels of productivity and sales, on the other.138 This is probably due to unsophisticated 137Latin America Region of the World Bank. Reducing the Gap in Technology and Innovation (World Bank, 2004). 138See footnote 129 on exports. Total Factor Productivity is also low -- TFP growth in the period 1991-2000 has been negative 3.95 percent. Fajnzylber, P. Gallego and Loayza (2003). 80 technological equipment, and to the labor force's low levels of education and technical skills, which constrain its capacity to adapt technology and push innovation. Other factors that may contribute are the lack of competitive pressure in the domestic market (as a function of its relatively small size and the internal barriers to entry), and local firms' limited export activity. 4.19 To encourage innovation and technology diffusion among Honduran private firms -- especially MSMEs and firms in industries in which Honduras has a comparative advantage -- the Government plans to develop a more efficient innovation infrastructure, including centers offering technology, training and advisory services. There are very few research institutions in Honduras that invest in applied research and support technology diffusion among private firms (see a preliminary list in Box D-2 on page 153 below). A preliminary diagnostic shows that there is potential for extending the activities of these centers to the wood and furniture industry (building on CUPROFOR's current programs), selected fast-growing agro industries (e.g., seafood and melons), textiles (especially to facilitate access to technologies for MSMEs and strengthen linkages between large firms and MSMEs), and handicrafts. These are all industries in which Honduras has -- or could easily develop -- a comparative advantage. Yet, firms operating in these sectors, especially MSMEs, do not seem to adopt and make productive use of advanced technologies.139 To address this shortcoming, the Honduran Government plans to support development of pilot centers for firms operating in the areas mentioned above.140 These centers would have the specific purpose of providing firms (especially MSMEs and firms in the clusters in the Government's plan) with specific training and advisory services (e.g., on equipment use and improving production processes) and with access to (and information on) new technologies (e.g., for raw material quality control, design, manufacturing, production and testing techniques, and packaging and labeling).141 4.20 The government is also considering developing a more proactive policy to attract selected Foreign Direct Investment -- a key conduit for technological innovation.142 In 2000, FDI accounted for 13.6 percent of gross capital formation in Honduras -- less than in Nicaragua (30.8 percent), but more than in Guatemala (7.2 percent) and in line with similar-income countries. In the last few years, thanks to the effort of the Government, the Foreign Investment and Export Promotion Agency (FIDE) and dynamic local private entrepreneurs, Honduras has succeeded in attracting foreign companies that have helped increase value added and diversify the country's production and export base. Further efforts in this direction are needed in the future. These should be specifically geared towards attracting higher value added FDI -- embedding technology in production processes and final products -- and creating linkages between foreign and private firms (as confirmed by the data presented in Figure 4-8 and Figure 4-12). Beyond the immediate benefits of increased production, employment generation and technology transfer, 139 Computerized equipment and licensing are almost non existent, for example, in the wood industry. Apparel is an exception here, but even in this sector, MSMEs are behind in adoption of advanced technologies and do not seem to be able to establish linkages with larger firms. Source: J. Fuster Background paper for the Trade Facilitation and Productivity Enhancement Project (World Bank, 2004). 140These initiatives are being funded with support from the World Bank and the Central American Bank for Economic Integration (CABEI). Note that the World Bank will finance CITEs for the wood and furniture and the handicrafts industries, while CABEI will finance CITEs for the textile sector. 141Firms using the centers will receive matching grants to pay for the use of the facilities (the maximum grant will cover 50 percent of the total expenses and will decrease with time). For a detailed description of the CITEs, see the World Bank Trade Facilitation and Productivity Enhancement Loan, Annex 2 (pages 64-70) and the background report by J. Fuster Honduras CITEs (World Bank, 2004). 142This is one of the key objectives of the Honduras Trade and Productivity Project (see the description of initiatives planned to improve attraction of selected FDI in Annex 2). 81 foreign companies can act as portals to new markets, facilitate the transfer of managerial and technical skills to domestic firms, and serve as incubators for small and medium enterprises associated with their productive chains. Specific initiatives that would help Honduras fully exploit the benefits offered by foreign companies include development and implementation of a National Promotional Strategy together with support of linkage programs combining different approaches (e.g., providing market or business information, matchmaking, and managerial and technical assistance to foreign and local firms operating in the same production chains). These actions should be coupled with strengthening of the specialized investment promotion agency, FIDE. 4.21 Additional measures that would help the Government intensify its efforts with regard to innovation include improving cooperation between the private and public sectors to stimulate productive innovation and supporting institutions working in the innovation field. The Government should consider allocating available research and innovation funds to research institutions that have clear linkages with the private sector -- e.g., secured contracts from local firms to carry out specific research or develop new technologies. To ensure transparency, funds should be allocated in a competitive manner. Given the limited resources for R&D and innovation, funds and incentives should be allocated selectively, starting with industries having higher comparative advantages. The experience of Chile, which has successfully developed innovation networks in the wood and fruit industries (among others), could provide a useful reference for Honduras.143 The Government may also want to consider introducing tax rebate schemes and matching grants to both domestic and foreign firms investing in innovation and industrial R&D. Finally, granting universities the right to file patents and own their research would help increase the patent rates and stimulate universities to conduct applied research with immediate benefits to the economy. Rather than playing a role in research execution, the Government -- which currently executes more than one third of the projects it finances -- could allocate research funds through transparent, competitive bidding processes that encourage private institutions and firms to participate. The selection criteria should be based on institutions' past performance, as well as on their ability and intent to collaborate with the private sector. 4.22 Strengthening the Directorate General of Intellectual Property (DGIP) at the Secretary of Industry and Trade is necessary to increase technology transfer from abroad (licensing) as well as to foster development of an indigenous technology sector.144 IPR protection in Honduras has been drafted in alignment with international practice (Laws of Copyright and several Decrees). Nevertheless, some complementary legislation needs to be enacted, and -- most importantly -- enforcement needs to be strengthened. Honduras' DGIP is currently understaffed and barely managing to carry out its vital functions of registration and control over intangible property rights. Strengthening the institutional capacity of the Directorate is critical to properly handling registration of patents, trademarks and copyrights, as well as any complaints regarding their infringement. These, in turn, are required to secure local R&D and license foreign technologies. 4.23 A strong body responsible for designing and coordinating implementation of a clear national innovation strategy is needed to ensure consistency between plans and actions. The Consejo Hondureño de Ciencia y Tecnología (COHCIT) is the Government institution responsible for design and implementation of the national innovation policy. Its limited staff (25) 143Chile New Economy Report (World Bank, 2003) and Closing the Gap in Technology and Education (World Bank, 2003). 144For a more detailed discussion on the current IPR Regime in Honduras, as well as on COHICIT's need for institutional strengthening, see P. Correa (2004). 82 and insufficient budget (US$0.4 million) prevents it from performing its functions completely. Partially as a result of this, the country lacks a clear national innovation strategy and the capacity to coordinate and evaluate existing initiatives and develop a vision and new programs for the future. COHCIT's innovation programs should be assessed, and the institution should be strengthened and reorganized. LABOR SKILLS145 4.24 The lack of a highly skilled labor force and of high quality training systems are key causes of Honduran firms' low use of quality certification and technologies, and their limited integration with productive and innovation networks. Quality certification, technological development and the skills of the labor force are highly interrelated variables -- i.e., it is impossible to develop one without the other. Firms are unlikely to introduce innovations, use quality certification schemes and license foreign technologies if their employees are not aware of the needs of the clients (especially the international market), do not know how to use advanced technologies and ignore what are the steps to receive international quality certifications. The section below completes this chapter by providing a summary of Honduran manufacturing firms' perception of local training systems and their suggestions for improvement to ensure that the Honduran labor force can improve productivity and catalyze innovation at the firm level. 4.25 Many Honduran firms offer formal training to their employees, but most of it is provided internally. The results of the survey show that Honduran firms do offer formal training to their employees. This is somewhat surprising, in light of the overall economic difficulties in Honduras and the general weaknesses in its training model and policies.146 About 49 percent of the private enterprises in Honduras provide formal employee training, and one-third of their workers receive formal training. As shown in Figure 4-13, the percentage of firms in Honduras providing formal training is considerably higher than in Pakistan, Bangladesh, and India; it is also somewhat higher than in Nicaragua. On the other hand, it is somewhat lower than in Guatemala, and significantly lower than in Brazil and China. As in many developing economies, there are marked differences in the incidence of formal training by firm size -- larger firms are more likely to train their workers than smaller firms (see Figure 4-14). There are also differences by sector -- firms in technology-intensive sectors in which Honduras actively competes (including chemicals and apparel) train more than firms in traditional sectors (such as furniture and non-metallic products). Like firms in the other Central American countries, Honduran firms that provide formal training rely more on internal training than on external sources of training (both public and private). About 49 percent of Honduran firms use in-house training, while only 29 percent use external training. 145 This section draws heavily from the Foreign Investment Advisory Service (FIAS) report Honduras: Reform of Labor Skills Training to Strengthen the Competitiveness of the Honduran Private Sector -- Internal Draft. (International Finance Corporation and World Bank, 2004). 146For a detailed description of the Honduran training model and policies, see FIAS (2004), pages 31 ff. 83 Figure 4-13: Percentage of firms providing formal Figure 4-14: Percentage of Honduran firms training, international comparison providing formal training, by firm size Pakistan 11.1 Overall 49.3 Bangladesh 26.6 India 27.2 Large 78.4 Nicaragua 37.4 Med 65.3 Honduras 49.3 Guatemala 55.0 Small 42.9 Brazil 67.1 Micro 23.9 China 69.6 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Source: Investment Climate Surveys Source: Investment Climate Surveys 4.26 Honduran firms rely excessively on learning-by-doing and informal training because of their complacency in the use of mature manufacturing technologies. As shown in Figure 4-15, the most frequent reason cited by Honduran firms that do not engage in formal training is that learning-by-doing is sufficient for their needs, and the second most frequent reason is that informal training from co-workers or supervisors is sufficient. Together, these facts suggest that a likely reason why Honduran firms do not train their workers is that the firms are complacent using mature technology instead of being forced to innovate in order to compete for market share.147 In those circumstances, workers can quickly become proficient through learning- by-doing or informal training, and there is no evident need for additional training. In general, the Honduran economy's traditional dependence on low-wage-based commodity exports has not created a high demand for human capital. These problems are especially acute for micro and small enterprises, which tend to use older, more manual equipment and lack the financing not only to purchase newer equipment but to train workers on its use. These factors "interact to create a vicious cycle of low levels of investment in human capital, low levels of productivity, and with limited resources, few incentives to train or adopt new technologies."148 Figure 4-15: Reasons cited by Honduran firms for not training their employees Primary reasons for not training Learning by Doing is Sufficient Informal training is sufficient Limited Resources Skilled workers can be readily hired Lack knowledge about training tech Labor Turnover Skills learnt in schools are adequate Skeptical about benefits of training 0% 5% 10% 15% 20% 25% 30% 35% 40% Source: Investment Climate Survey 147See the findings of the innovation section of this report, particularly with respect to the low level of computerized equipment, limited licensing of foreign technologies and lack of interaction with clients, suppliers or other institutions when introducing new equipment. 148FIAS (2004), p. 30. 84 4.27 Firms which provide formal training to their workers -- other than "on-the-job training" -- are more productive than those which do not provide such training. The econometric analysis indicates that training provided by firms to their employees beyond training "on the job" affects firm productivity positively and the results are highly significant at the 1 percent level in all six estimations -- see Table 1-3 in CHAPTER 1 on page 24 above. The estimated impact is quite large -- if a firm engages in a training program for its workers other than on-the-job training, its productivity is estimated to increase between 8.9 percent and 11.7 percent. The result is especially relevant for small firms. The country-by-country analysis displayed in Table 1-3 shows that the positive contribution to average country-wide productivity of the probability of a firm undertaking training of its employees is similar across the three Central American countries. 4.28 Secondary or higher education of firm employees is associated with higher productivity but the result is not significant. The econometric estimations displayed see Table 1-3 in CHAPTER 1 on page 24 above utilize the percentage of firm staff (out of total staff) with secondary education or higher. This variable exhibits a positive effect on productivity, but its impact is not significant in any of the six estimated regression specifications. The estimated size of the coefficients suggests that raising the fraction of staff (out of the total number of firm employees) with secondary or higher education by 1 percent will result in an increase in productivity between 0.03 and 0.06. Therefore, the analysis suggests that training is more important than secondary education -- a finding which is supported by other recent studies149. 4.29 The existing supply of training by public institutions is fragmented and inadequate. The primary provider of vocational and private sector training in Honduras, the Instituto Nacional de Formación Profesional (INFOP), had the capacity to train only 4.9 percent of the labor force in 2000, a shortfall of 83.5 percent from estimated demand.150 The political structure of INFOP and its high administrative costs151 make training provision inefficient. Although 60 percent of firms surveyed pay the 1 percent payroll levy that supports INFOP, less than a third of them actually use INFOP for training. The primary alternative to INFOP is the Advisory Committee for Human Resources Development (CADERH), a private non-profit organization that was established by the private sector and trade professionals in 1984 specifically to address perceived weaknesses with INFOP. CADERH manages a network of 31 independently operated and financed training centers that provide consistent, competency-based training. It maintains a high rate of institutional efficiency and lower per-student costs than INFOP. Some technical and occupational training for the private sector is also provided by technological institutes, universities, government ministries, and municipalities. However, the training provided by these diverse sources is fragmented -- there is little coordination among them -- and, even combined, meets only a tiny fraction of demand. 4.30 Large firms and firms operating in more advanced industries rely on private institutions for their training needs, suggesting that privately-offered training is superior to the public offerings. Overall, more firms who train their employees use private training schools (43.5 percent) than public training institutes (37.1 percent) or other providers. Furthermore, reliance on private sources of training increases with firm size and technological sophistication. As shown in Figure 4-16, use of private training schools correlates strongly with firm size in 149See Coulombe, Serge, Jean-François Tremblay and Sylvie Marchand, (2004), "Literacy Scores, Human Capital and Growth Across Fourteen OECD Countries", Published by Statistics Canada. 150FIAS (2004), p. 46. 151Administrative costs account for approximately 40 percent of INFOP's budget. FIAS (2003), p. 43. 85 Honduras -- only 11.1 percent of micro firms use private training, whereas 58.3 percent of large firms use it. As shown in Figure 4-17, use of private training schools is also strongly associated with sectors that require relatively advanced technologies--more than half of apparel firms and metal products firms take advantage of private training, whereas only about a third of food, tobacco and non-metallic minerals firms do so. Low wages152 and profit margins limit the abilities of many individual workers and smaller employers to invest in private training. The fact that the larger firms (who can better afford private training) and the more technologically advanced firms (who have a greater demand for high-quality training) turn to the private sector for their training needs suggests that privately-offered training is superior to the public offerings. Figure 4-16: Percentage of Honduran firms using Figure 4-17: Percentage of Honduran firms using private training schools, by firm size private training schools, by sector Apparel 56.3% Large 58.3% Metal Products 50.0% Med 39.3% Furniture and Wood 40.0% Small 25.9% Chemical and Rubber 40.0% Food and Tobacco 35.3% Micro 11.1% Non Metalic Minerals 33.3% 0% 10% 20% 30% 40% 50% 60% 70% 0% 10% 20% 30% 40% 50% 60% Source: Investment Climate Survey Source: Investment Climate Survey 4.31 In order to stimulate demand for and reform the supply of vocational training in Honduras, the government must first separate the governance and planning of training from the provision of training. Currently, INFOP is responsible for all aspects of training in Honduras, which creates conflicts of interest, inefficiencies, and insensitivity to market needs. There has been a succession of proposals for the reform of vocational training -- and particularly INFOP -- over the past few years. Proposals have come from INFOP, CADERH, the Interamerican Development Bank (IDB) and the Consejo Hondureño de la Empresa Privada (COHEP).153 Despite the differences in detail and emphasis between these proposals, there is a general consensus among them that training reform in Honduras requires reducing INFOP's monopoly. The government should establish a national training agency--separate from INFOP-- to plan and direct vocational training in Honduras. In order to ensure the alignment of the national training system with social and economic objectives, the training agency must have some government oversight. At the same time, in order to ensure a high level of private sector participation and support, the training agency must have some degree of autonomy from the government. To ensure the proper balance, the details of the agency charter and the public-private composition of its governing council should be determined through a planning and investigation process that engages the major public and private stakeholders. Through the same process, the government and the private sector also need to reach a consensus on the use of the payroll levy that balances the needs of the productive sector with the national priorities set by the government. Discontinuance of the training levy in the short to medium term would lead to the collapse of the vocational training system. The challenge, therefore, is to make the levy more effective. FIAS' recommendations in this respect include: (a) expand the supply of training courses and activities 15256.7 percent of workers receive poverty wages: FAIS (2003), p. 11. 153For a description of the various proposals, see FIAS (2004), pp. 66-71. 86 through more efficient use of the funds; (b) maximize the quality of this training; (c) allocate the funds to training programs that are most relevant to the needs of the private sector; and (d) build private sector confidence in the use and effectiveness of the levy.154 4.32 The Government should establish accreditation, standards and curriculum, and certification systems, but encourage private institutions to provide training. Once it is constituted, the proposed national training agency would be responsible for establishing, maintaining, and administrating a process for quality accreditation of public and private training providers, a set of skills standards for vocational training, and a coherent body of curricula and certificates. The training law in Honduras must also be revised in order to provide the legal basis for the division of the responsibilities currently held by INFOP, the changes in the use of the training levy, and the accreditation, standards and certification arrangements. Finally, to encourage private sector investment in training, the proposed national training agency should introduce a trial matching grants scheme. With a sound legal foundation, a set of standards administered by the government, and an incentive for increased demand for training, the private sector will have a solid basis for providing vocational training that meets the needs of the productive industries in Honduras. 154FIAS (2004), p.79. 87 CHAPTER 5. ACCESS TO AND COST OF CREDIT155 5.1 Financial constraints limit firms' investment and growth potential. A large body of research has demonstrated that firms' investment decisions depend on the availability of internal funds (e.g., retained earnings and new equity). Heavy reliance on internal funds is an indication of financial constraints, which lead firms to make sub-optimal investment decisions by affecting the timing and amount of their investments.156 In addition, in a recent cross-section study of 54 countries, Beck, Demirgüç-Kunt and Maksimovic (2002) show that financial constraints in terms of access and cost of funds influence firm growth and that smaller firms are most adversely affected by these constraints.157 5.2 This chapter summarizes the results of the finance section of the Honduras Investment Climate survey by focusing on two key issues: access to and costs of credit. The first section documents the main findings of the survey and provides a brief overview of the Honduran banking system. The second section provides preliminary policy recommendations, drawing from a background paper on access to and cost of finance prepared for the Investment Climate Assessment, the Honduras ROSC on the Insolvency and Credit Right System and the World Bank Financial Sector Technical Assistance Credit. This chapter mainly presents the perspective of Honduran firms on the financial system (demand analysis). The study could benefit from complementary assessments of the views of and constraints faced by banks and other financial institutions (supply analysis). THE PERSPECTIVE OF HONDURAN FIRMS 5.3 Larger Honduran firms are more likely to have loans. The share of Honduran firms with credit (51 percent of the total firms surveyed) is comparable to that in other Latin American countries -- see Figure 5-1. While the percentage of firms with a loan is positively correlated with size, similar percentages of Honduran large and micro companies have loans (47 percent and 43 percent, respectively) -- see Figure 5-2.158 This somewhat surprising finding is explained by the fact that only 58 percent of large firms demand loans,159 versus 85 percent of micros. To account for this and to develop a better idea of access issues, Figure 5-2 shows in parentheses the percentages of firms expressing demand for a loan that actually received one. Eighty-one percent of large firms that demand a loan have one, compared to 73 percent of medium, 66 percent of 155 This Chapter was prepared by Miguel Crivelli. Stefka Slavova summarized the results of the econometric analysis. Valuable inputs and advice were provided by Modibo Khane Camara, Aquiles Almansi, Michael Goldberg and Araceli de Leon (background paper). 156 For a review of the literature on financing constraints and investment, see Hubbard (1998). For applications to countries in Latin America see Galindo and Schiantarelli (2003). 157For a comprehensive review of the literature on finance and growth see Levine (1997) and World Bank Finance for Growth: Policy Choices in a Volatile World. 158For the purpose of this study firms were divided by size based on the number of permanent employees as follows: micro (1 to 9 employees -- 31 percent of the sample), small (10 to 25 employees -- 28 percent of the sample), medium (26 to 75 employees -- 16 percent of the sample) and large (over 76 employees -- 25 percent). See the description of the sample in Chapter One. 159 Firms that demand a loan are firms that: applied for and have a loan, applied for a loan and were rejected, or didn't apply because some constraint (e.g. interest rates, collateral requirements, or application procedures) prevented them from doing so. This classification follows Bigsten and Collier et al. (2000 and 2003). Rejection rates are very low in Honduras (3.6 percent of the sample) probably due to a self-selection mechanism that deters firms from applying. This phenomenon is observed in many countries. 88 small and 50 percent of micro firms. While worse off compared to larger firms, Honduran small and medium companies have a better chance of having credit than their Nicaraguan and Guatemalan counterparts. In Nicaragua, only 52 percent of medium and small firms that demand credit actually have it.160 Figure 5-1: Share of firms with loans, international Figure 5-2: Share of firms demanding and obtaining loans, comparison, percentage values. by size. Honduras, percentage values. 100 120 90 100 85 86 80 89 79 70 80 65 58 60 56 51.2 51 50.3 60 43 47 50 44.6 43.7 43.5 40 40 34.7 20 30 20 0 10 Micro Small Med (73%) Large Total 0 (50%) (66%) (81%) (64%) Honduras Ecuador Peru Nicaragua Guatemala Brazil Firms Demanding Loans Firms With Loans Source: Investment Climate Surveys. 5.4 Maquilas161 -- large firms operating in the apparel industry -- do not rely on the financial system to fund their activities. Large firms are far from a homogeneous group. In fact, once maquilas are treated as a different category, the percentage of firms demanding loans is similar across all sizes, with no exception for large firms (see Figure 5-3). This is due to the fact that only 31 percent of maquilas demand credit. The external linkages of maquilas -- which are mostly foreign-owned -- could help explain this pattern. Interestingly, about 92 percent of surveyed maquilas specified that they mainly manufacture products using strict specifications and exclusively for an individual client. This hints at the possibility of relatively long term contracts that would provide maquilas with a source of reliable hard currency funds. 160An in depth financial analysis of surveyed firms -- which is not possible at this stage -- would clarify whether limited access to credit is due to firms' own financial weaknesses. 161Maquilas included in the survey account for 10 percent of the total sample and 44 percent of the large firms. They are mainly large firms operating in the apparel industry, and 74.6 percent of them are foreign owned. See the description of the sample in Chapter One. 89 Figure 5-3: Share of firms demanding loans, maquila versus non-maquila. Honduras. Percentage values. 100 85 86 89 86 79 80 60 40 31 20 0 Micro Small Med Large Maquila Total Source: Honduras Investment Climate Survey. 5.5 Honduran firms, especially maquilas, rely mainly on retained earnings to finance their working capital needs. Bank financing ranks second as a source of credit, but is more common in Honduras than in comparator countries. Honduran firms, like firms in most comparator countries, finance almost half of their working capital needs with retained earnings -- see Figure 5-4. While not the main source of funding, bank financing accounts for a (relatively) large share of the total working capital financing of Honduran firms (26 percent of the total, versus 13 percent and 14 percent in Guatemala and Nicaragua respectively). The reliance on bank financing in Honduras is similar to that in more advanced countries (e.g., Ecuador and Brazil) -- see Figure 5-4. Large and medium firms in Honduras finance around 36 percent of their working capital needs with bank loans, an amount that decreases with firm size -- see Figure 5-5. On the other hand, maquilas rely heavily on retained earnings for working capital needs (78 percent of total needs), much more than any other group in the survey (average 49 percent). While maquilas depend on external suppliers for their inputs, they use little supplier credit (4 percent of the total).162 This somewhat surprising finding could be due to the fact that maquilas have other ways of financing their activities (e.g., off-balance sheet), that are not captured by the survey and are masked in the results. When financing long term capital, large Honduran firms rely on bank finance more than any other group of firms: they finance 45 percent of their long capital needs through this channel, versus an average of 28 percent. 162Note that an increasing number of maquilas offer "full package" services, i.e., they buy most of the needed supplies directly instead of merely providing labor to assemble the inputs. 90 Figure 5-4: Main Sources of finance for working capital Figure 5-5: Sources of finance for working capital needs, international comparison, percentage values. needs, by size, Honduras, percentage values. 70 100 59 60 58 90 78 49 80 50 44 43 70 60 50 40 42 50 4649 38 30 27 40 273635 25 26 22 26 18 30 20 20 16 16 14 13 14 15 13 20 919141713 1212 12 10 4 4 3 10 - 0 Retained Bank Supplier Other Retained earnings Banks Supplier Credit Earnings Finance Credit Ecuador Brazil Honduras Maquila Micro Small Med Large Total Guatemala Nicaragua Source: Investment Climate Surveys 5.6 Very high interest rates and collateral requirements were indicated as the main constraints for firms that do not apply for a loan.163 Eighty-two percent of the Honduran firms that did not apply to a loan did so because they perceived the interest rates as too high.164 Forty- five percent were discouraged by excessive collateral requirements and 33 percent by the banks' cumbersome bureaucratic procedures. These percentages are similar in Guatemala and Nicaragua -- see Figure 5-6. For Honduran micro and small firms, interest rates are more of a constraint than for large firms (80 percent of small firms and 100 percent of micros don't apply for a loan because of high interest rates). Collateral requirements in the application process also affect smaller firms more heavily: about 58 percent of micro firms and 47 percent of small firms find collateral requirements too high, compared to 31 percent of medium and 18.2 percent of large firms. Bureaucratic procedures do not seem excessive in Honduras. Information from visits indicates that in each city, banks have a number of attorney offices with which they work routinely and have pre-negotiated rates for their clients. This may account for the fact that, although judiciary tariff regulations allow for legal fees to reach up to 3 percent of the value of the loan, legal costs rarely exceed 2 percent of the value of the loan.165 163 Since few firms within the maquila group demand loans, in what follows we include them within the large firms group and not as a separate category. 164 Between 2000 and 2003, Honduran banks increased their holdings of government securities by 10 percent in real terms. Although this might exert upward pressure on the interest rates, the high level of liquidity in the system and the reluctance of banks to extend new credit to the private sector would tend to counteract it. 165This includes the costs of notarizing and legal "folios", insurance on the asset pledged if applicable (as in the case of transportation equipment or machinery), debt insurance (in some cases), and life insurance. 91 Figure 5-6: Constraints preventing firms from applying to a loan 100 90 82.4 78.6 80 67.6 70 57.1 60 47.2 50 27.1 45.1 40 27.8 33.3 30 20 10 - Cumbersome Collateral High Interest Rates Procedures Requirements Honduras Nicaragua Guatemala Source: Investment Climate Surveys 5.7 Interest rates are lower for larger firms and for firms obtaining loans in US dollars, but they are still high when compared to other countries in the region. Survey evidence shows that larger firms pay lower interest rates. This could be due to risk characteristics of the firms, but could also reflect high levels of asymmetric information in some segments of the market.166 Interest rates for loans in USD average 11.3 percent a year, less than the average interest rate charged for loans in Lempiras (24.7 percent)167 -- see Figure 5-7. This difference in interest rates reflects both currency risk and the fact that loans in USD are mainly granted to large companies. Survey results indicate that only 19.7 percent of firms have USD loans and that 77 percent of the firms that receive US loans are large- and medium-sized. Preliminary regressions168 show that, once size issues are accounted for, firms obtaining loans in Lempiras pay 11.5 percent more a year than comparable firms that have loans in USD. In international perspective, Figure 5-9 shows that large Honduran firms pay approximately the same as their Guatemalan counterparts and slightly less than firms in Nicaragua for their USD loans. However, interest rates are higher than those charged in El Salvador and Costa Rica -- see Figure 5-9. Regression analysis shows that micro firms in Honduras pay between 7 percent and 8 percent more than large firms, small firms between 3 percent and 3.5 percent more, and medium firms between 2.5 percent and 3 percent more. 166Information collected refers only to interest rates. Smaller firms relying on microfinance companies are likely to pay fees that may significantly increase their effective interest rate. 167Average inflation rates in Lempiras was 7.7 percent during 2003. Source: IMF Financial Statistics 168Regressions control for loan currency, firm size, sector, ownership, firm age, location, export orientation and collateral value. This analysis will be complemented by the results of the productivity analysis -- showing the links between the ICA variables and productivity measures -- in the next version of the report. (See the Introduction to the study). 92 Figure 5-7: Interest rates, percentage values. Figure 5-8: Interest Rates in USD for Large Firms 35.0 16.0 30.0 13.3 14.0 25.0 24.727.1 25.822.9 20.3 18.0 11.0 20.0 12.0 10.2 15.0 11.3 9.9 11.211.0 10.0 8.0 8.1 10.0 8.0 5.0 6.0 0.0 4.0 Lempiras USD 2.0 0.0 Total Micro Small Med Honduras Nicaragua Guatemala El Salvador* Costa Rica* Source: Investment Climate Surveys. Sources: Investment Climate Surveys for Honduras, Nicaragua and Guatemala. Central Banks for El Salvador and Costa Rica (ICS are not available for these countries). * Average figures for all firms. 5.8 Collateral requirements: real estate is what counts. On average, 90 percent of Honduran firms were asked for collateral or a deposit when requesting a loan. Collateral represented 154 percent of the loan amount -- a figure in line with comparable countries but above international standards. (See Figure 5-9.) Real estate is the most important form of collateral in Honduras. Firms reported pledging 60 percent of real estate belonging to the company for collateral purposes. In addition, many loans were secured with owner assets, which, in many cases, were also real estate -- especially for smaller companies. Machinery and equipment are 12 percent of the total value of collateral.169 This is somewhat surprising, given the lack of movable assets registries in the country. Preliminary regressions170 show that firms that own a building are 14 percent more likely to have a loan than firms that do not own real estate. Although this suggests that there is a strong statistical association between the probability of having a loan and the ownership of real estate, Honduran banks already have a large amount of real estate with low resale values, and may be less inclined to take additional real estate as collateral in the future. In fact, initial interviews and discussions show that they are already actively seeking new ways to secure loans. 169 Many lenders extend loans against movable property, keeping physical possession of the title or purchase papers until the loan is repaid. This seems to be accepted practice in many credit and savings cooperatives as well as banks and is generally complemented by an aval or fiduciary guarantee. 170 Following the methodology of Bigsten et al. (2000 and 2003), two step regressions were fitted identifying demand and supply equations. Regressions control for: firm size, firm sector, ownership, profitability, ownership of building, financial statement auditing and firm age. Although the magnitude of the relationship can be subject to discussion and there is some concern about the quality of information reported in this respect, the strong statistical association encountered points to a lack of other assets that could be used as collateral. If there were no differences among collateral types, property ownership would not matter in the probability of obtaining a loan and thus would not be statistically significant. 93 Figure 5-9: Average value of collateral (as percent Figure 5-10: Sources of collateral (as percentage of loan), of loan), international comparison Honduras. 90 250 218 80 70 60 6270 62 200 177 154 60 48 150 125 122 115 50 95 89 85 100 73 40 30 15 18 50 172427 20 12 13 6 11 0 10 5 2 2 1 0 4 5 9 11511 ilza ur 0 uadorcE Br Pe emala China Real Estate Machinery Intagible Owner Others Assets Assets Nicaragua Honduras uatG Malaysia Pakistan angladeshB Total Micro Small Med Large Source: Investment Climate Surveys 5.9 Having audited statements increases the chances of obtaining a loan, but auditing is (indirectly) costly. On average, only 43 percent of Honduran firms audit their financial statements (73 percent of large firms, 38 percent of small). Around 50 percent of the firms that obtained loans do not have audited financial statements. However, preliminary regressions171 show that firms with audited financial statements have a 12 percent greater chance of having a loan. Although this finding is interesting, from the firm's point of view, the decision to audit financial statements is related to the costs of formalization (and its tax implications). In fact, only 20 percent of the surveyed firms reported the cost of audits as a constraint for external auditing, supporting that there are drawbacks to formalizing operations offsetting the benefits it could have in terms of providing easier access to credit. 5.10 The econometric analysis indicate that improved financial disclosure -- through producing externally-audited financial statements -- raises firm productivity. The regional regression analysis indicates that having externally audited financial statements have a highly significant and positive impact on firm productivity. The size of the impact is large. By having its financial statements audited by an external auditor -- which presumably improves financial disclosure and alleviates information asymmetries in the financial market -- firms are estimated to increase their productivity between 11.6 percent and 17 percent -- see Table 1-3 in CHAPTER 1 on page 24 above. These findings are consistent with analyses and recommendations pointing to insufficient information disclosure in the financial market as causing inefficiencies. The country- by-country results do not reveal any major differences between the effects on average country productivity of the probability that firms have their financial statements audited by external auditors (see ANNEX C). 5.11 Besides the demand side issues outlined, the current state of the Honduran banking system is not conducive to private sector development. The liberalization of the banking sector in Honduras during the early 1990s led to a proliferation of small and weak institutions whose quality of assets deteriorated rapidly, partly due to a lax regulatory and supervisory framework. This deterioration was intensified after Hurricane Mitch, with the decline of private sector productive capacity. Many of the smaller intermediaries, including some banks, had to close, and it is expected that consolidation and reorganizations will continue for the next few years.172 171See footnote 170 for further details. 172The Honduran financial system appears to be relatively less concentrated than other banking systems in the region (see Table 5-1). The banking system is relatively large in relation to the level of intermediation 94 Currently, the Honduran financial system still faces a delicate situation, with a number of players in the system struggling with loan delinquency and non-performing loans. Although Honduras' private credit over GDP ratios are better than comparable countries' (see Table 5-1), this figure should be interpreted cautiously in light of the shortcomings of the data on private credit loans. In terms of the cost of finance, interest rates for savings and loans have fallen sharply during recent years, but spreads remain high (see Table 5-1). This may be caused by operational inefficiencies in the banking system and a high and stable level of non-performing loans in banks' portfolios, possibly a reflection of the weak financial state of the corporate sector in Honduras, as well as repayment culture and enforcement issues. Table 5-1: Key credit indicators, international comparison Country Private Credit Five bank concentration ratio Interest rate spread (Percent of GDP) (Percent) (Percent) Bolivia 52 57 24 Chile 61 79 5 Ecuador 30 69 10 Guatemala 18 44 11 Honduras 35 63 11 Nicaragua 48 87 11 Peru 26 80 14 Source: Doing Business 2003 5.12 Despite the recent increase in credit to the private sector, the level of intermediation in the banking system is low. Since 2000, credit to the private sector has increased by 4.3 percent in real terms. However, intermediation has not returned to pre-Mitch levels, as evidenced by the ratio of credit over deposits -- as of December 2003, it was 80 percent, compared to 90 percent at the end of 1998. The low level of intermediation is further evidenced by the high liquidity levels of many banks, which suggest a lack of lending opportunities and "conservative" bank behavior. An overall assessment of portfolio composition shows that banks are shifting away from agriculture and agribusiness lending and devoting a larger share of their portfolio to consumer lending and manufacturing firms. However, the portfolio of many banks remains heavily concentrated in specific sectors, posing problems in terms of low diversification of risks within the banking sector. INCREASING ACCESS TO AND REDUCING COSTS OF CREDIT FOR HONDURAN FIRMS: PRELIMINARY POLICY RECOMMENDATIONS173 5.13 Efforts must continue in order to enhance the soundness and efficiency of the financial system. As described in the previous section, the Honduran financial system faces several issues that affect its functioning and limit its potential to become an efficient intermediary of funds. The government has already embarked in a program to enhance the stability of the financial system, including efforts to improve the regulatory and legal environment and monitoring of systemic risks. However, further efforts are needed to enhance the efficiency of Honduran banks, which still have high operational costs. Honduras' ratio of administrative in the economy. As a result, administrative expenses over asset ratios are the highest in the region (see paragraph 5.13). 173The Honduran Government with the help of a Financial Sector Technical Assistance Credit from the World Bank is already pursuing various policy recommendation outlined below. 95 expenses over average assets is 6.4 percent, almost twice El Salvador's -- see Figure 5-11.174 A continuation of the consolidation trend observed may help increase efficiency of banks, providing a foundation upon which improvements in terms of access and cost of finance could be built. Consolidation will not be sufficient, however -- international evidence suggests that the direct effects of efficiency gains do not translate automatically into improved access for smaller firms. A deeper demand-side analysis is needed to understand the first-order effects of consolidation on access, especially for smaller firms. Figure 5-11: Administrative expenses over assets, international comparison 8 7 6.4 5.9 6 5.0 5 4.3 4 3.0 3 2 0.9 1 0 Rica emala El Hondura Costa uatG Nicaragua lvadoraS anamaP Source: Banco Central (BCH) and CNBS 2002 Source: De Leon, background paper 5.14 Improving judges' experience with and knowledge of commercial law cases would raise firms' trust in the judicial system, improve creditor rights, and increase use of insolvency procedures. This, in turn, may further reduce the cost of credit. Weak creditor rights might impose an additional barrier in the process of granting credit, given lower default recovery rates. Recent cross-country studies suggest that judicial efficiency (both for creating security titles and enforcing contracts) and inflation are the main sources of spread differentials.175 Honduras' judiciary scores for creditor rights protection are relatively low.176 The main problem seems to lie with judges' lack of practice and training in insolvency cases. It is common practice for troubled corporations to close without honoring their debts, rather than restructuring. Providing technical assistance to increase the experience of the judiciary on credit rights cases would enhance the confidence of the business sector in the judicial system, in turn increasing the use of insolvency procedures and (possibly) reducing the cost of credit.177 (See CHAPTER 2 on specific measures to improve the effectiveness and efficiency of the judicial system.) 5.15 The Government should improve the Immovable Assets Registry and support the creation of a Movable Assets Registry. Both measures would facilitate pledging of collateral and increase firms' access to credit. Despite the widespread use of real estate as collateral (60 percent of the total collateral pledged by firms, and up to 80 percent including owner assets), the Immovable Assets Registry is flawed. Recent studies found that 72 percent of the 174,000 real 174This is based on research carried out for the preparation of the Honduras Financial Sector Technical Assistance Loan. 175Laeven and Majononi (2003). Cross section model on the determinants of bank and aggregate level spreads. Sample of 106 countries. 176Doing Business (2003). 177For further details please refer to the Honduras ROSC on Insolvency and Creditor Rights Systems. 96 estate urban properties in Tegucigalpa (with a total estimated value of over US$2 billion) did not have registered titles.178 Registering land costs about 3 percent of the land's total value and may take up to 60 days -- compared with 0.15 percent of the land's value and one day for the same transaction in a developed economy.179 Other deficiencies are related to inconsistencies between information provided by different sources. The description of properties in the Immovable Assets Registry is often at odds with that in the National Cadastre, opening the door to disputes in authenticating rights. The government is in the process of approving a law to provide better registration and protection of property rights (Ley de Propiedad), which may integrate the National Cadastre with the Property Registry, reducing inconsistencies and disputes. Among other positive effects, the law will increase the availability of real state titles to be used as collateral. Possible further measures to improve the functioning of the Registry include enhancing its processing capabilities by automating and simplifying procedures and reviewing the fee structure for title registration.180 A second complementary measure that would help increase firms' opportunity to pledge more (and different) collateral is the creation of Movable Asset Registries. This would prove particularly useful for MSMEs, for which moveable assets could become a primary source of collateral to obtain credit from banks. Currently, moveable assets such as inventories and motor vehicles, for example, may not be used as pledges because they cannot be registered. In fact, the Motor Vehicle Tax Office may be the only place where firms can find any public records of sales or purchases of a vehicle. For this reason, cars or trucks are almost never used as collateral in credit transactions. 5.16 To further increase access to credit, the government should clarify the role of the national public registry (CNBS) as a monitoring tool, distinguish it from private credit registries, and prevent its use by commercial banks. Credit Bureaus are one of the most commonly used mechanisms to enhance access to credit and reduce the degree of asymmetric information between borrowers and lenders. Currently, Honduras has a public registry (Central de Riesgos -- Comisión Nacional de Banca y Seguros, or CNBS) and four relatively new private Credit Bureaus. The objective of CNBS is to monitor the stability of the banking system, while the objective of private registries is to improve bank decision-making processes by providing information on borrowers. Cross-country studies show that the existence of private credit registries is associated with lower perceived financial constraints and a higher proportion of bank financing, while the existence of public registries does not seem to significantly impact the financing constraints faced by firms.181 In Honduras, banks are allowed to access the public credit registry and use its information for credit granting purposes, a task for which it is not designed. Free access to the public registry discourages fee-based services that could prove more effective, as the international evidence shows. 5.17 Improve functioning of the Central de Riesgos -- CNBS. Currently, CNBS is part of the Central Bank regulatory bodies. Banks are mandated to supply data to CNBS only for credits in excess of L300, 000 (US$16,700). Once a lender has repaid its obligations to the bank, the CNBS erases that entry from its files and does not keep any other information on that debtor. Banks complain about the poor quality, accuracy, scope and timeliness of the information provided by CNBS. At times, information on individual debtors has been erroneously posted in the database. Banks often wait almost a month to get the latest credit information, and this may 178COHEP (2001). 179PAAR (2002) 180For further details on these topics please refer to the Honduras ROSC on Insolvency and Creditor Rights Systems. 181Love and Mylenko (2003). 97 delay loan processing by several weeks. Many of the problems may stem from the ambiguous role of the public registry, which could cause banks to delay submitting information, or avoid doing so altogether, to prevent its free use by competitors. 5.18 Improve the functioning of private credit registries. Private credit registries -- there are currently four, two of which are subsidiaries of international credit checking agencies -- concentrate on small loans (under L300, 000). They compile information on the credit history of individuals, in collaboration with banks and other organizations such as department stores, car dealerships and leasing companies. The main users of the private registries are financial companies, credit cooperatives and department stores. Potential users and providers of information have expressed fears of possible breaches of confidentiality or access to internal systems by unauthorized parties. To encourage the use of private credit registries, the government may want to consider providing an adequate framework for the use and protection of the information contained in the registries. 5.19 Studies should be conducted to improve understanding of the financial relationships of the maquilas. As the survey shows, the maquilas' financing patterns differ from other firms' primarily because of their intense links with foreign markets. The survey design makes it difficult to study many financing issues in depth, especially those involving off-balance sheet operations. Further studies could help identify maquilas' financing patterns in order to understand the behavior of these firms within the Honduran economy and gauge the needs and obstacles faced by this important and growing sector of the economy. 98 CHAPTER 6. CONCLUSIONS: MOVING FORWARD 6.1 This study has highlighted key microeconomic constraints on private sector development in Honduras and proposed policy recommendations for improving the country's investment climate to exploit the benefits of CAFTA and increase overall productivity and economic growth. Understanding the microeconomic determinants of growth and productivity has gained interest in recent years, and new instruments are available to quantify an increasing range of costs, barriers and risks faced by firms. This study presented the results of an Investment Climate Survey (ICS) of 450 Honduran manufacturing firms. The findings of the survey and suggested policy measures were organized by four key areas of the investment climate: (1) business-government relationships, with specific reference to red tape, procurement, the judiciary system, and crime; (2) infrastructure, including energy, information and communication technologies (ICT), and transport and logistics; (3) the capacity of Honduran firms to move up the value chain by innovating, adopting new technologies, certifying the quality of their processes and products, and improving the skills of their labor force; and (4) access to and cost of credit. The underlying motivation for the study is that improving the investment climate in Honduras would in turn improve its capacity to take advantage of the CAFTA trade agreements, help diversify its export base, increase firms' productivity, and spur economic growth. 6.2 The main objective of the study is to help the Government devise an Action Plan that can increase private sector productivity and growth. This report does not attempt to stipulate a comprehensive and final set of policy reforms. Rather, it should be considered as a first input for design of an Action Plan to improve the country's overall investment climate. The Summary Matrix (Table 1-3 in the Executive Summary in Volume 1) includes key findings of the diagnostic of the ICS and suggested policy recommendations. The Matrix is intended to be a first tool to support development of this Action Plan. 6.3 A policy agenda to improve the investment climate requires broad-based strategies and actions and close collaboration between the public and private sectors, as well as among different Government agencies. Improving the investment climate in Honduras will require cross sector strategies and actions geared towards aligning economic policy, public management, infrastructure, technology, FDI, innovation, training and finance policies. This can be done only by establishing close collaboration between the private and the public sectors and among various Ministries in the public sector to push forward coordinated initiatives that can unleash the private sector's potential for growth. Although the Honduran Government has already made significant progress in this respect -- the Competitive Commission is an excellent example of effective collaboration between the private and the public sectors -- more effort in this direction will be needed in the future. Ensuring that the Competitive Commission has sufficient capacity to implement reform initiatives agreed upon by its members and creating Technical Committees that can take responsibility for specific reforms is essential. The recent creation of a Technical Committee overseeing Competition Policy initiatives and the plans for a Committee for Administrative Simplification prove that the Government is moving in the right direction -- and that improving the Honduran investment climate is feasible. 6.4 A clear investment climate agenda and effective implementation mechanisms are necessary for Honduras to take full advantage of CAFTA and achieve high rates of productivity and growth. The 21st century reality is that Honduras is not being compared with -- and does not compete only with -- its Central American and Latin neighbors. In the context of a global economy, there is a price to be paid for complacency and a premium to be gained by being a first mover. Constant Government progress to reduce constraints on economic growth, as well as firms' dynamic pursuit of new markets and innovation, are both essential in 99 order for the country to take advantage of the opportunities offered by the new trade agreements, improve its growth performance and fulfill social expectations. 100 BIBLIOGRAPHY Amnesty International. 2003. "Honduras, Zero Tolerance...For Impunity. Extra Judicial Executions of Children and Youths since 1998." web.amnesty.org/library/eng-hnd/index Andinatel - www.andinatel.net Baffes, J. and Shah, A. 1998. "Productivity of Public Spending, Sectoral Allocation Choices and Economic Growth." Economic Development and Cultural Change 48(2). Barclay, G. and Tavares, C. 2002. "International Comparisons of Criminal Justice Statistics 2000." Issue 05/02. www.homeoffice.gov.uk/rds/pdfs2/hosb502.pdf.html Barro and Lee. 2001. "International Measures of Schooling Years and Schooling Quality." Inter- American Development Bank (IDB), Washington, D.C. Bigsten, A., Collier, P., Dercon, S., Fafchamps, M., Gauthier, B., Gunning, J.W., Soderbom, M., Oduro, A., Oostendorp, R., Patillo, C., Teal, F., Zeufack, A. 2003. "Credit Constraints in Manufacturing Enterprises in Africa," Journal of African Economies, 12 (1), 104-125. Consejo Hondureño de la Empresa Privada (COHEP). 1997. "II Congreso de la Empresa: Sobre el Desarrollo Social." Comisión Nacional de Telecomunicaciones de Honduras (CONATEL) www.conatel.hn. Cooperación Técnica Alemana (GTZ), INFOP, División de Planificación, 2001. "Estudio del Sistema de Formación Técnica Profesional y Requerimientos de Capacitación de Honduras." Correa, P. 2004. "A Preliminary Assessment of the Honduras National Innovation System." Mimeo, World Bank, Washington D.C. Coulombe, S., Tremblay J.F., Marchand, S. (2004). "Literacy Scores, Human Capital and Growth Across Fourteen OECD Countries". Statistics Canada. Chile's Government Procurement E-System. http://www.worldbank.org/publicsector/egov/eprocurement_chile.htm Devres Inc. 1994. "Final Evaluation of the Honduran Advisory Council for Human Resource Development." DevTech Systems, 2002. "Vocational Technical Education Project, USAID/Honduras." Final Report to USAID. Diamond, J. 1999. Guns, Germs, and Steel. New York: W.W. Norton & Company. Dollar, D., Hallward-Drieneir, M., and Mengistae, T. 2003. "Investment Climate, Infrastructure and Trade: A Comparison of Latin America and Asia." Development Economics, World Bank, Washington, D.C. Draft. Not for quotation. Ask for permission. 101 Dollar, D., Hallward-Drieneir, M., and Mengistae, T. 2003. "Investment Climate and Firm Performance in Developing Economies." Development Research Group, World Bank, Washington, D.C. Draft. Not for quotation. Ask for permission. Duryea, S. and Pages, C. 2002. "Human Capital Policies: What they can and cannot do for Productivity and Poverty Reduction in Latin America." Working Paper 468. Inter-American Development Bank, Washington DC. Easterly, W. and Levine R. "It's Not Factor Accumulation: Stylized Facts and Growth Models." World Bank, Washington, D.C. Easterly, W. and Levine, R. 2002. "Tropics, Germs and Crops: How Endowments Influence Economic Development." Working Paper 9106. Cambridge, MA: National Bureau of Economic Research. http://papers.nber.org/papers/W9106. Easterly, W. and Servén, L. 2003. "The Limits of Stabilization. Infrastructure, Public Deficits, and Growth in Latin America." Stanford University Press and the World Bank. Economic Commission for Latin America and the Caribbean (ECLAC). 2003. "Foreign Investment in Latin America in the Caribbean." Escribano A., Guasch J.L., (2004). "Assessing the Impact of the Investment Climate on Productivity using Firm Level Data: Methodology and The Case of Guatemala, Honduras and Nicaragua." World Bank, Universidad Carlos III de Madrid and Georgetown University. Draft. Foreign Investment Advisory Service (FIAS). 2004. "Honduras: Reform of Labor Skills Training to Strengthen the Competitiveness of the Honduran Private Sector." (Internal draft). IFC and World Bank, Washington DC. Fuster, J. 2004. "Centros de Innovación Tecnológicos para Honduras." Background Report for the Honduras Trade Facilitation and Productivity Enhancement Project. World Bank, Washington DC. Gaviria, A. and Pagés, C. 1999. "Patterns of Crime Victimization in Latin America." Working Paper #408, Inter-American Development Bank, Washington, D.C. Guasch, J.L. and Kogan, J. 2001. "Inventories in Developing Countries. Levels and Determinants--a Red Flag for Competitiveness and Growth." Policy Research Working Paper 2552, World Bank, Washington, D.C. Hanna, J. 2004. "Accelerating E-Business For Small Business Innovation and Growth." in Motta, et. al., Chile New Economy Study. World Bank, Washington, D.C. Hu, Z. and Khan, M. 2004. Why is China Growing so Fast? International Monetary Fund, Economic Issues 8, Washington, D.C. Instituto Centroamericano de Administración de Empresas (INCAE). 1998. "Turismo en Honduras, El Reto de la Competitividad." Instituto Costarricense de Electricidad (ICE). http://www.ice.go.cr/ 102 Inter-American Development Bank. 2003. "Honduras. Peace and Citizen Coexistence Project for the Municipalities of the Sula Valley." (HO-0205) Loan Proposal. ISO Management System Standards: http://www.iso.ch/iso/en/iso9000- 14000/basics/general/basics_1.html International Telecommunication Union (ITU). 2003-2004. "World Telecommunications Development Report." Geneva.. Jaramillo C.F. (2004). "Que sabemos de los efectos del CAFTA?" Revista Centroamérica en la economía mundial del siglo XXI. Johnson et al. (1999). "Why do firms hide ? Bribes and unofficial activity after communism." Public Economics 0308004, Economics Working Paper at WUSTL. Juan-Ramon, H. 1999. "Honduras Growth Performance during 1970-1997." Policy Discussion Paper (PDP/99/1), International Monetary Fund, Washington, D.C. Kaufmann, D. 2004. "Governance Redux: The Empirical Challenge." The Global Competitiveness Report 2003-2004. Chapter 2.5. World Economic Forum. Kaufmann, Kraay and Zoido-Lobatón (1999). "Governance Matters." World Bank, Washington, D.C. Kenny, C. and Motta, M. 2004. "Chile's Information and Communication Technology Sector" in Motta, et al., Chile New Economy Study. World Bank, Washington D.C. Knack, S. and Keefer, P. 1997. "Does Social Capital Have an Economic Payoff? A Cross Country Investigation." Quarterly Journal of Economics 112(4): 1251-1288. Leyva, H. 2001. "Delincuencia y Criminalidad en las Estadísticas de Honduras 1996-2000." Serie Sociedad y Cultura, Documento de Trabajo No. 3. Tegucigalpa: FIDE ­ PNUD (FOTOCOPIA). Laeven, L. and Majnoni, G. 2003. "Does Judicial Efficiency Lower the Cost of Credit?" Policy Research Working Paper No. 3159. World Bank, Washington, D.C. Love, I. and Mylenko, N. 2003. "Credit Reporting and Financing Constraints." Policy Research Working Paper No. 3142. World Bank, Washington, D.C. Martínez Espinoza, E. 1997. "Chile: Experiences in a Market-Oriented Training System". International Labor Office. Mauro, P. 1995. "Corruption and Growth." Quarterly Journal of Economics, 110(3): 681-712. Morales, O. M. 2000. "Competency-based Technical Education in Costa Rica in Competency Based Education and Training." A World Perspective, CONALEP. Olley G. S.and A. Pakes (1996) "The Dynamics of Productivity in the Telecommunications Equipment Industry". Econometrica, Vol. 64, 6, 1263-1297. 103 Solow R. M. (1957) "Technical Change and the Aggregate Production Function". The Review of Economics and Statistics, 39 (3), 312-320. Organization for Economic Cooperation and Development (OECD). 2002. "OECD Information Technology Outlook: ICT and the Information Economy". OECD Working Paper. Comisión Presidencial de Modernización del Estado 2002. " Propuesta de Diseño de Reforma del Subsector Eléctrico de Honduras." Public-Private Infrastructure Advisory Facility (PPIAF). 2003. "Private Solutions for Infrastructure in Honduras." A country Framework Report. World Bank, Washington, DC. Red Iberoamericana de Indicadores de Ciencia y Tecnología, 2004. Rodrik, D., Subramanian, A. and Trebbi, E. 2002. "Institutions Rule: The Primacy of Institutions over Geography and Integration in Economic Development." Working Paper 97, Cambridge, MA: Center for International Development, Harvard University. www.cid.harvard.edu/cidwp/097.htm. Rubio, M. 2002. "La Violencia en Honduras y la Región del Valle de Sula." Serie de Estudios Económicos y Sectoriales. Banco Interamericano de Desarrollo, Washington, D.C. Sachs, J. and Warner, A. 1997. "Natural Resource Abundance and Economic Growth." Cambridge, MA: Center for International Development and Harvard Institute for International Development. Santos, R. and De la Paz de Borjas, M. 1999. "Diagnóstico de la Educación Técnica en Honduras." Comisión Presidencial de Modernización del Estado. Honduras. Schneider and Enste (March 2000). "Shadow economies: Size, causes and consequences." Journal of economic literature. Pp. 77-114. Schneider, Fredrich (2002). "Size and measurement of the informal economy in 110 countries around the world." Presented at the Workshop of Australian National Tax Center, ANU, Canberra, Australia, July 17, 2002. Subsecretaría de Telecomunicaciones (SUBTEL). 2000. "Acceso Universal: Una Política de Acción", Universidad de Chile, Departamento de Ingeniería Industrial. Telemovil - www.telemovil.com The Economist Intelligence Unit. 2003. Honduras Country Profile, 2003. Thompson, D. 2003. " Improving the Competitiveness of Honduras through a National Quality System." Background Paper for the Honduras Trade Facilitation and Productivity Enhancement Project. World Bank. Washington D.C. World Bank. 2001. "Improving the Investment Climate in India." Pilot Investment Climate Assessment. Washington, D.C. Wellenius, B. 2002. "Closing the Gap in access to Rural Communications: Chile 1995-2002." 104 World Bank Discussion Paper 430, Washington, D.C. World Bank. 2003. "Improving the Investment Climate in China." Washington, D.C. World Bank. 2003. Enhancing Competitiveness: Trade Facilitation and Productivity Improvement Project. Project Appraisal Document. Washington D.C. World Bank. 2003. Honduras Financial Sector Technical Assistance Credit. Project Appraisal Document. Washington D.C. World Bank. 2002 and 2003. World Development Indicators. Washington, D.C. World Bank. 2003. "Closing the Gap in Education and Technology." Latin America and Caribbean Region. Washington D.C. World Bank. 2003. "Information and Communication Technology Strategy Paper." Washington, D.C. World Bank. 2003. "Country Assistance Strategy for the Republic of Honduras." Washington, D.C. World Bank. 2003. "Honduras Judicial Branch Modernization Project." Project Concept Note, World Bank Internal Document. Washington, D.C. World Bank/Inter-American Development Bank. 2003. "Honduras: Country Financial Accountability Assessment." Draft Report. Washington, D.C. World Bank. 2004. "Poverty Reduction Support Credit (PRSC)." Draft Matrix. Washington, D.C. World Bank. 2004. "Honduras Civil Service Reform." Honduras PRSC. Washington, D.C. World Bank. 2004. "Honduras Draft PRSC Matrix." Internal Document. Washington, DC. World Bank and Oxford University Press. 2004. "Doing Business in 2004: Understanding Regulation." Washington, D.C. World Bank. 2004. "Lessons from NAFTA for Latin America and the Caribbean countries: a summary of research findings." Washington, D.C. World Bank. 2004. "Honduras Insolvency Assessment and Creditor Rights ­ ROSC". Washington, DC. World Bank Institute. 2000. "Governance and Anti-Corruption in Honduras: An Input for Action Planning." Preliminary Draft Report. World Bank, Washington, D.C. World Business Environment Survey, 2000. World Economic Forum. 2004. "The Global Competitiveness Report 2003-2004." Geneva. 105 ANNEX A: SAMPLE DESCRIPTION The sample consists of 450 firms belonging to the manufacturing sector. To account for size characteristics, firms were divided into four groups based on the number of permanent employees, following standard ICA methodologies. Given the structure of the economy, the following size definitions were adopted (see Figure A-1): a. Micro firms: 1 to 9 employees -- 31 percent of the sample, b. Small firms: 10 to 25 employees -- 28 percent of the sample, c. Medium firms: 26 to 75 employees -- 16 percent of the sample, and d. Large firms: Over 76 employees -- 25 percent of the sample. The majority of the firms are domestically owned, with only 15.78 percent reporting foreign ownership. Foreign owned firms are concentrated within the large firm group where 50.5 percent of the firms are foreign. Figure A-1: Size Distribution of the Sample Figure A-2: Exporting Profile Large 25% Micro 31% Export 34% Mediu No m Small Exports 16% 28% 66% Thirty three percent of the firms surveyed report exporting182 some part of their production. Exporter firms tend to be concentrated within the large and medium firm groups, where 77 percent and 42 percent report exports respectively. For the purpose of the analysis, firms where divided into eleven industrial sectors. As Figure A-3 shows, Furniture and Wood, Food and Tobacco and Apparel are the three most important sectors surveyed, reflecting the composition of the Honduran manufacturing sector. In terms of their exporting profile, Figure A-4 shows that firms surveyed are more likely to be exporters if they are located in the Apparel, Textile, Food and Tobacco or Chemical and Rubber Industries. 182Exporting firms were defined as firms exporting 10 percent or more of their production 106 Figure A-3: Industrial Profile Figure A-4: Exporting Industries Metal Beverages 4.8 Products BeveragesTextile Non Metalic Minerals 6.3 7% 4% Furniture 5% and Wood Furniture and Wood 20.4 Chemical 24% and Rubber Metal Products 23.3 7% Chemical and Rubber 36.4 Non Metalic Minerals Food and Tobacco 41.5 11% Apparel 61.4 Food and Textile 62.5 Apparel Tobacco 18% 24% - 20.0 40.0 60.0 80.0 Firms in the sample belong to 14 departments in Honduras. Following standard classifications used in other studies firms were grouped in four regions as follows: Center South183 (49.6 percent), North Coast184 (37.6 percent), Olancho (8.2 percent) and West185 (4.7 percent). Within the three most important sectors surveyed (Apparel, Furniture and Wood, Food and Tobacco) different patterns of industrial location stand out: Apparel firms and (to a lesser extent) wood firms, are mainly located in the North Coast while Furniture and Wood and Food and Tobacco industries tend to be located in the Center South region of the country. Figure A-5: Geographical Distribution of the Figure A-6: Geographical Distribution of Selected Sample Industries West Center -South North - Coast Olancho West Olancho 5% 8% Food and Tobacco 49.1% 21.7% Center - Furniture and South Wood 56.6% 33.6% 49% North - Coast Apparel 28.9% 66.3% 38% 0% 20% 40% 60% 80% 100% The survey design also makes it possible to account for the fact that some firms surveyed are Maquilas. Given the information available, Maquilas were defined as firms benefiting either from the Ley de Zonas Libre or the Ley de Zonas de Procesamiento para exportación and exporting directly more than half of 183Department in the Center South Region: Choluteca, Comayagua, El Paraiso, Fco. Morazan, Valle and Yoro 184Departments in the North Coast Region: Atlantida, Colon, Cortes and Islas de la Bahia 185Departments in the West Region: Santa Barbara and La Paz 107 their production. The Maquilas subsample is mainly composed of large firms (48.6 percent of the large firm group is a Maquila) operating in the apparel industry and in 74.6 percent of the cases is foreign owned. Maquilas are located in 73 percent of the cases in the North coast region of the country. 108 ANNEX B: STANDARD TABLES Table B-1: Sample structure for Honduras Investment Climate Survey Sample population Sample population Firm size Firm activity Micro 139 Apparel 83 Small 127 Beverages 21 Medium-size 73 Chemical 33 Large 111 Food & Tobacco 106 Furniture, etc. 113 Nonmetal minerals 48 Textile 16 Market Metal Products 30 orientation Exporter 151 Nonexporter 299 Firm ownership Firm location Private, domestic 379 West 21 Private, foreign 71 Center-South 223 State 0 Olancho 37 North Coast 169 109 Table B-2: Foreign and domestic sales: Honduras investment climate survey r ict orte ityc all mes exp- Honduras Guatemala Nicaragua Micro Sm Medium Large Foreign Do Exporter Non Low-capa Disposition of sales Sold domestically 75.7 84.2 84.8 96.2 89.8 75.9 33.8 22.9 85.6 28.1 99.8 78.2 Exported directly 21.3 13.3 9.4 1.6 8.1 17.6 63.6 74.1 11.4 63.1 0.2 18.4 Exported indirectly 3.0 2.5 5.8 2.2 2.1 6.6 2.6 3.0 3.0 8.7 0.1 3.4 Source of inputs and supplies Domestic sources 67.6 67.6 64.3 89.7 73.9 70.2 30.4 75.6 25.2 45.1 79.0 71.4 Imported directly 25.0 22.2 15.3 2.7 18.1 21.9 63.6 17.0 67.7 50.1 12.4 21.2 Imported indirectly 7.3 10.2 20.4 7.5 7.9 7.9 6.0 7.4 7.2 4.8 8.6 7.4 110 Table B-3: Constraints on growth in Honduras, by type of firm ict ter ty ty all mes - Constraint Honduras Guatemala Nicaragua Micro Sm Mediume Large Foreign Do Exporter Non expor Low- capaci High- capaci Telecommunications 18.3 6.6 12.8 15.9 18.1 20.5 19.8 23.9 17.2 19.9 17.4 20.1 16.4 Electricity 36.4 26.6 34.1 33.8 32.3 43.8 39.6 35.2 36.7 34.4 37.5 37.1 35.8 Transportation 7.6 13.6 12.4 7.2 3.9 12.3 9.0 11.3 6.9 6.0 8.4 8.5 6.6 Access to Land 13.1 15.2 17.5 20.1 10.2 16.4 5.4 7.0 14.2 6.6 16.4 12.9 13.3 Tax rates 35.6 56.5 32.5 36.7 37.0 45.2 25.0 22.4 37.7 23.4 41.3 35.6 35.7 Tax Administration 23.2 34.8 16.6 24.5 24.6 29.2 15.3 18.3 24.0 17.4 25.9 25.0 21.4 Customs regulations 10.7 26.0 10.0 10.0 9.4 15.9 9.3 16.9 9.0 9.5 11.7 13.9 7.5 Trade regulations 8.8 12.6 6.8 6.5 7.9 15.0 7.5 9.9 8.5 9.5 8.2 10.9 6.5 Labor regulations 14.2 16.7 5.8 10.8 12.6 19.2 17.1 12.7 14.5 17.9 12.4 15.6 12.8 Skills and education of available Workers 26.4 31.4 15.3 20.9 26.8 30.1 30.6 22.5 27.2 26.5 26.4 30.4 22.6 Business Licensing, etc. 21.1 15.6 9.5 16.5 25.2 28.8 17.1 15.5 22.2 14.6 24.4 26.8 15.5 Availability of financing 49.0 30.8 49.6 49.6 59.5 50.7 34.0 22.2 53.4 39.9 53.4 59.0 38.8 Cost of financing 60.2 43.3 58.3 62.6 66.1 71.2 41.7 33.3 64.6 51.7 64.2 66.4 53.9 Access to financing 50.8 34.1 53.1 56.1 59.8 52.1 31.7 20.3 55.9 40.3 55.9 56.5 45.0 Economic and Reg. Policy Uncertainty 47.0 66.4 56.4 42.6 53.5 52.1 41.4 36.6 48.9 41.1 50.0 51.1 42.9 Macroeconomic Instability 51.6 60.9 47.8 53.3 56.7 60.3 37.8 33.8 54.9 39.1 57.9 54.1 49.1 Corruption 62.8 80.9 63.1 66.9 67.7 57.5 55.5 52.9 64.6 50.0 69.2 66.1 59.6 Crime & Violence 60.9 80.4 39.2 57.6 59.1 56.2 70.3 64.8 60.2 61.6 60.5 59.4 62.4 Anti-competitive practices 40.4 56.3 47.1 37.0 49.6 46.6 30.0 24.3 43.4 32.0 44.6 42.2 38.7 111 Table B-4: Infrastructure performance indicators in Honduras, by type of firm r ityc ict orte ityc all mes exp- Indicator Honduras Guatemala Nicaragua Micro Sm Medium Large Foreign Do Exporter Non Low-capa High-capa Share of firms that experienced a power cut 92.1 74.1 88.7 94.2 91.2 91.6 90.9 89.0 92.7 86.8 92.6 92.2 91.0 Frequency of power outages (times last year) 29.4 13.7 29.8 25.5 24.7 42.4 31.1 34.3 26.9 29.0 27.7 29.5 27.7 Hours lost due to power outages (total) 127.0 60.0 129.5 133 145.1 156.0 80.1 107.0 128.0 107.4 133.6 139 109.7 Sales lost due to power outages (percent) 3.0 2.3 3.9 3.3 3.5 3.0 2.0 2.0 3.2 2.6 3.2 3.1 2.9 Share of firms who own a generator 33.3 21.7 19.9 17.8 21.2 35.9 65.0 62.0 28.0 51.7 24.1 34.9 31.7 Days to obtain an electricity connection 32.5 62.3 31.4 22.4 38.6 45.5 33.7 30.4 32.8 25.4 37.1 26.2 40.4 Share of firms that experienced a phone cut 24.8 19.6 26.2 16.6 29.6 23.9 30.3 22.5 24.8 26.5 23.4 28.0 20.6 Frequency of phone cuts (times last year) 6.5 2.6 7.0 4.9 11.0 2.7 5.7 8.5 6.0 4.4 7.4 7.2 5.5 Total hours lost due to phone outages 56.0 17.5 79.5 40.2 49.5 48.5 88.1 61.3 54.8 65.3 49.4 72.2 38.3 Sales lost due to phone outages (percent) 0.3 0.2 0.7 0.3 0.5 0.3 0.2 0.3 0.4 0.3 0.4 0.4 0.2 Days to obtain a telephone connection 175.0 47.7 118.4 211 162.5 186.5 147.9 136.5 184.5 138.6 195.7 176 173.7 Share of Firms who communicate with clients through email 50.2 66.4 38.5 11.5 40.9 75.3 92.8 93.0 42.2 80.8 34.8 47.8 52.8 Share of firms who communicate with clients through a website 21.8 29.2 16.6 3.6 13.4 31.5 47.7 46.5 17.2 41.7 11.7 19.9 23.9 Days to obtain a water connection 54.5 68.1 54.6 61.6 70.7 7.0 17.3 17.5 57.7 14.9 69.9 74.7 18.6 112 Table B-5: Sources of finance for firms in Honduras, by type of firm all Honduras Guatemala Nicaragua Micro Sm Medium Large Share of firms with audited financial statements 43.2 34.2 30.2 18.1 38.1 54.2 73 Share of firms with overdraft or line of credit 42.7 49 30.7 22.5 34.9 58.3 66.7 Share of credit currently unused 61 60 55.5 60.1 56.4 57.9 65.9 Share of firms with a loan from a bank or other financial institution 51.2 43.5 43.7 42.8 56.4 65.3 46.9 For the most recent loan or overdraft Share of loans or overdraft requiring collateral 89.1 73.1 92.4 91.5 94.4 87.2 80.8 Average value of collateral required, as a share of loan 154 115.1 217.8 159 174 119 149 Average interest rate on US$ denominated loan 11.3 13.9 16.4 18 9 11.2 11 Average duration of US$ denominated loan (months) 38.5 41.7 32.5 24 36.8 38.3 42 113 Table B-6: Sources of finance for firms in Honduras, by firm type all Honduras Guatemala Nicaragua Micro Sm Medium Large Sources for working capital Retained earnings 49.2 59.2 57.6 50.1 42.5 40.0 61.3 Banks 25.9 13.0 14.0 21.0 26.8 35.4 24.6 Trade credit 13.0 18.3 16.3 8.9 19.3 13.3 10.9 Equity 0.6 1.4 0.4 0.6 1.1 0.9 0.1 Informal sources 1.5 .8 0.8 1.7 2.0 1.9 0.3 All others 9.8 7.4 10.9 17.6 8.4 8.5 2.8 Sources for new investments Retained earnings 52.7 59.4 68.6 50.0 52.5 48.2 58.5 Banks 27.9 19.5 17.0 27.3 23.6 32.4 30.2 Trade credit 5.7 8.4 3.7 5.2 10.9 3.1 2.6 Equity 2.2 1.4 0.8 1.0 2.6 0.0 4.5 Informal sources 1.4 .7 0.1 1.2 2.9 1.9 0.0 All others 10.1 10.7 9.8 15.3 7.7 14.4 4.1 114 Table B-7: Regulatory burden and administrative delays in Honduras, by type of firm ict ter ty ty all mes - Honduras Guatemala Nicaragua Micro Sm Medium Large Foreign Do Exporter Non expor Low- capaci High- capaci Confidence in the judiciary ( percent disagree) 38.5 47.7 41.0 36.8 37.6 46.6 36.1 31.4 39.8 38.9 38.2 39.5 37.4 Share of senior management's time spent dealing with regulations 8.6 11.16 7.7 4.1 8.6 9.5 13.4 9.8 8.4 11.5 6.8 6.7 8.9 Informal payments to officials to "get things done" as a share of revenue 2.5 3.7 1.7 3.5 3.2. 2.2 4.3 0.7 2.8 1.2 3.2 2.6 2.7 Share of revenue typically reported for tax purposes 68.4 77.3 66.4 61.5 70.2 72.1 72.8 73 67.5 72..3 66.1 69.5 67.4 Days spent in inspections or required meetings with officials (Mean) 12.1 9.9 21.3 6.5 10.9 10 21.7 18.25 10.9 15.3 10.2 13.9 10.3 Interpretations of regulations ( percent disagree) 45.9 71 42.2 46.4 43.3 50.1 45 41.7 46.7 45.3. 46.2 42.5 49.3 Share of disputes that go to court 2.1 1.4 2.1 0.5 2.8 3.2 1.9 1.9 3.5 3.2 1.3 2.4 1.8 Security Costs ( percent of total costs) 3.6 5.5 2.9 1.9 4.8 4.6 3.6 3.4 4.3 3.2 3.8 3.6 3.5 115 Table B-8: Technology indicators in Honduras by type of firm ict r te yt yt all mes - orp acip acip Indicator HondurasGuatemalaNicaragua Micro Sm Medium Large Foreign Do Exporter Non ex Low- ca High- ca Share of firms introducing a technological innovation in the last two years 67.3 80.9 67.9 59.0 68.5 69.9 74.8 74.6 66.0 73.5 64.2 67.0 67.7 Method of Introducing Innovations Embodied in new machinery or equipment 79.2 73.9 71.6 80.5 79.3 74.5 80.7 77.4 79.6 81.1 78.1 81.3 77.1 Hiring key personnel 49.5 49.5 45.3 52.4 56.3 45.1 42.2 43.4 50.8 42.3 53.6 49.3 49.7 Licensing or turnkey operations from international sources 3.3 4.1 1.7 1.2 5.7 3.9 2.4 3.8 3.2 3.6 3.1 4.0 2.6 Licensing or turnkey operations from domestic sources 1.7 1.4 0.7 2.4 0.0 0.0 3.6 1.9 1.6 1.8 1.6 2.0 1.3 Developed or adapted within the establishment locally 39.9 46.2 44.9 35.4 37.9 51.0 39.8 30.2 42.0 36.0 42.2 41.3 38.6 Transferred from parent company 7.6 5.2 4.4 1.2 1.1 3.9 22.9 34.0 2.0 17.1 2.1 6.0 9.2 Developed in cooperation with client firms 11.2 12.0 9.8 12.2 8.0 13.7 12.0 11.3 11.2 13.5 9.9 11.3 11.1 Developed with equipment or machinery suppliers 20.1 18.2 16.2 17.1 24.1 27.5 14.5 5.7 23.2 17.1 21.9 22.0 18.3 From a business or industry association 3.3 7.9 5.7 4.9 2.3 5.9 1.2 3.8 3.2 1.8 4.2 4.7 2.0 Trade fairs and/or study tours 18.2 16.8 16.9 14.6 24.1 25.5 10.8 5.7 20.8 12.6 21.4 19.3 17.0 Consultants 7.3 6.8 5.7 1.2 3.4 5.9 18.1 13.2 6.0 9.0 6.3 10.0 4.6 From universities, public institutions 3.0 3.0 4.4 3.7 4.6 2.0 1.2 3.8 2.8 1.8 3.6 2.0 3.9 Share of Firms with a Foreign License 15.2 19.6 9.1 2.2 11.9 16.7 34.2 38.0 10.8 27.2 9.0 13.8 16.4 Share of Firms with ISO Certification 5.4 3.3 3.3 1.5 3.2 5.6 12.6 14.1 3.4 8.6 3.3 3.1 7.1 Share of Firms with Other Quality Certification 17.0 11.9 20.3 2.2 10.3 18.1 42.3 39.4 12.9 33.8 8.7 17.1 18.1 Share of Firms with Computerized equipment 21.4 23.5 11.1 2.9 8.7 34.2 50.5 43.7 17.2 40.4 11.7 18.3 24.4 Share of Firms that Developed a major new product line since 2002 46.5 53.0 47.0 41.3 44.4 56.9 48.6 45.1 47.0 48.3 45.8 50.0 43.4 Share of Firms that Upgraded an existing product line since 2002. 72.3 81.5 85.1 68.1 72.2 73.6 76.6 80.3 70.7 73.5 71.6 67.9 76.5 Share of firms that Introduced new technology since 2002 45.2 43.3 52.5 39.9 43.7 47.2 52.3 50.7 44.6 48.3 44.1 42.4 48.7 116 Table B-9: Transport comparisons in Honduras by type of firm rs ne ict orte zo all mes exp- e zone fre Honduras Guatemala Nicaragua Micro Sm Medium Large Do Foreign Non Exporters No Free Transport Interruptions Share of firms experiencing interruptions 9.3 29.9 15.5 8.6 6.3 9.6 13.5 9.2 9.9 8.7 13.6 9.0 9.9 Number of interruptions (a) 4.1 7.0 7.4 3.9 4.8 3.7 4.0 3.8 4.6 4.3 3.1 4.1 4.0 Average duration of interruptions, hours (a) 10.6 8.4 18.2 8.3 13.4 10.9 10.9 8.6 14.0 10.9 8.7 9.8 13.5 Losses due to interruptions (a) (percent of sales) 2.9 2.8 6.8 2.4 7.0 0.9 1.6 2.5 3.6 3.1 1.5 3.2 1.3 Losses while transporting goods Share of firms that experienced losses 36.9 40.0 26.3 23.0 41.7 42.5 45.1 35.9 42.3 34.5 41.7 37.9 30.5 Losses (Share of consignment value) (b) 4.5 4.2 6.2 6.2 5.7 3.7 2.5 4.9 3.8 4.9 2.5 4.8 1.9 Customs clearance times (days, mean values) Imports, average period 5.1 9.4 5.2 15.0 4.8 6.5 3.9 6.1 4.4 6.4 2.7 6.1 2.6 Imports, longest period 9.6 19.3 10.3 27.4 9.8 11.9 7.1 14.5 6.6 12.8 4.2 12.1 4.1 Exports, average period 1.9 2.3 2.0 1.6 2.1 1.6 2.0 2.2 1.9 2.0 1.8 1.7 2.3 Exports, longest period 2.9 5.5 3.0 3.0 2.9 3.0 2.8 2.5 2.9 3.3 2.3 2.9 2.8 . Notes: (a) numbers apply to those firms having transport interruptions. Losses are as percent of sales. (b) Apply to those with positive losses 117 Table B-10: Labor and training in Honduras, by firm characteristic c Micro Small Medium Large Non- Low High Foreign Domesti Exporter Exporter Capacity Capacity Honduras Guatemala Nicaragua Labor Composition Share of workers that are permanent 86.8 97.0 88.2 82.9 83.3 90.3 93.4 94.2 85.4 87.8 86.3 85.9 87.7 Share of permanent workers that are female 26.4 28.8 24.8 17.2 25.1 21.3 93.4 46.6 22.9 36.7 21.4 22.8 30.0 Share of temporary workers that are female 24.0 26.4 17.9 10.7 32.0 24.7 33.3 48.5 21.9 31.4 20.4 19.6 28.8 Share of perm. skilled workers. that are foreign nationals 0.9 0.8 0.8 0.0 0.4 1.2 2.3 3.4 0.4 1.9 0.4 0.8 0.7 Labor Turnover New employees as a share of total 14.9 10.4 10.2 11.5 15.0 13.6 20.0 20.2 13.9 20.0 12.4 14.2 15.5 Employees that left as share of total 16.7 12.1 13.7 14.3 18.6 17.0 17.5 20.4 16.0 19.3 15.5 16.7 16.8 Avg. time to fill a skilled techno. vacancy (weeks) 4.3 5.8 4.2 1.9 4.9 4.1 4.8 3.6 4.5 5.2 3.4 6.0 2.9 Avg. time to fill a prod/service worker. vacancy (weeks) 1.6 2.1 1.7 1.8 1.7 1.5 1.3 1.3 1.6 1.6 1.5 1.5 1.7 Share of firms offering formal training 49.3 55.0 37.4 23.6 47.6 67.1 79.3 73.2 47.3 66.4 44.3 49.1 50.5 Desired Change in Workforce if there were not regulatory restrictions 8.6 8.1 5.8 14.2 10.7 .3 4.8 4.7 9.3 7.5 9.1 9.8 7.37 Labor Unrest Total days lost to labor disputes or civil unrest 0.3 1.3 1.0 0.3 0.5 0.2 0.3 0.2 0.4 0.3 0.3 0.5 0.2 118 ANNEX C: INVESTMENT CLIMATE DETERMINANTS OF PRODUCTIVITY IN HONDURAS: ECONOMETERIC METHODOLOGY186 INTRODUCTION This annex is a brief summary of the econometric methodology developed by Escribano and Guasch (2004) for the analysis of the investment climate (IC) determinants of productivity, in Guatemala, Honduras and Nicaragua. The analysis focuses on the impact of the investment climate on manufacturing firms performance for Guatemala, Honduras and Nicaragua. Our dataset is a short unbalanced panel of plants in the three countries resulting from surveys conducted by the World Bank in 2003 under the Investment Climate Initiative. We have observations for the years 2001 and 2002 for most of the variables. However for the investment climate (IC) variables (see Table C-1, Table C-2 and Table C-3) we have information only for 2002. This raises the first question: should we only use cross-sectional data (year 2002) or, can we also use data for the year 2001 even if there is no information on IC variables for that year? To answer this question, we assume that the investment climate does not change much from one year to next, which is very plausible. Then we will use the information on the IC variables in 2002 also for the year 2001. What may change from one year to the next is the reaction of the plant which depends on its perceptions of the impact of IC variables and on the time to be implemented. We want to use of as many observations as possible to benefit from the law of large numbers. This suggests to pool together the data for plants in the three countries analyzing later on their differences and similarities. This aspect is important since our data is very unevenly distributed over time and across plants, precluding us from doing separate country analysis of each industry. For example, if we were to do an analysis country by country for individual industries, we would have an industry in Honduras with only 9 observations, while if we pool the three countries together we have at least 38 observations in each industry, giving more reliable statistical results, see Table C-5. To have an order of magnitude of the data problems we are facing, in year 2001 we have only 441 observations pooling the three countries, while for year 2002 we have 1,020 observations -- see Table C-6.187 Therefore, if we were to measure productivity using rates of growth, we would have at most 441 plants which is a very small sample size to study differences by industry and by country. However, if we perform the analysis in levels or logs, we can use 1461 observations in total, which should provide us with more reliable statistical results. The three countries have relatively similar numbers of observations for the two year period: Guatemala has 468 plants, Honduras has 472 plants and Nicaragua has 521 plants -- see Table C-5. 186This Annex was prepared by Alvaro Escribano, Department of Economics, Universidad Carlos III de Madrid. It is based on a paper prepared by Escribano A. and Guasch J.L. The econometric analysis was carried out by Escribano A., Guasch J.L. and Heisnam Singh. 187This figure of 441 represents observations usable for our regressions of productivity on investment climate and other plant characteristics. 119 Any empirical evaluation of the impact of investment climate (IC) variables on productivity depends on the productivity measures considered. Therefore, the only way to obtain interesting results for policy analysis is to use econometric procedures that give consistent and robust conclusions that do not depend on the specific measure of productivity used. EMPIRICAL RESULTS ON INPUT-OUTPUT ELASTICITIES AND PRODUCTIVITY ESTIMATION We use several productivity measures searching for robust results. In particular, we focus on the measures recommended in options 4.2 and 4.3 of Escribano and Guasch (2004), under alternative levels of aggregation (country, industry, young and old plants, etc.). In order to control for differences in plant productivity at different levels of aggregation, we include eleven dummy variables (D) and a constant term. We include eight dummy variables to control for a constant industry effect (apparel, beverages, chemicals/rubber, food/tobacco, furniture/wood, leather/shoes, nonmetallic minerals, textiles, metal) leaving out apparel to avoid perfect multicolinearity with the constant term. Similarly, we add only the yearly dummy for 2002. To control for country effects we only include two dummies, one for Honduras and the other for Nicaragua, leaving out Guatemala. To reduce the simultaneous equation bias generated by the fact that ICit variables may be endogenous with respect to productivity and the risk of incurring in reverse causality problems, we use averages of the ICit variables taken at the region-industry cells ( IC ). jt The coefficients on investment climate variables ( IC ) and plant specific characteristics (Cit) are it generally estimated at the same level of aggregation (i.e., one coefficient per IC or per Cit for the whole it sample) -- see Table C-14. However, we allow the production function coefficients to be estimated at different levels of aggregation -- see Table C-9, Table C-10 and Table C-11 -- generating different productivity measures across several specifications. We consider two levels of aggregation for the input coefficients: restricted (same for all industries) and unrestricted (different for each industry). We consider two functional forms for the production function: Cobb-Douglas and Translog. And we consider two types of estimator: pooling OLS and random effects. Therefore, combining these various options, we obtain twelve different measures of productivity (Pit). The impact of the IC variables on each of these productivity measures could potentially be very different. If the sign of the impact of some IC variables on productivity change depending on how productivity is measured, those empirical results are not robust and thus it would be difficult to derive reliable policy prescriptions. However, we will show that this is not the case, and that our analysis is robust even if the correlations between productivity measures vary and are not always close to one. Table C-8 reports eight correlation coefficients between the productivity measures that we get from single-step estimation (described below in section 2.1 ii)) and from Solow residuals (described below in section 2.1 i)). In the restricted case, the correlations between Solow residuals and productivity measures are very similar in all cases, ranging from 0.99 to 0.83. In the unrestricted case, the correlations between productivity measures are smaller, ranging from 0.81 in the Cobb-Douglas case to 0.23 in the Translog case. The analysis done based on different productivity measures is explained in the sub-sections that follow. 120 Restricted Coefficient Estimation i) Two-step restricted estimation We can estimate the elasticity parameters of the IC variables from equation (1b) once we have a measure of productivity (Pit). For that purpose, we first obtain the Solow residuals and subsequently evaluate the impact of IC variables on that productivity measure. First-step, generate Solow residuals (see Solow (1967)) as residuals from equation (1a): logYit = sL log Lit + sM log Mit + sK log Kit + log^ Pit (1a) ,t ,t ,t where sr,t is the average cost share of each input r = L, M and K over the last two years given by sr = (s1 + s ) ,t r,t r,t-1 2 , and s rtis the average cost share of each input taken across the entire sample of plants from the three countries (restricted case) in year t. The input cost shares are presented in Table C-9 The labor cost share is 0.36, the intermediate materials cost share is 0.53 and the capital cost share is 0.11. Cost shares add up to 1, since we are imposing constant returns to scale (CRS) in the production function. Second-step, estimate equation (1b) by regression techniques: qIC qC 11 log^Pit = IC,rlog ICrit, + C,rlogCr,it + P + D,rDr,t + uit . (1b) r=1 r=1 r=1 The empirical results from estimating equation (1b) are shown in Table C-14. All the IC variables' elasticities that we discuss below were estimated simultaneously with the industry dummies, the country dummies, the year dummies, the constant term and other plant characteristics (C). In general the estimates obtained by OLS or random effects are very similar and therefore we focus only the interpretation of the OLS parameter estimates.188 ii) Single-step restricted estimation. If we assume that the coefficients on the three inputs (L,M and K) in the Cobb-Douglas (2) and Translog (3) production functions are constant for the whole manufacturing sector after pooling the observations from the three countries, L = L , M = M and K = K , then we can apply equations (2) and (3) ,i ,i ,i below, respectively. Each of the two equations is estimated in a single-step, which implies that the input coefficients (L,M and K) are estimated jointly with the IC, C and D coefficients. However, to make the empirical results more readable, we present the input coefficients and the IC and C coefficients in separate tables. 188 Note that we will be referring below to long lists of list of IC variables and plant characteristics that were included in our initial specifications from which we iteratively dropped the non-significant variables to reach the final specifications shown in the tables in the appendix. The long list of variables is shown in Tables A.1 and A.2. For the random effects estimates of IC variables on productivity see Escribano and Guasch(2004). 121 logYit = Llog Lit + Mlog Mit +Klog Kit + qIC qC 11 (2) + IC,rlogIC r,it+ C,rlogCr,it +P + D,rDr,t + uit r=1 r=1 r=1 The empirical results from OLS and random effects estimation are presented in Table C-9 and Table C-14. In particular, Table C-14 shows the estimated elasticities of IC and C variables on productivity while Table C-9 shows the input-output elasticities of the Cobb-Douglas production function specification. The empirical results in Table C-14 are very similar those obtained with the Solow residuals, as will be mentioned in section 5. In Table C-9 we test the null hypothesis of having constant returns to scale (CRS) with the Cobb-Douglas functional form of the production function and we reject the null hypothesis with a p-value of 0.0006. logYit =Llog Lit +Mlog Mit +Klog Kit + +LL(logLit)2 +MM(logMit)2 +KK(logKit)2 + +LM(logLit)(logMit)+LK(logLit)(logKit)+MK(logMit)(logKit)+ (3) qIC qC 11 + IC,rlogIC r,it+ C,rlogCr,it +P + D,rDr,t + uit r=1 r=1 r=1 The empirical results obtained from the estimation of the extended Translog production function, equation (3), by OLS and random effects are presented in Table C-9 and Table C-14 of the appendix. Specifically, the elasticities of the IC and C variables on productivity are shown in Table C-14 and the OLS and random effects input-output elasticities are shown in Table C-9. The estimated elasticities reported in Table C-14 are again very similar to the previous ones and all the signs of the coefficients are maintained. From the estimates of the Translog production function, we tested if a Cobb-Douglas specification is accepted by the data (null hypothesis) but the hypothesis was rejected with a 0.000 p-value. We also tested for CRS in the Translog specification and again reject it with a similar p-value. The productivity measures (Pit) that we obtain from equations (2) and (3) are not measures of the total factor productivity (TFPit) that would be obtained under Hicks technological change (or neutral technical change).189 Therefore, we prefer to use the term productivity and not TFP to designate the residual output not explained by the contribution of the inputs L, M and K. 189See Escribano and Guasch (2004) for more details. 122 Unrestricted Coefficient Estimation by Industry In the unrestricted case we allow the coefficients of the inputs (L, M and K) of the production function to vary by industry: i.e. L = L , M = M for all plants i of the three countries in ,i , j ,i , j and K = K ,i , j industry j, where j = 1, 2, ... , 9. The nine j industries are mentioned in Table C-5 and Table C-6. i) Two-step unrestricted estimation. For this unrestricted case we obtain the cost share ( srt ) of each input r = L, M and K, for each of the nine manufacturing industries pooling plants in the three countries in year t. First-step, generate the Solow residuals, Solow (1967), from equation (4a) logYit = sL log Lit + sM log Mit + sK log Kit + log^ Pit (4a) , jt , jt , jt where sr is the average industry j cost share of each input r = L, M and K over the last two years given , jt = (s1 by sr ) and s , jt + sr is the average cost share of each input taken across plants in the three 2 r, jt , jt-1 jt countries that belong to industry j in year t. The cost shares by industry are reported in Table C-10. There is some heterogeneity across the nine industries, but intermediate materials has always the highest share (around 50 percent), followed by the cost share of labor (near 40 percent) and the capital cost share (around 10 percent). Second-step, estimate equation (4b) by regression techniques, qIC qC 11 log^Pit = IC,rlog ICrit, + C,rlogCr,it + P + D,rDr,t + uit . (4b) r=1 r=1 r=1 The empirical results from the OLS estimation of equation (4b) are shown in Table C-14. The results are very similar to those of the unrestricted case. The potential correlation between IC variables and the errors (uit) in equation (4b) is mitigated by replacing the plant-level IC variables by a region-industry average ( IC ). If we believe that there is some remaining endogeneity in the IC variables due to the plants' perceptions about the general business climate in their country-industry, we could use information in the survey on perception variables (PE) -- see Table C-4. Plants were asked to rank a series of topics as obstacles to the growth and operation of their business in a scale ranging from no problem to a very severe problem. We consider that if the investment climate (IC) variables are endogenous, it is because in their answers to the questions, plants are taking into account their perceptions about that particular topic. So, we could add those perception variables into equation (4b) to correct for that endogeneity.190 190See Escribano and Guasch (2004) for an explanation of this econometric correction. 123 ii) Single-step unrestricted estimation by industry Each equation is estimated by OLS and by random effects. Again, we split the information on the input- output elasticities from the information on the IC elasticities to present the results more clearly. Equation (5) is the extended Cobb-Douglas specification used to estimate, in a single step, the production function parameters and the IC and C elasticities. The results are shown in Table C-10 and Table C-14. qIC qC logYit = L,jlog Lit + M,jlog Mit +K,jlog Kit + IC,rlog IC r, jt+ C,rlogCr,t + r=1 r=1 (5) qD +P + D,rDr,t + uit r=1 The results are again very similar for pooling OLS and random effects regressions. We test the null hypothesis of constant returns to scale (CRS) in the Cobb-Douglas specification and the null hypothesis is rejected for five out of nine industries. In the last column of Table C.3, one can see that beverages (p- value=0.74), furniture/wood (p-value=0.82), nonmetallic minerals (p-value=0.83) and textiles (p- value=0.64) fail to reject CRS. The estimated input-output elasticities are similar to the cost shares used to compute the Solow residuals for the two step procedure. Hence, the corresponding productivity measures will also be similar. The question of interest is whether these new productivity measures generate similar elasticity estimates of the investment climate (IC) variables. The results are shown in Table C-14 and the main conclusion is they are very similar to those in the previous sections. Similar single step procedures are now used to estimate a Translog specification obtaining different input-output coefficients for each industry. logYit =L log Lit +M log Mit +K log Kit + , j , j , j +LL (logLit)2 +MM (logMit)2 +KK (logKit)2 + , j , j , j +LM (logLit)(logMit)+LK (logLit)(logKit)+MK (logMit)(logKit)+ (6) , j , j ,i qIC qC qD + IC,rlogIC r, jt+ C,rlogCr,it +P + D,rDr,t + uit r=1 r=1 r=1 The estimated production function parameters for equation (6) are shown in Table C-11. We test the null hypothesis of constant returns to scale (CRS) in each industry and obtain almost the same results as with the Cobb-Douglas. Five out of the nine industries reject CRS. The four industries that fail to reject CRS correspond with the Cobb-Douglas production function technologies: beverages, furniture/wood, leather and textiles. The industries that fail to reject the Cobb-Douglas specification are the same that fail to reject CRS plus the (OLS) leather industry. The results on the empirical estimates of the IC and C elasticities are shown in Table C-14. The results are similar as those described earlier in terms of the sign of the coefficients but only some are significant in Table C-14. Again, the reason for this is that the Translog specification includes many nonlinear terms of inputs L, M and K by industry that compete in terms of explanatory power with IC and other C variables. 124 Country-by-country percentage productivity gains and losses of each IC variable, relative to log productivity. In the previous section we estimated equal elasticities for the three countries by pooling all the observations. If we want to have a country-by-country evaluation of the impact of investment climate variable on productivity we can decompose the average (log) productivity of each country among the contribution of each component (IC and C) from equation say (7). Following Escribano and Guasch (2004), the natural value for this evaluation is the sample mean of each variable since, by construction, the residuals have zero mean over the pool sample. Consider the following pool OLS equation estimated to evaluate the impact of IC variables on productivity, qIC qC log^Pit = IC,rlog ICr,it + C,rlogCr,it + P +uit . (7) r=1 r=1 This productivity equation estimated by least squares with a constant term, implies that we can evaluate (7) at the mean values of each variable without an error term, qIC 100 = ^IC,rlogICrit,100+ ^C,rlogCrit,100+ ^P qC (8) r=1 log^Pit r=1 log^Pit log^Pit 100 which represent the sum of the percent productivity gains and losses from all the explanatory variables of the regression, relative to log productivity. If we do this analysis country-by-country191 we get the results of Table C-15, Table C-16 and Table C-17, which are explained in section 5. FURTHER ANALYSIS OF THE IMPACT OF INVESTMENT CLIMATE VARIABLES AGE AND BY SIZE The purpose of this section is to evaluate the impact of investment climate (IC) variables on plant productivity at different levels of aggregation. In particular, we distinguish the impact of IC variables by plant age and by plant size. Since the analysis is robust to different productivity measures, we report the results only for the two step approach using Solow residuals as in equation (4b). We first analyze the differential effects across age categories. We allow all coefficients in equation (4b) (except the country, industry and year dummies) to vary across young plants (with less than 5 years of operation) and old plants (with more than 5 years in operation). The results are shown in Table C-12. It is clear that old plants' productivity is more negatively affected by IC variables. In Table C-13 we show the results from estimating equation (4b) allowing all coefficients (except the country, industry and year dummies) to differ across small (less than 25 workers) and large (more than 26 workers) plants. The negative impact of IC variables is more important for small plants' productivity than for large plants. For large plants the only IC variable with more significant negative elasticities than small plants is the one related to crime (number of criminal attempts suffered). 191Notice that the mean of the residuals of each country might not be equal to zero, but it will be a small term. 125 PRODUCTION ALLOCATION EFFICIENCY BY COUNTRY, BY INDUSTRY AND BY AGE AND SIZE Following Escribano and Guasch (2004), we complement the productivity analysis based on regression techniques with the analysis of allocation efficiency suggested by Olley and Pakes (1996). This analysis is particularly interesting when some of the industries in the countries under analysis have very small samples since one cannot credibly perform country by country and/or industry by industry regressions to evaluate the impact of IC variables on productivity. The Olley and Pakes decomposition provides additional insights on the efficiency of the allocation of resources within each country. The cross-sectional decomposition suggested by Olley and Pakes for industry (log) productivity Pkt in year t is given by: Nkt Pkt = Pkt + %k,itP%k,it . (9) t=1 Nkt The first term ( Pkt ) is the sample average (log) productivity of industry k and the second term s% P% k,it k,it t=1 measures the covariance between plant productivity and plant market share, where Nkt %k = (k -k )and P%k = (Pk -Pk ) represent deviations from the mean. Pkt = itPk,it is ,it ,it ,i ,it ,it ,t i=1 industry productivity obtained as a weighted average of plant productivity for plants in industry k and year t. Nkt is the number of firms in industry k and year t and k = 1, ... ,9 represents the industries in our sample listed in Table C-5. The weights it represent the share of plant i sales in total sales of industry k. If the covariance term is positive, there is an efficient allocation of resources where more productive plants have higher market shares. The larger the covariance, the higher is the share of sales made by more productive plants and the higher is industry productivity. If the covariance term is negative, there is an inefficiency in resource allocation. The more negative is the covariance term, the higher is the share of sales made by less productive plants, reducing therefore industry productivity. We perform three types of Olley and Pakes decompositions. First, we construct a measure of productivity for each of the countries and decompose it into average productivity and the covariance term. The results are shown in Figure C-1 and suggest that more productive plants have higher market shares in Guatemala, i.e. there is an efficient allocation of resources but that is not verified in Honduras nor in Nicaragua. Second, we construct a measure of productivity for each group of plants -- small and large - in each of the countries and decompose it into the two components. The results are shown in Figure C-2. In Guatemala, resources are efficiently allocated for small plants as well as for large plants. In Honduras and in Nicaragua, the less productive small and large plants have the higher market (inefficiency). Third, we construct a measure of productivity for each industry country-by-country and perform its decomposition.192 The results are shown in the three panels of Figure C-3. In Guatemala, we find an efficient allocation of resources in all the industries. In beverages and nonmetallic minerals the covariance terms are the highest. In Honduras, the allocation of resources is inefficient in beverages, food and tobacco and in textiles and efficient in the remaining industries. In Nicaragua, there is an inefficient 192 The details on the decomposition were described in equation (9). For more details see Escribano and Guasch (2004). 126 allocation of resources in five out of nine industries: apparel, beverages, leather and shoes, nonmetallic minerals and textiles. SUMMARY OF EMPIRICAL RESULTS ON INVESTMENT CLIMATE (IC) DETERMINANTS OF PRODUCTIVITY IN HONDURAS Four important groups of investment climate (IC) variables were identified: a) Red Tape, Corruption and Crime, b) Infrastructure, c) Quality, Innovation and Labor Skills, d) Finance and Corporate Governance. Within each group, all the IC variables always have the expected signs and the estimated elasticities or semi-elasticities were always within a reasonable range of values for the twelve productivity measures considered. In absolute terms the higher values of the elasticities correspond to the Solow residual or to the Cobb-Douglas specification, while the lowest usually correspond to the Translog production function. Therefore, we observe a trade-off between the role played by the inputs (labor, intermediate materials and capital) and the role played by the IC variables and other control variables. The robustness of these empirical results across productivity measures allows us to obtain consistent empirical evaluations of the IC determinants of productivity. The summary the main empirical results are divided in three parts: i) Productivity elasticities or semi-elasticities of IC variables and percentage R2 contribution of each IC variable for the pool of three countries (Guatemala, Honduras and Nicaragua) ii) Country-by-country percent productivity gains and losses of each IC variable, relative to log productivity iii) Olley and Pakes (1996) productivity decomposition for each country. Going by parts: i) Productivity elasticities or semi-elasticities of IC variables and percent R2 contribution of each IC variable for the pool of three countries (Guatemala, Honduras and Nicaragua), see Table D. Red Tape, Corruption and Crime a. If the plants dedicate one more day to inspection and regulation control activities the decrease in productivity will be between 5.8 percent and 10.7 percent. This is most relevant for old firms. b. A one percent increase in the fraction of undeclared sales to IRS decreases productivity between 0.42 and 0.77 percent. The impact is especially important in small plants. c. Those firms that win a competitive advantage by making payments to "speed up" bureaucratic issues will have an increases in productivity ranging from 1.3 to 3.3 percent. In terms of policy recommendations, there is room for improvements in the 127 administrative procedures followed in the three countries, so that no more arbitrary administrative gains in productivity exist from bribes of firms. d. Crime. For every new criminal attempt suffered by the firm the productivity is reduced between 1.8 and 3.2 percent. This is most relevant in old and large firms. Infrastructure e. Power. A one percent increase in the average duration of power outages decreases productivity between 0.02 and 0.1 percent. It mainly affects old plants. f. Transportation First, a one percent increase in the average number of days to clear customs will decrease productivity by a 0.1 percent. It specially affects young and small plants. Second, a one percent increase in the fraction of shipment losses due to breakage, theft, spoilage or other deficiencies of the transport mean used, will decrease productivity between 1.23 and 2.53 per cent. This is most important in old and small firms. g. Telecommunications. Firms with access to internet are between 11 percent and 15 percent more productive that those firms without it. The impact is especially relevant in old firms. Quality, Innovation and Labor Skills h. If the plants increase by one percent their fraction of computer controlled machinery the productivity increase will be 0.1 percent. This is relevant in old and large firms. If the plants increase the fraction of total staff dedicated to R&D activities by 1 percent the productivity will increase between 0.6 and 0.7 percent. The most important impact is on old and small firms. i. When the firm is engaged in a process of ISO quality certification its productivity will increase between 2.4 and 17.6 percent. This is specially relevant for large firms. j. A one percent increase in the fraction of total staff with secondary education of the firm implies an increase in productivity between 0.03 and 0.06 percent. k. If a firm engage in a training program for their workers, other than on the job training, they will experience an increase in productivity ranging from 8.9 to 11.7 percent. This is specially relevant for small plants. Finance and Corporate Governance. l. A firm by engaging in an external auditing of its financial statements can increase its productivity between 11.6 and 17 percent. m. By becoming a Sociedad Anonima, a firm increase its productivity between 11.5 and 15 128 percent. Other Control variables. n. If the experience of the firm in the market, measure by the age, increases by on percent then the productivity will increase between 0.03 and 0.05 percent. o. If the fraction of imported inputs for production is increased by one percent then the productivity of the firm will increase 0.1 percent. This is most relevant for old and small firms. ii) Country-by-country percent productivity gains and losses of each IC variable, relative to (log) productivity, see Tables E1 to E3. To understand how those IC productivity gains and losses are calculated see equation (8). Red Tape, Corruption and Crime a. The negative contribution of the average number of days spent in inspection and regulation related work is larger, in absolute value, in Nicaragua than in Guatemala and Honduras. b. The negative contribution of the average fraction of sales undeclared to IRS for tax purposes is smaller, in absolute value, in Guatemala than in Honduras and Nicaragua. c. The firms from Guatemala that win a competitive advantage by making payments to "speed up" bureaucratic issues have a larger contribution to the average productivity than those of Honduras and Nicaragua. d. Crime. The negative contribution of the average number of criminal attempts is larger, in absolute value, in Guatemala than in Honduras and Nicaragua. Infrastructure. e. Power. The negative contribution of the average duration of power outages is similar for Honduras and Nicaragua and is smaller, in absolute value, for Guatemala. f. Transportation. The negative contribution of the average number of days to clear customs for imports is larger, in absolute value, for Guatemala than for Honduras and Nicaragua. g. Transportation. The negative contribution to the country average productivity of the average fraction of shipment losses is similar for the three countries. h. Telecommunications. The positive contribution of the probability that a firm has access 129 to internet is larger for Guatemala than for Honduras and Nicaragua. Quality, Innovations and Labor Skills i. The positive contribution of the country average fraction of computer controlled machinery is similar in all countries. j. The positive contribution of the average fraction of total staff in R&D activities is similar for the three countries. k. The positive contribution of the probability of the firm engaging in a process of ISO quality certification is similar for the three countries. l. The positive contribution of the country average fraction of total staff with secondary education or more is similar for the three countries. m. The positive contribution of the probability of a firm engaging in a training program for their workers, other than on the job training is similar for the three countries. Finance and Corporate Governance. n. The contribution of the probability that the firms engage in an external auditing of its financial statements is similar for the three countries. o. The positive contribution of the probability of firms becoming Sociedades Anonimas is larger for Guatemala than for Honduras and Nicaragua. Other Control variables. p. The experience of the firm in the market, measured by the age, has a positive contribution to the country average productivity and its is of similar magnitude in the three countries. q. The positive contribution of the country average fraction of imported inputs for production is similar for the three countries. iii) Olley and Pakes (1996) productivity decomposition for each country, see Figure C-1, Figure C-2, Figure C-3 and equation (9). The Olley and Pakes decomposition provides additional insights on the efficiency of the allocation of resources in each country. The country aggregate productivity (weighted by the share of sales of each firm) is decomposed as the (unweighted) average productivity and the covariance productivity term. The larger is the positive covariance term, the higher is the share of sales made by more productive plants, increasing the industry productivity (efficiency). The more negative is the covariance term, the higher is the share of sales made by less productive plants, reducing industry productivity (inefficiency). 130 Figure 1 showed the decomposition of the three countries. The interpretation of the picture is clear; Guatemala has a covariance productivity term which is positive while Honduras and Nicaragua have a negative covariance. Therefore, in Guatemala the most productive firms have the highest share of sales while in Honduras and Nicaragua the less productive firms have the largest share of the market. This efficiency picture is maintained for Guatemala when we split the sample in small and large firms but it changes with Honduras and Nicaragua. In the case of Honduras the aggregate negative productivity covariance is mainly obtained because the negative productivity covariance of small firms. Therefore, in Honduras small firms allocate resources more inefficiently than large firms. For Nicaragua, the negative aggregate productivity covariance comes from the dominant negative covariance of large firms. Therefore, in Nicaragua small firms allocate resources more efficiently than large firms. From the analysis of the six different productivity measures we check the robustness of the covariance results. The signs of the covariance terms of Figure C-1 are maintained in each country for the restricted model in all the cases but in one of the productivity measures of Nicaragua -- see Table C7.1. The results of Figure C-2 are also maintained when we split the different productivity measures, of each country, in large and small firms, see Table C7.2. However, the efficiency results are not robust to alternative productivity measures in the unrestricted models of Honduras and Nicaragua, where we allow the elasticity of the inputs to vary sector by sector in the Cobb-Douglas and Translog specifications, see the last two rows of Table C7.1 and the last three rows of Tables C7.2. This is so because we have only few observations in some of the industries. 131 Table C-1: General Plant Information and Production Function Variables Industrial apparel, beverages, chemicals/rubber, food/tobacco, classification furniture/wood, leather/shoes, nonmetallic minerals, textiles, metal products. General Plant Regional - Guatemala: Guatemala City, Metropolitan area close to Information classification Guatemala city, Metropolitan area far from Guatemala city, Altiplano region, Coastal region, Northeast region - Honduras: Western region, Center-South region, Olancho region, North coast region - Nicaragua: Managua region, Pacific region. Sales used as the measure of output for the production function estimation. For all countries sales figures in local currency are converted into USD using IMF average exchange rates. Production Employment total number of permanent workers. Function Materials total costs of materials and raw materials used in production Variables (excluding fuel). For all countries materials figures in local currency are converted into USD using IMF average exchange rates. Capital stock book value of all fixed assets. Labor cost total expenditures with personnel. 132 Table C-2: Investment Climate (IC) Variables Power interruptions total duration of power outages suffered by the plant in hours (equals average duration times the total number of power outages). If the plant suffered no power outages the total duration of power outages is 0. Duration of power average duration of power outages suffered by the plant in hours. outages Power outages number of power outages suffered by a plant in 2002. Losses due to power value of the losses due to the power outages as a percentage of sales outages (conditional on the plant reporting power outages). Production dummy variable =1 if the plant had to stop production due to power interruption due to outages. power outages Phone interruptions total duration of phone interruptions suffered by the plant in hours Infrastructure (equals average duration times the total number of phone outages). If the plant suffered no phone interruptions the total duration of phone interruptions is 0. Wait for phone number of days the plant had to wait from the moment it requested a phone connection until the day it received the phone service. Days to clear average number of days that it took from the time the plant's imports customs for imports arrived to the point of entry until the time the plant could claim them from customs. Fraction of shipment fraction of the value of the plant's average cargo consignment that was losses lost in transit due to breakage, theft, spoilage or other deficiencies of the transport means used. Production stopped dummy variable =1 if the plant had to stop production due to transport due to transport interruptions interruptions Inspection days number of days spent by the plant in inspections and compulsory meetings with public officials from the following areas: tax inspectorate, labor and social security, fire and building. Percentage of time percentage of time in a typical week spent by management dealing with Bureaucracy/ spent dealing with bureaucracy/regulation. Corruption regulation Payments to deal payments to "speed up" bureaucratic issues as a percentage of sales. with bureaucracy "faster" Fraction of sales Fraction of total sales unreported to the IRS for tax purposes by a undeclared to the IRS typical firm in an area of activity as perceived by the reporting firm Losses due to crime as value of the plant's losses due to criminal activity as a percentage of sales. For a percentage of sales plants reporting no criminal activity the value of the variable is 0. Crime Number of criminal number of crimes is the number of criminal attempts suffered by the plant. attempts suffered Overdraft dummy variable =1 if the plant reports that it has an overdraft facility or a line of credit. Loan dummy variable =1 if the plant reports that it has a bank loan. Finance Constrained access to dummy variable =1 if i) the plant did not ask for a loan (except if the plant did loan not ask for a loan because it did not need it) or ii) the plant asked for a loan but its request was refused by the bank. 133 Table C-3: IC and Plant Characteristics (C) Variables Percentage of computer- percentage of computer-controlled machinery at the plant. controlled machinery Use of foreign technology dummy variable =1 if the plant uses technology obtained from foreign firms paying or not paying licenses for that use. Introduction of new products dummy variable =1 if the plant introduced new products in the last 3 years. R&D Fraction of total staff engaged en R&D activities Training dummy variable =1 if the plant provides training to its employees other than on the job Percentage of staff with percentage of workers (including management) with at least secondary education secondary education. Domestic market share how much the plant's sales represent out of total national market. Exporter dummy variable =1 if the plant exports at least 10 percent of its Other Plant output. Characteristics Foreign ownership dummy variable =1 if plant has any share of its capital that is foreign. Age difference between 2002 and the year that the firm started operations in the country. Capacity utilization percentage of capacity utilized Inverse number of 1/number of competitors of the plant. competitors Internet access dummy variable=1 if the plant uses email or a website in its interactions with clients or suppliers. ISO certification Firm has ISO Quality certification External audit Dummy variable=1 if the firm's annual financial statement is reviewed by an external auditor Share of imported inputs Fraction of material inputs that are imported Sociedad Anonima Dummy variable=1 if the firm is publicly listed 134 Table C-4: Investment Climate -- Perception (PE) Variables Electricity problem ranking by the plant of electricity as a problem for its operations and growth. Infrastructure Generator dummy variable =1 if the plant has a generator. Customs problem ranking by the plant of customs as a problem for its operations and growth. Export experience number of years since the plant first started exporting. Tax administration ranking by the plant of tax administration as a problem for its problem operations and growth. Trade regulation ranking by the plant of trade regulation as a problem for its problem operations and growth. Labor regulation ranking by the plant of labor regulation as a problem for its problem operations and growth. Bureaucracy/ Permit problem ranking by the plant of permit and business registration as a Corruption problem for its operations and growth. Regulatory policy ranking by the plant of regulatory policy uncertainty as a problem uncertainty problem for its operations and growth. Government dummy variable =1 if the plant answered that the government is inconsistency very inconsistent or inconsistent in interpreting laws and regulations affecting the plant. Corruption problem ranking by the plant of corruption a problem for its operations and growth. Crime problem ranking by the plant of crime a problem for its operations and Crime growth. Security cost expenditure on security related items by the plant as percentage of sales. Finance availability ranking by the plant of availability of finance as a problem for its Finance problem operations and growth. Cost of finance ranking by the plant of cost of finance as a problem for its problem operations and growth. Finance access problem ranking by the plant of access to finance as a problem for its operations and growth. Table C-5: Number of firms that enter in the IC regressions, by Industry and by Country INDUSTRY GUATEMALA HONDURAS NICARAGUA TOTAL Apparel 129 70 64 263 Beverages 8 19 17 44 Chemical/Rubber 61 35 67 163 Food/ Tobacco 102 134 68 304 Furniture/ Wood 56 126 127 309 Leather/ Shoes 6 0 45 51 Nonmetallic Minerals 36 44 69 149 Textiles 22 9 7 38 Metal Products 48 35 57 140 Total 468 472 521 1,461 135 Table C-6: Number of firms that enter in the IC regressions, by Industry and by Year INDUSTRY 2001 2002 TOTAL Apparel 71 191 263 Beverages 12 32 44 Chemical/Rubber 52 111 163 Food/ Tobacco 103 201 304 Furniture/ Wood 90 219 309 Leather/ Shoes 15 36 51 Nonmetallic Minerals 44 105 149 Textiles 8 30 38 Metal Products 45 95 140 Total 441 1,020 1,461 136 Table C-7: List of Significant ICA Variables and their Units of Measurement EXPLANATORY ICA VARIABLES UNITS OF MEASUREMENT Average duration Hours per day (in logs) of power outages Average number of days Days (in logs) to clear customs for imports Shipment Losses Fraction of total sales (per unit) Dummy for Internet Access 0 or 1 Number of days spent in Inspection Days and Regulation related work Fraction of sales declared to IRS for Fraction of Total Sales (per unit) tax purposes Payments to deal with bureaucracy Percentage of total sales (per 100) "faster" as a percentage of sales Number of criminal numbers attempts suffered Percentage of computer-controlled Fraction of total machinery (per machinery unit) Percentage of total staff engaged in R Fraction of total staff (per unit) & D Dummy for ISO quality certification 0 or 1 Percentage of total staff with Fraction of total staff ( per unit) secondary education or more Dummy for Training provided 0 or 1 beyond "on the job" Dummy for Sociedad Anonima 0 or 1 Dummy for external audit of 0 or 1 Financial statements Age of the firm Age in years (in logs) Share of Imported inputs Fraction of total inputs( per unit) * In all the regressions, outlier plants were excluded. Outlier plants were defined as those which had ratios of materials to sales larger than 1 or had ratios of labor costs to sales larger than 1. 137 Table C-8: Correlation between Solow residuals in levels and estimated productivity SOLOW RESIDUALS Restricted Unrestricted by Industry OLS 0.99 0.81 Cobb Douglas Estimated Random Effect 0.98 0.74 Productivity OLS 0.86 0.27 Translog Random Effect 0.83 0.23 Notes: a) Solow residuals in levels are obtained as sales (in logarithms or logs) minus a weighted average of labor, materials, capital (all in logs) where the weights are given by the share in total costs of each of the inputs. The cost of capital is assumed to be equal to 10 percent times the capital stock. (1) Restricted case: the cost shares are calculated as the averages of the plant-level cost shares across the entire sample of 3 countries in years 2002 and 2001. (2) Unrestricted by Industry case: the cost shares are calculated as the averages across plant-level cost shares in years 2002 and 2001 for each of 9 industries (for each industry all plants of the 3 countries that are not outliers are considered). (3) Outlier plants were defined as those which had ratios of materials to sales larger than 1 or had ratios of labor costs to sales larger than 1. b) Estimated Productivity in levels (logs) is obtained from Cobb-Douglas and Translog production functions of sales with inputs labor, materials, capital estimated by OLS and random effects under 2 different environments: (1) Restricted: a single set of production function coefficients is obtained using data on plants in the 3 countries, in all industries in years 2002 and 2001. (2) Unrestricted by Industry: a set of production function coefficients is obtained for each of 9 industries using data on all plants in the 3 countries in years 2002 and 2001 (excluding outliers). 138 Table C-9: Production function parameters from the restricted estimation Labor(L) Material Capital(K) L2 M2 K2 L*M L*K M*K Input(M) Cost Shares 0.36 0.53 0.11 - - - - - - SAL OLS 0.423*** 0.537*** 0.060*** - - - - - - Random 0.489*** 0.448*** 0.065*** - - - - - - OUGD Effects COBB Test for Constant Returns to Scale OLS Prob > F = 0.000 R.E. Prob > chi= 0.0006 OLS 1.329*** -0.11 -0.009 0.070*** 0.067*** 0.017** -0.141*** 0.016 -0.032** GO Random 1.545*** -0.315*** -0.164** 0.064*** 0.053*** 0.012** -0.121*** -0.014 0.001 SL Effects RANT Test for Cobb Douglas OLS Prob > F = 0.000 R.E. Prob >chi2 = 0.000 Test for Constant Returns to Scale OLS Prob > F = 0.000 R.E. Prob >chi2 = 0.000 Notes: (1) Significance is given by robust standard errors.* significant at 10%; ** significant at 5%; *** significant at 1%. (2) The cost shares of labor, materials and capital are calculated as averages of the plant-level cost shares of labor, materials and capital across all plants in the 3 countries in years 2002 and 2001 (excluding outliers). (3) The sample generating the sets of production function coefficients is constituted by all plants in the 3 countries in years 2002 and 2001 (excluding outliers). (4) 0 indicates that the coefficient on that input in that industry was not significant and hence is restricted to be equal to 0. 139 Table C-10: Production function parameters from the unrestricted estimation by industry: Cobb Douglas Specification. TABLE C.3: PRODUCTION FUNCTION PARAMETERS FROM THE UNRESTRICTED ESTIMATION BY INDUSTRY: COBB DOUGLAS SPECIFICATION Industry Coefficients Labor(L) Material Capital(K) Test for Input (M) Constant Returns to Scale ler Cost share 0.41 0.49 0.10 pa OLS 0.418*** 0.478*** 0.098*** Prob > F = 0.8959 Ap Random Effects 0.403*** 0.463*** 0.081*** Prob > chi2 = 0.0846 s Cost share 0.30 0.58 0.12 ge OLS 0.595*** 0.377*** 0.094 Prob > F = 0.5007 veraeB Random Effects 0.563*** 0.391*** 0.085 Prob > chi2 = 0.7420 / Cost share 0.33 0.59 0.08 alsicm reb OLS 0.387*** 0.597*** 0.085** Prob > F = 0.0179 Rub Random Effects 0.454*** 0.523*** 0.094** Prob > chi2 =0.1331 Che / o Cost share 0.35 0.55 0.11 od cc OLS 0.516*** 0.517*** 0.034 Prob > F = 0.0051 Fo obaT Random Effects 0.658*** 0.366*** 0.051** Prob > chi2 =0.0248 /We Cost share 0.35 0.53 0.12 d OLS 0.327*** 0.655*** 0.019 Prob > F = 0.9477 itur oo Random Effects 0.396*** 0.591*** 0.022 Prob > chi2 = 0.8243 Furn r/e Cost share 0.37 0.53 0.10 s oe OLS 0.866*** 0.285* 0.084*** Prob > F = 0.0467 Leath Sh Random Effects 0.790*** 0.286** 0.092 Prob > chi2 =0.2020 c Cost share 0.34 0.55 0.11 lli eta alser OLS 0.346*** 0.560*** 0.103*** Prob > F = 0.8921 nm Random Effects 0.473*** 0.436*** 0.106** Prob > chi2 =0.8308 Min No Cost share 0.36 0.51 0.13 tiles OLS 0.361*** 0.374*** 0.218* Prob > F = 0.5458 Tex Random Effects 0.428*** 0.310*** 0.221** Prob > chi2 = 0.6433 st Cost share 0.38 0.51 0.11 tal uc OLS 0.352*** 0.567*** 0.055 Prob > F = 0.5050 Me odrP Random Effects 0.402*** 0.506*** 0.06 Prob > chi2 = 0.5605 Notes: (1) Significance is given by robust standard errors. * significant at 10%; ** significant at 5%; *** significant at 1%. (2) The cost shares of labor, materials and capital are calculated as averages of the plant-level cost shares of labor, materials and capital for each industry using all plants in the 3 countries in years 2002 and 2001 (excluding outliers). 140 Table C-11: Production function parameters from the unrestricted by industry estimation: Translog specification. TABLE C.4: PRODUCTION FUNCTION PARAMETERS FROM THE UNRESTRICTED BY INDUSTRY ESTIMATION: TRANSLOG SPECIFICATION Labor(L) Material Capital L2 M2 K2 L*M L*K M*K Test for Test for Input(M) (K) CD* CRS p values p values Appar OLS 0.742*** -0.306 0.520*** 0.051** 0.126*** 0.022* -0.177*** 0.106*** -0.119*** 0.0000 0.0000 R.E. 0.537** -0.234 0.563*** 0.013 0.120*** 0.019* -0.151*** 0.123*** -0.123*** 0.0000 0.0000 Bev. OLS 0.598*** 0.378*** 0.092 0 0 0 0 0 0 CD 0.3602 R.E. 0.566*** 0.391*** 0.091 0 0 0 0 0 0 CD 0.5588 Chem. OLS 1.278*** 0.316 0.057 0.268*** 0.069*** 0 -0.152*** 0 0 0.0001 0.0025 R.E. 1.177*** 0.370* 0.069* 0.265*** 0.056** 0 -0.139*** 0 0 0.0214 0.0634 Food OLS 2.139*** -0.760*** 0.029 0.309*** 0.182*** 0 -0.234*** 0 0 0.0000 0.0000 R.E. 1.742*** -0.707*** 0.053** 0.188*** 0.144*** 0 -0.157*** 0 0 0.0000 0.0000 Furn. OLS 0.336*** 0.659*** 0.02 0 0 0 0 0 0 CD 0.7126 R.E. 0.407*** 0.597*** 0.023 0 0 0 0 0 0 CD 0.5302 Leather OLS 0.878*** 0.288* 0.086*** 0 0 0 0 0 0 CD 0.0207 R.E. 0.816*** 0.285*** 0.096 0 0 0 0 0 0 CD 0.1025 NonMet. OLS 0.305*** -0.356 0.096*** 0 0.086** 0 0 0 0 0.0288 0.0833 R.E. 0.415*** -0.508* 0.099** 0 0.090*** 0 0 0 0 0.0005 0.0022 Textiles OLS 0.359** 0.374*** 0.225* 0 0 0 0 0 0 CD 0.5901 R.E. 0.429*** 0.311*** 0.232** 0 0 0 0 0 0 CD 0.7078 Metal OLS 0.732** 0.036 0.216 0 0.102*** 0.026 -0.033 0 -0.043* 0.0294 0.0246 R.E. 0.846** -0.266 0.283 0 0.118*** 0.01 -0.041 0 -0.033 0.1375 0.1354 Notes: (1) Significance is given by robust standard errors. * significant at 10%; ** significant at 5%; *** significant at 1% (2) CD indicates that a Cobb-Douglas specification was estimated since it was preferred to the Translog specification. (3) 0 indicates that the coefficient on that input in that industry was not significant and hence restricted to be equal to 0. 141 Table C-12: Two step unrestricted by industry estimation for young and old firms. TABLE C 5.1: TWO STEP UNRESTRICTED BY INDUSTRY ESTIMATION FOR YOUNG AND OLD FIRMS Dependent Variable: Unrestricted by Industry Solow Residuals in Levels (logs) Young Plants Old Plants Number of days spent in Inspection and -0.077 -0.081* Regulation related work [0.067] [0.046] Fraction of sales undeclared to the tax -0.261 -0.481 Red Tape, Corruption and authority for tax purposes [0.582] [0.294] Crime Payments to deal with bureaucracy 0.037* 0.036** "faster", percent of sales [0.019] [0.014] -0.009 -0.029** Number of criminal attempts suffered [0.021] [0.012] -0.085 -0.097* Average duration of power outages (log) [0.098] [0.053] -0.168** -0.067 Days to clear customs for imports (log) [0.074] [0.043] Infrastructure -0.948 -2.321** Shipment Losses (fraction of Sales) [1.037] [1.181] 0.1 0.156*** Dummy for Internet Access [0.073] [0.044] Fraction of computer-controlled -0.191 0.218** machinery [0.193] [0.086] -0.369 0.773** Fraction of total staff engaged in R&D [0.567] [0.320] Quality, Innovation and 0.12 0.16 Dummy for ISO quality certification Labor Skills [0.196] [0.110] Fraction of total staff with secondary 0.09 0.06 education or higher [0.115] [0.075] Dummy for Training provided beyond "on 0.144** 0.093** the job" training [0.067] [0.042] 0.199** 0.132*** Dummy for publicly listed firm Finance and Corporate [0.087] [0.044] Governance Dummy for external audit of Financial 0.212*** 0.152*** statements [0.073] [0.049] 0.03 0.022 Age of the firm (log) [0.076] [0.030] Others Control Variables -0.016 0.133** Share of Imported inputs (fraction) [0.103] [0.055] Observations 1461 1461 R-squared 0.27 0.27 Test of joint significance Prob > F = 0.0004 Prob > F = 0.0004 of IC Variables Test of joint significance Prob > F = 0.0000 Prob > F = 0.0000 of ICVs and Plant level controls Notes: (1) All the coefficients in the table are obtained from a single regression. (2) Significance is given by robust standard errors. * significant at 10%; ** significant at 5%; *** significant at 1%. (2) The unrestricted by industry Solow Residual is obtained using cost shares for inputs (labor, materials and capital) obtained as averages for each of the industries using plants in the 3 countries in years 2001 and 2002. 142 Table C-13: Two step unrestricted by industry estimation for small and large firms. TABLE C 5.2: TWO STEP UNRESTRICTED BY INDUSTRY ESTIMATION FOR SMALL AND LARGE FIRMS Dependent Variable: Unrestricted by Industry Solow Residuals in Levels (logs) Small Plants Large Plants Number of days spent in Inspection and -0.075 -0.089 Regulation related work [0.048] [0.060] Fraction of sales undeclared to the tax -0.491* -0.322 Red Tape, Corruption and authority for tax purposes [0.283] [0.499] Crime Payments to deal with bureaucracy 0.033** 0.032* "faster", percent of sales [0.015] [0.017] -0.016 -0.039*** Number of criminal attempts suffered [0.016] [0.014] -0.077 -0.085 Average duration of power outages (log) [0.059] [0.080] -0.088* -0.06 Days to clear customs for imports (log) [0.045] [0.057] Infrastructure -1.733* -1.412 Shipment Losses (fraction of Sales) [1.008] [1.255] 0.125** 0.181*** Dummy for Internet Access [0.049] [0.060] Fraction of computer-controlled 0.06 0.174* machinery [0.138] [0.105] 0.530* -0.183 Fraction of total staff engaged in R&D [0.301] [1.395] Quality, Innovation and 0.082 0.206* Dummy for ISO quality certification Labor Skills [0.153] [0.111] Fraction of total staff with secondary 0.072 0.016 education or higher [0.078] [0.103] Dummy for Training provided beyond "on 0.138*** 0.069 the job" training [0.047] [0.053] 0.148*** 0.126** Dummy for publicly listed firm Finance and Corporate [0.053] [0.060] Governance Dummy for external audit of Financial 0.166*** 0.185*** statements [0.059] [0.057] 0.052* 0.055** Age of the firm (log) [0.029] [0.024] Others Control Variables 0.137** 0.072 Share of Imported inputs (fraction) [0.063] [0.077] Observations 1461 1461 R-squared 0.26 0.26 Test of joint significance Prob > F = 0.0000 Prob > F = 0.0000 of IC Variables Test of joint significance Prob > F = 0.0000 Prob > F = 0.0000 of ICVs and Plant level controls Notes: (1) All the coefficients in the table are obtained from a single regression. (2) Significance is given by robust standard errors. * significant at 10%; ** significant at 5%; *** significant at 1%. (2) The unrestricted by industry Solow Residual is obtained using cost shares for inputs (labor, materials and capital) obtained as averages for each of the industries using plants in the 3 countries in years 2001 and 2002. (4) The regressions include a constant, industry dummies, country dummies and year dummies. (5) Small plants are defined as those having up to 25 employees and large plants are defined as those having 26 or more 143 Table C-14: Elasticities or semi-elasticities and percentage of R-square Productivity Contribution of Each explanatory variable, after controlling for the other IC and Plant Control variables. Table D: Elasticities or Semi-elasticities and % R-square Productivity Contribution of Each Explanatory Variable, after Controling for the other IC and Plant Control Variables Restricted OLS Estimation Unrestricted OLS Estimation Solow Res. 1 step 1 step Cobb Douglas Translog Solow Res. Cobb Douglas Translog Red Tape, Corruption and Crime Number of days spent in Inspection and -0.097** -0.099** -0.058 -0.101** -0.107** -0.068* Regulation related work 0.76% 0.79% 0.26% 0.81% 0.97% 0.45% Fraction of sales undeclared to the tax -0.601** -0.612** -0.416* -0.676** -0.767*** -0.593** authority for tax purposes 0.77% 0.81% 0.65% 0.97% 1.31% 0.92% Payments to deal with bureaucracy "faster", 0.031** 0.033*** 0.015 0.031** 0.030** 0.013 percent of sales 0.76% 0.84% 0.18% 0.74% 0.75% 0.16% Number of criminal attempts suffered -0.029** -0.031*** -0.018 -0.032*** -0.029** -0.018 0.73% 0.85% 0.21% 0.88% 0.75% 0.36% Infrastructure Average duration of power outages (log) -0.095* -0.085* -0.072 -0.088* -0.075 -0.024 0.35% 0.29% 0.18% 0.30% 0.23% 0.03% Days to clear customs for imports (log) -0.097** -0.106** -0.125*** -0.105*** -0.105** -0.119*** 0.93% 1.11% 1.91% 1.08% 1.14% 1.67% Shipment Losses (fraction of Sales) -1.860** -2.119** -1.229 -1.948** -2.530*** -2.063** 0.20% 0.26% 0.18% 0.22% 0.38% 0.30% Dummy for Internet Access 0.147*** 0.144*** 0.119*** 0.139*** 0.128*** 0.111*** 1.51% 1.47% 4.99% 1.35% 1.28% 1.06% Quality, Innovation and Labor Skills Fraction of computer-controlled machinery 0.119 0.13 0.132* 0.117 0.131 0.084 0.14% 0.17% 0.47% 0.13% 0.18% 0.09% 0.594** 0.667** 0.581** 0.589** 0.607** 0.580** Fraction of total staff engaged in R&D 0.40% 0.51% 0.02% 0.39% 0.44% 0.47% 0.154 0.167 0.024 0.142 0.176* 0.105 Dummy for ISO quality certification 0.27% 0.32% 0.06% 0.23% 0.37% 0.16% Fraction of total staff with secondary 0.036 0.048 0.03 0.033 0.054 0.056 education or higher 0.03% 0.06% 0.02 0.03% 0.08% 0.10% Dummy for Training provided beyond "on 0.117*** 0.105*** 0.089*** 0.116*** 0.110*** 0.098*** the job" training 0.95% 0.78% 2.15% 0.94% 0.90% 0.81% Finance and Corporate Governance 0.150*** 0.140*** 0.115*** 0.146*** 0.132*** 0.117*** Dummy for publicly listed firm 1.48% 1.31% 5.19% 1.41% 1.21% 1.12% Dummy for external audit of Financial 0.168*** 0.159*** 0.121*** 0.170*** 0.150*** 0.116*** statements 1.80% 1.64% 4.06% 1.84% 1.52% 1.06% Others Control Variables 0.050** 0.049** 0.046** 0.051*** 0.043** 0.032* Age of the firm (log) 0.48% 0.46% 0.93 0.50% 0.37% 0.23% 0.113** 0.101** 0.096** 0.110** 0.074 0.101** Share of Imported inputs (fraction) 0.55% 0.45% 2.32% 0.53% 0.25% 0.55% R-Squared 0.2 0.2 0.36 0.26 0.51 0.94 Notes: (1) Robust standard errors shown in parentheses under coefficient estimates. (2) Significance is given by robust standard errors. * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. 144 Table C-15: Guatemala -- percent average (log) productivity gains and losses due to Investment Climate (IC) Table E.1: GUATEMALA % Average (log) Productivity Gains and Losses in Guatemala due to Investment Climate (IC) Restricted OLS Estimation Unrestricted OLS Estimation 1 Step 1 Step Solow Res. Cobb Douglas Translog Solow Res. Cobb Douglas Translog Red Tape, Corruption and Crime Number of days spent in Inspection and -5.89** -5.61** -1.97 -6.09** -6.23** -2.74* Regulation related work Fraction of sales undeclared to the tax authority -3.31** -3.16** -1.73* -3.71** -4.04*** -2.18** for tax purposes Payments to deal with bureaucracy "faster", 2.83** 2.79*** 0.78 2.8** 2.62** 0.78 percent of sales -2.13** -2.14*** -0.65 -2.32*** -2.0** -0.9 Number of criminal attempts suffered Cummulative Contribution -8.5 -8.12 -3.57 -9.32 -9.65 -5.04 Cummulative Absolute Contribution 14.16 13.7 5.15 14.92 14.89 6.6 Infrastructure -1.65* -1.40* -0.67 -1.53* -1.25 -0.27 Average duration of power outages (log) -5.39** -5.5** -4.42*** -5.82*** -5.58** -4.4*** Days to clear customs for imports (log) -0.74** -0.79** -0.4 -0.77** -0.96*** -0.55** Shipment Losses (fraction of Sales) 2.61*** 2.39*** 2.74*** 2.46*** 2.16*** 1.32*** Dummy for Internet Access Cummulative Contribution -5.17 -5.3 -2.75 -5.66 -5.63 -3.9 Cummulative Absolute Contribution 10.39 10.08 8.23 10.58 9.95 6.54 Quality, Innovation and Labor Skills 0.22 0.23 0.23* 0.22 0.24 0.11 Fraction of computer-controlled machinery Fraction of total staff engaged in R&D 0.7** 0.74** 0.08** 0.69** 0.69** 0.46** Dummy for ISO quality certification 0.13 0.13 0.03 0.12 0.14* 0.06 Fraction of total staff with secondary education 0.33 0.42 0.14 0.31 0.49 0.35 or higher Dummy for Training provided beyond "on the 1.57*** 1.33*** 1.35*** 1.56*** 1.42*** 0.88*** job" training Cummulative Contribution 2.95 2.85 1.83 2.9 2.98 1.86 Finance and Corporate Governance 2.1*** 1.84*** 2.27*** 2.04*** 1.76*** 1.1*** Dummy for publicly listed firm Dummy for external audit of Financial 1.52*** 1.35*** 1.31*** 1.54*** 1.3*** 0.7*** statements Cummulative Contribution 3.62 3.19 3.58 3.58 3.06 1.8 Others Control Variables Age of the firm (log) 3.35** 3.07** 2.66** 3.42*** 2.73** 1.42* Share of Imported inputs (fraction) 0.94** 0.8** 1.1** 0.92** 0.59 0.57** Cummulative Contribution 4.29 3.87 3.76 4.34 3.32 1.99 Grand Total Contribution -3.35 -3.51 2.85 -4.16 -5.83 -3.32 Grand Total Absolute Contribution 35.41 33.69 22.53 36.32 34.2 18.79 Note: The asterisks correspond to the significance level of the variables in the corresponding OLS regression: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. 145 Table C-16: Honduras -- percent average (log) productivity gains and losses due to Investment Climate (IC) Table E.2: HONDURAS % Average (log) Productivity Gains and Losses in Honduras due to Investment Climate (IC) Restricted OLS Estimation Unrestricted OLS Estimation 1 Step 1 Step Solow Res. Cobb Douglas Translog Solow Res. Cobb Douglas Translog Red Tape, Corruption and Crime Number of days spent in Inspection and -6.34** -6.03** -2.15 -6.66** -6.74** -2.9* Regulation related work Fraction of sales undeclared to the tax -4.48** -4.28** -2.37* -5.13** -5.51*** -2.91** authority for tax purposes Payments to deal with bureaucracy 1.84** 1.81*** 0.51 1.85** 1.72** 0.5 "faster", percent of sales -0.82** -0.83*** -0.25 -0.91*** -0.78** -0.34 Number of criminal attempts suffered Cummulative Contribution -9.8 -9.33 -4.26 10.85 -11.31 4.97 Cummulative Absolute Contribution 13.48 12.95 5.28 14.55 14.75 6.65 Infrastructure -2.89* -2.45* -1.19 -2.73* -2.21 -0.47 Average duration of power outages (log) -3.73** -3.8** -3.09** -4.1*** -3.89** -2.99*** Days to clear customs for imports (log) -0.74** -0.79** -0.41 -0.79** -0.97*** -0.54** Shipment Losses (fraction of Sales) 1.74*** 1.6*** 1.85*** 1.7*** 1.45*** 0.87*** Dummy for Internet Access Cummulative Contribution -5.62 -5.44 -2.84 -5.92 -5.62 -3.13 Cummulative Absolute Contribution 9.1 8.64 6.54 9.32 8.52 4.87 Quality, Innovation and Labor Skills 0.2 0.21 0.22* 0.2 0.22 0.1 Fraction of computer-controlled machinery Fraction of total staff engaged in R&D 0.4** 0.42** 0.4** 0.4** 0.39** 0.25** Dummy for ISO quality certification 0.18 0.18 0.05 0.18 0.2* 0.08 Fraction of total staff with secondary 0.29 0.37 0.12 0.27 0.42 0.3 education or higher Dummy for Training provided beyond "on 1.34*** 1.13*** 1.16*** 1.35*** 1.22*** 0.73*** the job" training Cummulative Contribution 2.41 2.31 1.95 2.4 2.45 1.46 Finance and Corporate Governance 0.91*** 0.79*** 1*** 0.9*** 0.77*** 0.47*** Dummy for publicly listed firm Dummy for external audit of Financial 1.71*** 1.52*** 1.49*** 1.76*** 1.47*** 0.77*** statements Cummulative Contribution 2.62 2.31 2.49 2.66 2.24 1.24 Others Control Variables 2.95** 2.69** 2.37** 3.06*** 2.42** 1.23* Age of the firm (log) 0.78** 0.66** 0.92** 0.77** 0.49** 0.46** Share of Imported inputs (fraction) Cummulative Contribution 3.73 3.35 3.29 3.83 2.91 1.69 Grand Total Contribution -6.66 -6.8 0.63 -7.88 -9.33 -4.39 Grand Total Absolute Contribution 31.34 29.56 19.55 32.76 30.87 15.91 Note: (1) The asterisks correspond to the significance level of the variables in the corresponding OLS regression: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. 146 Table C-17: Nicaragua -- percent average (log) productivity gains and losses due to Investment Climate (IC) Table E.3 : NICARAGUA % Average (log) Productivity Gains and Losses in Nicaragua due to Investment Climate (IC) Restricted OLS Estimation Unrestricted OLS Estimation 1 Step 1 Step Solow Res. Cobb Douglas Translog Solow Res. Cobb Douglas Translog Red Tape, Corruption and Crime Number of days spent in Inspection and -7.9** -7.5** -2.64 -8.31** -8.41** -3.82* Regulation related work Fraction of sales undeclared to the tax -5.41** -5.16** -2.82* -6.18** -6.65*** -3.71** authority for tax purposes Payments to deal with bureaucracy "faster", 1.49** 1.47*** 0.41 1.5** 1.39** 0.43 percent of sales -0.72** -0.72*** -0.22 -0.8*** -0.68** -0.31 Number of criminal attempts suffered Cummulative Contribution -12.54 -11.91 -5.27 -13.79 -14.25 -7.41 Cummulative Absolute Contribution 15.52 14.85 6.09 16.79 17.13 8.27 Infrastructure -2.83* -2.39* -1.14 -2.66* -2.15 -0.49 Average duration of power outages (log) -3.91** -3.98** -3.2*** -4.29*** -4.07** -3.31*** Days to clear customs for imports (log) -0.86** -0.92** -0.47 -0.92** -1.13*** -0.67** Shipment Losses (fraction of Sales) 1.54*** 1.41*** 1.61*** 1.46*** 1.29*** 0.81*** Dummy for Internet Access Cummulative Contribution -6.06 -6.86 -3.2 -6.41 -6.06 -3.66 Cummulative Absolute Contribution 9.14 9.68 6.42 9.33 8.64 5.28 Quality, Innovation and Labor Skills 0.12 0.13 0.13* 0.13 0.13 0.06 Fraction of computer-controlled machinery Fraction of total staff engaged in R&D 0.53** 0.56** 0.06** 0.53** 0.52** 0.36** Dummy for ISO quality certification 0.19 0.19 0.05 0.17 0.2* 0.09 Fraction of total staff with secondary 0.36 0.46 0.15 0.34 0.53 0.4 education or higher Dummy for Training provided beyond "on 1.16*** 0.98*** 0.99*** 1.17*** 1.06*** 0.67*** the job" training Cummulative Contribution 2.36 2.32 1.38 2.34 2.44 1.58 Finance and Corporate Governance 1.35*** 1.18*** 1.47*** 1.34*** 1.15*** 0.74*** Dummy for publicly listed firm Dummy for external audit of Financial 1.3*** 1.16*** 1.12*** 1.34*** 1.12*** 0.62*** statements Cummulative Contribution 2.65 2.34 2.59 2.68 2.27 1.36 Others Control Variables 3.51** 3.2** 2.78** 3.63*** 2.88** 1.54* Age of the firm (log) 1.06** 0.89** 1.24** 1.05** 0.67 0.67** Share of Imported inputs (fraction) Cummulative Contribution 4.57 4.09 4.02 4.68 3.55 2.21 Grand Total Contribution -9.02 -10.02 -0.48 -10.5 -12.05 -5.92 Grand Total Absolute Contribution 34.24 33.28 20.5 35.8 34.03 18.7 Note: (1) The asterisks correspond to the significance level of the variables in the corresponding OLS regression: * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. 147 Figure C-1: Olley-Pakes decomposition of Solow residual by country Olley-Pakes Decomposition of Solow Residual 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Guatemala Honduras Nicaragua -0.5 Weighted TFP Unweighted TFP Covariance 148 Figure C-2: Olley-Pakes decomposition of Solow residual by firm size Olley-Pakes Decomposition of Solow Residual 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Small Large Small Large Small Large -0.5 Guatemala Honduras Nicaragua Weighted TFP Unweighted TFP Covariance 149 Figure C-3: Olley-Pakes decomposition of industry Solow residual Olley-Pakes Decomposition of Industry Solow Residual - Guatemala 6 5 4 3 2 1 0 el Chemicals/Rub ApparBeveragber/PlaFood Tobrniture/Leather/ es stics acco Wood Shoes erals xti les / nmetallicMin Te Fu MetalProducts No Weighted TFP Unweighted TFP Covariance Olley-Pakes Decomposition of Industry Solow Residual - Nicaragua 5 4 3 2 1 0 -1parel es co od Ap Beveragber/PlaFood stics es es ls/Rub /Tobacture/Wother/ShoicMinerals TextiltalProducts rni Fu Lea etall Me Chemica Nonm Weighted TFP Unweighted TFP Covariance Olley-Pakes Decomposition of Industry Solow Residual - Honduras 5 4 3 2 1 0 -1 parel Ap Beverageer/Plastics s ls/Rubb Food /Tobaccoture/WoodcMinerals Textiles Products Furni etalli Metal Chemica Nonm Weighted TFP Unweighted TFP Covariance 150 ANNEX D: QUALITY SYSTEMS AND R&D IN HONDURAS Box D-1: Main Components of a Typical National Quality System National Quality Standards include standards for specific sectors (e.g., IEC standards for electronics, ASTM standards for steel and metal) and general standards related to management processes (e.g., ISO quality management system standards for manufacturing and service industries). The aim of National Quality Standards is to protect human health, safety, and the environment; they also facilitate interoperability, increase efficiencies, and reduce the production of defective parts and materials. National standards should be defined on the basis of the World Trade Organization's (WTO) Code of Good Practice for the Preparation, Adoption and Application of Standards, included in the WTO's Agreement on Technical Barriers to Trade (TBT). Metrology Systems include scientific and legal metrology systems. Scientific systems include centers responsible for maintaining national measurement standards (e.g., for length, mass and volume), as well as calibration services that ensure precision of equipment for public and private sector laboratories. Legal systems include centers that have responsibility for checking that measurement standards used by firms are consistent with national measurement standards (e.g., ensuring that supermarkets' scales are correct). Certification Institutions certify compliance of firms' products and services with National Quality Standards. Certification institutions can be public or private and can be accredited by a National Accreditation Authority. Accreditation Institutions certify compliance of laboratories' functions with National Quality Standards. Like certification institutions, accreditation organizations can be public or private and can be accredited by a National Accreditation Authority. Quality Information Centers help firms and consumers identify requirements for both domestic and export markets. Many National Information Centers also serve as the Enquiry Points required under the WTO TBT Agreement. Among its duties, the Enquiry Point responds to requests for information from other WTO members concerning technical regulations, standards and certification systems in effect in its country. In return, information from the other 144 WTO members is available at the Enquiry Point. Source: Improving the Competitiveness of Honduras through a National Quality System, background report prepared by Thompson Consulting, Inc. for the World Bank. 151 Figure D-1: Institutional Setting of a Typical National Quality System MINISTRY OF TRADE QUALITY COORDINATION COUNCIL ACCREDITATION AUTHORITY ACCREDITATION BODIES Metrology: definition Development Quality of standards for of Quality Information Labs' Firms' measurement Standards Centers accreditation certification Figure D-2: Institutional Setting of the Honduran National Quality System Secretaría de Industria Comisión y Comercio Interinstitucional COHCIT Dirección General de Producción y Consumo de Normalización Comités Sectoriales Dirección General Comités Sectoriales Comités Sectoriales Subdirección General Comercialización Protección al Comercio Medicinas Metrología y Consumidor Interno Normalización Inspectoría Inspectoría Ferias Auditores de Asesoría Legal Calidad Estadísticas y Auditorias Estudios Técnicos Económicos 152 Box D-2: Institutions investing in R&D and supporting technology diffusion among Honduran firms Universities: · Universidad Autónoma de Honduras (UNAH). · Escuela Nacional de Ciencias Forestales (ESNACIFOR), carries out research on forestry and delivers consultancy services on agribusinesses SMEs through its staff and students. Government Research Centers: · Consejo Hondureño de Ciencias Forestales (CUPROFOR). · Dirección de Ciencia y Tecnología Agropecuniaria (DICTA) -- focuses on coffee, banana, shrimp, lobster, melon, wood, sugar and tobacco products. One of its main objectives is to transfer knowledge to the private sector. · Instituto Hondureño del Café (IHCAFE) -- includes a Department of Technology Transfer and Training, as well as a Laboratory for Coffee Quality Control. · Instituto Hondureño de Investigaciones Medico Veterinarias. Private and Non Profit Centers: · Fundación Hondureña de Investigación Agrícola (FHIA), a private non-profit research center. · Escuela Agrícola Panamericana (Zamorano) -- conducts applied agricultural research. Sources: Correa, P. (World Bank, 2003) and Fuster, J. Background Paper for the Honduras Trade and Productivity Project. (World Bank, 2003) 153